Fuel Cell &

Electrolyzer Module
User’s Guide
Fuel Cell & Electrolyzer Module User’s Guide
© 1998–2024 COMSOL
Protected by patents listed on [Link]/patents, or see Help > About COMSOL Multiphysics on
the File menu in the COMSOL Desktop for less detailed lists of U.S. Patents that may apply. Patents
pending.
This Documentation and the Programs described herein are furnished under the COMSOL Software License
Agreement ([Link]/sla) and may be used or copied only under the terms of the license
agreement.
COMSOL, the COMSOL logo, COMSOL Multiphysics, COMSOL Desktop, COMSOL Compiler,
COMSOL Server, and LiveLink are either registered trademarks or trademarks of COMSOL AB. All other
trademarks are the property of their respective owners, and COMSOL AB and its subsidiaries and products
are not affiliated with, endorsed by, sponsored by, or supported by those trademark owners. For a list of such
trademark owners, see [Link]/trademarks.
Version: COMSOL 6.3

Contact Information
Visit the Contact COMSOL page at [Link]/contact to submit general inquiries
or search for an address and phone number. You can also visit the Worldwide Sales Offices
page at [Link]/contact/offices for address and contact information.

If you need to contact Support, an online request form is located on the COMSOL Access
page at [Link]/support/case. Useful links:

• Support Center: [Link]/support

• Product Download: [Link]/product-download
• Product Updates: [Link]/product-update
• COMSOL Blog: [Link]/blogs
• Discussion Forum: [Link]/forum
• Events: [Link]/events
• COMSOL Video Gallery: [Link]/videos
• Support Knowledge Base: [Link]/support/knowledgebase
• Learning Center: [Link]

Part number: CM025101

C o n t e n t s

Chapter 1: Introduction

About the Fuel Cell & Electrolyzer Module 24

What Can the Fuel Cell & Electrolyzer Module Do? . . . . . . . . . 24
Fuel Cell & Electrolyzer Module Physics Interface Guide . . . . . . . . 25
Common Physics Interface and Feature Settings and Nodes . . . . . . 29
Where Do I Access the Documentation and Application Libraries? . . . . 30

Overview of the User’s Guide 34

Chapter 2: Modeling with Electrochemistry

Introduction to Electrochemistry Modeling 38

What Is Electrochemistry? . . . . . . . . . . . . . . . . . . . 38
Electrochemical Applications . . . . . . . . . . . . . . . . . . 39
Fundamentals of Electrochemistry Modeling . . . . . . . . . . . . 39
Current Distribution Cases and Choosing the Right Interface to
Model an Electrochemical Cell . . . . . . . . . . . . . . . . 41
Understanding the Different Approximations for Conservation of
Charge in Electrolytes . . . . . . . . . . . . . . . . . . . 42
Modeling Electrochemical Reactions . . . . . . . . . . . . . . . 46
Double Layer Capacitance . . . . . . . . . . . . . . . . . . . 53
Porous Electrodes . . . . . . . . . . . . . . . . . . . . . . 54
Boundary Conditions for Running and Controlling Electrochemical
Cells . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Modeling Cyclic Voltammetry . . . . . . . . . . . . . . . . . . 55
Common Simplifications When Modeling Electrochemical Cells . . . . . 56
Before You Start Building Your Model . . . . . . . . . . . . . . . 58
Meshing Advice . . . . . . . . . . . . . . . . . . . . . . . 59
Solving Electrochemical Models . . . . . . . . . . . . . . . . . 60
Postprocessing Your Solution . . . . . . . . . . . . . . . . . . 64

CONTENTS |3
 Chapter 3: Fuel Cell and Electrolyzer Interfaces

The Hydrogen Fuel Cell and Water Electrolyzer Interfaces 68

Electrolyte Phase . . . . . . . . . . . . . . . . . . . . . . . 73
Electronic Conducting Phase . . . . . . . . . . . . . . . . . . 74
Gas Phase . . . . . . . . . . . . . . . . . . . . . . . . . 74
Membrane . . . . . . . . . . . . . . . . . . . . . . . . . 77
Separator. . . . . . . . . . . . . . . . . . . . . . . . . . 78
Current Collector . . . . . . . . . . . . . . . . . . . . . . 78
Gas Diffusion Layer . . . . . . . . . . . . . . . . . . . . . . 79
Gas-Electrolyte Compartment . . . . . . . . . . . . . . . . . . 79
Gas Diffusion Electrode . . . . . . . . . . . . . . . . . . . . 80
Gas Flow Channel . . . . . . . . . . . . . . . . . . . . . . 81
Initial Values . . . . . . . . . . . . . . . . . . . . . . . . 82
Initial Values, O2 Domains . . . . . . . . . . . . . . . . . . . 83
Initial Values, H2 Domains . . . . . . . . . . . . . . . . . . . 83
Water Condensation-Evaporation . . . . . . . . . . . . . . . . 84
Water Gas Shift Reaction . . . . . . . . . . . . . . . . . . . 84
Reaction Sources . . . . . . . . . . . . . . . . . . . . . . . 85
Gas Diffusion Electrode Reaction. . . . . . . . . . . . . . . . . 85
Porous Double Layer Capacitance . . . . . . . . . . . . . . . . 87
Overview of the Boundary Nodes in the Hydrogen Fuel Cell and
Water Electrolyzer Interfaces. . . . . . . . . . . . . . . . . 87
Insulation . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Electrolyte Potential . . . . . . . . . . . . . . . . . . . . . 88
Electrolyte Current Density. . . . . . . . . . . . . . . . . . . 89
Internal Electrolyte Contact Resistance . . . . . . . . . . . . . . 89
Electric Ground . . . . . . . . . . . . . . . . . . . . . . . 89
Electric Potential . . . . . . . . . . . . . . . . . . . . . . . 89
Electrode Current . . . . . . . . . . . . . . . . . . . . . . 90
Electrode Current Density . . . . . . . . . . . . . . . . . . . 90
Electrode Power . . . . . . . . . . . . . . . . . . . . . . . 91
Internal Electrode Contact Resistance . . . . . . . . . . . . . . . 91
No Flux . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Inlet . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Outlet . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

4 | CONTENTS
 Water Absorption-Desorption . . . . . . . . . . . . . . . . . 94
Water Flux . . . . . . . . . . . . . . . . . . . . . . . . . 95
Water Chemical Potential . . . . . . . . . . . . . . . . . . . 95
Thin Gas Diffusion Electrode . . . . . . . . . . . . . . . . . . 95
Electrode Surface. . . . . . . . . . . . . . . . . . . . . . . 96
Internal Electrode Surface . . . . . . . . . . . . . . . . . . . 96
Thin Gas Diffusion Electrode Reaction. . . . . . . . . . . . . . . 96
Electrode Reaction . . . . . . . . . . . . . . . . . . . . . . 97
Periodic Condition . . . . . . . . . . . . . . . . . . . . . . 97

Theory for the Hydrogen Fuel Cell and Water Electrolyzer

Interfaces 98
Electronic Conducting Phase Theory . . . . . . . . . . . . . . . 98
Electrolyte Phase Theory . . . . . . . . . . . . . . . . . . . . 99
Gas Phase Theory . . . . . . . . . . . . . . . . . . . . . . 99
Calculation of Built-in Equilibrium and Thermoneutral Potentials . . . . 102
Membrane Crossover Theory . . . . . . . . . . . . . . . . . 104
Electroosmotic Water Drag Theory . . . . . . . . . . . . . . 109
References . . . . . . . . . . . . . . . . . . . . . . . . 113

Chapter 4: Electrochemistry Interfaces

The Primary and Secondary Current Distribution Interfaces 116

The Primary Current Distribution and Secondary Current Distribution
Interfaces . . . . . . . . . . . . . . . . . . . . . . . 116
Electrolyte . . . . . . . . . . . . . . . . . . . . . . . . 119
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 120
Porous Electrode. . . . . . . . . . . . . . . . . . . . . . 120
Periodic Condition . . . . . . . . . . . . . . . . . . . . . 122
Thin Electrolyte Layer . . . . . . . . . . . . . . . . . . . . 122
Edge Electrode. . . . . . . . . . . . . . . . . . . . . . . 123
Electrode Line Current Source . . . . . . . . . . . . . . . . 124
Electrolyte Line Current Source . . . . . . . . . . . . . . . . 124
Electrode Symmetry Axis Current Source . . . . . . . . . . . . 124
Electrolyte Symmetry Axis Current Source . . . . . . . . . . . . 125
External Current Source . . . . . . . . . . . . . . . . . . . 125

CONTENTS |5
 Electrode Point Current Source . . . . . . . . . . . . . . . . 125
Electrolyte Point Current Source. . . . . . . . . . . . . . . . 125
Electrode Current . . . . . . . . . . . . . . . . . . . . . 126

The Tertiary Current Distribution, Nernst–Planck Interface 127

The Tertiary Current Distribution, Nernst–Planck Interface . . . . . 127
Electrolyte . . . . . . . . . . . . . . . . . . . . . . . . 132
Porous Electrode. . . . . . . . . . . . . . . . . . . . . . 133
Separator. . . . . . . . . . . . . . . . . . . . . . . . . 134
Reactions. . . . . . . . . . . . . . . . . . . . . . . . . 134
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 135
Ion Exchange Membrane . . . . . . . . . . . . . . . . . . . 135
Ion Exchange Membrane Boundary . . . . . . . . . . . . . . . 136
Thin Electrolyte Layer . . . . . . . . . . . . . . . . . . . . 136

Shared Physics Features in the Current Distribution Interfaces

138
Domain, Boundary, Pair, Edge, and Point Nodes for the
Electrochemistry Interfaces . . . . . . . . . . . . . . . . 138
Current Conductor . . . . . . . . . . . . . . . . . . . . . 140
Highly Conductive Porous Electrode . . . . . . . . . . . . . . 140
Electrode Current Source . . . . . . . . . . . . . . . . . . 140
Electrolyte Current Source . . . . . . . . . . . . . . . . . . 141
Porous Electrode Reaction . . . . . . . . . . . . . . . . . . 141
Porous Matrix Double-Layer Capacitance . . . . . . . . . . . . 142
Insulation . . . . . . . . . . . . . . . . . . . . . . . . . 142
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 143
Electrode Surface. . . . . . . . . . . . . . . . . . . . . . 143
Electrode Reaction . . . . . . . . . . . . . . . . . . . . . 148
Double-Layer Capacitance . . . . . . . . . . . . . . . . . . 153
Perforated Electrode Surface . . . . . . . . . . . . . . . . . 154
Internal Electrode Surface . . . . . . . . . . . . . . . . . . 154
Electrolyte Potential . . . . . . . . . . . . . . . . . . . . 155
Electrolyte Current . . . . . . . . . . . . . . . . . . . . . 155
Electrolyte Current Density. . . . . . . . . . . . . . . . . . 156
Thin Electrode Layer . . . . . . . . . . . . . . . . . . . . 156
Electrode-Electrolyte Boundary Interface. . . . . . . . . . . . . 156
Electric Ground . . . . . . . . . . . . . . . . . . . . . . 157

6 | CONTENTS
Electric Potential . . . . . . . . . . . . . . . . . . . . . . 158
Electrode Current Density . . . . . . . . . . . . . . . . . . 158
Electrode Current . . . . . . . . . . . . . . . . . . . . . 158
Electrode Power . . . . . . . . . . . . . . . . . . . . . . 159
Harmonic Perturbation . . . . . . . . . . . . . . . . . . . 160
Electrode Potential . . . . . . . . . . . . . . . . . . . . . 160
External Short . . . . . . . . . . . . . . . . . . . . . . . 161
Initial Values for Adsorbing–Desorbing Species . . . . . . . . . . 161
Initial Values for Dissolving–Depositing Species . . . . . . . . . . 161
Nonfaradaic Reactions . . . . . . . . . . . . . . . . . . . . 162
Reference Electrode . . . . . . . . . . . . . . . . . . . . 162
Electric Reference Potential . . . . . . . . . . . . . . . . . . 162
Circuit Terminal . . . . . . . . . . . . . . . . . . . . . . 163

The Concentrated Electrolyte Transport Interface 164

Electrolyte . . . . . . . . . . . . . . . . . . . . . . . . 166
Diffusion Coefficients . . . . . . . . . . . . . . . . . . . . 167
Separator. . . . . . . . . . . . . . . . . . . . . . . . . 167
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 168
Species Sources . . . . . . . . . . . . . . . . . . . . . . 168
Reference Electrode . . . . . . . . . . . . . . . . . . . . 168
No Flux . . . . . . . . . . . . . . . . . . . . . . . . . 169
Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Inflow . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Outflow . . . . . . . . . . . . . . . . . . . . . . . . . 169
Electrolyte Potential . . . . . . . . . . . . . . . . . . . . 169

The Electrode, Shell Interface 170

Boundary, Edge, Point, and Pair Nodes for the Electrode, Shell
Interface. . . . . . . . . . . . . . . . . . . . . . . . 171
Electrode . . . . . . . . . . . . . . . . . . . . . . . . . 172
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 173
External Current Density . . . . . . . . . . . . . . . . . . 173
Current Source . . . . . . . . . . . . . . . . . . . . . . 173
Normal Current Density . . . . . . . . . . . . . . . . . . . 174
Electric Insulation . . . . . . . . . . . . . . . . . . . . . 174
Boundary Current Source . . . . . . . . . . . . . . . . . . 174
Ground . . . . . . . . . . . . . . . . . . . . . . . . . 174

CONTENTS |7
 Electric Potential . . . . . . . . . . . . . . . . . . . . . . 175

The Electroanalysis Model Wizard Entry 176

Theory for the Current Distribution Interfaces 177

The Nernst–Planck Equations . . . . . . . . . . . . . . . . . 177
Domain Equations for Primary and Secondary Current Distributions . . 179
Electrochemical Reactions and the Difference Between a Primary and
a Secondary Current Distribution . . . . . . . . . . . . . . 180
Domain Equations for Tertiary Current Distributions Using the
Nernst–Planck Equations and Electroneutrality . . . . . . . . . 183
Mass Fluxes and Sources Due to Electrochemical Reactions . . . . . 184
Deposition-Dissolution Rates, Growth Velocities, and Thicknesses on
an Electrode Surface . . . . . . . . . . . . . . . . . . . 185
Stoichiometric Coefficients for Double Layer Capacitive Charging . . . 186
Film Resistance . . . . . . . . . . . . . . . . . . . . . . 187
Equilibrium Potentials and the Nernst Equation . . . . . . . . . . 187
Electrode Kinetics Expressions . . . . . . . . . . . . . . . . 188
Theory for Specific Current Distribution Feature Nodes . . . . . . . 193
Reference . . . . . . . . . . . . . . . . . . . . . . . . 201

Theory for the Concentrated Electrolyte Transport Interface 202

Dependent Variables, Species, and Electrolyte Components . . . . . 202
Volumetric Equations of State . . . . . . . . . . . . . . . . . 203
Flux Definitions and Material Balances . . . . . . . . . . . . . . 204
Chemical Potentials and the Stefan–Maxwell–Onsager Relations . . . . 205
The Reference Electrode . . . . . . . . . . . . . . . . . . . 206
Reference . . . . . . . . . . . . . . . . . . . . . . . . 207

Theory for Electrochemical Heat Sources 208

The First Law of Thermodynamics . . . . . . . . . . . . . . . 209
Electrochemical Contributions to The Heat Equation . . . . . . . . 209
Joule Heating Due to Charge Transport . . . . . . . . . . . . . 210
Heating Due to Electrochemical Reactions . . . . . . . . . . . . 211
Heating Due to Heat of Mixing . . . . . . . . . . . . . . . . 213
Total Overpotential Calculation for Concentration Dependent Kinetics . 213
References . . . . . . . . . . . . . . . . . . . . . . . . 214

8 | CONTENTS
 Theory for the Electrode, Shell Interface 215
Governing Equations . . . . . . . . . . . . . . . . . . . . 215
Coupling to Other Physics Interfaces . . . . . . . . . . . . . . 215

Theory for Electroanalysis 217

Electroanalytical Methods . . . . . . . . . . . . . . . . . . 217
Supporting Electrolyte . . . . . . . . . . . . . . . . . . . . 217
Domain Equations for the Electroanalysis Case . . . . . . . . . . 218
Electrode Boundary Conditions in the Electroanalysis Model . . . . . 220
The Electroanalytical Butler–Volmer Equation . . . . . . . . . . . 222
Counter Electrodes and Overall Charge Balance . . . . . . . . . . 223

Electrode Potentials and Reference Electrodes 224

Reference Electrodes . . . . . . . . . . . . . . . . . . . . 224
Boundary Conditions Using Reference Electrode Potentials. . . . . . 225
Nodes for Handling Electrode Potentials and Reference Electrodes. . . 225

Chapter 5: Chemical Species Transport Interfaces

The Transport of Diluted Species Interface 228

The Transport of Diluted Species in Porous Media Interface . . . . . 232
Domain, Boundary, and Pair Nodes for the Transport of Diluted
Species Interface. . . . . . . . . . . . . . . . . . . . . 234
Prescribing Conditions on Fluid-Solid Interfaces . . . . . . . . . . 236
Species Properties . . . . . . . . . . . . . . . . . . . . . 236
Fluid . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Solid . . . . . . . . . . . . . . . . . . . . . . . . . . 239
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 240
Mass-Based Concentrations . . . . . . . . . . . . . . . . . . 240
Reactions. . . . . . . . . . . . . . . . . . . . . . . . . 241
No Flux . . . . . . . . . . . . . . . . . . . . . . . . . 242
Inflow . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Outflow . . . . . . . . . . . . . . . . . . . . . . . . . 243
Concentration . . . . . . . . . . . . . . . . . . . . . . . 244
Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 246

CONTENTS |9
 Flux Discontinuity . . . . . . . . . . . . . . . . . . . . . 246
Partition Condition . . . . . . . . . . . . . . . . . . . . . 246
Periodic Condition . . . . . . . . . . . . . . . . . . . . . 247
Line Mass Source . . . . . . . . . . . . . . . . . . . . . . 249
Point Mass Source . . . . . . . . . . . . . . . . . . . . . 249
Open Boundary . . . . . . . . . . . . . . . . . . . . . . 250
Thin Diffusion Barrier . . . . . . . . . . . . . . . . . . . . 250
Thin Impermeable Barrier . . . . . . . . . . . . . . . . . . 251
Equilibrium Reaction . . . . . . . . . . . . . . . . . . . . 251
Surface Reactions . . . . . . . . . . . . . . . . . . . . . 252
Surface Equilibrium Reaction . . . . . . . . . . . . . . . . . 252
Fast Irreversible Surface Reaction . . . . . . . . . . . . . . . 253
Porous Electrode Coupling . . . . . . . . . . . . . . . . . . 253
Reaction Coefficients . . . . . . . . . . . . . . . . . . . . 254
Electrode Surface Coupling . . . . . . . . . . . . . . . . . . 254
Porous Medium . . . . . . . . . . . . . . . . . . . . . . 255
Fluid (Porous Medium) . . . . . . . . . . . . . . . . . . . 255
Porous Matrix . . . . . . . . . . . . . . . . . . . . . . . 257
Dispersion . . . . . . . . . . . . . . . . . . . . . . . . 258
Unsaturated Porous Medium . . . . . . . . . . . . . . . . . 258
Liquid . . . . . . . . . . . . . . . . . . . . . . . . . . 259
Gas . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Adsorption . . . . . . . . . . . . . . . . . . . . . . . . 262
Volatilization . . . . . . . . . . . . . . . . . . . . . . . 264
Species Source. . . . . . . . . . . . . . . . . . . . . . . 264
Hygroscopic Swelling . . . . . . . . . . . . . . . . . . . . 265
Out-of-Plane Flux . . . . . . . . . . . . . . . . . . . . . 266
Fracture . . . . . . . . . . . . . . . . . . . . . . . . . 268
Fluid (Fracture) . . . . . . . . . . . . . . . . . . . . . . 268
Fracture Material . . . . . . . . . . . . . . . . . . . . . . 269

The Transport of Diluted Species in Fractures Interface 270

Boundary, Edge, Point, and Pair Nodes for the Transport of Diluted
Species in Fractures Interface . . . . . . . . . . . . . . . . 272
Adsorption . . . . . . . . . . . . . . . . . . . . . . . . 273
Concentration . . . . . . . . . . . . . . . . . . . . . . . 275
Dispersion . . . . . . . . . . . . . . . . . . . . . . . . 275
Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . 276

10 | C O N T E N T S
Inflow . . . . . . . . . . . . . . . . . . . . . . . . . . 276
No Flux . . . . . . . . . . . . . . . . . . . . . . . . . 277
Outflow . . . . . . . . . . . . . . . . . . . . . . . . . 277
Reactions. . . . . . . . . . . . . . . . . . . . . . . . . 277
Species Source. . . . . . . . . . . . . . . . . . . . . . . 278

The Transport of Concentrated Species Interface 280

The Transport of Concentrated Species in Porous Media Interface . . . 286
Domain, Boundary, and Pair Nodes for the Transport of
Concentrated Species Interface . . . . . . . . . . . . . . . 287
Prescribing Conditions on Fluid-Solid Interfaces . . . . . . . . . . 288
Species Properties . . . . . . . . . . . . . . . . . . . . . 288
Fluid . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Porous Medium . . . . . . . . . . . . . . . . . . . . . . 292
Fluid (Porous Medium) . . . . . . . . . . . . . . . . . . . 292
Porous Matrix . . . . . . . . . . . . . . . . . . . . . . . 296
Electrode Surface Coupling . . . . . . . . . . . . . . . . . . 296
Reaction . . . . . . . . . . . . . . . . . . . . . . . . . 297
Reaction Sources . . . . . . . . . . . . . . . . . . . . . . 299
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 299
Mass Fraction . . . . . . . . . . . . . . . . . . . . . . . 300
Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Inflow . . . . . . . . . . . . . . . . . . . . . . . . . . 302
No Flux . . . . . . . . . . . . . . . . . . . . . . . . . 303
Out-of-Plane Flux . . . . . . . . . . . . . . . . . . . . . 303
Outflow . . . . . . . . . . . . . . . . . . . . . . . . . 305
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 306
Flux Discontinuity . . . . . . . . . . . . . . . . . . . . . 306
Open Boundary . . . . . . . . . . . . . . . . . . . . . . 307
Equilibrium Reaction . . . . . . . . . . . . . . . . . . . . 307
Surface Equilibrium Reaction . . . . . . . . . . . . . . . . . 308

The Chemistry Interface 310

Feature Nodes Available for the Chemistry Interface . . . . . . . . 315
Reaction . . . . . . . . . . . . . . . . . . . . . . . . . 315
Species . . . . . . . . . . . . . . . . . . . . . . . . . 323
Electrode Reaction . . . . . . . . . . . . . . . . . . . . . 326
Electrode Reaction Group . . . . . . . . . . . . . . . . . . 328

CONTENTS | 11
 Reversible Reaction Group . . . . . . . . . . . . . . . . . . 328
Equilibrium Reaction Group. . . . . . . . . . . . . . . . . . 330
Species Group . . . . . . . . . . . . . . . . . . . . . . . 331
Reaction Thermodynamics . . . . . . . . . . . . . . . . . . 331
Species Activity . . . . . . . . . . . . . . . . . . . . . . 331
Species Thermodynamics. . . . . . . . . . . . . . . . . . . 332

The Nernst–Planck–Poisson Equations Interface 333

The Electrophoretic Transport Interface 335

Common Settings for the Species Nodes in the Electrophoretic
Transport Interface . . . . . . . . . . . . . . . . . . . 338
Diffusion and Migration Settings . . . . . . . . . . . . . . . . 340
Domain, Boundary, and Pair Nodes for the Electrophoretic Transport
Interface. . . . . . . . . . . . . . . . . . . . . . . . 341
Solvent . . . . . . . . . . . . . . . . . . . . . . . . . 342
Porous Matrix Properties . . . . . . . . . . . . . . . . . . 342
Fully Dissociated Species . . . . . . . . . . . . . . . . . . . 342
Uncharged Species . . . . . . . . . . . . . . . . . . . . . 342
Weak Acid . . . . . . . . . . . . . . . . . . . . . . . . 343
Weak Base . . . . . . . . . . . . . . . . . . . . . . . . 343
Ampholyte . . . . . . . . . . . . . . . . . . . . . . . . 343
Protein . . . . . . . . . . . . . . . . . . . . . . . . . 344
Current Source . . . . . . . . . . . . . . . . . . . . . . 344
Initial Potential. . . . . . . . . . . . . . . . . . . . . . . 344
Current . . . . . . . . . . . . . . . . . . . . . . . . . 344
Current Density . . . . . . . . . . . . . . . . . . . . . . 344
Insulation . . . . . . . . . . . . . . . . . . . . . . . . . 345
Potential . . . . . . . . . . . . . . . . . . . . . . . . . 345
Species Source. . . . . . . . . . . . . . . . . . . . . . . 345
Initial Concentration . . . . . . . . . . . . . . . . . . . . 346
Concentration . . . . . . . . . . . . . . . . . . . . . . . 346
No Flux . . . . . . . . . . . . . . . . . . . . . . . . . 346
Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
Inflow . . . . . . . . . . . . . . . . . . . . . . . . . . 347
Outflow . . . . . . . . . . . . . . . . . . . . . . . . . 347

12 | C O N T E N T S
The Surface Reactions Interface 348
Boundary, Edge, Point, and Pair Nodes for the Surface Reactions
Interface. . . . . . . . . . . . . . . . . . . . . . . . 349
Surface Properties . . . . . . . . . . . . . . . . . . . . . 350
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 351
Reactions. . . . . . . . . . . . . . . . . . . . . . . . . 351
Surface Concentration . . . . . . . . . . . . . . . . . . . . 352

The Reacting Flow Interfaces 353

The Reacting Laminar Flow Interface . . . . . . . . . . . . . . 353
The Reacting Flow Coupling Feature . . . . . . . . . . . . . . 354
Physics Interface Features . . . . . . . . . . . . . . . . . . 358

The Reacting Flow in Porous Media Interfaces 360

The Reacting Flow in Porous Media, Transport of Diluted Species
Interface. . . . . . . . . . . . . . . . . . . . . . . . 360
The Reacting Flow in Porous Media, Transport of Concentrated
Species Interface. . . . . . . . . . . . . . . . . . . . . 361
The Reacting Flow, Diluted Species Coupling Feature . . . . . . . . 362
The Reacting Flow Coupling Feature . . . . . . . . . . . . . . 362
Physics Interface Features . . . . . . . . . . . . . . . . . . 362

The Nonisothermal Reacting Flow Multiphysics Interfaces 363

The Nonisothermal Reacting Laminar Flow Interface . . . . . . . . . 363
The Nonisothermal Reacting Turbulent Flow, k- Interface. . . . . . 364
The Nonisothermal Reacting Turbulent Flow, k- Interface . . . . . 365
The Nonisothermal Reacting Turbulent Flow, SST Interface . . . . . 367
The Nonisothermal Reacting Turbulent Flow, Low Re k- Interface . . 368
The Reacting Flow Coupling Feature . . . . . . . . . . . . . . 369
Physics Interface Features . . . . . . . . . . . . . . . . . . 369

Theory for the Transport of Diluted Species Interface 371

Mass Balance Equation . . . . . . . . . . . . . . . . . . . . 372
Equilibrium Reaction Theory . . . . . . . . . . . . . . . . . 373
Material Balance Formulation . . . . . . . . . . . . . . . . . 375
Solving a Diffusion Equation Only . . . . . . . . . . . . . . . 375
Mass Sources for Species Transport . . . . . . . . . . . . . . 376
Adding Transport Through Migration . . . . . . . . . . . . . . 378

CONTENTS | 13
 Supporting Electrolytes . . . . . . . . . . . . . . . . . . . 379
Crosswind Diffusion . . . . . . . . . . . . . . . . . . . . 380
Danckwerts Inflow Boundary Condition . . . . . . . . . . . . . 381
Mass Balance Equation for Transport of Diluted Species in Porous
Media . . . . . . . . . . . . . . . . . . . . . . . . . 382
Convection in Porous Media . . . . . . . . . . . . . . . . . 383
Diffusion in Porous Media . . . . . . . . . . . . . . . . . . 385
Dispersion . . . . . . . . . . . . . . . . . . . . . . . . 386
Adsorption . . . . . . . . . . . . . . . . . . . . . . . . 388
Reactions. . . . . . . . . . . . . . . . . . . . . . . . . 390
Mass Transport in Fractures . . . . . . . . . . . . . . . . . 391
References . . . . . . . . . . . . . . . . . . . . . . . . 392

Theory for the Transport of Concentrated Species Interface 394

Multicomponent Mass Transport . . . . . . . . . . . . . . . . 394
Multicomponent Gas Diffusion: Maxwell–Stefan Description . . . . . 395
Multicomponent Diffusivities . . . . . . . . . . . . . . . . . 397
Multicomponent Diffusion: Mixture-Averaged Approximation . . . . . 399
Multispecies Diffusion: Fick’s Law Approximation . . . . . . . . . 401
Multicomponent Thermal Diffusion . . . . . . . . . . . . . . . 402
Pore Wall Interactions. . . . . . . . . . . . . . . . . . . . 402
Regularization of Reaction Rate Expression . . . . . . . . . . . . 403
References for the Transport of Concentrated Species Interface . . . . 404

Theory for the Electrophoretic Transport Interface 406

Theory for the Surface Reactions Interface 412

Governing Equations for the Surface Concentrations . . . . . . . . 412
Governing Equations for the Bulk Concentrations . . . . . . . . . 413
ODE Formulations for Surface Concentrations . . . . . . . . . . 415
Surface Reaction Equations on Deforming Geometries . . . . . . . 416
Reference for the Surface Reactions Interface . . . . . . . . . . . 417

Theory for the Coupling of Mass Transport to

Electrochemical Reactions 418
Molar Sources and Sinks . . . . . . . . . . . . . . . . . . . 418
Mass Sources and Sinks . . . . . . . . . . . . . . . . . . . 419

14 | C O N T E N T S
 Theory for the Reacting Flow Interface 420
Pseudo Time Stepping for Mass Transport . . . . . . . . . . . . 420
The Stefan Velocity . . . . . . . . . . . . . . . . . . . . . 420
The Chemical Reaction Rate . . . . . . . . . . . . . . . . . 422
Nonisothermal Reacting Flow . . . . . . . . . . . . . . . . . 423
References for the Reacting Flow Interfaces . . . . . . . . . . . 426

Chapter 6: Fluid Flow Interfaces

The Laminar Flow and Creeping Flow Interfaces 429

The Creeping Flow Interface . . . . . . . . . . . . . . . . . 429
The Laminar Flow Interface . . . . . . . . . . . . . . . . . . 430
Domain, Boundary, Pair, and Point Nodes for Single-Phase Flow . . . . . 436
Fluid Properties . . . . . . . . . . . . . . . . . . . . . . 436
Volume Force . . . . . . . . . . . . . . . . . . . . . . . 438
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 438
Wall . . . . . . . . . . . . . . . . . . . . . . . . . . 439
Inlet . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
Outlet . . . . . . . . . . . . . . . . . . . . . . . . . . 446
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 449
Open Boundary . . . . . . . . . . . . . . . . . . . . . . 449
Boundary Stress . . . . . . . . . . . . . . . . . . . . . . 450
Periodic Flow Condition . . . . . . . . . . . . . . . . . . . 452
Porous Interface . . . . . . . . . . . . . . . . . . . . . . 454
Flow Continuity . . . . . . . . . . . . . . . . . . . . . . 454
Pressure Point Constraint . . . . . . . . . . . . . . . . . . 455
Point Mass Source . . . . . . . . . . . . . . . . . . . . . 456
Line Mass Source . . . . . . . . . . . . . . . . . . . . . . 456
Gravity . . . . . . . . . . . . . . . . . . . . . . . . . 457

The Darcy’s Law Interface 458

Domain, Boundary, Edge, Point, and Pair Nodes for the Darcy’s
Law Interface . . . . . . . . . . . . . . . . . . . . . . 460
Porous Medium . . . . . . . . . . . . . . . . . . . . . . 462
Fluid . . . . . . . . . . . . . . . . . . . . . . . . . . 464
Porous Matrix . . . . . . . . . . . . . . . . . . . . . . . 466

CONTENTS | 15
 Unsaturated Porous Medium . . . . . . . . . . . . . . . . . 469
Dual Porosity Medium . . . . . . . . . . . . . . . . . . . . 471
Dual Permeability Medium . . . . . . . . . . . . . . . . . . 473
Unsaturated Dual Permeability Medium . . . . . . . . . . . . . 474
Macropores . . . . . . . . . . . . . . . . . . . . . . . . 476
Micropores . . . . . . . . . . . . . . . . . . . . . . . . 478
Mass Source . . . . . . . . . . . . . . . . . . . . . . . 480
Contributing Velocity . . . . . . . . . . . . . . . . . . . . 481
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 482
Gravity . . . . . . . . . . . . . . . . . . . . . . . . . 483
Cross Section . . . . . . . . . . . . . . . . . . . . . . . 484
Thickness. . . . . . . . . . . . . . . . . . . . . . . . . 485
Porous Electrode Coupling . . . . . . . . . . . . . . . . . . 485
Electrode Surface Coupling . . . . . . . . . . . . . . . . . . 486
Pressure . . . . . . . . . . . . . . . . . . . . . . . . . 487
Mass Flux. . . . . . . . . . . . . . . . . . . . . . . . . 488
Line Mass Source . . . . . . . . . . . . . . . . . . . . . . 489
Point Mass Source . . . . . . . . . . . . . . . . . . . . . 490
Inlet . . . . . . . . . . . . . . . . . . . . . . . . . . . 490
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 492
No Flow . . . . . . . . . . . . . . . . . . . . . . . . . 492
Flux Discontinuity . . . . . . . . . . . . . . . . . . . . . 493
Outlet . . . . . . . . . . . . . . . . . . . . . . . . . . 494
Precipitation . . . . . . . . . . . . . . . . . . . . . . . 495
Interior Wall . . . . . . . . . . . . . . . . . . . . . . . 495
Thin Barrier. . . . . . . . . . . . . . . . . . . . . . . . 496
Pressure Head . . . . . . . . . . . . . . . . . . . . . . . 496
Hydraulic Head . . . . . . . . . . . . . . . . . . . . . . 497
Atmosphere/Gauge . . . . . . . . . . . . . . . . . . . . . 497
Pervious Layer . . . . . . . . . . . . . . . . . . . . . . . 498
Well . . . . . . . . . . . . . . . . . . . . . . . . . . 500
Fracture . . . . . . . . . . . . . . . . . . . . . . . . . 500
Fluid (Fracture) . . . . . . . . . . . . . . . . . . . . . . 503
Fracture Material . . . . . . . . . . . . . . . . . . . . . . 504

The Free and Porous Media Flow, Brinkman Interface 507

Domain, Boundary, Point, and Pair Nodes for the Free and Porous
Media Flow, Brinkman Interface . . . . . . . . . . . . . . . 509

16 | C O N T E N T S
Fluid Properties . . . . . . . . . . . . . . . . . . . . . . 510
Porous Medium . . . . . . . . . . . . . . . . . . . . . . 511
Fluid . . . . . . . . . . . . . . . . . . . . . . . . . . 512
Porous Matrix . . . . . . . . . . . . . . . . . . . . . . . 513
Mass Source . . . . . . . . . . . . . . . . . . . . . . . 514
Volume Force . . . . . . . . . . . . . . . . . . . . . . . 514
Porous Electrode Coupling . . . . . . . . . . . . . . . . . . 515
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 515
Gravity . . . . . . . . . . . . . . . . . . . . . . . . . 515
Electrode-Electrolyte Interface Coupling . . . . . . . . . . . . . 516
Wall . . . . . . . . . . . . . . . . . . . . . . . . . . 516

The Free and Porous Media Flow, Darcy Interface 518

Free and Porous Media Flow Coupling (Multiphysics Coupling) . . . . 518

The Brinkman Equations Interface 522

Domain, Boundary, Point, and Pair Nodes for the Brinkman
Equations Interface . . . . . . . . . . . . . . . . . . . . 525
Porous Medium . . . . . . . . . . . . . . . . . . . . . . 526
Fluid . . . . . . . . . . . . . . . . . . . . . . . . . . 528
Porous Matrix . . . . . . . . . . . . . . . . . . . . . . . 529
Mass Source . . . . . . . . . . . . . . . . . . . . . . . 531
Volume Force . . . . . . . . . . . . . . . . . . . . . . . 531
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 532
Fluid Properties . . . . . . . . . . . . . . . . . . . . . . 532
Gravity . . . . . . . . . . . . . . . . . . . . . . . . . 534

The Multiphase Flow in Porous Media Interface 535

Multiphase Flow in Porous Media (Multiphysics Coupling) . . . . . . 536
Well (Multiphysics Coupling) . . . . . . . . . . . . . . . . . 537
Thin Barrier (Multiphysics Coupling) . . . . . . . . . . . . . . 540

The Multiphase Free and Porous Media Flow Interface 543

Coupling Features . . . . . . . . . . . . . . . . . . . . . 544

The Bubbly Flow Interface 545

The Laminar Bubbly Flow Interface . . . . . . . . . . . . . . . 545
Domain and Boundary Nodes for Bubbly Flow . . . . . . . . . . 548

CONTENTS | 17
 Fluid Properties . . . . . . . . . . . . . . . . . . . . . . 549
Mass Transfer . . . . . . . . . . . . . . . . . . . . . . . 552
Gravity . . . . . . . . . . . . . . . . . . . . . . . . . 553
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 554
Wall . . . . . . . . . . . . . . . . . . . . . . . . . . 554
Inlet . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
Outlet . . . . . . . . . . . . . . . . . . . . . . . . . . 558
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 559
Periodic Flow Condition . . . . . . . . . . . . . . . . . . . 559
Interior Wall . . . . . . . . . . . . . . . . . . . . . . . 560
Flow Continuity . . . . . . . . . . . . . . . . . . . . . . 562
Gas Boundary Condition Equations . . . . . . . . . . . . . . . 562

The Mixture Model Interface 564

The Mixture Model, Laminar Flow Interface. . . . . . . . . . . . 564
Domain and Boundary Nodes for the Mixture Model Interface . . . . 568
Mixture Properties . . . . . . . . . . . . . . . . . . . . . 569
Mass Transfer . . . . . . . . . . . . . . . . . . . . . . . 572
Gravity . . . . . . . . . . . . . . . . . . . . . . . . . 573
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 574
Wall . . . . . . . . . . . . . . . . . . . . . . . . . . 575
Inlet . . . . . . . . . . . . . . . . . . . . . . . . . . . 578
Outlet . . . . . . . . . . . . . . . . . . . . . . . . . . 579
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 580
Periodic Flow Condition . . . . . . . . . . . . . . . . . . . 581
Interior Wall . . . . . . . . . . . . . . . . . . . . . . . 581
Mixture Continuity . . . . . . . . . . . . . . . . . . . . . 583

The Euler-Euler Model Interface 584

The Euler–Euler Model, Laminar Flow Interface . . . . . . . . . . 584
Domain, Boundary, Point, and Pair Nodes for the Euler–Euler Model
Interfaces . . . . . . . . . . . . . . . . . . . . . . . 587
Phase Properties . . . . . . . . . . . . . . . . . . . . . . 588
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 592
Mass Transfer . . . . . . . . . . . . . . . . . . . . . . . 593
Gravity . . . . . . . . . . . . . . . . . . . . . . . . . 593
Volume Force . . . . . . . . . . . . . . . . . . . . . . . 593
Wall . . . . . . . . . . . . . . . . . . . . . . . . . . 594

18 | C O N T E N T S
Inlet . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
Outlet . . . . . . . . . . . . . . . . . . . . . . . . . . 597
Interior Wall . . . . . . . . . . . . . . . . . . . . . . . 598
Flow Continuity . . . . . . . . . . . . . . . . . . . . . . 599

Theory for the Laminar Flow and Creeping Flow Interfaces 601
General Single-Phase Flow Theory . . . . . . . . . . . . . . . 602
Compressible Flow . . . . . . . . . . . . . . . . . . . . . 604
Weakly Compressible Flow . . . . . . . . . . . . . . . . . . 604
The Mach Number Limit . . . . . . . . . . . . . . . . . . . 604
Incompressible Flow . . . . . . . . . . . . . . . . . . . . 605
The Reynolds Number. . . . . . . . . . . . . . . . . . . . 606
Gravity . . . . . . . . . . . . . . . . . . . . . . . . . 607
Theory for the Wall Boundary Condition . . . . . . . . . . . . 608
Prescribing Inlet and Outlet Conditions . . . . . . . . . . . . . 612
Mass Flow . . . . . . . . . . . . . . . . . . . . . . . . 614
Fully Developed Flow (Inlet) . . . . . . . . . . . . . . . . . 616
Fully Developed Flow (Outlet). . . . . . . . . . . . . . . . . 617
No Viscous Stress . . . . . . . . . . . . . . . . . . . . . 618
Normal Stress Boundary Condition . . . . . . . . . . . . . . . 619
Pressure Boundary Condition . . . . . . . . . . . . . . . . . 619
Porous Interface . . . . . . . . . . . . . . . . . . . . . . 622
Mass Sources for Fluid Flow . . . . . . . . . . . . . . . . . 625
Numerical Stability — Stabilization Techniques for Fluid Flow . . . . . 627
Solvers for Laminar Flow . . . . . . . . . . . . . . . . . . . 629
Pseudo Time Stepping for Laminar Flow Models . . . . . . . . . . 631
Nitsche Constraints. . . . . . . . . . . . . . . . . . . . . 633
Particle Tracing in Fluid Flow . . . . . . . . . . . . . . . . . 633
References for the Single-Phase Flow, Laminar Flow Interfaces . . . . 634

Theory for the Darcy’s Law Interface 637

About Darcy’s Law . . . . . . . . . . . . . . . . . . . . . 637
Darcy’s Law — Equation Formulation . . . . . . . . . . . . . . 638
Storage Models . . . . . . . . . . . . . . . . . . . . . . 639
Average Linear Velocity . . . . . . . . . . . . . . . . . . . 640
Dual Porosity and Dual Permeability Theory . . . . . . . . . . . 640
References for the Darcy’s Law Interface. . . . . . . . . . . . . 642

CONTENTS | 19
 Theory for the Free and Porous Media Flow, Brinkman
Interface 644
Reference for the Free and Porous Media Flow, Brinkman Interface . . 644

Theory for the Free and Porous Media Flow, Darcy Interface
645
References . . . . . . . . . . . . . . . . . . . . . . . . 647

Theory for the Brinkman Equations Interface 648

About the Brinkman Equations . . . . . . . . . . . . . . . . 648
Brinkman Equations Theory. . . . . . . . . . . . . . . . . . 648
References for the Brinkman Equations Interface. . . . . . . . . . 650

Theory for the Multiphase Flow in Porous Media Interface 651

Pressure-Saturation Formulation . . . . . . . . . . . . . . . . 651
Mass Source and Boundary Mass Source . . . . . . . . . . . . . 653
Boundary Conditions . . . . . . . . . . . . . . . . . . . . 654
Thin Barrier (Multiphysics Coupling) . . . . . . . . . . . . . . 654
References . . . . . . . . . . . . . . . . . . . . . . . . 656

Theory for the Bubbly Flow Interface 657

The Bubbly Flow Equations . . . . . . . . . . . . . . . . . . 657
Turbulence Modeling in Bubbly Flow Applications . . . . . . . . . 661
References for the Bubbly Flow Interfaces . . . . . . . . . . . . 663

Theory for the Mixture Model Interface 664

The Mixture Model Equations . . . . . . . . . . . . . . . . . 664
Dispersed Phase Boundary Condition Equations . . . . . . . . . . 668
Turbulence Modeling in Mixture Models . . . . . . . . . . . . . 670
Slip Velocity Models. . . . . . . . . . . . . . . . . . . . . 671
References for the Mixture Model Interfaces . . . . . . . . . . . 673

Theory for the Euler-Euler Interface 674

The Euler–Euler Model Equations . . . . . . . . . . . . . . . 674
Some Notes on Boundary Conditions . . . . . . . . . . . . . . 684
Turbulent Two-Phase Flow Modeling . . . . . . . . . . . . . . 684
References for the Euler–Euler Model Interfaces . . . . . . . . . . 689

20 | C O N T E N T S
 Theory for the Coupling of Fluid Flow to Electrochemical
Reactions 690
Momentum Sources and Sinks . . . . . . . . . . . . . . . . . . 690

Chapter 7: Heat Transfer Interfaces

Chapter 8: Thermodynamics

Using Thermodynamic Properties 696

Workflow for Thermodynamics Property Calculations . . . . . . . 696
Thermodynamics . . . . . . . . . . . . . . . . . . . . . . 698
Thermodynamic System . . . . . . . . . . . . . . . . . . . 699
External Thermodynamic Packages . . . . . . . . . . . . . . . 707
External Thermodynamic System . . . . . . . . . . . . . . . . 708
Predefined System . . . . . . . . . . . . . . . . . . . . . 712
Exporting and Importing Thermodynamic Systems . . . . . . . . . 714
Species Property . . . . . . . . . . . . . . . . . . . . . . 716
Mixture Property. . . . . . . . . . . . . . . . . . . . . . 722
Equilibrium Calculation . . . . . . . . . . . . . . . . . . . 724
Generate Chemistry . . . . . . . . . . . . . . . . . . . . 728
Generate Material . . . . . . . . . . . . . . . . . . . . . 730
Coupling with the Reaction Engineering and the Chemistry Interfaces . . 735
Evaluating a Property Function in a Physics Interface . . . . . . . . 739
User-Defined Species . . . . . . . . . . . . . . . . . . . . 740
References . . . . . . . . . . . . . . . . . . . . . . . . 746

Thermodynamic Models and Theory 747

Introduction . . . . . . . . . . . . . . . . . . . . . . . 747
Thermodynamic Models . . . . . . . . . . . . . . . . . . . 747
Selecting the Right Thermodynamic Model . . . . . . . . . . . . 762
Species Property References . . . . . . . . . . . . . . . . . 763
Thermodynamic Properties Definitions . . . . . . . . . . . . . 763
Standard Enthalpy of Formation and Absolute Entropy Terms . . . . . 767
Reference State . . . . . . . . . . . . . . . . . . . . . . 769
Transport Properties . . . . . . . . . . . . . . . . . . . . 769

CONTENTS | 21
 Surface Tension . . . . . . . . . . . . . . . . . . . . . . 795
References . . . . . . . . . . . . . . . . . . . . . . . . 795

Chapter 9: Multiphysics Coupling Nodes

Electrochemical Heating . . . . . . . . . . . . . . . . . . . 804
Potential Coupling . . . . . . . . . . . . . . . . . . . . . 804
Space Charge Density Coupling . . . . . . . . . . . . . . . . 804
Reacting Flow, H2/O2 Gas Phase. . . . . . . . . . . . . . . . 805

Chapter 10: Glossary

Glossary of Terms 808

Index 813

22 | C O N T E N T S
 1

Introduction

This guide describes the Fuel Cell & Electrolyzer Module, an optional add-on
package for COMSOL Multiphysics® designed to assist you in building detailed
models of the various reactions and transport processes in fuel cells and
electrolyzers.

This chapter introduces you to the capabilities of the module. A summary of the
physics interfaces and where you can find documentation and model examples is
also included. The last section is a brief overview with links to each chapter in this
guide.

• About the Fuel Cell & Electrolyzer Module

• Overview of the User’s Guide

23
 About the Fuel Cell & Electrolyzer
Module
In this section:

• What Can the Fuel Cell & Electrolyzer Module Do?

• Fuel Cell & Electrolyzer Module Physics Interface Guide
• Common Physics Interface and Feature Settings and Nodes
• Where Do I Access the Documentation and Application Libraries?

What Can the Fuel Cell & Electrolyzer Module Do?

The Fuel Cell & Electrolyzer Module extends the COMSOL Multiphysics
environment with customized physics interfaces for modeling of fuel cells and
electrolyzers. These physics interfaces provide tools for building detailed models of the
configuration of the various parts constituting electrochemical energy converters. The
interfaces include descriptions of the electrochemical reactions and the transport
properties that influence the performance in terms of fuel or electricity consumption,
energy efficiency and cooling requirements. With this module, you have the perfect
tools to investigate the influence of using different materials, geometric configurations,
and operating conditions.

The core functionality of the module are the electrochemistry interfaces. Generic
electrochemistry interfaces for various types (primary, secondary, and tertiary) current
distributions are available, which may be used for modeling both solid nonporous as
well as for porous electrodes. In addition to these generic electrochemistry interfaces,
the module also contains dedicated Hydrogen Fuel Cell and Water Electrolyzer
interfaces.

The tailored physics interfaces mentioned above are also complemented with extended
functionality in other physics interfaces for chemical species transport, heat transfer,

24 | CHAPTER 1: INTRODUCTION
 and fluid flow. In particular, the fluid flow interfaces are extended to also include
functionality to model two-phase flow.

A selection of suitable materials are also available in the Fuel Cell and
Electrolyzer Material Library. For more detailed information about
materials, see Materials in the COMSOL Multiphysics Reference
Manual.

Finally, the Fuel Cell & Electrolyzer Application Library includes a number of tutorials
(model examples) for various types of fuel cell and electrolyzers.

See Where Do I Access the Documentation and Application Libraries? to locate and
use these examples and tutorials as a starting point to your own investigations.

Fuel Cell & Electrolyzer Module Physics Interface Guide

The table lists the physics interfaces available with this module in addition to the core
physics included with COMSOL Multiphysics.

In the COMSOL Multiphysics Reference Manual:

• Studies and Solvers

• The Physics Interfaces
• For a list of all the core physics interfaces included with a COMSOL
Multiphysics license, see Physics Interface Guide.

PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE

DIMENSION

Chemical Species Transport

Transport of Diluted tds all dimensions stationary; time dependent

Species
Transport of tcs all dimensions stationary; time dependent
Concentrated Species
Chemistry chem all dimensions stationary; time dependent

ABOUT THE FUEL CELL & ELECTROLYZER MODULE | 25

 PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE
DIMENSION

Nernst–Planck–Poisson tds+es all dimensions stationary; time dependent;

Equations stationary source sweep;
small-signal analysis,
frequency domain
Electrophoretic Transport el all dimensions stationary; stationary with
initialization; time
dependent; time dependent
with initialization
Transport of Diluted tds all dimensions stationary; time dependent
Species in Porous Media
Transport of tcs all dimensions stationary; time dependent
Concentrated Species in
Porous Media
Surface Reactions sr all dimensions stationary (3D, 2D, and 2D
axisymmetric models only);
time dependent
Transport of Diluted dsf 3D, 2D, 2D stationary; time dependent
Species in Fractures axisymmetric
Reacting Flow

Laminar Flow, Diluted — 3D, 2D, 2D stationary; time dependent

Species axisymmetric

Laminar Flow, — 3D, 2D, 2D stationary; time dependent

Concentrated Species axisymmetric
Reacting Flow in Porous Media

Transport of Diluted — 3D, 2D, 2D stationary; time dependent

Species axisymmetric
Transport of — 3D, 2D, 2D stationary; time dependent
Concentrated Species axisymmetric
Nonisothermal Reacting Flow

Brinkman Equations — 3D, 2D, 2D stationary; time dependent

axisymmetric

26 | CHAPTER 1: INTRODUCTION
PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE
DIMENSION

Laminar Flow — 3D, 2D, 2D stationary; time dependent

axisymmetric
Electrochemistry

Primary Current cd all dimensions stationary; stationary with

Distribution initialization; time
dependent; time dependent
with initialization; AC
Secondary Current
impedance, initial values;
Distribution
AC impedance, stationary;
AC impedance, time
dependent
Tertiary Current tcd all dimensions stationary; stationary with
Distribution, Nernst– initialization; time
Planck (Electroneutrality, dependent; time dependent
Water-Based with with initialization; AC
Electroneutrality, impedance, initial values;
Supporting Electrolyte, AC impedance, stationary;
Poisson) AC impedance, time
dependent
Concentrated Electrolyte cet all dimensions stationary; stationary with
Transport initialization; time
dependent; time dependent
with initialization;
Electroanalysis tcd all dimensions stationary; time dependent;
AC impedance, initial
values; AC impedance,
stationary; AC impedance,
time dependent; cyclic
voltammetry
Electrode, Shell els 3D, 2D, 2D stationary; time dependent
axisymmetric
Hydrogen Fuel Cell fc all dimensions stationary; time dependent;
Interfaces frequency domain
Water Electrolyzer fc all dimensions stationary; time dependent;
Interfaces frequency domain

ABOUT THE FUEL CELL & ELECTROLYZER MODULE | 27

 PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE
DIMENSION

Fluid Flow

Multiphase Flow

Bubbly Flow

Laminar Bubbly Flow bf 3D, 2D, 2D stationary; time dependent

axisymmetric
Mixture Model

Mixture Model, mm 3D, 2D, 2D stationary; time dependent

Laminar Flow axisymmetric
Euler–Euler Model

Euler–Euler Model, ee 3D, 2D, 2D stationary; time dependent

Laminar Flow axisymmetric
Phase Transport Mixture Model

Phase Transport, — 3D, 2D, 2D stationary; time dependent

Mixture Model, axisymmetric
Laminar Flow
Phase Transport

Phase Transport phtr all dimensions stationary; time dependent

Phase Transport in phtr all dimensions stationary; time dependent

Porous Media
Phase Transport in phtr all dimensions stationary; time dependent
Free and Porous Media
Flow
Porous Media and Subsurface Flow

Brinkman Equations br 3D, 2D, 2D stationary; time dependent

axisymmetric

28 | CHAPTER 1: INTRODUCTION
PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE
DIMENSION

Darcy’s Law dl all dimensions stationary; time dependent

Multiphase Flow in — all dimensions stationary; time dependent

Porous Media
Multiphase Free and — 3D, 2D, 2D stationary; time dependent
Porous Media Flow axisymmetric
Free and Porous Media fp 3D, 2D, 2D stationary; time dependent
Flow, Brinkman axisymmetric
Free and Porous Media — 3D, 2D, 2D stationary; time dependent
Flow, Darcy axisymmetric
Nonisothermal Flow

Brinkman Equations — 3D, 2D, 2D stationary; time dependent;

axisymmetric stationary, one-way NITF;
time dependent, one-way
NITF
Heat Transfer

Heat Transfer in ht all dimensions stationary; time dependent

Porous Media

Common Physics Interface and Feature Settings and Nodes

There are several common settings and sections available for the physics interfaces and
feature nodes. Some of these sections also have similar settings or are implemented in
the same way no matter the physics interface or feature being used. There are also some
physics feature nodes that display in COMSOL Multiphysics.

ABOUT THE FUEL CELL & ELECTROLYZER MODULE | 29

 In each module’s documentation, only unique or extra information is included;
standard information and procedures are centralized in the COMSOL Multiphysics
Reference Manual.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for

links to common sections and Table 2-5 to common feature nodes.
You can also search for information: press F1 to open the Help
window or Ctrl+F1 to open the Documentation window.

Where Do I Access the Documentation and Application Libraries?

A number of online resources have more information about COMSOL, including
licensing and technical information. The electronic documentation, topic-based (or
context-based) help, and the Application Libraries are all accessed through the
COMSOL Desktop.

If you are reading the documentation as a PDF file on your computer,

the blue links do not work to open an application or content
referenced in a different guide. However, if you are using the Help
system in COMSOL Multiphysics, these links work to open other
modules, application examples, and documentation sets.

THE DOCUMENTATION AND ONLINE HELP

The COMSOL Multiphysics Reference Manual describes the core physics interfaces
and functionality included with the COMSOL Multiphysics license. This book also has
instructions on how to use COMSOL Multiphysics and how to access the electronic
Documentation and Help content.

Opening Topic-Based Help

The Help window is useful as it is connected to the features in the COMSOL Desktop.
To learn more about a node in the Model Builder, or a window on the Desktop, click
to highlight a node or window, then press F1 to open the Help window, which then

30 | CHAPTER 1: INTRODUCTION
displays information about that feature (or click a node in the Model Builder followed
by the Help button ( ). This is called topic-based (or context) help.

To open the Help window:

• In the Model Builder, Application Builder, or Physics Builder, click a node

or window and then press F1.
• On any toolbar (for example, Home, Definitions, or Geometry), hover the
mouse over a button (for example, Add Physics or Build All) and then
press F1.
• From the File menu, click Help ( ).
• In the upper-right corner of the COMSOL Desktop, click the Help ( )
button.

Opening the Documentation Window

To open the Documentation window:

• Press Ctrl+F1.
• From the File menu, select Help>Documentation ( ).

THE APPLICATION LIBRARIES WINDOW

Each model or application includes documentation with the theoretical background
and step-by-step instructions to create a model or application. The models and
applications are available in COMSOL Multiphysics as MPH-files that you can open
for further investigation. You can use the step-by-step instructions and the actual
models as templates for your own modeling. In most models, SI units are used to
describe the relevant properties, parameters, and dimensions, but other unit systems
are available.

Once the Application Libraries window is opened, you can search by name or browse
under a module folder name. Click to view a summary of the model or application and
its properties, including options to open it or its associated PDF document.

The Application Libraries Window in the COMSOL Multiphysics

Reference Manual.

ABOUT THE FUEL CELL & ELECTROLYZER MODULE | 31

 Opening the Application Libraries Window
To open the Application Libraries window ( ):

From the File menu, select Application Libraries.

To include the latest versions of model examples, from the File>Help menu
select ( ) Update COMSOL Application Libraries.

From the File or Windows menu, select Application Libraries.

To include the latest versions of model examples, from the Help menu
select ( ) Update COMSOL Application Libraries.

CONTACTING COMSOL BY EMAIL

For general product information, contact COMSOL at info@[Link].

COMSOL ACCESS AND TECHNICAL SUPPORT

To receive technical support from COMSOL for the COMSOL products, please
contact your local COMSOL representative or send your questions to
support@[Link]. An automatic notification and a case number will be sent to you
by email. You can also access technical support, software updates, license information,
and other resources by registering for a COMSOL Access account.

32 | CHAPTER 1: INTRODUCTION
COMSOL ONLINE RESOURCES

COMSOL website [Link]

Contact COMSOL [Link]/contact
COMSOL Access [Link]/access
Support Center [Link]/support
Product Download [Link]/product-download
Product Updates [Link]/product-update
COMSOL Blog [Link]/blogs
Discussion Forum [Link]/forum
Events [Link]/events
COMSOL Application Gallery [Link]/models
COMSOL Video Gallery [Link]/videos
Learning Center [Link]/support/learning-center
Support Knowledge Base [Link]/support/knowledgebase

ABOUT THE FUEL CELL & ELECTROLYZER MODULE | 33

 Overview of the User’s Guide
The Fuel Cell & Electrolyzer Module User’s Guide gets you started with modeling
using COMSOL Multiphysics. The information in this guide is specific to this module.
Instructions how to use COMSOL in general are included with the COMSOL
Multiphysics Reference Manual.

As detailed in the section Where Do I Access the Documentation and

Application Libraries? this information can also be searched from the
COMSOL Multiphysics software Help menu.

TABLE OF CONTENTS, GLOSSARY, AND INDEX

To help you navigate through this guide, see the Contents, Glossary of Terms, and
Index.

MODELING WITH ELECTROCHEMISTRY

The Modeling with Electrochemistry chapter discusses a variety of topics, including
Fundamentals of Electrochemistry Modeling, Modeling Electrochemical Reactions,
Modeling Cyclic Voltammetry, and Postprocessing Your Solution.

THE FUEL CELL AND WATER ELECTROLYZER INTERFACES

The Fuel Cell and Electrolyzer Interfaces chapter describes the Hydrogen Fuel Cell
and the Water Electrolyzer interfaces.

THE ELECTROCHEMISTRY INTERFACES

The Electrochemistry Interfaces chapter describes the Primary Current Distribution;
Secondary Current Distribution; Tertiary Current Distribution, Nernst–Planck;
Electrode, Shell; and Electroanalysis interfaces.

THE CHEMICAL SPECIES TRANSPORT INTERFACES

The Chemical Species Transport Interfaces chapter describes the Transport of Diluted
Species, Transport of Concentrated Species, the Transport of Diluted Species in
Porous Media, the Surface Reaction and the Reacting Flow interfaces.

THE FLUID FLOW INTERFACES

The Fluid Flow Interfaces chapter describes, among others, the Darcy’s Law and Free
and Porous Media Flow interfaces, included under the Porous Media and Subsurface

34 | CHAPTER 1: INTRODUCTION
Flow branch, as well as the Bubbly Flow, Mixture Model and Euler-Euler interface,
included under the Multiphase Flow branch.

The Laminar Flow interface is also available and is described in the COMSOL
Multiphysics Reference Manual.

THE HEAT TRANSFER INTERFACES

The Heat Transfer Interfaces chapter describes how electrochemical heat sources can
be coupled to heat transfer, and the theory for how these are calculated.

THERMODYNAMICS
This chapter describes the functionality of the Thermodynamics node to define
thermo-physical and transport properties.

THE MULTIPHYSICS COUPLING NODES

The Multiphysics Coupling Nodes chapter describes the coupling nodes available
under the Multiphysics node for coupling various electrochemical interfaces using
flow, potential, and temperature couplings, for example.

OVERVIEW OF THE USER’S GUIDE | 35

36 | CHAPTER 1: INTRODUCTION
 2

Modeling with Electrochemistry

In this chapter:
• Introduction to Electrochemistry Modeling

37
 Introduction to Electrochemistry
Modeling
In this section:

• What Is Electrochemistry?
• Electrochemical Applications
• Fundamentals of Electrochemistry Modeling
• Current Distribution Cases and Choosing the Right Interface to Model an
Electrochemical Cell
• Understanding the Different Approximations for Conservation of Charge in
Electrolytes
• Modeling Electrochemical Reactions
• Double Layer Capacitance
• Porous Electrodes
• Boundary Conditions for Running and Controlling Electrochemical Cells
• Modeling Cyclic Voltammetry
• Common Simplifications When Modeling Electrochemical Cells
• Before You Start Building Your Model
• Meshing Advice
• Solving Electrochemical Models
• Postprocessing Your Solution

What Is Electrochemistry?
An electrochemical process is one that either converts electric energy to chemical
energy or converts chemical energy to electric energy.

In electrochemical reactions, chemicals react by gaining or losing electrons. Chemicals

can exchange these electrons with an electrical conductor or circuit. The chemical
potential — or thermodynamic driving force — for a chemical reaction is thus
converted into an electric potential or voltage, which can be exploited to do useful
work when an electric current is drawn.

38 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

Electrochemical Applications
Electrochemical systems may be classified in different ways.

One approach is to contrast those electrochemical systems involving desirable reactions

from those involving undesirable reactions. Synthetic electrolysis is an example of a
desirable reaction because the chemical process is promoted to generate a desired
product. Or, for example, when designing a battery we might want to set up conditions
to promote an electrochemical reaction in order to extract useful energy.

Corrosion is an example of undesirable electrochemistry because in this case

spontaneous electrochemical reactions destroy structural materials and so harm the
integrity of a system: we aim to minimize the rate of these reactions.

Electrochemical systems can also be classified into systems that output energy or
systems that consume energy. Batteries and fuel cells are energy extraction devices —
an electrochemical reaction is used to convert the energy in chemical system into a
voltage. Such cells are also called galvanic cells. By contrast, in electrolysis, the system
consumes energy to promote an electrochemical reaction for synthesis. Similar
electrochemical systems needing energy input include manufacturing processes such as
electroplating. Electrochemical reactions may also be driven for electroanalysis, to
quantify or otherwise explore the chemical constituents or reactivity of a system.

The different electrochemical modules in COMSOL Multiphysics are designed to

offer tailored physics interfaces provision for each of these electrochemical situations.
The underlying physical description provided by the general physics interfaces
described below is common to all the modules — the differences between the modules
lies in the particular expected system inputs and outputs and the intended behavior of
the system.

Fundamentals of Electrochemistry Modeling

Electrochemical systems consist of electrically conducting media. These may be
classified as electrodes or electrolytes. An electrode carries current by transport of
electrons; normally the electrode is a conventional electrical conductor such as a metal.
The electrolyte carries current by transport of charged chemical species (ions).
Electrolytes are often salt solutions in water but may include salt solutions in other
liquid solvents, as well as solids, such as concrete, which can conduct by transport of
oxide ions. The electric conductivity of an electrode is normally several orders of
magnitude larger than the electric conductivity of an electrolyte.
 THE ELECTRODE-ELECTROLYTE INTERFACE
At the electrode-electrolyte interface, conventional electric current in the electrode is
converted into ionic current in the electrolyte. According to the overall conservation
of charge, these currents must balance here. The conversion between the two types of
currents may arise due to electrochemical reaction (electrolysis) or capacitive charging.

Electrolysis occurs when a chemical species in the electrolyte exchanges one or more
electrons with the electrode. Capacitive charging occurs when the potential of an
electrode is changing, so that ions in the electrolyte are either attracted or repelled
from the surface, drawing a current.

Batteries and fuel cells can also involve porous electrodes, in which an electrode
material has a micro- or nanostructure that is permeable to an electrolyte solution. The
advantage of such a material is the great increase in the area of the electrode-electrolyte
interface.

Note that all current must move in circuits. An isolated electrode-electrolyte interface
cannot draw a net current, but a system with two such interfaces can. An
electrochemical system with two or more electrodes in contact with electrolyte is called
an electrochemical cell.

In an electrochemical cell with two electrodes, these electrodes are identified as an

anode, at which the electrochemical reaction transfers electrons from electrolyte to
electrode, and a cathode, at which electrons are transferred from the electrode to the
electrolyte. Note that it is the direction of the current that will determine if an
electrode reaction is anodic or cathodic. For a battery, for instance, the location of the
anode and cathode will change depending on whether the battery is charged or
discharged. (The general habit in the battery community to always denounce the
positive electrode as the “cathode” is hence strictly only correct during battery
discharge.)

Conventional electric current is the flow of positive charge, which is then from anode
to cathode through the electrolyte. A closed circuit, conserving overall system charge,
is formed by the flow of electric current in the electrode domains (and any electrical
circuitry) from cathode to anode, and by the transport of ions through the electrolyte
domains from anode to cathode.

OUTPUTS OF INTEREST FROM A MODEL

The experimentally measurable features of an electrochemical system are the external
(lumped) current and voltage acting between the two electrodes. It is, in general,
impossible to measure local current densities or potentials at different points on the

40 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

surface of a conducting electrode, or at arbitrary points within an electrochemical
system.

The advantage of physical modeling is the ability to investigate the full

space-dependent behavior of the system under a wide range of circumstances to
identify the reasons for the observed current-voltage relationship, and so to identify
the appropriate physical conditions to optimize current and voltage as required.

The overall current-voltage curve of an electrochemical cell is also known as a

polarization curve or, in an analytical context, a voltammogram. These curves are not
unique but rather depend on the means by which the current or voltage is altered, since
these lumped parameters are related to multiple physical effects with different length
and time scales. Hysteresis in practical polarization curves is not uncommon.

Polarization curves are frequently nonlinear. The combination of nonlinearity and

hysteresis means that electrochemical cells do not necessarily resemble “ideal”
electrical components (such as a circuit of resistors and capacitors) in the sense of
giving a predictable and linear current-voltage response.

• Potential Variables
• Current Variables and Calculating the Total Cell Current

Current Distribution Cases and Choosing the Right Interface to

Model an Electrochemical Cell
COMSOL Multiphysics has many built-in physics interfaces aimed at specific
applications. In many generic cases, however, the modeler has to make some basic
assumptions about his system before starting to build his model.

Under the assumption of a linear relation of current density to electric field, Ohm’s law
is obeyed for the electrolyte current. This is the assumption of primary current
distribution, where one also assumes infinitely fast electrodes kinetics, resulting in
negligible potential drops over the electrode-electrolyte interfaces. If the electrode
reaction kinetics proceed at a finite rate, then the system has a secondary current
distribution. In cases where more advanced nonlinear charge conservation equations
and concentration-dependent electrode polarization are required, the system is
described as obeying tertiary current distribution.

In some applications, especially within the field of electroanalysis, the potential

gradients in the electrolyte are so small that the spatial distribution of current in the
 electrolyte is not solved for. Such models are instead centered around the interplay of
electrode kinetics and transport (by diffusion) of the reacting species in the vicinity of
the electrode.

A rule-of-thumb flowchart for selecting a proper current distribution interface for a

model is shown below. The recommendations here are indicative and the modeler may
sometimes want to deviate from the suggested route; in particular, simpler descriptions
are recommended at an earlier stage of model development. In the following sections
we explain some of the theory behind this flowchart with regards to the description of
the electrolyte charge transport and electrode reactions.

Understanding the Different Approximations for Conservation of

Charge in Electrolytes
Deciding how to model the charge transport in the electrolyte is usually the starting
point when setting up an electrochemical model since this will determine what physics

42 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

interface you will use when starting to build your model. Different theoretical
descriptions of the electrolyte current density are included in COMSOL Multiphysics.
They are applicable in different circumstances as discussed in this section.

The electric displacement field in a medium is related to the local charge density
according to Gauss’s law, one of Maxwell’s equations:

  D = v

In electrolytes, we can normally assume that the electric permittivity is constant and
equal to a bulk value:

D =  0  s E = –  0  s V

Hence

2 v
 V + ---------
- = 0
0 s

In an electrolyte with ionic charge carriers, the charge density can be written as:

v = F  zi ci
i

Hence

F
0 s 
2
 V + ---------- zi ci = 0
i

This is the Poisson equation relating the electrolyte potential to the distribution of
charge carriers within the electrolyte. In its derivation we assumed that the only charge
carriers are ions, and that the solvated ions and electric field do not alter the
permittivity of the medium.

The mass transport of the charge carriers in aqueous systems is normally given by the
Nernst–Planck equations. These equations neglect ion-ion interactions, and so they
are only exact for infinitely dilute solutions:

N i = – D i c i – z i u m i Fc i  l + c i u

Note that concentrated electrolyte systems, such as those in many batteries, use an
extended concentrated species flux definition, based on the Maxwell–Stefan set of
 equations. This will result in a different set of equations to solve for, but the general
principles and conclusion in this section will be the same.

Substituting the Nernst–Einstein relation for the electrical mobility of an ion we get:

ziF
N i = – D i  c i + -------- c i  l + c i u
RT

The above expressions for the n species i, together with the Poisson equation, give a
set of n1 equations in n1 unknowns. These are the Nernst–Planck–Poisson
equations. They can be defined in COMSOL Multiphysics by coupling Transport of
Diluted Species with Electrostatics, or by using the Tertiary Current Distribution,
Nernst-Planck interface with Charge conservation model: Poisson, but they are highly
nonlinear and difficult to converge. Most often, further approximations can simplify
the problem without compromising accuracy.

• Theory for the Transport of Diluted Species Interface

• Theory of Electrostatics in the COMSOL Multiphysics Reference
Manual

An important dimensional quantity occurring in the Poisson equation is:

RT  0  s
xD = ------------------
-
2
F I

This is the length across which electric fields are screened. It is called the Debye length.
This is a very short length in electrolyte solutions: for a typical ionic strength, it is of
the order of 1 nm. Electroneutrality holds at distances much larger than 1 nm from a
charged surface:

 zi ci = 0

The constraint of electroneutrality can be used as a condition determine the electric

potential in the Nernst–Planck equations, in place of the full Poisson equation. The
Nernst–Planck equations with electroneutrality are used to describe current flow in the
Tertiary Current Distribution, Nernst-Planck interface.

The Tertiary Current Distribution, Nernst–Planck Interface

44 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

The current flow itself is given by

il = F  zi Ni
From substitution of the Nernst–Planck expressions for Ni, the laws of conservation
of mass and charge combine to automatically satisfy conservation of current.

We can simplify the system further by considering the arising expression for il in more
detail:

2
F
  zi  zi ci
2
il = – F D i z i c i – --------  l Di ci + u
RT

Clearly, the rightmost term is zero: that is, convection of an electroneutral solution
does not cause current flow. The leftmost term (diffusion current) also vanishes due to
electroneutrality if the gradients of the charge carrying species are zero.

Even if this is not the case, however, this term is often much smaller than the central
term (migration current), so long as the concentrations of the current-carrying ions do
not vary markedly through the solution. Under conditions where the composition of
the electrolyte can be considered nearly constant and current-carrying ions are not
significantly depleted, the diffusion current can be assumed to contribute negligibly.

Hence, it follows that

i l = –  l  l

This expression for current density is used in the Secondary Current Distribution
interface, and also the Primary Current Distribution interface. The difference between
these interfaces lies in the treatment of the electrode-electrolyte interfaces (see Kinetics
of Electrochemical Reactions below). From the above, the conductivity of the
electrolyte l is given as:

2
F
 zi
2
 l = -------- Di ci
RT

So long as this quantity does not vary markedly through the solution, the
approximation of zero diffusion current is good. If the diffusivities and concentrations
can be taken as constant, we can approximate that:

2
2F ID mean
 l  -----------------------------
RT
 The advantage of the ohmic expression for current density is that it is a linear relation
of current density to electrolyte potential. It is only weakly nonlinear if l is allowed to
depend on a concentration solved for in a species transport interface. By comparison,
the Nernst–Planck equations with electroneutrality can be highly nonlinear.

The approximations used to derive the secondary current distribution expression place
tighter constraints on the allowed system configurations, however. The ionic strength
of the solution must remain near-constant for the constant conductivity approximation
to be valid. Usually this is only the case for relatively high conductivity solutions.

When the conductivity is large with respect to the current drawn, the electric field
becomes negligible in solution. For negligible electric fields, a diffusion-only
approximation may be used, where E0. This converts the Nernst–Planck equations
into Fick’s laws, with a term for convective transport where necessary. Fick’s laws with
convection and electrochemical boundary conditions are solved for in the
Electroanalysis interface.

Even if you think a problem will involve the full Nernst–Planck equations,
it is best to set the model up in Secondary Current Distribution first, in order
to identify any other possible complications in the system while using a
simpler electrochemical model.

Theory for the Current Distribution Interfaces

Modeling Electrochemical Reactions

Electrochemical reactions are defined by using Electrode Reaction or Porous Electrode
Reaction nodes. An electrode reaction is defined by its thermodynamics, kinetics, and
stoichiometry. The latter describes the mass fluxes, sources, and sinks arising due to a
certain current density of the reaction.

THERMODYNAMICS OF ELECTROCHEMICAL REACTIONS

An electrolytic reaction involves the exchange of electrons with the electrode. Such a
reaction is written as reduction, even if the reaction occurs predominantly in the
oxidative direction. For example:

+ -
Ag (aq)+e  Ag(s)

46 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

This reaction is called a “half-cell” reaction, since it will occur at a specific
electrode-electrolyte interface. It cannot occur in isolation, but only when coupled to
another half-cell reaction within a two-electrode electrochemical cell. Each reaction
has a characteristic Gibbs energy change that determines whether or not it is
thermodynamically favorable. A negative Gibbs energy change means that the reaction
proceeds spontaneously — it is thermodynamically “downhill”.

The Gibbs energy change is related to the equilibrium potential difference from the
electrode to the electrolyte according to:

G m
E eq m = – ------------
-
nm F

where Eeq,m is the potential difference on some external reference scale for which the
reaction is at equilibrium (G0). This is called the equilibrium potential or
reduction potential (or in corrosion, corrosion potential) of the electrochemical
reaction, and its absolute value depends on the choice of reference electrode.

Equilibrium potentials should always be quoted versus a specific

reference. If a different reference scale is used in different parts of your
model, the thermodynamics of the system may not behave as expected.

From the standard thermodynamic relation

G = – RT ln K

it follows that

RT
E – E eq = -------- ln K
nF

This is the Nernst equation which is a universal thermodynamic expression. It is always

true of systems at thermodynamic equilibrium; it does not necessarily apply to systems
not at equilibrium.

Assuming that the species are ideal and that activity effects can be treated as constant,
then for the conversion between unimolecular reduced and oxidized species:

[Ox] = [Red] exp   --------   s –  l – E f 

nF
  RT 
 Hence at equilibrium, the concentrations of reactants and products at the
electrolyte-electrode surface are related by an expression which depends on the
potential difference between the two phases, and two reaction parameters: n, the
number of electrons transferred per molecule reduced; and Ef, the formal reduction
potential of the reaction measured on the same potential scale as the
electrode-electrolyte potential difference.

The quantity

 m =  s –  l – E eq m

is known as the overpotential and is particular to a specific reaction occurring at the

interface.

m depends on both the electric potential in the electrode s and the

electrolyte potential l. Where there is substantial resistance to current
flow through a solution, the corresponding potential difference in l
called ohmic drop, alters the position of the electrochemical equilibrium.
Additional applied potential in the electrical circuit may then be needed
to drive an equivalent overpotential.

KINETICS OF ELECTROCHEMICAL REACTIONS

The Nernst equation tells us the position of equilibrium of a reaction. However, it tells
us nothing about how fast the system may get there. If there is a kinetic limitation —
that is, if the reaction proceeds slowly — the equilibrium condition may never be
observed. As a familiar example, diamond is thermodynamically unstable with respect
to reacting to form graphite at room temperature and pressure. However, this reaction
is kinetically limited by a vast activation energy for the reorientation of atoms, such that
it is never in practice observed, and diamond is technically described as metastable.

We encounter the same issue in many electrochemical contexts. Reactions are

prevented from proceeding to their equilibrium by kinetic limitations. Indeed,
overcoming the natural kinetic sluggishness of the surface reactions of small, nonpolar
molecules such as hydrogen and oxygen is key to much fuel cell research.

There are two important expressions describing the current density due to an
electrochemical reaction as a function of the overpotential and the concentrations of
reactant and product. It should be noted that the validity of these expressions is not
general and can never replace experimental kinetic data if such is available.

48 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

The first is the Tafel law which describes an irreversible anodic or cathodic process:

log  ---- = A
i
 i 0

The constant A is the Tafel slope and has units 1/V. It is usually close to a half-integer
multiple of F/RT and is less than or equal to nF/RT. Note that a reference exchange
current density i0 must be specified for the reaction. This is by definition the current
density drawn at zero overpotential.

The Tafel law assumes that a reaction is irreversible. If the reverse reaction
might occur in practice, Tafel kinetics will not be correct.

The second expression is the Butler–Volmer equation, which describes a reversible

process, so that either anodic or cathodic current may flow depending on the sign and
magnitude of the overpotential:

 a F –  c F
i = i 0  exp  --------------- – exp  ----------------- 
RT RT

The Butler–Volmer equation is the most general description of electrode kinetics. It is

highly adaptable because:

• i0 is an empirical quantity.
• It agrees with the Nernst equation when i = 0, so for a very fast reaction (i0)
then the Butler–Volmer equation gives the same potential difference as the Nernst
equation. This is equally true under high resistance conditions.
• It agrees with the Tafel equation when either the anodic or cathodic term
dominates. For highly irreversible reactions (very low i0), appreciable current is only
drawn for large overpotential, so this is typically the case.

For a discussion on the Butler–Volmer expression and concentration changes of the

participating species, see the Defining Concentration Dependent Butler–Volmer
Kinetics section below.

For a reversible reaction at very low overpotential ( of order RTF ~ 25 mV), the
exponentials in the Butler–Volmer equation can be linearized:

  a +  c F
i loc = i 0  ---------------------------- 
 RT 
 Using a linearized Butler–Volmer reaction can be advantageous when
investigating convergence issues of a model.

The linearized Butler–Volmer equation is not correct for applied

overpotentials larger than (RT/F). This is about 25 mV at room
temperature. It is not suitable outside this range and therefore its use is
confined to electrochemical processes occurring exclusively at low current
density, such as electroplating or electrochemical impedance spectroscopy.

Electrode Kinetics Expressions

FLUXES AND SOURCES/SINKS DUE TO ELECTRODE REACTIONS

Electrode reactions will result in a molecular flux of reacting species to or from the
electrode surface. If you are including mass transport in your model it is common to
couple the flux of a reacting species on a boundary to the electrode reaction current
density (by the Faraday’s law of electrolysis).

The coupling of chemical flux to electric current density is automated in some of the
Electrochemistry interfaces by defining the reaction stoichiometry in the Electrode
Reaction and Porous Electrode Reaction nodes. In the Chemical species transport
interfaces the coupling however needs to be set up manually by the Electrode Surface
Couplingnodes. When modeling porous electrodes, the corresponding coupling node
to create a source/sink in a domain is the Porous Electrode Coupling node.

The mathematical treatment can be summed up by the expression:

 jm i m
N j = – ---------------
-
nm F

This means that the flux Nj of chemical species j into the surface is proportional to the
current density im due to reaction m drawn at an electrode-electrolyte interface. The
constant of proportionality is the stoichiometric number for the species divided by nm,
the number of electrons transferred in reaction m, in a reductive direction and F the
Faraday constant (96,485 C/mol, the absolute charge on a mole of electrons).

50 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

As such, knowledge of the stoichiometry of an electrochemical reaction allows the local
flux of a chemical species to be coupled to the current density contributed by that
reaction. Note that multiple reactions may take place simultaneously at an electrode,
and their contributions to the current density are simply summed.

The stoichiometric coefficients of the chemical species in an

electrochemical reaction are conventionally written with the forward
reaction in the direction of reduction, even if in practice within the system
the reaction predominantly proceeds in the other direction.

COMSOL Multiphysics uses the standard convention that anodic

(oxidative) current is positive. Cathodic (reductive) current is negative.

DEFINING CONCENTRATION DEPENDENT BUTLER–VOLMER KINETICS

If the concentrations at an electrode surface change, this will have an impact on the
local kinetics. The Butler–Volmer kinetics expression is derived by considering the rate
of a redox reaction

-
Ox + ne  Red (2-1)

as the sum of the forward and backward rates according to:

i loc  a FE  c FE
- = k fwd c R exp  --------------- – k rwd c O exp  – --------------
r = ------- - (2-2)
nF  RT   RT 

where kfwd and krwd are reaction rate constants and cO and cR are the activities of the
oxidized and reduced species of the redox couple, respectively. The potential E is here
defined as

E = s – l (2-3)

and the transfer coefficients are equal the sum of electrons in the charge transfer
reaction according to

a + c = n (2-4)

By defining an equilibrium potential at which the forward reaction and backward

reaction rates are equal, Equation 2-1 can be shown to be equivalent to the commonly
used Butler–Volmer equation:
  a F  c F 
i loc = i 0  exp  --------------- – exp  – -------------
- (2-5)
  RT   RT  

In this formulation, assuming the stoichiometric coefficients of the oxidized and

reduced species to be 1 and 1, respectively, the exchange current density, i0, is defined
as:

cR c  n cO a  n
i 0 = i 0 ref ------------- -------------- (2-6)
c R ref c O ref

where i0, ref is the exchange current density at some chosen reference conditions, and
the overpotential is defined as

 = E – E eq (2-7)

where Eeq is the equilibrium potential is defined by the Nernst equation as

RT c R  c R ref
E eq = E eq ref – -------- ln -----------------------
- (2-8)
nF c O  c O ref

and Eeq, ref is the equilibrium potential at the same reference conditions.

Typically standard conditions are used as reference concentrations, for instance

concentrations of 1 M for ions in aqueous solutions.

Note that in Equation 2-5 both i0 and Eeq are concentration dependent. This has some
numerical drawbacks when modeling electrochemical cells including mass transport,
since for low concentrations of the participating species (that is, when cO0 or
cR0), the factor lncRcO may become undefined during the solution process. An
expression of the form of Equation 2-2 is more desirable since this expressions contains
a simple linear dependence on the species activities.

A common solution to this issue is to rewrite the Butler–Volmer expression by defining

the overpotential with respect to a fixed reference state for the activities cR, ref and cO,
ref (typically corresponding to the inlet or initial concentrations), resulting in

cR  a F ref cO  c F ref 
i loc = i 0 ref  ------------- exp  -------------------- – -------------- exp  – ------------------- (2-9)
 c R ref  RT  c O ref  RT  

where

 ref = E – E eq,ref (2-10)

52 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

Note that Equation 2-9 now contains a linear dependence on the activities cO and cR.

Double Layer Capacitance

The Tafel and Butler–Volmer expressions can be used to calculate the magnitude of the
current due to electrolysis: this is called faradaic current. However, the
electrode-electrolyte interface may also draw capacitive current due to the attraction
or repulsion of ions creating an oppositely charged layer of ions to form in solution
close to the electrode.

The layer of charge on the electrode and layer of opposite charge in the adjacent
electrolyte is called the double layer and can be thought of as behaving like a parallel
plate capacitor, since the absolute amount of charge it separates varies with the charge
density on the electrode, and hence with its voltage. The physics of double layer
structure and formation are highly complex and are not yet well understood. One of
the simplest empirical methods to account for the observed influence of capacitance
on polarization curves is to introduce a constant ideal capacitance across the
electrode-electrolyte interface.

This effect can be added to via the Double Layer Capacitance condition. The capacitor
stores a surface charge density QCdsl, and contributes a dynamic charging
current density (nonfaradaic current) equal to iNFdQ/dt. The total current
recorded in a real experiment equals:

i tot = i Far + i NF

Sometimes in electroanalysis a blank scan with no electrolysis is run to subtract

nonfaradaic current. If double layer capacitance is ignored in your model, you should
make sure that you are comparing it to experimental data recorded under the same
conditions.

The ability of COMSOL Multiphysics to model electrostatics and

transport of charged species make it tempting to incorporate the double
layer explicitly to add more detail to a model. These coupled equations are
always extremely nonlinear and convergence is not trivial except in the
simplest cases. Although the diffuse double layer can be modeled in
isolation (see example), this is not recommended in conjunction with a
fuller electrochemical model.
 Double-Layer Capacitance

Porous Electrodes
A porous electrode is one in which the three-dimensional structure of the electrode is
permeable to electrolyte. The electrode-electrolyte interface then extends over a much
larger surface area. This specific surface area (“SSA”, area per unit volume, units 1/m)
is a key property of a porous electrode. Additionally, such an electrode can conduct
electric current independently through its electrode and electrolyte domains.

Most electrochemical interfaces in COMSOL Multiphysics include Porous Electrode

domain nodes that can be used to model the above situation. In a Porous Electrode,
current density is carried in both the electrode and in the pores which are assumed to
contain electrolyte. For gas diffusion electrodes where the solid is conducting but the
pore phase is not, the Electrode condition should be used.

It is common to combine an electrochemical interface with other physics, in order to

model the mass transport, heat transfer and flow properties of a porous medium. For
instance, Secondary Current Distribution might be coupled to Transport of Diluted
Species in Porous Media, Heat Transfer in Porous Media, and Darcy’s Law.

Homogenized transport parameters such as diffusion coefficients are

commonly obtained for a bulk solution of the material. In porous media
these must be corrected for tortuosity and porosity. Carefully consider
such effects when defining your material properties.

Porous Electrode Theory

Boundary Conditions for Running and Controlling Electrochemical

Cells
An electrochemical cell is typically operated by controlling either the cell voltage
(potentiostatic control) or the cell current (galvanostatic control). In a lab experiment
this is accomplished by using a potentiostat.

In a model, potentiostatic control is defined by using fixed potential boundary

conditions. This can be done by setting the Electric Potential with respect to ground to

54 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

control the cell potential, or the Electrode Potential with respect to a reference potential
somewhere in the system. Galvanostatic control can be accomplished using Total
Current or Average Current Density boundary conditions. For both cases, one electrode
in the cell needs to be grounded.

Mathematically, a Total Current or Average Current Density condition implies setting the
potential of a boundary to be equal to an additional extra global potential degree of
freedom (floating potential) to comply with the specified current condition. For this
reason, solving for galvanic control is numerically slightly more complex.

Note that explicitly prescribing the current density distribution on an electrode

boundary is hard to accomplish in a real world experiment. In higher dimensions than
1D, Electrode Current Density and Electrolyte Current Density boundary conditions
should be used with care.

SHORT-CIRCUITING A CELL AND GALVANIC CORROSION

The simplest way to model a short-circuited cell is to set both electrodes to the same
(electronic) potential, using a fixed potential boundary condition. The charge balance
equation of the electrolyte will make sure the total current over all electrode surfaces
sums up to zero.

Note that many galvanic corrosion situations are practically equivalent to a short circuit
of two electrodes consisting of different metals. In such models, the two metals are set
to the same potential. Usually this potential is chosen to be zero (ground).

• Electrode Potential
• Reference Electrode

Modeling Cyclic Voltammetry

In a cyclic voltammetry experiment, a potentiostat is used to sweep the voltage at the
working electrode linearly from a start potential to a vertex potential, and back again.
Typically, the potential window is chosen to span the formal potential of a redox couple
of interest, allowing both the kinetics and transport properties of the chemical species
to be investigated. The target species reacts under kinetic control to the point of
depletion where the current becomes transport-controlled.

The built-in Cyclic Voltammetry study step in the Electroanalysis interface can be used
to automatically set up the voltage sweep in a time-dependent study.
 Common Simplifications When Modeling Electrochemical Cells

SHOULD I MODEL IN 3D OR CAN I MODEL IN 1D?

Many electrochemical systems use repetitive unit cells of high aspect ratio. In these
cases, the lumped effect of edge phenomena in the system will be small. For example,
in fuel cells or batteries it is often possible to use one-dimensional geometries with little
loss of detail. Consider also any mirror or rotational symmetries. Reducing the
geometry dimension where possible, either exactly or approximately, will reduce the
meshing and solving time significantly.

Starting with a 1D model helps to understand the influence of different

reactions and phenomena in an electrochemical system, and gives a good
first estimate of current-voltage behavior.

NEGLECTING THE CHARGE TRANSFER AND POTENTIAL GRADIENTS IN

ELECTRODES
The voltage over the surface of an electrode is typically constant — or very nearly
constant — since it is a good conductor. This implies that it is either the intrinsic
resistance of the electrolyte or the rate of the electrochemical reaction at the
electrode-electrolyte interface that controls the amount of current drawn in an
electrochemical cell. On the assumption of a constant surface voltage, the electrode
domain itself need not be modeled, and the coupling of charge and mass transport at
its surface treated by an Electrode Surface node.

Electrode Surface

However, when modeling porous and gas diffusion electrodes the metal phase
potential is need typically to be included since the conductivity of the metal phase
potential can be much lower in this type of electrodes. This is done in the Porous
Electrode nodes.

HALF-CELL MODELS
Often, an investigator is only interested in the chemistry taking place at one electrode
in a cell. A model of one electrode is called a “half-cell model”.

56 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

This would correspond to an experimental situation where the electrode of interest —
be it anode or cathode — is classified as the working electrode, and the other electrode
is called the counter electrode.

One usually ignores the kinetics of the counter electrode in a model; commonly it is
represented by a constant potential boundary condition. Such a model is only valid if
the counter electrode can draw arbitrarily large amounts of current compared to the
working electrode, so that it never limits the current flow in the electrochemical cell.

REPLACING THIN LAYERS WITH BOUNDARY CONDITIONS

When considering a layer that is very thin compared to other dimensions in the
geometry, it is usually correct to treat that layer using a physical condition, rather than
meshing it.

One important example is the catalyst layer in a fuel cell. Since this layer is only
nanometers in size, transport across it is very fast compared to other parts of the
system. Hence, it is not necessary to resolve a distribution of concentrations or
potentials through the layer.

Another example is the passivation layer on an oxidized electrode surface, for which
the “Thin Film Resistance” setting can be used. Because the layer is much thinner than
its surroundings, the electric field through it is almost constant. Therefore, an ohmic
expression can be substituted to create a boundary condition with a potential drop.
This is much more efficient than meshing a geometrically narrow layer.

Film Resistance

INFINITE ELEMENTS FOR STEADY-STATE ELECTROANALYSIS

For microelectrodes, the diffusion profile may approach a steady state whose size is
much larger than that of the working electrode of interest. In such an example, it is
preferable to circumscribe the simulation space to a region not more than about
twenty-five times the electrode size - assuming that the surrounding electrolyte is
uniform. Applying a bulk boundary condition such as a fixed concentration at this
finite distance causes inaccuracy in the solution, however.

It is better to use a layer of Infinite Elements around the finite simulation space to
project the simulation space to infinity, eliminating any error from artificially limiting
the simulation space. This is a typical approximation when the electrolyte domain is a
 few orders of magnitude larger than the electrode: for example, a microelectrode in a
cm-scale reaction vessel.

Infinite Elements, Perfectly Matched Layers, and Absorbing Layers in the

COMSOL Multiphysics Reference Manual

Before You Start Building Your Model

INTRODUCING COMPLEXITY ONE STEP AT A TIME

The true art of modeling is knowing what phenomena to include and what phenomena
to neglect. Building complexity in a model by adding one additional piece of physics
or geometric feature at a time is usually the fastest way to reach the modeling goal.
Here are some general tips:

• Start thinking about your cell in the lowest possible dimension. Starting with a 1D
model helps to understand the influence of different reactions in an electrochemical
system, and gives a good first estimate of current-voltage behavior. Go from 1D to
2D, then from 2D to 3D.
• Every electrode reaction adds numerical nonlinearities to your model. If you have
multiple electrode reactions, add them one at a time.
• Start with a simple description of the electrolyte current, such as Secondary Current
Distribution. Analyze the results to ensure that the electrochemical model is
consistent. Switch only to more complex electrolyte models, or add extra physics
such as mass transfer, heat transfer or flow, only if deemed necessary and when
satisfied with the results from a simpler case.
• If you are including flow in your model, solve for the flow field first before coupling
flow and electrochemistry together.

MODELING CHECKLIST
• Identify which domains are electrode and electrolyte. How will their conductivity
be assessed?
• What is happening on the electrode-electrolyte interfaces? Do both the anode and
the cathode need to be modeled? Do either need to be modeled as domains, or can
they be treated as boundaries?
• What electrochemical reactions take place at the electrode surfaces to cause charge
transfer? Can you parameterize their thermodynamics? Do you know the

58 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

 equilibrium potentials? Can you parameterize their kinetics — and are the kinetics
ever going to be important? If not, ignore them.
• What is the system reference potential used to quote equilibrium potentials? Where
is the system ground?
• Are charge carriers in the electrolyte plentiful with respect to the drawn current
density, or is charge depletion important? Be aware of nonlinear effects that may
make convergence more difficult.
• If you are performing a time-dependent study, do your initial conditions have a
consistent current-voltage relationship?
• What other physics interfaces need to be coupled? How does charge transfer
influence these physics interfaces?

Meshing Advice
The default triangular (2D) or tetrahedral (3D) mesh is normally suitable for solving
the equations describing conservation of charge and mass for a general electrochemical
problem.

Electrochemical models involving mass transport generally benefit from a finer mesh
at the electrode surfaces, and at singularities such as the boundary between an
electrode surface and an insulating surface. This may be accomplished by adding
additional Size mesh nodes for these boundaries only. Also, consider refining the
“element growth rate”, and/or using boundary layer meshing in 3D.

Sometimes, a regular Mapped mesh can be more appropriate in a rectangular (2D)

domain; such domains are common in cutaway schematic studies of batteries and fuel
cells. Similarly the Swept mesh is often appropriate for very thin layers in a 3D model.

For fluid domains, the default physics-controlled mesh should be used, with boundary
layers as required.

For some problems with a stationary flow velocity field and time-dependent
convection of electrochemically reacting species, it may improve convergence to set up
a refined mesh without boundary layers for the species transport study step.

Meshing in the COMSOL Multiphysics Reference Manual

 Solving Electrochemical Models
Due to the highly nonlinear nature of electrode kinetics, some electrochemical models
can be difficult to solve. This section includes some general tips and tricks to facilitate
model solving, trouble shooting, and to improve solution accuracy.

GENERAL CURRENT DISTRIBUTION PROBLEMS

Start with the following suggestions if you encounter difficulty solving a problem.

• Make sure that the potential levels are “boot-strapped” somewhere in the model,
preferably by grounding one electrode. If there is no potential level defined
anywhere in the model, your model may have infinitely many solutions, and the
model will not converge.
• Consider using a Stationary with Initialization or a Time Dependent with Initialization
study. Both these studies will use a Current Distribution Initialization study step as a
first step to solve for the potentials only. If you run into problems solving for the
second step in this study you may have to change the “Current distribution type”
setting to Secondary on the Current Distribution Initialization study step node, and
also review the Initial Values as described in the next bullet.
• Review the Initial Values, especially the potentials. Suitable initial potential values can
usually be derived making a “potential walk” through the geometry, starting at the
grounded boundary. Compute electric and electrolyte potentials in other domains
by assuming equilibrium potential differences between electrode and electrolyte for
the main electrode reactions.
• Switch to Linearized Butler-Volmer kinetics (or a Primary current distribution) while
troubleshooting. This can be useful to help achieve a solution for a model that does
not solve with nonlinear kinetics, thereby indicating suitable initial values for the
nonlinear problem.
• If your model contains porous electrodes, try refining the mesh resolution in these
domains, especially toward the electrolyte boundaries.

• Electrochemistry Interfaces
• Specifying Initial Values and Inspecting and Troubleshooting Meshes
in the COMSOL Multiphysics Reference Manual

60 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

ELECTROCHEMISTRY COUPLED TO MASS TRANSPORT
If the model involves electrochemistry coupled to mass transport, here are a few things
to try to help improve model convergence.

• Review the Initial Values for the concentration values. Zero initial concentration
values can be unsuitable for tertiary current distribution problems and battery
simulations, since they could imply that no charge carriers or no reacting material is
present.
• If steep concentration gradients are expected close to electrode surfaces, use
boundary layer meshing or finer mesh Size settings at these boundaries.
• When setting up user-defined kinetics expressions, avoid evaluating negative
concentrations by using expressions such as max(c, eps^2), where eps is the
machine epsilon (a very small but finite number).
• Try to solve for low currents and low overpotentials first, then increase the cell load
(for stationary problems this can be done using an auxiliary sweep with
continuation).
• If a problem involving mass transport is hard to solve for high currents, but solves
for low currents, it might be due to mass transport limitations. In this case, review
the transport parameter values and check that the current magnitudes are
reasonable. If the current densities are unreasonably high, review the electrode
reaction settings.
• For time-dependent problems that run into convergence problems after a certain
time, review the solution at the last time-step. If the solution of a reactant reaches
zero or a maximum value (for insertion electrodes in batteries) when the
convergence issues occur, the current load of the model is too high in relation to
your initial concentrations or mass transport properties.

• Electrochemistry Interfaces
In the COMSOL Multiphysics Reference Manual:

• 2D and 3D Boundary Layer Meshes

• Stationary and Parametric Sweep

SETTING UP A STUDY SEQUENCE FOR MULTIPHYSICS PROBLEMS

For multiphysics problems, try to adjust the study sequence.

• Solve certain physics interfaces in a sequence. This can in many cases reduce
computational time and improve convergence. Analyzing the results when solving a
 physics interface separately can also help when troubleshooting a nonconverging
model.
• A good strategy is often to solve for the potentials only (that is, disable mass
transport and flow interfaces), using a stationary study step, before solving the full
model in the study sequence. In this way the stationary solution is used as initial
values for the following steps. This can be manually by modifying the settings of the
study node, or in an automated way by using the Stationary with Initialization or Time
Dependent with Initialization study sequences, as described above.
• In many models the flow profile is only slightly (or not at all) affected by changes in
current density. Therefore it can be a good strategy to solve separately for the flow
early in the study sequence, and then solve for the other physics interfaces in the
subsequent steps. (If the flow is not affected at all by the current distribution,
solving for the flow can be disabled entirely in consecutive steps.)

• Electrochemistry Interfaces
In the COMSOL Multiphysics Reference Manual:

• Studies and Solvers

• Analyzing Model Convergence and Accuracy
• Achieving Convergence When Solving Nonlinear Equations

TIME-DEPENDENT PROBLEMS WITH LOAD STEPS

For time-dependent problems, try the following to address accuracy and convergence
issues with regards to sudden current or potential load steps.

• Use smoothed current or potential load functions in order to avoid instantaneous

or discrete load steps.
• Add a double layer capacitance to the model, which can improve the numerical
stability.
• Reduce the Maximum step taken by the solver if you want to prevent the solver from
“missing” short square load steps, or change the Steps taken by solver setting from
Free to Strict, or Intermediate, to control the time steps using the Times text field.

62 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

 Using the Events interface can also be an option in certain cases if the load cycle
itself varies dynamically.

• Double-Layer Capacitance
In the COMSOL Multiphysics Reference Manual:

• Step and Time Dependent

• Building a COMSOL Multiphysics Model and The Events Interface

SOLVER SETTINGS
Try adjusting the solver settings.

• In rare cases, try to increase the Maximum number of iterations.

• If you know the order of magnitude of the dependent variables beforehand, setting
the scaling method to Manual for these can improve convergence and the accuracy
of the results. This may be of special importance when solving for concentrations
that are initially very small but where the final order of magnitude is know a priori.
This could be the case for the active species in a battery model, for example.

In the COMSOL Multiphysics Reference Manual:

• Studies and Solvers

• Analyzing Model Convergence and Accuracy
• Achieving Convergence When Solving Nonlinear Equations

Electrochemistry Studies and Study Steps and Cyclic Voltammetry in the

COMSOL Multiphysics Reference Manual

In multicomponent systems it is rarely straightforward to predict the rest

potential and direction of current flow. The Current Initialization step in
COMSOL Multiphysics does this for you but is only correct if
concentrations and formal potentials for all electrode reactions are
specified precisely and with respect to a common reference scale.
 Electrochemistry Studies and Study Steps in the COMSOL Multiphysics
Reference Manual

Studies and Solvers and Building a COMSOL Multiphysics Model in the

COMSOL Multiphysics Reference Manual

Postprocessing Your Solution

POTENTIAL VARIABLES
Several different potential variables are available for postprocessing and during
computation. The most common ones are described in Table 2-1.
TABLE 2-1: COMMON POTENTIAL VARIABLES.

VARIABLE DESCRIPTION DEFINED AT

phil Electrolyte phase potential Electrolyte and Porous

Electrode domains
phis Electrode phase potential Electrode and Porous
Electrode domains
[Link] Electrode phase potential Electrode Reaction boundaries
to Electrolyte domains
xxx.Eeq_yy Equilibrium potential Electrode Reaction boundaries
and Porous Electrode Reaction
domains
xxx.eta_yy Overpotential Electrode Reaction boundaries

The unit is V for all the above variables.

xxx denotes the tag of the physics interface. For instance, cd for the Secondary Current Distribution inter-
face.
yy denotes the tag of the (Porous) Electrode Reaction node, for instance, er1 for an Electrode Reaction
node.

CURRENT VARIABLES AND CALCULATING THE TOTAL CELL CURRENT

The current density may vary between locations on an electrode surface. This effect is
very important to understand through modeling because it cannot be directly
measured. Experimental measurement can only return the total current drawn at an
electrode, although the variation in current density can be inferred by other means.

64 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

To calculate the total current in COMSOL Multiphysics, it is necessary to integrate the
total drawn current density along the boundary or over the domain of the electrode.
Some common current density variables are shown in Table 2-2.
TABLE 2-2: COMMON CURRENT DENSITY VARIABLES.

VARIABLE UNIT DESCRIPTION DEFINED AT

2
[Link] A/m Electrolyte current density in Boundaries to Electrolyte and
normal direction Porous Electrode domains
[Link] A/m2 Electrode current density in Boundaries to Electrolyte and
normal direction Porous Electrode domains
xxx.iloc_yy A/m2 Local current density of Electrode Reaction boundaries
electrode reaction
[Link] A/m2 Total interface current density Electrode Reaction boundaries
(local sum of all xxx.iloc_yy)
xxx.iv_yy A/m3 Volumetric current density of a Porous Electrode domains
Porous Electrode Reaction
[Link] A/m3 Total volumetric current Porous Electrode domains
densities (local sum of all
xxx.iv_yy)
[Link] A/m2 Electrolyte current density Electrolyte domains
magnitude (L2 norm)
[Link] A/m2 Electrode current density Electrode domains
magnitude (L2 norm)

xxx denotes the tag of the physics interface. For instance cd for the Secondary Current Distribution inter-
face.
yy denotes the tag of the (Porous) Electrode Reaction node, For instance er1 for an Electrode Reaction
node.

In 1D or 1D axisymmetric, the electrode is implicitly uniform so you can just multiply

by the area of the electrode. In 2D or 2D axisymmetric, compute the appropriate line
integral. In 3D, compute the surface integral.

You can also define your own total current variable by using an Integration nonlocal
coupling across the electroactive boundaries. This variable can also be used during the
 computation. In 1D axisymmetric and 2D axisymmetric components, make sure to
select the Compute integral in revolved geometry checkbox.

In the COMSOL Multiphysics Reference Manual:

• Nonlocal Couplings and Coupling Operators and Integration

Avoid evaluating the [Link]/[Link] variables on boundaries. These

are generally less accurate than the [Link]/[Link] variables.

66 | CHAPTER 2: MODELING WITH ELECTROCHEMISTRY

 3

Fuel Cell and Electrolyzer Interfaces

This chapter describes the physics interfaces found under the Electrochemistry >
Hydrogen Fuel Cells ( ) and > Water Electrolyzers ( ) branches.

In this chapter:

• The Hydrogen Fuel Cell and Water Electrolyzer Interfaces

• Theory for the Hydrogen Fuel Cell and Water Electrolyzer Interfaces

67
 The Hydrogen Fuel Cell and Water
Electrolyzer Interfaces
The Hydrogen Fuel Cell (fc) and Water Electrolyzer (we) physics interfaces are used to
model the different electron and ion-conducting layers of various types of hydrogen
fuel cells, or water electrolyzer cells. Optionally, gas phase transport and Darcy’s law
for convection in porous media and flow channels may be enabled and also solved for
by the interfaces.

The interface supports simulation in 1D, 2D, and 3D as well as for axisymmetric
components in 1D and 2D.

Note that once added to a model, the Hydrogen Fuel Cell (fc) and the Water
Electrolyzer (we) interfaces are identical in terms of functionality and only differ by
some default parameter values (see below), the default name (fc/we), and the icon
( / ).

The Electrochemistry > Hydrogen Fuel Cell ( ) and the Electrochemistry > Water
Electrolyzers ( ) branches in the Model Wizard, or in the Select physics tab, contain
a number of different entries for adding either a Hydrogen Fuel Cell (fc) or a Water
Electrolyzer (we) physics interface, of different electrolyte types, to a model.

Depending on the selected entry, different species will be included by default in the
electrode gas mixtures, and gas phase diffusion will be enabled by default if there are
more than one species present a mixture. See also the H2 Gas Mixture and O2 Gas
Mixture section below.

The chosen Model Wizard entry will also have an effect on the default stoichiometry
and kinetics parameter values used when adding electrode reactions to the model. See
the Electrode Reaction Settings below.

SETTINGS AT THE INTERFACE LEVEL (FC/WE)

The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

68 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

The default Name (for the first physics interface in the model) is fc for the Hydrogen
Fuel Cell interface and we for the Water Electrolyzer interface.

DOMAIN SELECTION
If any part of the model geometry should not partake in the fuel cell/electrolyzer
model, remove that part from the selection list.

H2 GAS MIXTURE AND O2 GAS MIXTURE

These sections specify what the species are present in the Gas Phase domains on the
respective sides of the cell. Apart from hydrogen and oxygen, which are always
assumed to be present on each side, you may here choose to include one or several of
the species H2O, N2, CO2, and CO on each side. The species CH4 may be included on the
H2 side. An additional, arbitrarily defined, gaseous species can be included on either
side using Auxiliary species. The settings for the auxiliary species are found in the
corresponding Gas Phase domain nodes.

If you expect concentration gradients in the gas diffusion electrodes (GDEs), gas
diffusion layers (GDLs) or the flow channels, you may choose to Include gas phase
diffusion in these sections. Gas phase diffusion can only be enabled if there are more
than one species present in the gas mixture.

Use Darcy’s law for momentum transfer for solving for the gas phase pressure using
Darcy’s Law in gas phase domains. This can be enabled individually for each gas
mixture in these sections. Note, however, that instead of using the built-in Darcy’s law
of the fc/we interface, you can instead couple the interface to any Fluid Flow
interface, or an analytical expression, by specifying the velocity explicitly in the Gas
Phase nodes or by using an Reacting Flow, H2/O2 Gas Phase multiphysics coupling
nodes.

If Use H2O(l) in Reaction Stoichiometry is enabled, this will allow for setting the
stoichiometric coefficient of liquid water in electrode reactions, which in turn may be
used for the built-in calculations of equilibrium and thermoneutral potentials. Unit
activity of H2O(l) will be assumed in all expressions.

ELECTRODE REACTION SETTINGS

In this section you can control the default parameter settings that electrode reactions
will get when added to the model, and how the potential of the global hydrogen
reference electrode will be calculated.

The Charge-carrying ion setting (Proton, Hydroxide, Carbonate, Oxide or Generic) will
impact the default stoichiometry of electrode reactions and a built-in hydrogen

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 69

 reference electrode used for the calculation of equilibrium and thermoneutral
potentials of the electrode reactions.

For all types except Generic, the Reference hydrogen electrode equilibrium potential,
Eeq,RHE (V), will be calculated using an internal database of reaction enthalpies and
entropies, in combination with the Reference hydrogen electrode temperature, which
should be a global value with no spatial dependency.

The Generic option will add no default stoichiometric coefficients, and for this case the
Eeq,RHE and the Charge of the charge-carrying ion will have to be specified explicitly.

Note that the Entropy of the charge-carrying ion has no impact on the equilibrium
potentials vs Eeq,RHE, but will have an effect the calculation of the thermoneutral
potentials. (See also Calculation of Built-in Equilibrium and Thermoneutral
Potentials.)

The following reaction stoichiometry will be used for calculating Eeq,RHE, and as
default for the electrode reactions on the H2 and O2 sides, respectively.

Proton:

+ -
2H + 2e  H 2 (g)

+ -
O 2 (g) + 4H + 4e  2H 2 O(g)

Hydroxide:

- -
2H 2 O(g) + 2e  H 2 (g) + 2OH

- -
O 2 (g) + 2H 2 O(g) + 4e  4OH

Carbonate:

- 2-
H 2 O(g) + CO 2 (g) + 2e  H 2 (g) + CO 3

- 2-
O 2 (g) + 2CO 2 (g) + 4e  2CO 3

Oxide:

- 2-
H 2 O(g) + 2e  H 2 (g) + O

70 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

 - 2-
O 2 (g) + 4e  2O

The Operation mode setting (Fuel cell or Electrolyzer) will impact the default values of
the kinetics settings of added electrode reactions.

MEMBRANE TRANSPORT
In this section, transport of additional species (apart from the charge carrying ion) in
Membrane domains may be enabled. The section is available if Electrolyte type setting
is set to either Proton exchange or Hydroxide exchange.

The Crossover species include H2, O2, and N2. Note that the crossover species N2 is
available only if species N2 has been selected in both the H2 Gas Mixture and the O2 Gas
Mixture.

Electroosmotic water drag is available if the species H2O has been selected in both the
H2 Gas Mixture and the O2 Gas Mixture.

REFERENCE PRESSURE LEVEL

This section is available if the Use Darcy’s law for momentum transport checkbox has
been selected. In this section, you specify the Reference pressure level.

CONSISTENT STABILIZATION
This section is available when Stabilization is selected in the Show More Options dialog
shown when the Show More Options button ( ) is selected.

The stabilization settings are applicable separately for the H2 Gas Mixture and the O2
Gas Mixture and are available if the corresponding Include gas phase diffusion checkbox
is selected.

In convection-dominated parts of a cell where the mesh is not fine enough to resolve
the concentration gradients, for instance in a boundary layer, stabilization may be
added in order to avoid spurious oscillations in the solution, and to improve
convergence. In fuel cells and electrolyzer cells, stabilization is more likely to be
needed in the gas flow channels. In gas diffusion layers and gas diffusion electrodes
stabilization is usually not needed due to the lower flow velocities. The stabilization
expressions, particularly the crosswind diffusion expressions, may be costly to
compute. Therefore, it is recommended to add stabilization only if needed, for
instance when spurious gas mass fractions oscillations are observed in a solution.

There are two consistent stabilization methods available — Streamline diffusion (active
by default) and Crosswind diffusion (inactive by default). The Residual setting applies to

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 71

 both the consistent stabilization methods. Approximate residual is the default setting
and it means that derivatives of the diffusion tensor components are neglected. This
setting is usually accurate enough and computationally faster. If required, select Full
residual instead.

By default, Streamline diffusion stabilization, using an Approximate residual, is added

only to gas flow channels. Crosswind diffusion stabilization, and computation of the Full
residual may be turned on as required, in all domains.

ADVANCED SETTINGS
From the Regularization list, select On (default) or Off. When turned On, regularized
mass fractions are selected such that they lie between 0 and 1. The regularized mass
fractions are used in the calculation of composition-dependent material properties,
such as density.

This section is available when the Advanced Physics Options is selected in the Show More
Options dialog shown when the Show More Options button ( ) is selected, in
combination with the Include gas phase diffusion checkbox being selected.

PHYSICS VS. MATERIALS REFERENCE ELECTRODE POTENTIAL

The Physics vs. Materials Reference Electrode Potential setting on the physics interface
node can be used to combine material library data for current densities and equilibrium
potentials with an arbitrary reference electrode scale in the physics. The setting affects
the electrode potentials used for model input into the materials node as well as all
equilibrium potential values output from the materials node.

Note that the setting will only impact how potentials are interpreted in communication
between the physics and the Materials node. If the From material option is not in use
for equilibrium potentials or electrode kinetics, the setting has no impact.

This section is available when the Advanced Physics Options is selected in the Show More
Options dialog shown when the Show More Options button ( ) is selected.

OVERVIEW OF THE DOMAIN NODES IN THE HYDROGEN FUEL CELL AND

WATER ELECTROLYZER INTERFACES
The following phase nodes are added as default nodes:

• Electrolyte Phase
• Electronic Conducting Phase
• Gas Phase (H2 and O2)

72 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

These default phase nodes define the physics in the different phases of the cell, and
contain settings for, for instance, electrolyte conductivity and gas diffusion coefficients.
Initial Values for the dependent variables defined by the phase nodes are available as
subnodes. Boundary conditions for, for instance, potential, current densities, inflow
gas compositions, and so on, are added as subnodes to the above phase nodes.

The default phase nodes have their selections locked to all domains, but may have their
selections overridden by the following nodes, which may be added by the user:

• Membrane
• Separator
• Current Collector
• Gas Diffusion Electrode (H2 or O2)
• Gas Diffusion Layer (H2 or O2)
• Gas-Electrolyte Compartment (H2 or O2)
• Gas Flow Channel (H2 or O2)

These nodes define properties that do not pertain to the pure electrolyte or gas phases
such as porosities, permeabilities, and electronic conductivities. Electrode reactions are
added as subnodes to the gas diffusion electrode nodes.

Electrolyte Phase
The Electrolyte Phase node defines the governing domain equations for the electrolyte
phase.

Note that this node is added by default, and that you cannot alter the selection of the
node by settings on the node itself. The node will be active on domains selected by
Membrane, Separator, Gas Diffusion Electrode or Gas-Electrolyte Compartment
nodes.

ELECTROLYTE CHARGE TRANSPORT

The Electrolyte conductivity, l (SI unit: S/m), parameter will define how the current
in the domain depends on the gradient of the potential.

See also Electrolyte Phase Theory.

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 73

 Electronic Conducting Phase
The Electronic Conducting Phase node defines the governing equations for the
electronic conducting (electrode) phase.

Note that the node is added by default, and that you cannot alter the selection of the
node by settings on the node itself. The node will be active on domains selected by
Current Collector, Gas Diffusion Electrode or Gas Diffusion Layer nodes. The
electronic conductivity, Electric conductivity, s (SI unit: S/m), used in the domain
equations, is defined on these domains.

See also Electronic Conducting Phase Theory.

Gas Phase
The H2 Gas Phase and O2 Gas Phase nodes define the governing domain equations for
gas phase transport.

Note that these nodes are added by default, and that you cannot alter the selection of
the nodes by settings on the nodes themselves. The H2 Gas Phase and O2 Gas Phase are
active on domains selected by the corresponding H2/O2 Gas Diffusion Electrode,
Gas Diffusion Layer, Gas-Electrolyte Compartment or Gas Flow Channel nodes.

The H2 Gas Phase and O2 Gas Phase subnodes are only visible if the Include gas phase
diffusion checkbox, or if the Use Darcy’s law for momentum transport checkbox, are
enabled in the corresponding Gas Mixture section on the interface top node.

COMPOSITION
This section is only visible if any of the Include species checkboxes has been selected
and the Include gas phase diffusion checkbox has not been selected in the physics
interface top node.

Use Mixture specification to specify mixture composition in Mole fractions or Humidified

mixture for all gas phase domains. The O2 Gas Phase node has an additional option of
Humidified air for O2 gas phase domains.

The option Mole fractions is always available and text fields are used to specify the
mixture composition for all included species.

Humidified mixture allows for specifying a gas mixture using values for a dry gas
mixture, which is subsequently humidified to a specified relative humidity. The setting
is available if H2O has been selected in the corresponding Gas Mixture section on the
interface top node. Text fields are used to specify the mixture composition for all

74 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

included species except H2O. Additionally, the Relative humidity, Humidification
temperature and Humidification absolute pressure of the humidifier are also specified.

Humidified air is available if H2O and N2 have been selected in the O2 Gas Mixture
section on the interface top node. Relative humidity, Humidification temperature, and
Humidification absolute pressure of the humidifier are also specified. The dry mole
fractions of the included species are set internally as follows: 0.79 (N2), 4e-4 (CO2)
and 1e-6(CO).

Given that the humidifier (inlet) temperature (Thum), absolute pressure (pA,hum) and
relative humidity (RHhum) of the humidifier (inlet) gases are known, the conversion
from a set of dry (containing no water vapor) molar fraction to a corresponding
humidified mixture may be calculated as follows. The water mole fraction is defined as

p vap  T hum 
x H2O = ------------------------------ RH hum
p A hum

The wet mole fractions for all species except water are then calculated from the dry
mixture molar fractions according to

x i = x i dry  1 – x H2O 

MIXTURE PROPERTIES
In this section, the gas phase values for the Density and Dynamic viscosity are specified.

The mixture properties will for instance be used by the interface when modeling
momentum transport by Darcy’s law.

Note: Depending on the settings on the interface top node, the mixture property
values may or may not be used by the interface itself. However, the interface will
define mixture property variables that may be used when coupling to other physics
interfaces, either manually or by multiphysics nodes such as Reacting Flow, H2/O2 Gas
Phase. Additionally, the interface also defines liquid water properties (density and
dynamic viscosity) in the H2/O2 Gas Phase domain nodes that may be used when
coupling to other physics interfaces, if H2O has been selected in the corresponding
Gas Mixture section on the interface top node.

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 75

 WATER VAPOR PRESSURE
This section will only be visible if H2O checkbox in the Include species or Include H2O(l)
in reaction stoichiometry checkbox has been selected in the corresponding Gas Mixture
section on the interface top node.

In this section, the values for the Vapor pressure and Heat of vaporization are specified.
The Vapor pressure is for instance used when specifying inlet and initial values for the
gas composition in terms of relative humidity. The Heat of vaporization is used for
automatic calculation of equilibrium and thermoneutral electrode reaction potentials
involving liquid water, as well as heat sources added by the Water
Condensation-Evaporation nodes and the Water Absorption-Desorption nodes.

CONVECTION
This section will only be visible if the Use Darcy’s law for momentum transport checkbox
has been disabled in the interface top node.

Specify the Velocity field of the gas. The velocity field can be specified explicitly, or
using an analytical expression, or coupled from any Fluid Flow interface, or by using
an Reacting Flow, H2/O2 Gas Phase multiphysics coupling node.

DIFFUSION
This section will only be visible if the Include gas phase diffusion checkbox has been
selected in the interface top node.

Specify the Binary diffusion coefficients for all pairs of species in the gas.

See also Gas Phase Theory.

AUXILIARY SPECIES
This section will only be visible if Auxiliary species checkbox has been selected in the
interface top node.

Specify the properties of the auxiliary species, such as Molar Mass, Molar standard
enthalpy of formation, Molar standard entropy, and Molar heat capacity. The Fuller
diffusion volume is available if the Include gas phase diffusion checkbox has been selected
in the interface top node and if using Built-in Binary diffusion coefficients. The Dynamic
viscosity, Thermal conductivity and Boiling point are used for calculating the built-in gas
phase dynamic viscosity and thermal conductivity, and are available if using Built-in
Dynamic viscosity.

76 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

Membrane
The Membrane node specifies a domain with only an Electrolyte Phase node being
active.

The electrolyte conductivity of the membrane is defined on the Electrolyte Phase

node.

HYDROGEN/OXYGEN/NITROGEN CROSSOVER
These sections specifies the permeation (diffusion) of gases between the hydrogen and
oxygen gas phase mixtures, across the membrane.

The Crossover sections for H2, O2, and N2, available if the corresponding crossover
species has been selected in the Membrane Transport section on the interface top node,
can be used to specify the Permeation coefficient and the Henry’s law coefficient of the
corresponding species. The Heat of reactions, arising due to the parasitic reactions of
H2 and O2 species, is available if H2 or O2 has been selected as crossover species. The
Thermoneutral voltage is calculated automatically when the Built in option is used.

See also Membrane Crossover Theory.

ELECTROOSMOTIC WATER DRAG

This section specifies the water transport properties of the membrane. Water transport
may occur both due to migration in the electric field and diffusion.

This section is visible if Electroosmotic water drag has been selected in the Membrane
Transport section on the interface top node. The Water transport model can be selected
as either Vapor equilibrated (default for fuel cells) or Liquid equilibrated (default for
electrolyzers). The Electroosmotic coefficient, the Water transport coefficient, and the
Standard condition gas pressure are specified in this section. The Heat of water
absorption is available if Electroosmotic water drag has been selected.

See also Electroosmotic Water Drag Theory.

WATER VAPOR PRESSURE

This section will only be visible if Electroosmotic water drag has been selected in the
Membrane Transport section on the interface top node.

The Vapor pressure defined on this node is used for computing a domain water activity
(relative humidity) variable which corresponds to the computed water chemical
potential.

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 77

 Note: The computed water activity variable is typically used as model input to the
Materials node, setting initial values of the water activity in terms of relative humidity,
or during post-processing.

ELECTROLYTE WATER ACTIVITY FOR MATERIAL MODEL INPUT

This section is visible if Electroosmotic water drag has not been selected in the
Membrane Transport section on the interface top node. The water activity specified in
this section is used by the Electrolyte Phase node when using an electrolyte material that
requires Water activity (relative humidity) as model input.

Separator
The Separator node specifies a domain with only an Electrolyte Phase node being
active. This node is typically used to separate two Gas-Electrolyte Compartment
domains of a liquid-electrolyte-based electrolyzer.

EFFECTIVE ELECTROLYTE CHARGE TRANSPORT

The bulk electrolyte conductivity defined in the Electrolyte Phase node will be
multiplied by a correction factor to compensate for the effect of the Electrolyte volume
fraction being less than 1. This can be done by either using a Bruggeman, Tortuosity, or
User defined correlation. For Tortuosity, select either Isotropic to define a scalar value,
or Diagonal or Symmetric to define anisotropic tensor values.

Current Collector
The Current Collector node specifies a domain with only the Electronic Conducting
Phase node being active.

ELECTRODE CHARGE TRANSPORT

The electronic conductivity, Electric conductivity, s (SI unit: S/m) parameter will
define how the electric current in the corresponding Electronic Conducting Phase
domain depends on the gradient of the potential.

78 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

Gas Diffusion Layer
The H2 Gas Diffusion Layer and O2 Gas Diffusion Layer nodes specify domains where both
an Electronic Conducting Phase and a Gas Phase node are active.

ELECTRODE CHARGE TRANSPORT

The effective electronic conductivity, Effective electric conductivity, s (SI unit: S/m)
parameter will define how the electric current in the corresponding Electronic
Conducting Phase domain depends on the gradient of the potential.

GAS TRANSPORT
This section will only be visible if the Include gas phase diffusion checkbox or the Use
Darcy’s law for momentum transport checkbox has been selected in the corresponding
Gas Mixture section on the interface top node.

If the Include gas phase diffusion checkbox has been selected, the gas diffusivity defined
in the Gas Phase node will be multiplied by a correction factor to compensate for the
effect of the Gas pore volume fraction being less than 1. This can be done by either using
a Bruggeman, Tortuosity, or a User defined correlation.

If the Use Darcy’s law for momentum transport checkbox has been selected, the Gas
permeability is defined in this node.

By checking Include pore-wall interaction, additional friction forces between the

individual gas molecules and the walls are accounted for when computing the gas
fluxes (given that Include gas phase diffusion is active). The pore-wall friction forces
make use of a Wall diffusivity, which is set to Knudsen by default. The Knudsen
diffusivity can be derived from kinetic gas theory by relating the mean-free path of the
gas molecules to the Pore diameter of the pores.

Gas-Electrolyte Compartment
The H2 Gas-Electrolyte Compartment and O2 Gas-Electrolyte Compartment nodes specify
domains where both an Electrolyte Phase, and a Gas Phase node are active. This node
is typically used to model a hydrogen or oxygen evolving electrolyte compartment in
a liquid-electrolyte-based electrolyzer.

EFFECTIVE ELECTROLYTE CHARGE TRANSPORT

The bulk electrolyte conductivity defined in the Electrolyte Phase node will be
multiplied by a correction factor to compensate for the effect of the Electrolyte volume
fraction being less than 1 (caused by the evolving gas). This can be done by either using

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 79

 a Bruggeman, Tortuosity, or User defined correlation. For Tortuosity, select either
Isotropic to define a scalar value, or Diagonal or Symmetric to define anisotropic tensor
values.

GAS TRANSPORT
This section will only be visible if the Include gas phase diffusion checkbox or the Use
Darcy’s law for momentum transport checkbox has been selected in the corresponding
Gas Mixture section on the interface top node.

If the Include gas phase diffusion checkbox has been selected, the gas diffusivity defined
in the Gas Phase node will be multiplied by a correction factor to compensate for the
effect of the Gas pore volume fraction being less than 1 (caused by the evolving gas).
This can be done by either using a Bruggeman, Tortuosity, or User defined correlation.
For Tortuosity, select either Isotropic to define a scalar value, or Diagonal or Symmetric
to define anisotropic tensor values.

If the Use Darcy’s law for momentum transport checkbox has been selected, the Gas
permeability is defined in this node.

Gas Diffusion Electrode

The H2 Gas Diffusion Electrode and O2 Gas Diffusion Electrode nodes specify three phase
domains where an Electronic Conducting Phase, an Electrolyte Phase, and a Gas Phase
node are all active.

ELECTRODE CHARGE TRANSPORT

The effective electronic conductivity, Effective electric conductivity, s (SI unit: S/m)
parameter will define how the electric current in the corresponding Electronic
Conducting Phase domain depends on the gradient of the potential.

EFFECTIVE ELECTROLYTE CHARGE TRANSPORT

The bulk electrolyte conductivity defined in the Electrolyte Phase node will be
multiplied by a correction factor to compensate for the effect of the Electrolyte volume
fraction being less than 1. This can be done by either using a Bruggeman, Tortuosity, or
User defined correlation. For Tortuosity, select either Isotropic to define a scalar value,
or Diagonal or Symmetric to define anisotropic tensor values.

80 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

GAS TRANSPORT
This section will only be visible if the Include gas phase diffusion checkbox or the Use
Darcy’s law for momentum transport checkbox has been selected in the corresponding
Gas Mixture section on the interface top node.

If the Include gas phase diffusion checkbox has been selected, the gas diffusivity defined
in the Gas Phase node will be multiplied by a correction factor to compensate for the
effect of the Gas pore volume fraction being less than 1. This can be done by either using
a Bruggeman, Tortuosity, or User defined correlation. For Tortuosity, select either
Isotropic to define a scalar value, or Diagonal or Symmetric to define anisotropic tensor
values.

If the Use Darcy’s law for momentum transport checkbox has been selected, the Gas
permeability is defined in this node.

By checking Include pore-wall interaction, additional friction forces between the

individual gas molecules and the walls are accounted for when computing the gas
fluxes (given that Include gas phase diffusion is active). The pore-wall friction forces
make use of a Wall diffusivity, which is set to Knudsen by default. The Knudsen
diffusivity can be derived from kinetic gas theory by relating the mean-free path of the
gas molecules to the Pore diameter of the pores.

Gas Flow Channel

The H2 Gas Flow Channel and O2 Gas Flow Channel nodes specify domains with only a
Gas Phase node being active.

This node will be available if the Include gas phase diffusion checkbox or the Use Darcy’s
law for momentum transport checkbox has been selected in the corresponding Gas
Mixture section on the interface top node.

GAS TRANSPORT
If the Include gas phase diffusion checkbox has been selected, there are no settings on
this node. The gas diffusivity defined in the Gas Phase node.

If the Use Darcy’s law for momentum transport checkbox has been selected, the Gas
permeability is defined in this node.

The Gas permeability option of Straight channels can be used to specify the permeability
for a straight channel with a uniform cross section, for which analytical correlations
exist for expressing the flow rate as proportional to the volumetric flow. For a

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 81

 rectangular cross section, with height H and width W, the corresponding Darcy
permeability g (derived by Boussinesq in 1868) is


2
16W
3
1 cosh   n H  – 1
g = W
-------- – --------------
-
 -----------------------
- ---------------------------------------
 2n – 1  sinh   n H 
12 5 5
 H
n=1

with n being a help variable equal to

 n = ------------------------
2n – 1 -
W

In the COMSOL implementation of the above expression, the sum is truncated to the
first ten terms. It should be noted that the analytical expression renders a correct
average flow rate for a straight channel with no slip conditions on the channel walls.
To resolve the local flow field close to a wall, or for curved channels, solving for the
full set of the Navier–Stokes equations (as defined by the Laminar Flow, Brinkman
Equations, or Free and Porous Media Flow, Brinkman interfaces) is typically required.
However for cases were full accuracy of the local flow field is not required, the above
approximation may save substantial computational resources in terms of memory and
computation time.

Initial Values
Use this subnode to specify the initial values used by the solver of the dependent
variables defined by the parent node.

In the Initial Values subnode to the H2 Gas Phase and O2 Gas Phase nodes, Mixture
specification in the Initial Composition section can be used to specify mixture
composition in Mole fractions or Humidified mixture for all gas phase domains. The O2
Gas Phase node has an additional option of Humidified air for O2 gas phase domains.

The option Mole fractions is always available and text fields are used to specify the
mixture composition for all included species.

Humidified mixture allows for specifying a gas mixture using values for a dry gas
mixture, which is subsequently humidified to a specified relative humidity. The setting
is available if H2O has been selected in the corresponding Gas Mixture section on the
interface top node, and text fields are used to specify the mixture composition for all
included species except H2O. Additionally, the Relative humidity, Humidification
temperature, and Humidification absolute pressure of the humidifier are also specified.

82 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

Humidified air is available if H2O and N2 have been selected in the O2 Gas Mixture
section on the interface top node. Relative humidity, Humidification temperature, and
Humidification absolute pressure of the humidifier are also specified. The dry mole
fractions of the included species are set internally as follows: 0.79 (N2), 4e-4 (CO2)
and 1e-6(CO).

Given that the humidifier (inlet) temperature (Thum), absolute pressure (pA,hum) and
relative humidity (RHhum) of the humidifier (inlet) gases are known, the conversion
from a set of dry (containing no water vapor) molar fraction to a corresponding
humidified mixture may be calculated as follows. The water mole fraction is defined as

p vap  T hum 
x H2O = ------------------------------ RH hum
p A hum

The wet mole fractions for all species except water are then calculated from the dry
mixture molar fractions according to

x i = x i dry  1 – x H2O 

In the Initial Values subnode to the Membrane node, Specify water chemical potential as
can be used to specify water Activity, Liquid water pressure, or water Chemical potential.
The default From parent option specifies water activity for Vapor equilibrated model and
liquid water pressure for Liquid equilibrated model, respectively, chosen on the parent
node.

Initial Values, O2 Domains

Use this subnode to specify the initial values used by the solver of the dependent
variables defined by the parent Electronic Conducting Phase node.

The selection is locked to the intersection of the parent node and the O2 Gas Phase
node.

Initial Values, H2 Domains

Use this subnode to specify the initial values used by the solver of the dependent
variables defined by the parent Electronic Conducting Phase node.

The selection is locked to the intersection of the parent node and the H2 Gas Phase
node.

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 83

 Water Condensation-Evaporation
This subnode is available to the Gas Phase nodes when H2O has been selected, along
with Include gas phase diffusion or Use Darcy’s law for momentum transport in the
corresponding Gas Mixture section on the interface top node.

In low-temperature fuel cells or electrolyzers, gaseous water vapor may condensate as

liquid water if the partial pressure of vapor exceeds the (equilibrium) vapor pressure as
follows

H 2 O(g)  H 2 O(l)

The node may be used for adding an additional mass source due to water
condensation–evaporation. Note that mass source may also be added to liquid water if
Include H2O(l) in reaction stoichiometry is active in the corresponding Gas Mixture
section on the interface top node. The Condensation-Evaporation Rate may be defined
using the From vapor pressure, or by a User defined expression. The node contains text
fields for specifying the Rate constant and Reference pressure, in case of the default From
vapor pressure option. The rate expression is added to the rate variable for gaseous
water (based on the reaction stoichiometry), which is in turn used in the domain
equations for diffusion and/or Darcy’s law if these are active on the interface top node.

The Heat of reaction is calculated automatically when the Built in option is used.

The default selection is all domains of the parent node, but may be changed.

Water Gas Shift Reaction

This subnode is available to the H2 Gas Phase node when H2O, CO, and CO2 have
all been selected, along with Include gas phase diffusion in the H2 Gas Mixture on the
interface top node.

The water gas shift reaction defines the reaction of carbon monoxide and water vapor
to form carbon dioxide and hydrogen according to

CO(g) + H 2 O(g)  CO 2 (g) + H 2 (g)

The water gas shift reaction is of particular interest for fuel cells operating on fuel
stemming from hydrocarbon reformation, containing significant amounts of carbon
dioxide and carbon monoxide. The reaction is also of general interest in molten
carbonate fuel cells and electrolyzers, due to the presence of carbon dioxide in the
hydrogen fuel/product fluid stream.

84 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

The node may be used for adding an additional mass source due to water gas shift
reaction, such as in a molten carbonate fuel cell. The Water Gas Shift Reaction Rate may
be defined using the From mass action law or by a User defined expression. The node
contains text fields for specifying the Rate constant and Reference pressure, in case of
the default From mass action law option. The rate expression is added to the rate
variables for the gaseous species (based on the reaction stoichiometry), which are in
turn used in the domain equations for diffusion. Since the reaction does not involve
phase changes, the total mass source from the water gas shift reaction is zero, and no
contribution is hence made to the momentum equation.

The Heat of reaction is calculated automatically when the Built in option is used.

The default selection is all domains of the parent node, but may be changed.

Reaction Sources
This subnode is available to the Gas Phase nodes when the Advanced Physics Options is
selected in the Show More Options dialog shown when the Show More Options button
( ) is selected, in combination with the Include gas phase diffusion or Use Darcy’s law
for momentum transport checkbox being selected in the interface top node for the
corresponding gas mixture.

The node may be used for adding additional mass sources in a model, typical stemming
from nonfaradaic reactions such as water condensation-evaporation or water drag in a
low-temperature fuel cell, or homogeneous reactions such as the water gas shift
reaction in a molten carbonate fuel cell. The node contains text fields for specifying the
mass or molar reaction sources of all active gas phase species, including liquid water.

The Heat Source is calculated automatically when the Built in option is used.

The default selection is all domains of the parent node, but may be changed.

Gas Diffusion Electrode Reaction

The H2 Gas Diffusion Electrode Reaction and O2 Gas Diffusion Electrode Reaction are
available as subnodes to the corresponding Gas Diffusion Electrode parent nodes.

The node specifies the stoichiometry, thermodynamics, and kinetics of a gas diffusion
electrode reaction.

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 85

 STOICHIOMETRIC COEFFICIENTS
Specify the coefficients using negative signs for oxidized species and positive signs for
reduced species. The number of electrons should always be positive.

EQUILIBRIUM POTENTIAL
The equilibrium potential may be defined using the Nernst equation, From material, or
by a User defined expression.

The Built in option for the Reference equilibrium potential is based on the reaction
stoichiometry and the enthalpies and entropies of the reacting species.

See also Calculation of Built-in Equilibrium and Thermoneutral Potentials and

Electrode Reaction in Shared Physics Features in the Current Distribution Interfaces.

REFERENCE PRESSURES
Both the exchange current density and the equilibrium potential of an electrode
reaction depend on the partial pressures of the reacting species. By the use of a
reference state (a given set of partial pressures) and the Reference equilibrium potential
and Reference exchange current density, analytical expressions such as the Nernst
equation (for the equilibrium potential) and Mass action law or Lumped multistep (for
the Butler–Volmer kinetics) may be used to calculate the corresponding partial-pressure
dependent values.

Specify the species partial pressures for the reference state used for defining the
Reference exchange current density and the Reference equilibrium potential. The default
value (1 atm) refers to standard conditions.

If the Reference equilibrium potential is set to Built in, the settings of this section only
applies to the Reference exchange current density.

ELECTRODE KINETICS
The settings of this section define how the overpotential of the electrode reaction
relates to the local current density on the (catalyst) electrode surface.

See also Electrode Reaction in Shared Physics Features in the Current Distribution
Interfaces.

ACTIVE SPECIFIC SURFACE AREA

Specify the active surface (catalyst) area available for the reaction.

86 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

HEAT OF REACTION
The settings of this section define the heat source of the reaction.

The Built in option for the Thermoneutral voltage is based on the reaction stoichiometry
and the entropies of the reacting species.

See also Calculation of Built-in Equilibrium and Thermoneutral Potentials and

Electrode Reaction in Shared Physics Features in the Current Distribution Interfaces.

Porous Double Layer Capacitance

See Porous Matrix Double-Layer Capacitance in Shared Physics Features in the
Current Distribution Interfaces.

Overview of the Boundary Nodes in the Hydrogen Fuel Cell and

Water Electrolyzer Interfaces
Boundary nodes applying constraints or flux contributions to one phase only are added
as subnodes. The following boundary subnodes may be added to Electronic
Conducting Phase:

• Insulation (default)
• Electric Ground
• Electric Potential
• Electrode Current
• Electrode Current Density
• Electrode Power
• Internal Electrode Contact Resistance

The following boundary subnodes may be added to Electrolyte Phase:

• Insulation (default)
• Electrolyte Potential
• Electrolyte Current Density
• Internal Electrolyte Contact Resistance

The following boundary subnodes may be added to H2 or O2 Gas Phase:

• No Flux (default)
• Inlet

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 87

 • Outlet
• Flux

The following boundary subnodes may be added to Membrane:

• Water Absorption-Desorption, H2/O2 side (default)

• Water Flux
• Water Chemical Potential
The following additional boundary nodes, with contributions to multiple phases, can
be added as nodes directly under the interface node:

• Thin Gas Diffusion Electrode

• Electrode Surface
• Internal Electrode Surface
• Periodic Condition

Insulation
The Insulation boundary condition describes the walls of a cell or the boundaries of the
cell that do not face a conductor. The boundary condition imposes the following
equation:

ik  n = 0

where ik denotes the current density vector and k = l, s is an index for the electrolyte
and electrode, respectively.

The node can be added as a subnode to either Electrolyte Phase or Electronic

Conducting Phase.

Electrolyte Potential
Add the Electrolyte Potential node as a subnode to the Electrolyte Phase for a boundary
with a fixed potential at a position in the electrolyte.

The node sets the potential in the electrolyte, l, to be equal to the Boundary electrolyte
potential, l,bnd (SI unit: V).

88 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

Electrolyte Current Density
Use the Electrolyte Current Density node as a subnode to the Electrolyte Phase to
specify the current density distribution along a boundary.

Note that using this node in 2D or 3D may result in an uneven potential distribution
along the boundary.

Internal Electrolyte Contact Resistance

Use the Internal Electrolyte Contact Resistance node as a subnode to the Electrolyte
Phase to specify a Contact resistance at an internal boundary of the domain selection of
the Electrolyte Phase. The contact resistance is modeled by adding slit boundary
conditions. The slit boundary defines separate degrees of freedom for the electrolyte
potential at the up and down sides of the slit boundary, and then adds current density
(Neumann) conditions at both the up and down side.

Electric Ground
The node can be added as a subnode to Electronic Conducting Phase. This node sets
the electric potential to zero. The node is typically used to ground the voltage at the
current collector of the hydrogen side of the cell.

CONTACT RESISTANCE
Enabling Include contact resistance allows for a Contact resistance Rc (SI unit: ·m2)
to be specified and the following current density condition is used:

–
– n  i s = -------s-
Rc

Electric Potential
The node can be added as a subnode to Electronic Conducting Phase. This node sets
the electric potential in the electrode (or a porous electrode), s, to a value, sbnd
according to the following:

 s =  s bnd

The node is typically used to set the cell voltage at the current collector of the oxygen
side of the cell.

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 89

 CONTACT RESISTANCE
Enabling Include contact resistance allows for a Contact resistance Rc (SI unit: ·m2)
to be specified and the following current density condition is used:

 s bnd –  s
– n  i s = ---------------------------
-
Rc

HARMONIC PERTURBATION
Enable Include harmonic perturbation to specify the Perturbation amplitude value. This
setting will be used by the Frequency-Domain Perturbation study step when running an
AC Impedance study.

Electrode Current
Use the Electrode Current node to set the total current or average current density over
an external at an external boundary of an Electronic Conducting Phase domain —
typically at the interface between the electrode and the current collector or current
feeder. The condition sets the total inward current without imposing the current
density distribution. The potential along the boundary is calculated in order to satisfy
the total value of the current.

When using the Total current option in 1D or 2D, the boundary area is based either
on the Cross-sectional area (1D), or the Out-of-Plane thickness (2D) parameters.

CONTACT RESISTANCE
Enabling Include contact resistance allows for a Contact resistance Rc (SI unit: ·m2)
to be specified.

HARMONIC PERTURBATION
Enable Include harmonic perturbation to specify the Perturbation amplitude value. This
setting will be used by the Frequency-Domain Perturbation study step when running an
AC Impedance study.

Electrode Current Density

The Electrode Current Density boundary condition can be added as a subnode to and
applied at an external boundary of an Electronic Conducting Phase domain.

Note that using this node in 2D or 3D may result in an uneven potential distribution
along the boundary. To mitigate such effects, use the Electrode Current node instead.

90 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

HARMONIC PERTURBATION
Enable Include harmonic perturbation to specify the Perturbation amplitude value. This
setting will be used by the Frequency-Domain Perturbation study step when running an
AC Impedance study.

Electrode Power
The Electrode Power boundary condition sets the power drawn from, or inserted to,
the cell at an external boundary of an Electronic Conducting Phase domain

When using the Total power option in 1D or 2D, the boundary area is based either on
the Cross-sectional area (1D), or the Out-of-Plane thickness (2D) parameters.

CONTACT RESISTANCE
Enabling Include contact resistance allows for a Contact resistance Rc (SI unit: ·m2)
to be specified.

Internal Electrode Contact Resistance

Use the Internal Electrode Contact Resistance node as a subnode to the Electronic
Conducting Phase to specify a Contact resistance at an internal boundary of the domain
selection of the Electronic Conducting Phase. The contact resistance is modeled by
adding slit boundary conditions. The slit boundary defines separate degrees of
freedom for the electric potential at the up and down sides of the slit boundary, and
then adds current density (Neumann) conditions at both the up and down side.

No Flux
The No Flux node is the default boundary condition available for exterior boundaries
of Gas Phase nodes. It should be used on boundaries across which there is no mass flux,
typically exterior solid walls where no surface reactions occur. The condition applied
for each species corresponds to

–n  ji = 0

where n denotes the outward pointing normal of the boundary.

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 91

 Inlet
The H2 Inlet and O2 Inlet nodes adds a boundary condition for an inflow boundary,
defining the composition of the mixture and the flow boundary condition.

This node is available for exterior boundaries to Gas Phase nodes.

The Mixture specification section is available if the Include gas phase diffusion checkbox
has been selected on the interface top node. The mixture composition can be specified
using the following quantities in:

• From initial values (default)

• Mole fractions: xi  x0,i
• Humidified mixture
• Humidified air
• Mass fractions: i  0,i
• Mole fractions: xi  x0,i
• Molar concentrations: ci  c0,i
• Number densities, which describes the number of particles per volume: ni  n0,i
• Densities: i  0,i
• Mass flow rates: J0,i

The section requires input for all selected gas species.

Humidified mixture allows for specifying a gas mixture using values for a dry gas
mixture, which is subsequently humidified to a specified relative humidity. The setting
is available if H2O has been selected in the corresponding Gas Mixture section on the
interface top node, and text fields are used to specify the mixture composition for all
included species except H2O. Additionally, the Relative humidity, Humidification
temperature, and Humidification absolute pressure of the inlet humidifier are also
specified.

Humidified air is available for the O2 Inlet node, if H2O and N2 have been selected in
the O2 Gas Mixture section on the interface top node. Relative humidity, Humidification
temperature, and Humidification absolute pressure of the inlet humidifier are also
specified. The dry mole fractions of the included species are set internally as follows:
0.79 (N2), 4e-4 (CO2) and 1e-6(CO).

Given that the humidifier (inlet) temperature (Thum), absolute pressure (pA,hum) and
relative humidity (RHhum) of the humidifier (inlet) gases are known, the conversion

92 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

from a set of dry (containing no water vapor) molar fraction to a corresponding
humidified mixture may be calculated as follows. The water mole fraction is defined as

p vap  T hum 
x H2O = ------------------------------ RH hum
p A hum

The wet mole fractions for all species except water are then calculated from the dry
mixture molar fractions according to

x i = x i dry  1 – x H2O 

The Flow boundary condition section is available if the Use Darcy’s law for momentum
transport checkbox has been selected on the interface top node. The flow inlet
condition can be specified using the From initial values, Pressure, or Total mass flow rate
options.

Outlet
The H2 Outlet and O2 Outlet nodes are the preferred boundary conditions at outlets
where the species are to be transported out of the model domain. It is useful at outlets
where it is assumed that convection is the dominating effect driving the mass flow.

This node is available for exterior boundaries to Gas Phase nodes.

If the Include gas phase diffusion checkbox has been selected on the interface top node,
the boundary condition is applied to all selected gas species and corresponds to:

– n   i  Dik dk = 0
k

If the Use Darcy’s law for momentum transport checkbox has been selected on the
interface top node, the outlet Pressure can be specified.

Flux
This subnode is available to the Gas Phase nodes when the Advanced Physics Options is
selected in the Show More Options dialog shown when the Show More Options button
( ) is selected, in combination with the Include gas phase diffusion or Use Darcy’s law
for momentum transport checkbox being selected in the interface top node for the
corresponding gas mixture.

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 93

 The node may be used for adding additional boundary mass fluxes in a model, typical
stemming from nonfaradaic reactions such as water condensation-evaporation on a
channel wall or diffusion of gaseous species into adjacent media. The node contains
text fields for specifying the inward mass fluxes of all active gas phase species.

This node is available for exterior boundaries to Gas Phase nodes.

Water Absorption-Desorption
The Water Absorption-Desorption, H2 side and Water Absorption-Desorption, O2 side
nodes are available if Electroosmotic water drag has been selected in the Membrane
Transport section on the interface top node. These nodes are added by default and the
selection is locked to all intersection boundaries of the Membrane and the respective
Gas Phase nodes.

The node sets up the boundary conditions that define the absorption-desorption water
flux between the membrane and the neighboring gas phase domain. The node also
contributes to the boundary heat source variable, based on the Heat of water absorption
parameter on the parent node.

The Absorption-Desorption Condition can be specified as Vapor absorption-desorption (by

specifying an Absorption-desorption rate constant), Liquid water pressure, water
Chemical potential, or User-defined rate. The default From parent option is identical to
the Vapor absorption-desorption option for Vapor equilibrated model and Liquid water
pressure option Liquid equilibrated model, respectively, chosen on the parent node.

The Water absorbs-desorbs as setting controls if the node should contribute to the rate
variables for gaseous or liquid water. This setting is available for Chemical potential or
User-defined rate options. If Gas is selected, and if the Include gas phase diffusion
checkbox or the Use Darcy’s law for momentum transport checkbox has been selected
in the corresponding Gas Mixture section on the interface top node, the corresponding
absorption-desorption mass fluxes are added to these equations. If Liquid is selected,
the corresponding absorption-desorption mass flux is added to the flux variable for
H2O(l), if Include H2O(l) in reaction stoichiometry is active in the corresponding Gas
Mixture section on the interface top node. Note that Vapor absorption-desorption will
absorb-desorb water as for the Gas setting and Liquid water pressure as for the Liquid
setting.

94 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

Water Flux
This subnode is available to the Membrane node when the Advanced Physics Options is
selected in the Show More Options dialog shown when the Show More Options button
( ) is selected, in combination with the Electroosmotic water drag being selected in
the Membrane Transport section on the interface top node.

This node is available for exterior boundaries to Membrane node.

Water Chemical Potential

This subnode is available to the Membrane node when the Advanced Physics Options is
selected in the Show More Options dialog shown when the Show More Options button
( ) is selected, in combination with the Electroosmotic water drag being selected in
the Membrane Transport section on the interface top node.

This node is available for exterior boundaries to Membrane node.

Thin Gas Diffusion Electrode

Use the Thin H2 Gas Diffusion Electrode or the Thin O2 Gas Diffusion Electrode nodes to
define a thin gas diffusion electrode located at a boundary between an Electrolyte
Phase, an Electronic Conducting Phase and a Gas Phase node.

The electrode is assumed to be so thin that any current and species distribution effects
along the electrode thickness can be neglected.

The volumetric current density and corresponding mass sources/fluxes of the Thin
Gas Diffusion Electrode Reaction subnodes are multiplied by the Electrode Thickness
parameter, and applied as a boundary conditions.

FILM RESISTANCE
Use a film resistance if you want to include an additional potential drop due to an
ohmic resistance at the interface between the electrode and the electrolyte (a contact
resistance), for instance due to build-up of insulating deposits.

Specify either a Surface resistance Rfilm (SI unit: ·m2) directly or choose the Thickness
and conductivity option to calculate the surface resistivity based on a depositing film
thickness.

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 95

 Electrode Surface
Use the H2 Electrode Surface or the O2 Electrode Surface nodes to define an electrode
surface located at external boundaries of a H2/O2 Gas-Electrolyte Compartment
domain that are also external to the domain selection of the interface. Set the electric
potential of the electrode or specify a current condition that the potential of the
electrode shall fulfill.

FILM RESISTANCE
Use a film resistance if you want to include an additional potential drop due to an
ohmic resistance at the interface between the electrode and the electrolyte (a contact
resistance), for instance due to build-up of insulating deposits.

Specify either a Surface resistance Rfilm (SI unit: ·m2) directly or choose the Thickness
and conductivity option to calculate the surface resistivity based on a depositing film
thickness.

Specify the electrode reaction using Electrode Reaction child nodes.

Internal Electrode Surface

Use the Internal H2 Electrode Surface or the Internal O2 Electrode Surface nodes to define
an internal electrode surface located at boundaries adjacent to both an external
boundary of a H2/O2 Gas-Electrolyte Compartment domain and an external
boundary of an Electronic Conducting Phase domain.

FILM RESISTANCE
Use a film resistance if you want to include an additional potential drop due to an
ohmic resistance at the interface between the electrode and the electrolyte (a contact
resistance), for instance due to build-up of insulating deposits.

Specify either a Surface resistance Rfilm (SI unit: ·m2) directly or choose the Thickness
and conductivity option to calculate the surface resistivity based on a depositing film
thickness.

Specify the electrode reaction using Electrode Reaction child nodes.

Thin Gas Diffusion Electrode Reaction

This node can be added as a subnode to a Thin Gas Diffusion Electrode node.

Regarding the settings of the node, see Gas Diffusion Electrode Reaction.

96 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

Electrode Reaction
This node can be added as a subnode to Electrode Surface and Internal Electrode
Surface nodes.

Regarding the settings of the node, see Gas Diffusion Electrode Reaction.

Periodic Condition
Use the Periodic Condition to define a periodic relation between two boundaries — for
instance, in a model describing a repetitive unit cell.

Use the Apply for electrolyte phase and Apply for electronic conducting phase checkboxes
to enable the periodic condition for the electrolyte and electrode phase potentials,
respectively. Additionally, the periodic condition for the chemical potential is enabled
when Electroosmotic water drag checkbox is selected under Membrane Transport
section.

By specifying a Potential difference, an offset in potential (typically per unit cell) can be
prescribed.

For more information, see Periodic Boundary Conditions in the COMSOL

Multiphysics Reference Manual.

THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 97

 Theory for the Hydrogen Fuel Cell
and Water Electrolyzer Interfaces
In this chapter:

• Electronic Conducting Phase Theory

• Electrolyte Phase Theory
• Gas Phase Theory
• Calculation of Built-in Equilibrium and Thermoneutral Potentials
• Membrane Crossover Theory
• Electroosmotic Water Drag Theory

For general theory on electrode reactions, kinetics, as well as mass and heat sources,
see also

• Electrode Kinetics Expressions

• Equilibrium Potentials and the Nernst Equation
• Mass Fluxes and Sources Due to Electrochemical Reactions
• Joule Heating Due to Charge Transport
• Heating Due to Electrochemical Reactions

in Theory for the Current Distribution Interfaces.

For general theory on thermodynamic models, see also

• Thermodynamic Models and Theory

Electronic Conducting Phase Theory

The Electronic Conducting Phase node solves for the electric potential s(V) as
dependent variable.

The current density in the phase, is (A/m2), is defined as (Ref. 1)

i s = –  s eff  s (3-1)

where s, eff is the effective electric conductivity (S/m).

The charge balance equation in the phase is defined as

98 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

   i s = – i v tot (3-2)

where iv,tot (A/m3) is the sum of a volumetric current density contributions of the
electrode reactions occurring in the Gas Diffusion Electrode domains.

For Current Collector and Gas Diffusion Layer domains,

  is = 0

Electrolyte Phase Theory

The Electrolyte Phase node solves for the electric potential l (V) as dependent
variable.

The current density in the phase, il (A/m2), is defined as (Ref. 1)

i l = –  l eff  l (3-3)

where l, eff is the effective electrolyte conductivity (S/m).

The charge balance equation in the phase is defined as

  i l = i v tot (3-4)

where iv,tot (A/m3) is the sum of a volumetric current density contributions of the
electrode reactions occurring in the Gas Diffusion Electrode domains.

For Membrane, Separator and Gas-Electrolyte Compartment domains

  il = 0

Gas Phase Theory

GAS PHASE DIFFUSION

When gas phase diffusion is enabled, the gas phase nodes solve for a set of mass fraction
variables, i, where i is the index of the species. The governing equations are defined
using the Maxwell–Stefan set of equations, as described in Theory for the Transport of
Concentrated Species Interface.

DARCY’S LAW
When Darcy’s law is enabled, the gas phase nodes solve for a pressure variable p (Pa).

THEORY FOR THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 99
 The Darcy’s law equations are documented in Theory for the Darcy’s Law Interface.

BUILT-IN BINARY DIFFUSION COEFFICIENTS

The calculation of the built-in binary diffusion coefficients is done using the Fuller–
Schettler–Giddings (FGS) model (Ref. 2):

–2 1.75 1- + ------ 1-
1.01325  10 T ------
0 Mi Mj
D i j = -------------------------------------------------------------------------------- (3-5)
1 1 2
 --- ---
 3   3
P   v +  v 
 
  
 i  
j

where T denotes the temperature (K), Mi the molecular weight of species i (g/mol),
P is the pressure (Pa), and vi are the atomic diffusion volumes (Fuller diffusion
volume) (cm3).
TABLE 3-1: DIFFUSION VOLUMES USED FOR CALCULATING BINARY DIFFUSION COEFFICIENTS.

SPECIES FULLER DIFFUSION

VOLUME

H2 6.12
O2 16.3
H2O 13.1
N2 18.5
CO 18
CO2 26.9
CH4 25.15

BUILT-IN DYNAMIC VISCOSITY

The built-in mixture viscosity of the gas phase is based on kinetic theory by Brokaw
(Ref. 3):

 
 x 
 ---------------------
i i v 
v =  
-

i
 
 x j  i j

j

where

100 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

  i v
 i j = A i j ---------
- (3-6)
 j v

and the interaction parameter is defined as

 
 i j  RM i j – RM i0.45 j


A i j = -------------------- 1 + ----------------------------------------------------------------------------------------- (3-7)
RM i j  1 + RM i0.45 
2  1 + RM i j  + ------------------------------------------------ j
   i j   1 + RM i j 

with

 4M i M j  1 / 4 M
 i j =  ----------------------------2 , RM i j = -------i (3-8)
  Mi + M   Mj
j

where i,v is the vapor viscosity of each individual species, and Mi are the molecular
weights.

BUILT-IN THERMAL CONDUCTIVITY

The gas phase mixture thermal conductivity, which may be used in heat transfer
simulations, is calculated by the Hydrogen Fuel Cell and Water Electrolyzer interfaces
according to (Ref. 4)

 
 x 
 --------------------
i i v 
v =  
-

(3-9)
i
 
 x j  i j

j

2
 T + 3   
 --- T b i   9
 T + S ij  --4- T b i T b j
 M 3 / 4  2  
= ---  1 + ----------  -------  ------------------------------   ------------------------------------------------------
1 i v j
 i j (3-10)
4  j v M i T + 3  
--- T   T +  --- T b i
3
   
  2 b j    2  

In the above formula, the summations are done for all active species, where i,v and
i,v are the individual vapor thermal conductivity and dynamic viscosity correlations as
a functions of temperature for the pure gases, respectively. The normal boiling points
Tb,i and the species molecular weights are also considered. The factor Sij equals 0.773
if either i or j represents the index of water (steam). For all other cases Sij =1.

THEORY FOR THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 101
 Since the above equation relies on the individual species viscosities, the thermal
conductivity is only calculated in gas phase domains using built-in (that is, not
user-defined) dynamic viscosities.

BUILT-IN HEAT CAPACITY

The Hydrogen Fuel Cell and Water Electrolyzer interfaces also calculate the heat
capacity of the gas mixture which may be used in heat transfer simulations. The heat
capacity is calculated as

 i c p i
C p mix =  --------------
Mi
-
i

where Cp,mix is the heat capacity of the fluid mixture in J/(kg·K), cp,i is the species
heat capacity in J/(mol·K).

Calculation of Built-in Equilibrium and Thermoneutral Potentials

The change of enthalpy for an electrode reaction is (Ref. 1)

r H =  i Hi
i

where Hi = Hi(T) are the enthalpies of formation of the reacting species of index i, and
i the stoichiometric coefficients.

Similarly, the change of entropy is

r S =  i Si
i

Where Si = Si(T) is the (absolute) molar entropy.

Hi and Si are generally known for liquids and gases, whereas for ions the values will
depend on the nature of the electrolyte in use, and are hence hard to measure
experimentally.

The corresponding change of Gibbs free energy of a reaction is

 r G =  r H – T r S

The equilibrium potential for an electrode reaction is

102 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

 –r G
E eq = -------------
-
nF

and the thermoneutral voltage, required to compute electrode reaction heat sources, is

–r H
E therm = -------------
-
nF

where n is the number of electrons included in the electrode reaction.

Temperature dependent thermodynamic expressions for gaseous species of Hi(T) and

Si(T), at standard partial pressures of 1 atm, are based on Ref. 6.

For the charge-conducting ion in the electrolyte, entropy, and enthalpy values are
however generally not known and need to be set by the user. The stoichiometric
coefficient of the charge-carrying ion is derived from the number of electrons
participating in the charge transfer reaction and the charge carrying ion charge as

n-
 ion = – --------
z ion

For convenience, the built-in equilibrium potentials are defined versus a reversible
reference hydrogen electrode (RHE) for the same electrolyte type, for a specified
reference temperature. This will have the effect that the HOR/HER will have an
equilibrium potential in a model close to 0 V, and the ORR/OER reaction will have
an equilibrium potential ranging from around 1.2 (at room temperature) to 1.0 V (at
800°C). This shift facilitates setting for instance initial values for the potential
dependent variables in a model.

The reference equilibrium and thermoneutral potentials used by the electrode reaction
node, versus the RHE, are defined as

–  r G ref
E eq ref = -------------------
- – E eq RHE
nF

and

–r H
E therm = -------------
- – E eq RHE
nF

where

 r G ref =  r H – T r S ref

THEORY FOR THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 103
 with Sref being the change in entropy at the chosen reference conditions (typically
1 atm).

Note that due to cancellation in the equation for Eeq, ref, the values of the ion entropy
(0 by default) has no impact on the equilibrium potentials values (nor the overall cell
voltage and energy balance), but it will impact how the thermoneutral voltage Etherm
and corresponding heat sources for the individual electrode reactions are computed.

LIQUID WATER THERMODYNAMIC EXPRESSIONS AND RELATIVE

HUMIDITY
The fc/we interface defines the water vapor pressure, pvap (Pa), and the enthalpy of
vaporization Hvap (J/mol), by the use of an experimental correlations (interpolation)
functions vs temperature. The pvap(T) correlation is based on Ref. 5, whereas Hvap(T)
is defined using the same source data as in the Thermodynamics node (See Using
Thermodynamic Properties).

The enthalpy of formation of liquid water, HH2O(l) (J/mol), is then defined as

H H2O(l) = H H2O(g) – H vap

Assuming that the vapor behaves as an ideal gas, and that the liquid water is
incompressible, the following equation for the entropy of liquid water can be derived

p vap H vap
S H2O(l) = – R ln ----------
- – ------------- + S H2O(g)
p ref T

The relative humidity, RH (1), is defined as

p vap
RH = ----------
-
p ref

Membrane Crossover Theory

In a membrane-based fuel cell or electrolyzer, the reacting gases may dissolve into the
membrane and diffuse through the membrane to the other side of the cell. This
cross-membrane transport of species is termed crossover.

For hydrogen and oxygen, these species will be quickly oxidized/reduced when
reaching the other side of the cell. These reactions give rise to small parasitic current
densities which will have an effect on, for instance, the open circuit voltage of the cell.
For nitrogen, typically diffusing from the oxygen to the hydrogen side in air-operated
fuel cells, the crossover may result in a dilution effect of the hydrogen fuel.

104 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

HYDROGEN CROSSOVER REACTIONS
On the hydrogen side of the membrane, hydrogen dissolves into the membrane
according to

H 2 (g)  H 2 (mem)

The dissolved hydrogen then diffuses through the membrane, and in the gas diffusion
electrode on oxygen side, the hydrogen gets oxidized according to (depending on
charge-carrying ion):

Proton exchange:

+ -
H 2 (mem)  2H + 2e

Hydroxide exchange:

- -
H 2 (mem) + 2OH  2H 2 O(g) + 2e

Due to the highly anodic overpotential potential for the hydrogen oxidation reaction
on the oxygen side, the reaction operates at a limiting current density, completely
governed by the molar flux of hydrogen though the membrane.

OXYGEN CROSSOVER REACTIONS

Similarly as for hydrogen, on the oxygen side of the membrane, oxygen dissolves into
the membrane according to

O 2 (g)  O 2 (mem)

The dissolved oxygen then diffuses through the membrane and in the gas diffusion
electrode on hydrogen side, oxygen gets reduced according to (depending on the
charge-carrying ion):

Proton exchange:

+ -
O 2 (mem) + 4H + 4e  2H 2 O(g)

Hydroxide exchange:

- -
O 2 (mem) + 2H 2 O(g) + 4e  4OH

THEORY FOR THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 105
 Due to the highly cathodic overpotential potential for the oxygen reduction reaction
on the hydrogen side, the reaction operates at a limiting current density, completely
governed by the molar flux of oxygen though the membrane.

NITROGEN CROSSOVER REACTIONS

On either side of the membrane, nitrogen may dissolve into and out of the membrane
according to

N 2 (g)  N 2 (mem)

MEMBRANE CROSSOVER TRANSPORT MODEL

The membrane crossover model defines either one or several thermodynamical
activities of hydrogen and oxygen in the membrane, aH2, aO2, and aN2, respectively,
as dependent variables. The governing equation for the membrane species is based on
Fick’s law of diffusion in combination with Henry’s law

c i mem
- +   J i mem = 0
-------------------
t

J i mem = –  i p ref a i

where Ji,mem (mol/(m2·s)) is the molar flux of the species and i (mol/(m·s Pa)) is
the corresponding permeation coefficient. pref is a reference pressure, arbitrarily set to
1 atm.

The species concentration in the membrane ci,mem (mol/m3) used in the time
derivative term is calculated based on Henry’s law according to

p ref
c i mem = a i --------- H i
RT

where Hi (1) is the Henry’s law coefficient of the species.

Note that, for isothermal conditions, the transport equation is equivalent to Fick’s law
of diffusion according to

c i mem
- +    – D i c i mem  = 0
-------------------
t

with the diffusion coefficient Di (m2/s) defined as

106 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

  i RT
D i = ---------------
Hi

Hydrogen boundary conditions:

On Membrane boundaries adjacent to hydrogen Gas Phase domains, the hydrogen

activity aH2 is set as follows

p H2  g 
a H2 = ----------------
p ref

where pH2(g) is the hydrogen partial pressure in the adjacent gas phase. Hydrogen is
assumed to be instantly oxidized when it reaches an electrode on the oxygen gas side.
On Membrane boundaries adjacent to oxygen Gas Diffusion Electrode domains (or
oxygen Thin Gas Diffusion Electrode boundaries), the hydrogen activity is set to

a H2 = 0

Oxygen boundary conditions:

Similarly, on Membrane boundaries adjacent to oxygen Gas Phase domains, the

oxygen activity aO2 is set as follows

p O2(g)
a O2 = --------------
-
p ref

where pO2(g) is the oxygen partial pressure in the adjacent gas phase. Oxygen is
assumed to be instantly reduced when it reaches an electrode on the hydrogen gas side.
On Membrane boundaries adjacent to hydrogen Gas Diffusion Electrode domains (or
hydrogen Thin Gas Diffusion Electrode boundaries), the oxygen activity is set to

a O2 = 0

Nitrogen boundary conditions:

On Membrane boundaries adjacent to any Gas Phase domain, the nitrogen activity
aN2 is set to

p N2(g)
a N2 = --------------
-
p ref

THEORY FOR THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 107
 CONTRIBUTIONS TO CURRENT DISTRIBUTION MODEL
For the adjacent oxygen gas diffusion electrode boundaries where aH2,mem is set to 0,
the corresponding oxidation charge transfer reaction will give rise to a local crossover
current density iloc according to

nF  n  J H2 mem 
i loc = -------------------------------------------
-
 H2

Similarly, the local crossover current density iloc on adjacent hydrogen gas diffusion
electrode boundaries becomes

nF  n  J O2 mem 
i loc = -------------------------------------------
-
 O2

The local crossover current density is added as normal current density contributions
for the electronic and electrolyte potentials, respectively:

n  i l = i loc

n  i s = – i loc

Parasitic reactions not involving the H2/O2 species are not added by default, but may
be added as ordinary electrode reactions in gas diffusion electrodes.

HEAT OF REACTIONS
The parasitic reactions of the H2/O2 species also give rise to boundary heat sources.
The thermoneutral potential Etherm is defined as

–r H
E therm = -------------
- – E eq,RHE
nF

where n is the number of electrons included in the electrode reaction, and the
corresponding heat source Qb is defined as

Q b = i loc   s –  l – E therm 

CONTRIBUTIONS TO GAS DIFFUSION AND MOMENTUM TRANSFER MODEL

The species crossover flux will have an impact on the gas phase flux. For hydrogen, on
the Membrane boundary adjacent to the hydrogen Gas Phase domain, the following
flux contribution is added. Similarly, for oxygen, on the Membrane boundary adjacent
to the oxygen Gas Phase domain, the following flux contribution is added For

108 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

nitrogen, on Membrane boundaries adjacent to any Gas Phase domain, the following
flux contribution is added

n   j i + u s  i n  = n   M i J i mem 

The above contributions are also added to the momentum transfer equation and Stefan
velocity as

n  u =  n   Mi Ji mem 
i

u s = – n    ji + us i n 
i

For any gas species participating in the parasitic hydrogen oxidation or oxygen
reduction reaction, the corresponding mass flux contribution, on Membrane
boundaries adjacent to H2/O2 Gas Diffusion Electrode domains or H2/O2 Thin Gas
Diffusion Electrode boundaries, is

M i  i i loc
n   j i + u s  i n  = – --------------------
-
nF

The above contribution is also added to the momentum transfer equation and Stefan
velocity as

M i  i i loc
n  u =  – --------------------
nF
-
i

u s = – n    ji + us i n 
i

Electroosmotic Water Drag Theory

In polymer electrolytes, ion transport and water transport are strongly intercoupled.
As a result of this intercoupling, the water content of a polymer electrolyte impacts
electrolyte conductivity. Accurate models of the electrolyte water content are hence
paramount for polymer electrolyte-based fuel cell or electrolyzer modeling.

THEORY FOR THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 109
 MEMBRANE WATER TRANSPORT MODEL
A complication when modeling the water-ion-polymer system is that the water-ion
friction forces need to be taken into account in the transport equations both for the
ions as well as the water molecules. The Weber-Newman model (Ref. 7) is based on
concentrated solution theory, taking into account all three binary interactions between
the charge-carrying ion, the immobilized polymer matrix and the water molecules.
The model introduces the chemical potential of water in the polymer 0 (J/mol) as
dependent variable, and defines the polymer water flux, N0 (mol/(m2·s)), as

2
l   l  
N 0 = – --------  l +   + ----------
-  0
F  F 
2

where  (dimensionless) is the electroosmotic coefficient and  (mol2/(J·m·s)) the

water transport coefficient, with the mass balance

  N0 = 0

in the domain. The electrolyte phase charge balance equations gets a contribution
from the water flux by adding a water flux term to the current vector as follows

l 
i l = –  l  l –  --------  0
F

CHEMICAL POTENTIAL
The chemical potential driving force may generally be defined as

 0 = RT ln  a 0  + V H2O p H2O(l)

where a0 is the thermodynamic activity of water, VH2O is the partial molar volume of
water, and pH2O(l) is the liquid water pressure. VH2O is defined as

V H2O = M H2O   H2O

where MH2O is the molar mass and H2O is the density of water at the operating
temperature.

By integration, the chemical potential 0 is then defined as

 0 =  0 ref + RT ln  a 0  + V H2O p H2O(l)

where 0,ref is a chosen reference potential, for convenience set to 0.

110 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

WATER ACTIVITY CORRELATIONS AND POLYMER ELECTROLYTE
PROPERTIES
Many polymer electrolyte systems exhibit “Schröder’s paradox,” which refers to the
phenomenon of strong hysteresis (path dependence) in water uptake when operating
close to the dew point. Typically, an electrolyte membrane immersed in liquid water
(liquid equilibrated) features significantly higher water uptake and conductivity
compared to a previously dried membrane that has been hydrated in contact with water
vapor at 100% relative humidity (vapor equilibrated).

For a vapor-equilibrated membrane, the parameters l, , and typically all depend on
the polymer water content. When measuring water uptake and other polymer
electrolyte properties (such as l, , and ), it is common to report these versus the
water activity, aw, which is not the same as the thermodynamic activity a0.

In the gas phase, aw is defined as follows

p H2O(g) x H2O(g) p g
a w = -------------------- = ------------------------
-
p vap  T  p vap  T 

where is the pH2O(g) is the partial water pressure, pvap is the vapor pressure, and pg is
the gas pressure. At an interface between a polymer electrolyte and the gas phase,
assuming the polymer phase to be in equilibrium with the gas phase (and ideal gas
conditions), the following condition holds

pg p vap  T 
a 0 = x H2O(g) -----0- = --------------------
0
aw
pg pg
0
where p g is the standard condition gas pressure (typically 1 atm or 1 bar). The
corresponding water activity inside the polymer electrolyte for a vapor-equilibrated
polymer, for which the liquid pressure pH2O(l) is 0, may hence be defined as

0 0
pg pg 0
a w = -------------------- a 0 = -------------------- exp  --------
p vap  T  p vap  T   RT

For a liquid-equilibrated system aw is equal to 1.

Gas diffusion electrodes typically have a large electrolyte-gas phase interface area, in
combination with fairly short diffusion lengths in the transport direction perpendicular
to the pore walls. The water activity in the polymer of the gas diffusion electrode may
be approximated to always be in equilibrium with the adjacent gas phase in the pores.

THEORY FOR THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 111
 Hence the expression for aw in the gas phase may be used also for the polymer in gas
diffusion electrodes, and 0 need not to be solved for explicitly.

ABSORPTION-DESORPTION BOUNDARY CONDITIONS

As a result of the definition of the chemical potential, both the liquid pressure and the
thermodynamic activity of water need to be considered when defining boundary
conditions for water absorption-desorption at a Membrane boundary.

For a membrane in contact with vapor, the liquid pressure pH2O(l) is 0 Pa, and the
chemical potential 0 relates to the thermodynamic activity a0 as

 0 = RT ln  a 0 

so that

0
a 0 = exp  --------
 RT

Experimentally, the interface between a membrane and the vapor phase is often seen
to exhibit a mass transfer resistance, where the absorption rate rabs,dsp follows a
relationship based on the difference in the thermodynamic water activity between the
two phases, as follows

 p H2O(g) 
r abs,dsp = k abs,dsp  ------------------
- – a 0
 p 0 
g

where kabs,dsp (mol/(m2 s)) is a rate constant and the activity of water in the gas phase
has been assumed to equal the partial pressure pH2O(g) divided by the standard
0
condition pressure p g (ideal gas conditions).

The rate may then be added as a flux boundary condition according to

– n  N 0 = r abs,dsp

For a membrane in contact with liquid water, aw is equal to 1. Interfaces of this kind
have been empirically found to exhibit way lower mass transfer resistances than for the
vapor case. For this case, the boundary condition may be formulated based on the
liquid pressure pH2O(l) as

 p vap  T 
 0 = RT ln  -------------------
- + V H2O p H2O(l)
 p 0g 

112 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

For this cases the absorption rate rabs,dsp can be evaluated at the boundary as

r abs,dsp = – n  N 0

The absorbed-desorbed water adds to the boundary mass flux of the neighboring gas
phase according to

n   j H2O + u s  H2O n  = – M H2O r abs,dsp

The water may be transported into the gas phase domain in either the liquid water
phase or as vapor. The above contribution is also added to the momentum transfer
equation and Stefan velocity as

n  u = – M H2O r abs,dsp

u s = – n   j H2O + u s  H2O n 

HEAT OF WATER ABSORPTION

The boundary heat source Qb due to the absorption-desorption is defined as

Q b = r abs,dsp  r H abs(g)

where rHabs(g) is the heat of absorption for vapor into the polymer electrolyte. If the
water transport occurs in the liquid phase, the heat source becomes

Q b = r abs,dsp   r H abs(g) +  r H vap  T  

References
1. J. Newman and K. Thomas-Alyea, Electrochemical Systems, John Wiley & Sons
2004.

2. E.N. Fuller, P.D. Schettler, and J.C. Giddings, “A new method for prediction of
binary gas-phase diffusion coefficients,” Ind. Eng. Chem. Res., vol. l 5, pp. 19–27,
1966.

3. R.S. Brokaw, “Approximate formulas for the viscosity and thermal conductivity of
gas mixtures. ii,” [Link]. Phys. vol. 42, no. 4, pp. 1140–1147, 1965.

4. A.L. Lindsay and L.A. Bromley, “Thermal conductivity of gas mixtures,” Ind. Eng.
Chem. Res., vol. 42, no. 8, pp. 1508–1511, 1950.

THEORY FOR THE HYDROGEN FUEL CELL AND WATER ELECTROLYZER INTERFACES | 113
 5. HJ Kretzschmar and W. Wagner, International Steam Tables, Properties of Water
and Steam based on the Industrial Formulation IAPWS-IF97, 3rd Edition, 2019,
Springer-Verlag GmbH Germany.

6. MW Chase, “NIST-JANAF Thermochemical Tables, Fourth Edition,” J. Phys.

Chem. Ref. Data, Monograph 9, 1998, 1-1951.

7. A. Weber and J. Newman, “Transport in Polymer-Electrolyte Membranes II.

Mathematical Model,” J. Electrochem. Soc., vol. 151, no. 2, pp. A311–325, 2004.

114 | CHAPTER 3: FUEL CELL AND ELECTROLYZER INTERFACES

 4

Electrochemistry Interfaces

This chapter describes the physics interfaces found under the Electrochemistry
branch ( ).

In this chapter:

• The Primary and Secondary Current Distribution Interfaces

• The Tertiary Current Distribution, Nernst–Planck Interface
• Shared Physics Features in the Current Distribution Interfaces
• The Concentrated Electrolyte Transport Interface
• The Electrode, Shell Interface
• The Electroanalysis Model Wizard Entry
• Theory for the Current Distribution Interfaces
• Theory for the Concentrated Electrolyte Transport Interface
• Theory for Electrochemical Heat Sources
• Theory for the Electrode, Shell Interface
• Theory for Electroanalysis
• Electrode Potentials and Reference Electrodes

115
 The Primary and Secondary Current
Distribution Interfaces
In this section:

• The Primary Current Distribution • Electrode Line Current Source

and Secondary Current Distribution • Electrolyte Line Current Source
Interfaces
• Electrode Symmetry Axis Current
• Electrolyte Source
• Initial Values • Electrolyte Symmetry Axis Current
• Porous Electrode Source
• Periodic Condition • Electrode Point Current Source
• Thin Electrolyte Layer • Electrolyte Point Current Source
• Edge Electrode

Only the physics interface-specific nodes are described here. All other
nodes in the Primary Current Distribution and Secondary Current
Distribution interfaces are described in Shared Physics Features in the
Current Distribution Interfaces

The Primary Current Distribution and Secondary Current

Distribution Interfaces
The Primary Current Distribution (cd) interface ( ) and the Secondary Current
Distribution (cd) interface ( ) are found under the Electrochemistry branch ( )
when adding a physics interface.

Primary Current Distribution Interface

The Primary Current Distribution interface defines the transport of charged ions in an
electrolyte of uniform composition as well as current conduction in electrodes using
Ohm’s law in combination with a charge balance. The physics interface neglects
activation overpotentials due to charge transfer reactions.

116 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Use this physics interface to estimate the ohmic losses in simplified models of
electrochemical cells, where the overpotentials of the electrode reactions are small
compared to the ohmic voltage drops in the electrolyte and electrodes.

Secondary Current Distribution Interface

The Secondary Current Distribution interface is similar to the Primary Current
Distribution interface, except that it also accounts for activation overpotentials. The
relation between charge transfer and overpotential can be described using arbitrary
kinetic expressions, such as Butler–Volmer and Tafel equations.

Use this physics interface for generic modeling of electrochemical cells. It can be
combined with interfaces modeling mass transport to describe concentration
dependent (tertiary) current distributions.

Use the Current Distribution Type setting on the physics interface node, described
below, to switch between a Primary Current Distribution and a Secondary Current
Distribution interface.

• Introduction to Electrochemistry Modeling

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is cd.

DOMAIN SELECTION

Domains that do not conduct current should be omitted from the

Domain Selection: for example, the gas channels in a fuel cell.

THE PRIMARY AND SECONDARY CURRENT DISTRIBUTION INTERFACES | 117

 OUT-OF-PLANE THICKNESS

For 2D components, enter a value or expression for the out-of-plane

Thickness d (SI unit: m). The value of d determines the size of the domain
perpendicular to the modeled 2D cross section. This value yields, for
example, the correct total current when the current density is obtained
from a 2D simulation.

CROSS-SECTIONAL AREA

For 1D components, enter a Cross-sectional area Ac (SI unit: m2) to define

a parameter for the area of the geometry perpendicular to the 1D
component. The value of this parameter is used, among other things, to
automatically calculate the total current from the current density vector.
The analogy is valid for other fluxes. The default is 1 m2.

CURRENT DISTRIBUTION TYPE

The Current Distribution Type selected in the list is based on the choice made when
adding a physics interface — Primary or Secondary. The Primary Current Distribution
interface changes to a Secondary Current Distribution interface if the choice is
changed to Secondary, and vice versa.

The selection between Primary or Secondary governs how electrode reactions are
modeled on interfaces between electrodes and electrolytes.

• Models using a Primary current distribution type use potential constraints (Dirichlet
boundary conditions), according to the equilibrium potential setting.
• Secondary current distribution models use current flux conditions (Neumann
boundary conditions) according to the sum of all electrode reaction current
densities.

PHYSICS VS. MATERIALS REFERENCE ELECTRODE POTENTIAL

The Physics vs. Materials Reference Electrode Potential setting on the physics interface
node can be used to combine material library data for current densities and equilibrium
potentials with an arbitrary reference electrode scale in the physics. The setting affects
the electrode potentials used for model input into the materials node, as well as all
equilibrium potential values output from the materials node.

118 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Note that the setting will only impact how potentials are interpreted in communication
between the physics and the Materials node. If the From material option is not in use
for equilibrium potentials or electrode kinetics, the setting has no impact.

DEPENDENT VARIABLES
This physics interface defines dependent variables (fields) for the Electrolyte potential
and Electric potential. The names can be changed but the names of fields and
dependent variables must be unique within a model.

DISCRETIZATION
The interface uses Linear elements by default and this setting is recommended for most
problems. Certain types of problems, such as models using porous electrodes, or
current distribution problems in two dimensions or higher, may benefit in terms of
solution accuracy from using Quadratic elements.

In multiphysics models it is generally recommended to use the same element order in

all coupled interfaces.

To see all settings in this section, click the Show More Options button ( ) and select
Advanced Physics Options in the Show More Options dialog.

• Electrochemical Reactions and the Difference Between a Primary and

a Secondary Current Distribution
• Domain Equations for Primary and Secondary Current Distributions
• Shared Physics Features in the Current Distribution Interfaces
• Theory for the Current Distribution Interfaces

Electrolyte
Use the Electrolyte node to define an electrolyte domain that only conducts current in
the ion conducting phase.

Note that electrolyte in this case does not refer to the pore electrolyte in porous
electrodes (which should be defined by a Porous Electrode node instead).

THE PRIMARY AND SECONDARY CURRENT DISTRIBUTION INTERFACES | 119

 The Electrolyte conductivity, l (SI unit: S/m), parameter will define how the current
in the domain depends on the gradient of the potential.

• Domain Equations for Primary and Secondary Current Distributions

• Electrolyte Theory

Initial Values
Use this node to specify the Initial Values of the electrolyte potential and the electric
potential for the solver.

For many electrochemical problems that use nonlinear electrode kinetics, such as
Butler–Volmer kinetics, providing reasonable initial values can significantly improve
solver convergence.

A good value for the Electric potential (SI unit: V) in electrode and porous electrode
domains can usually be derived from the boundary conditions. For instance, if a
boundary has been grounded or set to a cell potential, use that value as the initial value
also in the adjacent domain. For the Electrolyte potential (SI unit: V) a good initial
value is often the negative of the equilibrium potential of the grounded electrode.

Note: For Primary current distributions, or for models solved using a Current
Distribution Initialization study step set to Primary, the settings of this section usually
have limited impact on the solver convergence.

Porous Electrode
The Porous Electrode node sets up charge balances for the electrode and the pore
electrolyte in a porous electrode. Note that the node should be used for porous
domains that conduct current in both an electrolyte and an electrode phase. For the
case of domains that do not contain a pore electrolyte — for instance, a current
collector/feeder or a gas diffusion layer (GDL) in a fuel cell — use an Electrode node
instead.

Use Porous Electrode Reaction subnodes to define the charge transfer reactions that
occur on the interface between the electrolyte and electrode phases within the porous
electrode. For the Secondary Current Distribution interface, the Porous Matrix
Double-Layer Capacitance subnode is also available.

120 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

The conductivities are taken From material by default. From the respective material list
you may any material in the model, if present. By default they are set to the Domain
material (which is the material applied to the active domain in the Materials node).

You may use the Effective conductivity correction factors to account for the lowered
effective conductivities of the electrode and electrolyte phases due to the lower volume
fractions of each phase, and the tortuosity of the porous matrix.

The Electrode volume fraction is used to calculate the effective electric conductivity of
the porous matrix when the correction factor is set to Bruggeman or Tortuosity.
Additionally, it is used in calculating the active specific surface area of the porous matrix
when Particle-based area option is selected in the child nodes.

DISSOLVING-DEPOSITING SPECIES
Use the settings of this section to define species that participate in
dissolution-deposition electrode reactions within the porous electrode, for instance
metal deposition/dissolution or oxide formation.

Use the Add ( ) and Delete ( ) buttons as needed in the tables to control the
number of species.

Dependent variables for the volumetric molar concentration are added for each
dissolving-depositing species. These variables can be used to keep track of the amount
of reacted material in the porous electrode. The total molar dissolution/deposition
rate depends on the reaction rates and stoichiometry, defined in the Porous Electrode
Reaction subnodes.

The Density and Molar mass determine the electrode growth velocity and the resulting
dissolved/deposited layer thickness. By multiplying by the electrode surface area (in
the case of multiple electrode reaction the average surface area is used), the change in
electrode and electrolyte volume fractions can be also be calculated. By use of the Add
volume change to electrode volume fraction (not available for Separator node of Tertiary
Current Distribution interface) and Subtract volume change from electrolyte volume
fraction checkboxes you may define how these volume changes should be included in
the model.

Thickness variables, based on the surface area, are also defined that you for instance
can use to couple to the Film Resistance (see below).

ADSORBING-DESORBING SPECIES
See the Electrode Surface node.

THE PRIMARY AND SECONDARY CURRENT DISTRIBUTION INTERFACES | 121

 FILM RESISTANCE
See the Electrode Surface node.

• Domain Equations for Primary and Secondary Current Distributions

• Porous Electrode Theory

Periodic Condition
Use the Periodic Condition to define a periodic relation between two boundaries — for
instance, in a model describing a repetitive unit cell.

Use the Apply for electrolyte phase and Apply for electrode phase checkboxes to enable
the periodic condition for the electrolyte and electrode phase potentials, respectively.

By specifying a Potential difference, an offset in potential (typically per unit cell) can be
prescribed.

For more information, see Periodic Boundary Conditions in the COMSOL

Multiphysics Reference Manual.

Thin Electrolyte Layer

The Thin Electrolyte Layer node can be used to model a thin insulating or resistive
sheet, located on an interior boundary in an electrolyte domain. The node can be used
as an alternative to drawing the actual layer domain in the model geometry, which may
significantly reduce meshing and solver time, especially in 3D models.

Thin insulating sheets are commonly inserted in the electrolyte in various types of
electrochemical cells. For example they may be used for optimizing the current
distribution in a corrosion protection application, of for optimizing the local
deposition rate in a deposition bath.

The layer may be set to be either Insulating or Resistive.

For the case of Resistive, the Surface Resistance can either be set directly, or calculated
from Thickness and Conductivity values.

ADVANCED SETTINGS
Enable Slit electric potential in porous electrodes to make the layer fully insulating for
the electrode phase potential along the selected boundaries.

122 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Edge Electrode
The Edge Electrode can be used in 3D problems to define electrodes such as long pipes
and thin wires where the electric potential variation within the electrode in the normal
direction to the electrode surface is negligible. This assumption allows for the thin
electrode domain to be replaced by a lumped one-dimensional partial differential
equation formulation on the edge, describing an electrode surface along the edge with
a given Edge electrode radius. In this way the problem size can be reduced, and
potential problems with mesh anisotropy in the thin layer can be avoided.

The electric current conduction in the tangential direction of an edge can be described
by Ohm’s law or a Fixed electric potential or a Floating potential assuming infinite
conductivity of the edge or an External short electric potential which allows to connect
two electrodes over an external connector with a given bulk resistance or a Connection
point electric potential which allows to define a location for connections to other
locations in the model.

An Edge Electrode can only be applied to edges within, or adjacent to, Electrolyte
domains.

A default Electrode Reaction subnode is added by default to the feature. Double Layer
Capacitance, External Current Source, Electric Ground, Electric Potential, and Electrode
Current subnodes can also be added to the feature.

FILM RESISTANCE
See the Electrode Surface node. The section is only available when a Secondary current
distribution has been selected on the parent node.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

The mesh size in the perpendicular direction to the edge should

preferably be in the same order of magnitude as the Edge electrode
diameter in order to achieve good numerical accuracy.

THE PRIMARY AND SECONDARY CURRENT DISTRIBUTION INTERFACES | 123

 Electrode Line Current Source

For 3D components, use the Electrode Line Current Source on edges.

For 2D components use it on points.

For 2D axisymmetric components, it can only be selected on points

outside the symmetry axis.

ELECTRODE CURRENT SOURCE

Enter the Electrode current source ql,s (SI unit: A/m). The default is 0 A/m.

Electrolyte Line Current Source

For 3D components, use the Electrolyte Line Current Source on edges.

For 2D components use it on points.

For 2D axisymmetric components, it can only be selected on points

outside the symmetry axis.

ELECTROLYTE CURRENT SOURCE

Enter the Electrolyte current source ql,l (SI unit: A/m). The default is 0 A/m.

Electrode Symmetry Axis Current Source

For 2D axisymmetric components apply an Electrode Symmetry Axis

Current Source along the symmetry axis boundary.

ELECTRODE CURRENT SOURCE

Enter the Electrode current source ql,s (SI unit: A/m). The default is 0 A/m.

124 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Electrolyte Symmetry Axis Current Source

For 2D axisymmetric components apply an Electrolyte Symmetry Axis

Current Source along the symmetry axis boundary.

ELECTROLYTE CURRENT SOURCE

Enter the Electrolyte current source ql,l (SI unit: A/m). The default is 0 A/m.

External Current Source

EXTERNAL CURRENT SOURCE

Enter an External current source ql,s (SI unit: A/m) at the Edge Electrode. The default
is 0 A/m.

Electrode Point Current Source

For 3D components, use the Electrode Points Current Source on edges.

For 2D axisymmetric components, it can only be selected on points

located on the symmetry axis.

ELECTRODE CURRENT SOURCE

Enter the Electrode current source qp,s (SI unit: A). The default is 0 A.

Electrolyte Point Current Source

For 3D components, use the Electrolyte Points Current Source on edges.

For 2D axisymmetric components, it can only be selected on points

located on the symmetry axis.

ELECTROLYTE CURRENT SOURCE

Enter the Electrolyte current source qp,l (SI unit: A). The default is 0 A.

THE PRIMARY AND SECONDARY CURRENT DISTRIBUTION INTERFACES | 125

 Electrode Current
Use this node to define a current source in a point of an Edge Electrode node.

This node is available as a subnode to the Edge Electrode node, when Ohm’s Law has
been selected as the electric potential model.

126 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

The Tertiary Current Distribution,
Nernst–Planck Interface
In this section:

• The Tertiary Current Distribution, • Separator

Nernst–Planck Interface • Reactions
• Electrolyte • Thin Electrolyte Layer
• Porous Electrode • Initial Values

Various nodes are also available and described for the Transport of Diluted Species
interface. See Domain, Boundary, and Pair Nodes for the Transport of Diluted Species
Interface

All other nodes in the Tertiary Current Distribution, Nernst–Planck

interface are described in Shared Physics Features in the Current
Distribution Interfaces

The Tertiary Current Distribution, Nernst–Planck Interface

The Tertiary Current Distribution, Nernst-Planck (tcd) interface ( ), found under the
Electrochemistry branch ( ) when adding a physics interface, describes the current
and potential distribution in an electrochemical cell taking into account the individual
transport of charged species (ions) and uncharged species in the electrolyte due to
diffusion, migration, and convection using the Nernst–Planck equations. The physics
interface supports different descriptions of the coupled charge and mass transport in
the electrolyte (see Electrolyte Charge Conservation below). The electrode kinetics for
the charge transfer reactions can be described by using arbitrary expressions or by using
the predefined Butler–Volmer and Tafel expressions.

Ohm’s law is used in combination with a charge balance to describe the flow of
currents in the electrodes. The charge transfer reactions can be defined as boundary

THE TERTIARY CURRENT DISTRIBUTION, NERNST–PLANCK INTERFACE | 127

 conditions or as sources or sinks within a domain in order for the case of porous
electrodes.

Introduction to Electrochemistry Modeling

SETTINGS
The Label is the physics interface node name that will be shown in the model builder
tree.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is tcd.

DOMAIN SELECTION
The domains that do not conduct current should be omitted from the selection list,
for example, the gas channels in a fuel cell.

OUT-OF-PLANE THICKNESS
For 2D components, the Thickness field (default value: 1 m) defines a parameter for the
thickness of the geometry perpendicular to the two-dimensional cross-section. The
value of this parameter is used, among other things, to automatically calculate the total
current from the current density vector. The analogy is valid for other fluxes.

CROSS-SECTIONAL AREA
For 1D components, enter a Cross-sectional area Ac (SI unit: m2) to define a parameter
for the area of the geometry perpendicular to the 1D component. The value of this
parameter is used, among other things, to automatically calculate the total current
from the current density vector. The analogy is valid for other fluxes. The default is
1 m2.

ELECTROLYTE CHARGE CONSERVATION

The physics interface features five different descriptions of the coupled charge and
mass transport in the electrolyte.

128 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Use the Electroneutrality or the Electroneutrality, water based charge conservation
option to model cells with significant concentration gradients of the current-carrying
species (ions). The electroneutrality condition implicitly assumes that all major
current-carrying ions are included in the model. In addition to the electroneutrality
condition, the Electroneutrality, water based option also adds the water auto-ionization
equilibrium condition, including proton and hydroxide transport, when defining the
electrolyte equations. Note that this option adds the concentration variables for
protons ([Link]) and hydroxide ([Link]) automatically, and that dependent variables
for these two species should not be added under Dependent Variables below. With this
setting, to control the initial pH in a simulation, set the initial concentrations of the
other ions in the simulation such that the matching concentration of protons and
hydroxide ions matches the pH desired. For pH less than 7, add and set the
concentration of anions. For pH more than 7, add and set the concentration of cations.
For example, for a water-based system with only Cl-, an initial concentration for Cl- of
105 M will result in an initial pH of 5. This is analogous to acidifying the solution
using HCl.

A Supporting electrolyte describes a situation where the major part of the charge is
transferred by ions whose concentration can be described as constant.

Use the Electroanalysis (no potential gradients) option to model electroanalytical

problems with electrolyte solutions containing a large quantity of inert supporting
electrolyte, with a conductivity so high that ohmic losses can be assumed to be
negligible. The electroanalysis option will not solve for the electrolyte potential as a
dependent variable, setting it to a constant value of 0. Migration effects are hence
neglected. Domain and boundary nodes only applicable to the electrolyte phase
potential will be disabled when using electroanalysis.

The Poisson option couples the Nernst–Planck equations for mass transport to the
Poisson equation for describing the potential distribution in the electrolyte, without
any assumption of electroneutrality. This option is typically used when modeling
problems where charge separation effects are of interest, typically within nanometers
from an electrode surface.

For the Electroneutrality option, the From electroneutrality list sets the species that is
calculated from the corresponding condition. Note that the choice of species to be
taken from electroneutrality affects the specific boundary conditions that can be set on
the eliminated species. For example, flux and concentration settings cannot be set for
the eliminated species, and initial values cannot be provided. The choice can also have
an impact on the numerics of the problem.

THE TERTIARY CURRENT DISTRIBUTION, NERNST–PLANCK INTERFACE | 129

 A general advice is to choose a relatively inert ion with high mole fraction to be taken
from electroneutrality for best numerical results.

PHYSICS VS. MATERIALS REFERENCE ELECTRODE POTENTIAL

The Physics vs. Materials Reference Electrode Potential setting on the physics interface
node can be used to combine material library data for current densities and equilibrium
potentials with an arbitrary reference electrode scale in the physics. The setting affects
the electrode potentials used for model input into the materials node, as well as all
equilibrium potential values output from the materials node.

Note that the setting will only impact how potentials are interpreted in communication
between the physics and the Materials node. If the From material option is not in use
for equilibrium potentials or electrode kinetics, the setting has no impact.

ADVANCED SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.
Normally these settings do not need to be changed. Select a Material Balance Form —
Nonconservative (the default) or Conservative. The selection affects the formulation of
the convective term and, for porous domain nodes, also the accumulation term in the
material balance. For compressible electrolytes the conservative form should be used.
In the absence of convection and electrolyte volume fraction changes (porosity), the
choice between the conservative and nonconservative form is arbitrary. The
nonconservative form is usually preferred due to a slighter lower computational load.

DEPENDENT VARIABLES
This physics interface defines these dependent variables (fields), the Concentrations of
the species, the Electrolyte potential, and the Electric potential.

The names can be changed but the names of fields and dependent variables must be
unique within a model.

DISCRETIZATION
Concentrations basis function orders higher than Quadratic are not recommended if
transport by convection is dominating in the model.

To see all settings in this section, click the Show More Options button ( ) and select
Advanced Physics Options from the Show More Options dialog.

130 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

CONSISTENT STABILIZATION AND INCONSISTENT STABILIZATION
To display these sections, click the Show More Options button ( ) and select
Stabilization from the Show More Options dialog. There are two consistent stabilization
methods available and selected by default — Streamline diffusion and Crosswind
diffusion. There is one inconsistent stabilization method, Isotropic diffusion, which is
not selected by default. Any settings unique to this physics interface are listed below.

• When the Crosswind diffusion checkbox is selected, a weak term that reduces spurious
oscillations is added to the transport equation. The resulting system is nonlinear.
There are two options for Crosswind diffusion type:
- Do Carmo and Galeão — the default option. This type of crosswind diffusion
reduces undershoot and overshoot to a minimum but can in rare cases give
equations systems that are difficult to fully converge.
- Codina. This option is less diffusive compared to the Do Carmo and Galeão
option but can result in more undershoot and overshoot. It is also less effective
for anisotropic meshes. The Codina option activates a text field for the Lower
gradient limit glim (SI unit: mol/m4). It defaults to 0.1[mol/m^3)/[Link],
where [Link] is the local element size.
• For both consistent stabilization methods, select an Equation residual. Approximate
residual is the default setting and it means that derivatives of the diffusion tensor
components are neglected. This setting is usually accurate enough and is faster to
compute. If required, select Full residual instead.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

• The Nernst–Planck Equations

• Domain Equations for Tertiary Current Distributions Using the
Nernst–Planck Equations and Electroneutrality
• Shared Physics Features in the Current Distribution Interfaces
• Theory for the Current Distribution Interfaces

THE TERTIARY CURRENT DISTRIBUTION, NERNST–PLANCK INTERFACE | 131

 Electrolyte
Use the Electrolyte node to define an electrolyte domain that only conducts current in
the ion conducting phase. The combined charge and mass transfer in the electrolyte is
defined by the node.

What settings are available in this node depends on the Electrolyte Charge
Conservation setting, available on the top node. The Electrolyte conductivity (SI unit:
S/m) setting is only available for the Supporting Electrolyte option. Diffusivity and
mobility settings for H+ and OH- are only available for the Electroneutrality,
water-based option.

The Convection section is available when the Convection checkbox is selected on the
interface top node. The Velocity field u (SI unit: m/s) of the solvent is specified as a
feature input. Select the source of velocity field from the velocity field list.

By default the Mobility (SI unit: s·mol/kg) for each species is set to be calculated based
on the diffusion coefficients (SI unit: m2/s) specified in the Diffusion section and the
temperature using the Nernst–Einstein relation.

The mobility setting will only have an impact on the transport by migration of charged
species, as defined by the Charge number zc (dimensionless, specify negative charges
using a minus sign). For the Electroneutrality charge conservation model you need at
least one positively and one negatively charged species (ion) in the electrolyte.

Specify the temperature (if you are using mobilities based on the Nernst-Einstein
relation) in the Model Inputs section.

The Water Self-ionization section is available for the Electroneutrality, water-based

option. The section allows the water self-ionization constant to be set. Using the Built
in option, the constant will be set to account for temperature, giving the pH of the
simulation the correct temperature dependence.

Note that the electrolyte in this case does not refer to the pore electrolyte in porous
electrodes (which should be defined by a Porous Electrode node instead). For porous
separators, use the Separator instead.

• Electrolyte Theory
• Domain Equations for Tertiary Current Distributions Using the
Nernst–Planck Equations and Electroneutrality

132 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Porous Electrode
The Porous Electrode node sets up charge balances for the electrode and the pore
electrolyte in a porous electrode, and as well as the mass balance for the species in the
electrolyte.

Note that the node should be used for porous domains that conduct current in both
an electrolyte and an electrode phase. For the case of domains that do not contain a
pore electrolyte — for instance, the gas diffusion layer (GDL) in a PEMFC electrode
— use an Electrode node instead.

Use Porous Electrode Reaction child nodes to define the charge transfer reactions that
occur on the interface between the electrolyte and electrode phases within the porous
electrode. The Porous Matrix Double Layer Capacitance subnode is also available.

See the Electrolyte node for more information about the Diffusion and Migration in
Electric Field settings of this node.

Correction factors may be specified in the Effective Transport Parameter Correction

section to account for the lowered effective conductivities of the electrode and
electrolyte phases due to the lower volume fractions of each phase, and the tortuosity
of the porous matrix.

The Electrode volume fraction is used to calculate the effective electric conductivity of
the porous matrix when the correction factor is set to Bruggeman or Tortuosity.
Additionally, it is used in calculating the active specific surface area of the porous matrix
when Particle-based area option is selected in the child nodes.

DISSOLVING-DEPOSITING SPECIES
See the Porous Electrode node of The Primary and Secondary Current Distribution
Interfaces

ADSORBING-DESORBING SPECIES
See the Electrode Surface node.

THE TERTIARY CURRENT DISTRIBUTION, NERNST–PLANCK INTERFACE | 133

 FILM RESISTANCE
See the Electrode Surface node.

• Porous Electrode Theory

• Porous Electrode Reactions Theory
• Domain Equations for Tertiary Current Distributions Using the
Nernst–Planck Equations and Electroneutrality

Separator
Use a Separator node to model electrolyte charge and mass transport in an
electronically isolating porous matrix. Use correction factors to account for the
lowered diffusion coefficients in the electrolyte and the lowered conductivities of the
electrode, due to the lower volume fractions of each phase and the tortuosity of the
porous matrix.

See also the Electrolyte node for more information about the Diffusion and Migration in
Electric Field settings of this node.

DISSOLVING-DEPOSITING SPECIES
See the Porous Electrode node of The Primary and Secondary Current Distribution
Interfaces

Reactions
Use the Reactions node to define nonelectrochemical reactions in an electrolyte
domain.

The node is found from the Additional Sources submenu.

REACTING VOLUME
When specifying reaction rates in the Rc2 (SI unit: mol/m3·s) fields for a species in a
Porous Electrode domain, the specified reaction rate expression may either refer to the
total volume or the pore (electrolyte) volume. For nonporous domains the settings of
the Reacting Volume section has no impact.

For Total volume the reaction expressions are used as specified (multiplied by unity).

For Pore volume this results in the specified reaction expressions being multiplied by
the domain electrolyte volume fraction l. (l equals unity for nonporous domains).

134 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Initial Values
Use this node to specify the Initial Values of the concentration, electrolyte potential and
electric potential dependent variables to be used by the solver.

For many electrochemical problems that use nonlinear electrode kinetics, such as
Butler–Volmer kinetics, providing reasonable initial values can significantly improve
solver convergence.

A good value for the Electric potential (SI unit: V) in electrode and porous electrode
domains can usually be derived from the boundary conditions. For instance, if a
boundary has been grounded or set to a cell potential, use that value as the initial value
also in the adjacent domain. For the Electrolyte potential (SI unit: V) a good initial
value is often the negative of the equilibrium potential of the grounded electrode.

For the Concentration initial values, at least one positive and one negative charged
species should have a nonzero and positive initial value (after considering the
electroneutrality condition). The initial value for the ion calculated from the
electroneutrality condition cannot be set explicitly.

Ion Exchange Membrane

Use the ion exchange membrane node to specify a domain with a permanent fixed
space charge. The node is typically used for defining permselective membranes, used
for various applications such as dialysis cells or flow batteries.

The node models the transport of all species added at the interface top node, and adds
a fixed space charge to the electroneutrality condition.

The Fixed space charge specifies the charge ions fixed in the membrane polymer matrix.
Use negative space charges for cation selective membranes and positive charges for
anion selective membranes, respectively.

Select the Apply Donnan Boundary Conditions checkbox to enable Donnan equilibrium
conditions on all interior boundaries between the domain selected by the node and all
adjacent domains selected by the interface (except Electrode nodes). The boundary
conditions are applied for all species and the electrolyte potential dependent variable.
This option is not available for the Poisson charge conservation model option.

For the remaining settings of this node, see the Electrolyte and Separator nodes.

THE TERTIARY CURRENT DISTRIBUTION, NERNST–PLANCK INTERFACE | 135

 Ion Exchange Membrane Boundary
Use the Ion Exchange Membrane boundary node to specify an external boundary over
which the flux of a charge-carrying species is continuous but the electrolyte potential
shifts with a specified value. This condition is typically used in electrochemical cells
containing both free electrolytes and ion exchange membranes, for instance in dialysis
or flow battery problems. The node is typically used to couple two separate current
distribution interfaces, where one of the interfaces models free electrolyte domain, and
the other interface models the ion-exchange membrane domain.

Note that using the Ion Exchange Membrane domain node often is a more convenient
modeling approach if the transported species in the free electrolyte and ion exchange
membrane are the same.

The node is applicable to external boundaries to electrolyte domains. For defining

ion-exchange membranes on an interior boundary between two electrolyte domains,
use the Thin Electrolyte Layer node.

The choice of Charge-carrying species concentration species specifies that the current
flowing over the boundary will be carried by this species (which must have a nonzero
charge number). Use the Membrane potential setting to set the electrolyte potential on
the membrane side of the boundary. Note that if this potential is set to the electrolyte
potential of a Primary or Secondary Current Distribution interface, no additional
settings are needed in that interface to set up the correct boundary condition.

The potential condition may be either Donnan, which will calculate the potential shift
over the boundary based on the membrane charge carrying species concentration, or
can be User defined.

Thin Electrolyte Layer

Use this node to specify a thin layer on an interior boundary between two electrolyte
domains. The node can be used as an alternative to drawing the actual layer as a
domain in the model geometry, which may significantly reduce meshing and solver
time, especially in 3D models.

The layer may either be Insulating, Resistive (supporting electrolytes only), or an Ion
exchange membrane.

For Resistive or Ion exchange membrane, the potential drop over the membrane is
determined either from the Surface resistivity or the Thickness and conductivity.

136 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

For Ion exchange membrane, the choice of Charge-carrying species concentration species
specifies that the current flowing over the layer will be carried by this species (which
must have a nonzero charge number). The potential condition may be either Donnan,
which calculates the potential shift over the boundary based on the membrane charge
carrying species concentration, or User defined.

THE TERTIARY CURRENT DISTRIBUTION, NERNST–PLANCK INTERFACE | 137

 Shared Physics Features in the
Current Distribution Interfaces
Domain, Boundary, Pair, Edge, and Point Nodes for the
Electrochemistry Interfaces
This chapter describes various domain, boundary, edge, point, and pair nodes that are
common to several of the Electrochemistry Interfaces interfaces.

The nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (macOS or Linux users), or right-click to access the context menu (all
users).The nodes and features described in this section are available for all the
Electrochemistry branch interfaces unless otherwise indicated.

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by (right) clicking
the parent node and selecting it from the Attributes menu.

138 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

• Circuit Terminal • Highly Conductive Porous
• Double-Layer Capacitance Electrode

• Electric Ground • Initial Values for Adsorbing–

Desorbing Species
• Electric Potential
• Initial Values for Dissolving–
• Electric Reference Potential
Depositing Species
• Current Conductor
• Insulation
• Electrode Current
• Perforated Electrode Surface
• Electrode Current Density
• Internal Electrode Surface
• Electrode Current Source
• Line Mass Source1
• Electrode-Electrolyte Boundary
• Nonfaradaic Reactions
Interface
• Point Mass Source1
• Electrode Potential
• Porous Electrode Reaction
• Electrode Reaction
• Porous Matrix Double-Layer
• Electrode Surface
Capacitance
• Electrolyte Current Density
• Reference Electrode
• Electrolyte Current Source
• Symmetry
• Electrolyte Potential
• Thin Electrode Layer
• External Short
• Harmonic Perturbation
1
For the Tertiary Current Distribution, Nernst–Planck interface, these nodes are
available and described for the Transport of Diluted Species interface.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Theory for the Current Distribution Interfaces

SHARED PHYSICS FEATURES IN THE CURRENT DISTRIBUTION INTERFACES | 139

 Current Conductor
Use the Current Conductor node to define an electrode domain that only conducts
current in the electron conducting phase.

The node is typically used for modeling solid metal electrodes, current collectors,
current feeders, gas diffusion layers and gas backings.

The Electric conductivity s (SI unit: S/m) parameter will define how the current in the
domain depends on the gradient of the potential.

Electrode Theory

Highly Conductive Porous Electrode

The Highly Conductive Porous Electrode node defines a charge balance of the pore
electrolyte in a porous electrode, and, for interfaces solving for electrolyte species, the
corresponding mass balance for the species in the electrolyte. Use this node when
electric conductivity is high enough to assume uniform potential in the electron
conducting phase of the porous electrode.

See the Electrode Surface node for a description of the Electrode Phase Potential
Condition and Harmonic Perturbation sections.

See the node for a description of the remaining settings.

Electrode Current Source

Use the Electrode Current Source node to define contributions to the current density
from electron current sources. The Current source, Qs (SI unit: A/m3), is added
according to the following equation:

  is = Qs

where

i s = –  s  s

and s denotes the electrode’s conductivity and s the electric potential.

140 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

To use this feature, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog. Then add the node from the Additional
Sources submenu.

Electrolyte Current Source

Use the Electrolyte Current Source node to define contributions to the current density
in the electrolyte from, for example reactions, or other effects.

The Current source, Ql (SI unit: A/m3), is added according to the following equation:

  il = Ql

To use this feature, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog. Then add the node from the Additional
Sources submenu.

Porous Electrode Reaction

The Porous Electrode Reaction node defines the electrode kinetics for a charge transfer
reaction that occurs at the interface between the pore electrolyte and the electrode
matrix a porous electrode. The node can be added as subnode to a Porous Electrode
node. Add multiple nodes to the same Porous Electrode node to model multiple
reactions, for instance in mixed potential problems.

See the Electrode Reaction node for a description of the Equilibrium Potential,
Electrode Kinetics, Stoichiometric Coefficients and Heat of Reaction sections.

ACTIVE SPECIFIC SURFACE AREA

The Active specific surface area, av (SI unit: m2/m3) specifies the area of the
electrode-electrolyte interface that is catalytically active for this porous electrode
reaction. av is multiplied by iloc, defined in the Electrode Kinetics section, to produce
a current source in the domain.

This section is not available for primary current distribution interfaces.

• Porous Electrode Reactions Theory

• Electrode Kinetics Expressions.

SHARED PHYSICS FEATURES IN THE CURRENT DISTRIBUTION INTERFACES | 141

 Porous Matrix Double-Layer Capacitance
Use the Porous Matrix Double-Layer Capacitance subnode to define a nonfaradaic
double-layer current density at the interface between the porous electrode matrix and
the electrolyte. The resulting double-layer current source in the Porous Electrode
domain depends on the time derivative of the potentials and is proportional to both
the Electrical double-layer capacitance Cdl (SI unit: F/m2) and the Double-layer area
av,dl (SI unit: 1/m). The node can be added as subnode to a Porous Electrode node.

Use this node to simulate transient analysis techniques such as AC-impedance analysis
and current interrupt studies. Note that for stationary problems the double-layer
current is zero.

This node is not available for the Primary Current Distribution interface.

STOICHIOMETRIC COEFFICIENTS
Use the settings of the Stoichiometric Coefficients section (not available in the Primary
or Secondary Current Distribution interfaces) to control what species are participating
in the double-layer charging — that is, the mass exchange between the double layer
and the electrolyte outside the double layer.

The number of participating electrons should be chosen so that the mass exchange is
scaled correctly to the double layer charging current.

The stoichiometric coefficients, i, should be chosen considering the situation when
the porous electrode is being cathodically polarized. In other words, use a negative
stoichiometric coefficient for ions (species Si) that are transported into the double layer
of the electrode that is being cathodically polarized (typically the cations), and positive
stoichiometric coefficients for ions (species Si) that are transported out of the double
layer (typically the anions).

Insulation
The Insulation boundary condition describes the walls of a cell or the boundaries of the
cell that do not face a conductor. The boundary condition imposes the following
equation:

ik  n = 0

where ik denotes the current density vector and k = l, s is an index for the electrolyte
and electrode, respectively.

142 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Symmetry
For the Primary Current Distribution and Secondary Current Distribution interfaces,
the Symmetry boundary condition is identical to the Insulation condition.

For the Tertiary Current Distribution, Nernst–Planck interface, the Symmetry

boundary condition imposes a no flux condition for the molar flux of species at a
boundary.

Symmetry Theory

Electrode Surface
Use the Electrode Surface node to model an electrochemical electrode-electrolyte
interface between an electrolyte domain and an electrode boundary where the
electrode is not included explicitly as a domain in the model geometry. Set the electric
potential of the electrode or specify a current condition that the potential of the
electrode shall fulfill, and use subnodes to specify the Electrode Reaction and the
Double-Layer Capacitance at the interface.

This node can only be applied on outer boundaries to electrolyte domains. For interior
boundaries between electrolyte and electrode domains, use the Internal Electrode
Surface node. For interior boundaries to electrolyte domains, use the Perforated
Electrode Surface node.

DISSOLVING–DEPOSITING SPECIES
Use the settings of this section to define species that participate in
dissolution-deposition electrode reactions, for instance metal deposition/dissolution
or oxide formation.

Use the Add ( ) and Delete ( ) buttons as needed in the table to control the
number of species.

The Density and Molar mass, in conjunction with the reaction rates and stoichiometry,
defined in the Electrode Reaction subnodes, determine the normal electrode growth
rate.

When the Solve for species concentrations variables checkbox is checked, dependent
variables for the molar surface concentration of the dissolving–depositing species are
added. These can be used to model the thickness of a dissolving/depositing layer in a

SHARED PHYSICS FEATURES IN THE CURRENT DISTRIBUTION INTERFACES | 143

 time-dependent simulation where the resulting deformation in the model geometry is
small and will have negligible impact on the current distribution. In stationary studies,
the molar surface concentration variables will be defined but not solved for by default.

When solving for the species concentration variables, corresponding thickness variables
are defined that you for instance can use to couple to the Film Resistance (see below).

Deposition-Dissolution Rates, Growth Velocities, and Thicknesses on an

Electrode Surface

ADSORBING–DESORBING SPECIES
Use the settings of this section to define species that may adsorb or desorb on the
electrode surface during charge transfer, or nonfaradaic, reactions.

The Density of sites parameter will be used to set up the local surface mass balance
equations for the fractional surface coverages.

Use the Add ( ) and Delete ( ) buttons as needed in the table to control the
number of species. The name of each adsorbing–desorbing Species may be edited in
the first column of the table. Use a Site occupancy number larger than unity for species
that occupy more than one site at the electrode surface.

According to the settings of the table, the fractional surface coverage variable names
will be defined according to xxx.theta_yyy_zzz where xxx is the interface name,
yyy the tag of the electrode surface node, and zzz the species name. The variable
xxx.thetafree_yyy represents the fraction of free surface sites. These variable names
may be used when defining kinetics parameters in Electrode Reaction subnodes.

Use The Surface Reactions Interface to model surface diffusion.

FILM RESISTANCE
Use a film resistance if you want to include an additional potential drop due to an
ohmic resistance at the interface between the electrode and the electrolyte, for instance
due to build-up of insulating deposits.

Specify either a Surface resistance Rfilm (SI unit: ·m2) directly or choose the Thickness
and conductivity option to calculate the surface resistivity based on a depositing film
thickness.

144 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

HARMONIC PERTURBATION
Use this section in conjunction with AC Impedance study types to control the
perturbation amplitude in the frequency domain.

The perturbation parameter is either Electric potential, Electrode potential, Total

current, or Average current density, based on the Boundary condition selected in the next
section.

The frequency spectrum is specified in the study node.

ELECTRODE PHASE POTENTIAL CONDITION

This section specifies the potential in the electrode phase of the electrolyte-electrode
interface. The electrode potential is used (via the overpotential) by the Electrode
Reaction subnodes.

Use the Electric potential option to set the value of the potential explicitly with respect
to ground whereas the Electrode potential will set the potential value with respect to a
reference potential. Total current, Average current density, and External short all add an
extra global degree of freedom for the potential in the electrode phase, set to comply
with the chosen condition. When using the Total current option in 1D or 2D, the
boundary area is based either on the Cross-sectional area (1D), or the Out-of-Plane
thickness (2D) properties, set on the physics interface top node.

In the Lithium-Ion Battery and the Battery with Binary Electrolyte interfaces, if the
Define cell state of charge (SOC) and initial charge inventory checkbox has been ticked in
the Cell Settings section on the physics interface top node, the Total current can be
specified either explicitly or as a multiple of the C-rate current variable
(xxx.I_1C_cell, where xxx is the physics interface tag) that is defined by the global
SOC and Initial Charge Distribution node.

See also the documentation for the Electrode Potential and External Short nodes for
further information about these boundary condition.

SHARED PHYSICS FEATURES IN THE CURRENT DISTRIBUTION INTERFACES | 145

 Cyclic Voltammetry
The Cyclic voltammetry setting varies the electric potential linearly in time as follows
when used in conjunction with a Cyclic Voltammetry study step:

End potential
Cycle 1 Cycle 2

Vertex 1

Vertex 2

Start potential

Figure 4-1: Electric potential vs time generated by the cyclic voltammogram boundary
condition. The linear sweep rate is 100 mV/s, the number of cycles is 2. Potentials levels are
also shown.

More advanced waveforms can be obtained using the Electric potential option with a
parameter setting based on Functions found in the Definitions menu.

Counter Electrode (Electroanalysis Only)

This boundary condition is only available for the Electroanalysis charge conservation
model in the Tertiary Current Distribution, Nernst-Planck (tcd) interface.

The Counter electrode option will set a potential to ensure an overall charge balance of
the cell so that the integral of all electrode reaction currents of all electrode surface
node sums up to zero.

See also Counter Electrodes and Overall Charge Balance

ADVANCED CYCLIC VOLTAMMETRY SETTINGS

To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

146 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

If Cyclic voltammetry is selected as the Boundary condition, the checkboxes Add events
at vertex potentials in cyclic voltammetry and Smoothing of cyclic voltammetry wave
functions are both selected by default.

When Smoothing of cyclic voltammetry wave functions is enabled, smoothing is applied

on the triangular wave around the vertex potentials. The Smoothing factor defaults to
1·103. The smoothing zone corresponds to the product of the smoothing factor with
half the duration of one period of the triangular wave.

When Add events at vertex potentials in cyclic voltammetry is enabled, solver events are
added at the vertex potentials for the cyclic voltammogram. When the vertex potentials
are reached, these events cause the solver to stop, reinitialize some variables, and then
continue. This typically improves solver convergence. In addition, by forcing the
time-dependent solver to resolve the time steps when the vertex potentials are reached,
the solution accuracy is improved, especially if the solver is set to output solutions
before and after events.

EQUILIBRIUM POTENTIAL HANDLING (PRIMARY CONDITION)

This setting only has an effect if there are multiple Electrode Reaction subnodes
present and if either a Current Distribution Initialization study step (using a Primary
Current distribution type) or the Primary Current Distribution interface is used.

The setting determines which equilibrium potential value will be used for defining the
primary current distribution constraint. When the First reaction has been selected, the
first electrode reaction subnode must be active in the model.

To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

CONSTRAINT SETTINGS
For primary current distributions, the use of weak constraints will in some cases give a
more accurate value of the local current density during the solver process. This may in
turn render more accurate results when coupling to the local current density variable
to describe other phenomena in the model, for instance when modeling geometry
deformation due to electrode dissolution/deposition.

The section is available in the Primary Current Distribution and Secondary Current
Distribution interfaces when the Current Distribution Model property has been set to
Primary.

This section is only available in the Primary Current Distribution and Secondary
Current Distribution interfaces when the Current Distribution Model property has been

SHARED PHYSICS FEATURES IN THE CURRENT DISTRIBUTION INTERFACES | 147

 set to Primary. To display this section, click the Show More Options button ( ) and
select Advanced Physics Options in the Show More Options dialog.

• Electrode Reactions Theory

• Film Resistance

Electrode Reaction
The Electrode Reaction subnode defines the electrode kinetics for a charge transfer
reaction that occurs on an electrolyte-electrode interface boundary. Use multiple
nodes to model multiple reactions, for instance in mixed potential problems.

The parent node may be either an Internal Electrode Surface or an Electrode Surface.

Note that all settings described below are not available for all Electrochemistry
interfaces.

• Electrode Reactions Theory

EQUILIBRIUM POTENTIAL
The Equilibrium potential, Eeq (SI unit: V), is used in the electrode kinetics expressions
in the Electrode Kinetics section (via the definition of the overpotential), or for setting
up primary current distribution potential constraints.

The equilibrium potential may be defined either in the Materials node (From material),
by using the Nernst Equation, or by using a User defined expression.

If the Nernst Equation is used, the concentration dependence is calculated

automatically based on the Reference equilibrium potential Eeq, ref (V).

For all interfaces except the Tertiary Current Distribution interface, the concentration
dependence is based on the user-defined Reduced species expression CR (unitless) and
Oxidized species expression CO (unitless) parameters. CR and CO should be defined so
that the quotient between them is 1 for the reference state (for which Eeq  Eeq,ref).

In the Tertiary Current Distribution interface, the concentration dependence of the

Nernst Equation is based entirely on the settings in the Stoichiometric Coefficients and
the Reference Concentrations sections.

148 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

When using Nernst Equation, additional options are available in the Butler-Volmer
expression type in the Electrode Kinetics section.

• Equilibrium Potentials and the Nernst Equation

REFERENCE CONCENTRATIONS
This section is only available in the Tertiary Current Distribution interface, if the
equilibrium potential has been selected to be defined by the Nernst Equation.

The reference concentrations define the reference state for which Eeq  Eeq,ref.

ELECTRODE KINETICS
The settings of this section will define the local current density, iloc (SI unit: A/m2),
at the interface between the electrolyte and the electrode. Note that iloc for all built-in
kinetics expression types will depend on the overpotential, which in turn depend on
the Equilibrium potential defined in the previous section.

The Local current density expression, iloc,expr (SI unit: A/m2), can be defined either in
the Materials node (From material), by using the From kinetics expression, or by using
a User defined expression.

For all kinetic expressions the Exchange current density i0 (SI unit: A/m2) is a measure
of the kinetic activity. The exchange current density is typically concentration
dependent.

Most kinetic expression types feature the Limiting Current Density option in order to
impose an upper limit on the local current density magnitude. The feature can be used
to model additional mass transport limitations that are not already included in the local
current density expression. For Limiting Current Density enter a value for ilim
(SI unit: A/m2).

In the Tertiary Current Distribution interface, the Linearize concentration dependence

for low concentrations option is used to set a Concentration linearization limit clim
(SI unit: mol/m3) for linearizing the concentration dependence of kinetics for low
concentrations, in order to improve convergence for nonunit stoichiometries. Note
that this option is available for Nernst Equation equilibrium potential and Butler-Volmer
kinetics with either Mass action law or Lumped multistep selected as the exchange
current density type.

SHARED PHYSICS FEATURES IN THE CURRENT DISTRIBUTION INTERFACES | 149

 Butler–Volmer or Linearized Butler–Volmer
The Butler–Volmer kinetics expression is the most common way to define
electrochemical kinetics. The Linearized Butler–Volmer is valid when the overpotentials
of the reactions are small (<<25 mV). The linearized version can also be used to
troubleshoot a model with convergence problems.

When using the Nernst Equation for defining the equilibrium potential (see above), the
concentration dependence of the Exchange current density i0 can be defined in a
thermodynamically consistent way in accordance with the Nernst equation, in
combination with a Reference exchange current density i0,ref (A/m2), which is the
exchange current density when Eeq  Eeq,ref.

For all interfaces except the Tertiary Current Distribution interface, the concentration
dependence when using From Nernst Equation will use CR and CO as preexponential
factors for the anodic and cathodic terms, respectively. In the Tertiary Current
Distribution interface, the Lumped multistep option can be used to define i0 by the use
of either Generic exponentials, or Anodic or Cathodic reaction orders. The Mass action
law will define the reaction orders according to the reaction stoichiometry and the law
of mass action.

The Anodic transfer coefficient, a (dimensionless), and Cathodic transfer coefficient, c

(dimensionless), parameters will impact how much iloc will change upon changes in the
overpotential. In order to ensure thermodynamic consistency, c cannot be user
defined when i0 is calculated From Nernst Equation (or by Mass action law in the Tertiary
Current Distribution interface). For this case, c is defined automatically, based on the
number of participating electrons in the reaction, defined in the stoichiometry section.

Anodic Tafel Equation

This kinetics expression type neglects the cathodic (negative) term in the Butler–
Volmer equation. It is only valid for electrode reactions with high anodic
overpotentials (>>100 mV).

The Anodic Tafel slope, a (SI unit: V), defines the required increase in overpotential
to result in a tenfold increase in the current density.

Cathodic Tafel Equation

This kinetics expression type neglects the anodic (positive) term in the Butler–Volmer
equation. It is only valid for electrode reactions with significant cathodic overpotentials
(<<-100 mV).

150 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

The Cathodic Tafel slope, c (SI unit: V), describes the required decrease in
overpotential to result in a tenfold increase in the current density magnitude. c
should be a negative value.

Concentration Dependent Kinetics

This expression type is not available if Nernst equation has been selected in the
Equilibrium Potential section.

Note that the combination of Nernst equation and the Butler-Volmer kinetics type will
in most cases render identical kinetics as for the Concentration Dependent Kinetics. It
is recommended to always use Nernst Equation + Butler–Volmer whenever possible,
since this combination is guaranteed to be thermodynamically consistent.

The Concentration Dependent Kinetics expression type may be used in concentration

dependent (tertiary) current distribution problems. One or both of the Oxidizing
species expression CO (dimensionless) and Reducing species expression CR
(dimensionless) parameters may be concentration dependent, and should typically be
defined so that COCR at equilibrium.

• Electrode Kinetics Expressions

Fast Irreversible Electrode Reaction

This kinetics expression type is typically used in tertiary current distribution problems
for reactions occurring far away from the equilibrium potential.

The kinetics expression type defines an irreversible electrode reaction where the
kinetics is so fast that the only factor limiting the reaction rate is the transport of a
species to the reacting surface.

The node will set the Rate limiting species concentration to zero at the boundary, and
balance the fluxes of the species participating in the reaction and the current densities
according to the Stoichiometric Coefficients settings.

Thermodynamic Equilibrium (Primary Condition)

This choice imposes a zero overpotential for the electrode reaction by applying a
constraint on the potential variables in order to comply with the equilibrium potential.
Use this kinetics for very fast reactions.

In the Secondary Current Distribution interface the condition set by this expression
type is mathematically identical to what is applied when a Primary Current

SHARED PHYSICS FEATURES IN THE CURRENT DISTRIBUTION INTERFACES | 151

 Distribution is chosen on the interface top node. The expression type can hence be
used to mix primary and secondary current distributions on different electrodes. The
Thermodynamic equilibrium (primary condition) cannot not be used when defining
the kinetics for multiple electrode reactions at the same electrode in the Secondary
Current Distribution interface.

STOICHIOMETRIC COEFFICIENTS
Specify the Number of participating electrons nm in the electrode reaction and the
Stoichiometric coefficient (vc1, vc2, and so forth) for each of the involved species
according to the following generic electrochemical reaction:

 ox Sox + ne  red Sred

-
 (4-1)
ox red

Set i as positive (red) for the reduced species and negative (ox) for the oxidized
species in an electrochemical reaction. The number of participating electrons, n,
should be positive.

If the concentration of a species in the charge conservation model for the electrolyte is
based on an algebraic expression (such as the electroneutrality condition, or the water
auto ionization), the stoichiometric coefficient for this species cannot be set explicitly.
The stoichiometric coefficient will instead be set implicitly, based on the number of
electrons and the stoichiometric coefficients of the other species participating in the
reaction.

An easy way to determine the stoichiometric coefficients for a reaction is

to write the reaction as a reduction reaction (with the electrons on the
left), irrespectively on the expected actual direction of the reaction in the
model. The species on the left side then have negative coefficients and the
species on the right have positive coefficients.

• Mass Fluxes and Sources Due to Electrochemical Reactions

HEAT OF REACTION
The Heat of Reaction section provides two options: Temperature derivative and
Thermoneutral voltage to calculate the reversible heat source of the electrode reaction,
which in turn can be used for coupling to heat transfer physics.

152 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

The Temperature derivative of equilibrium potential parameter, dEeqdT (SI unit:
V/K), can be specified in case of Temperature derivative selection. Note that dEeqdT
parameter value has no impact on the equilibrium potential variable.

The Thermoneutral voltage parameter, Etherm (SI unit: V), can be specified in case of
Thermoneutral voltage selection.

• Heating Due to Electrochemical Reactions

Double-Layer Capacitance
The Double-Layer Capacitance node describes the nonfaradaic currents due to charge
and discharge of the double layer situated at the interface between the electrode and
the electrolyte. The resulting double-layer current density on the boundary depends
on the time derivative of the potentials and is proportional to the Electrical double-layer
capacitance Cdl (SI unit: F/m2). The parent node may be either an Internal Electrode
Surface or an Electrode Surface.

Use this node to simulate transient analysis techniques such as AC-impedance analysis
and current interrupt studies. Note that for stationary problems the double-layer
current is zero.

This node is not available for the Primary Current Distribution interface.

STOICHIOMETRIC COEFFICIENTS
Use the settings of the Stoichiometric Coefficients section (not available in the Primary
or Secondary Current Distribution interfaces) to control what species are participating
in the double-layer charging — that is, the mass exchange between the double layer
and the electrolyte outside the double layer.

The number of participating electrons should be chosen so that the mass exchange is
scaled correctly to the double layer charging current.

The stoichiometric coefficients, i, should be chosen considering the situation when
the porous electrode is being cathodically polarized. In other words, use a negative
stoichiometric coefficient for ions (species Si) that are transported into the double layer
of the electrode that is being cathodically polarized (typically the cations), and positive
stoichiometric coefficients for ions (species Si) that are transported out of the double
layer (typically the anions).

SHARED PHYSICS FEATURES IN THE CURRENT DISTRIBUTION INTERFACES | 153

 Perforated Electrode Surface
The Perforated Electrode Surface is applicable to internal boundaries of an electrolyte
domains. The node may be used to model a thin and highly conductive perforated
(mesh) electrode, immersed in an electrolyte.

Apart from the applicable selection, the node is identical to the Electrode Surface
node.

Internal Electrode Surface

The Internal Electrode Surface node defines the electrochemical electrode-electrolyte
interface between an electrode domain and the electrolyte domain. Use this node as a
parent node for Electrode Reactions and the Double-Layer Capacitance of an
electrolyte-electrode interface when you explicitly model the electron conducting
electrode as a domain in your model.

Electrode domain
s
Electrolyte domain
l

Internal Electrode Surface

The node can be applied at interior boundaries between electrode and electrolyte
domains, and to external boundaries of porous electrode domains. Note that, due to
the high conductivity of many electrode materials, a Current Conductor domain can
many times be replaced an external condition at the electrolyte boundary. For such
cases, consider using the Electrode Surface instead.

154 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

For a description of the Film Resistance, Dissolving–Depositing Species and
Constraint Settings sections, see the Electrode Surface node.

• Electrode Reactions Theory

Electrolyte Potential
Add the Electrolyte Potential node from Electrolyte submenus for boundaries, edges,
and points to set a fixed potential at a position in the electrolyte. This node can be used
to model half-cells, or to set the electrolyte potential at the position of, for example, a
reference electrode.

The node sets the potential in the electrolyte, l, to be equal to the Boundary electrolyte
potential, l, bnd (SI unit: V).

By right-clicking this node you may add a Harmonic Perturbation subnode.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

Electrolyte Current
The Electrolyte Current boundary condition sets the total current or average current
density of a boundary. The condition sets the total inward current without imposing
the current density distribution. It will set a constant electrolyte potential along the
given boundary, that satisfies the current value setting.

When using the Total current option in 1D or 2D, the boundary area is based either
on the Cross-sectional area (1D) or the Out-of-Plane thickness (2D) properties, set on
the physics interface top node.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

• Electrolyte Current Theory

SHARED PHYSICS FEATURES IN THE CURRENT DISTRIBUTION INTERFACES | 155

 Electrolyte Current Density
Use the Electrolyte Current Density node to specify the current density distribution
along a boundary.

Note that using this node in 2D or 3D may result in an uneven potential distribution
along the boundary. To mitigate such effects you may use the Electrode Current node
instead.

By right-clicking this node you may enable Harmonic Perturbation. This means that the
node will only be active when solving for Frequency Domain study steps (typically used
in AC Impedance studies). The frequency spectrum is specified in the study node.

Thin Electrode Layer

The Thin Electrode Layer node can be used to model a thin insulating or resistive sheet,
located on an interior boundary in an electrode domain. The node can be used as an
alternative to drawing the actual layer domain in the model geometry, which may
significantly reduce meshing and solver time, especially in 3D models.

A thin electrode layer can be used to model, for instance, a contact impedance between
two electronic conductors.

The layer may be set to be either Insulating or Resistive.

For the case of Resistive, the Surface Resistance can either be set directly, or calculated
from Thickness and Conductivity values.

ADVANCED SETTINGS
Enable Slit electrolyte potential in porous electrodes to make the layer fully insulating
for the electrolyte phase potential along the selected boundaries.

Electrode-Electrolyte Boundary Interface

Note that this is an advanced physics feature that is normally not used. To display this
feature in the context menu, click the Show More Options button ( ) and then select
Advanced Physics Options in the Show More Options dialog. Then add the
Electrode-Electrolyte Boundary Interface node from the Electrode or Pairs > Electrode
submenu.

Use this node to model an electrochemical interface between an electrolyte and an

electrode where the electrolyte domain is not included explicitly in the model. The

156 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

node can only be applied on outer boundaries to electrode domains. Set the potential
of the electrolyte either explicitly, or specify a current condition that the potential of
the electrode shall fulfill.

Electrode Reaction and Double-Layer Capacitance subnodes are available from the
context menu (right-click the parent node) or from the Physics toolbar, Attributes
menu.

This node is available for the Secondary Current Distribution and Tertiary Current
Distribution, Nernst–Planck interfaces.

BOUNDARY CONDITION
This section specifies the potential of the electrolyte phase for the electrolyte-electrode
interface. The electrolyte potential is used (via the overpotential) by the Electrode
Reaction subnodes.

The Electrolyte potential will set the potential value directly, whereas Total current or
Average current density both add an extra global degree of freedom for the potential in
the electrolyte phase, set to comply with the chosen condition.

When using the Total current option in 1D or 2D, the boundary area is based either
on the Cross-sectional area (1D) or the Out-of-Plane thickness (2D) properties, set on
the physics interface top node.

Electric Ground
This node to sets the electric potential to zero.

The node is typically used to ground the voltage at an external boundary in a model
that contains either electrode or porous electrode domains.

CONTACT RESISTANCE
Enabling Include contact resistance allows for a Contact resistance Rc (SI unit: ·m2)
to be specified and the following current density condition is used:

–
– n  i s = -------s-
Rc

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

SHARED PHYSICS FEATURES IN THE CURRENT DISTRIBUTION INTERFACES | 157

 Electric Potential
This node sets the electric potential in the electrode (or a porous electrode), s, to a
value, sbnd according to the following:

 s =  s bnd

The node is typically used to set the cell voltage at an external boundary in a model
that contains either electrode or porous electrode domains.

CONTACT RESISTANCE
Enabling Include contact resistance allows for a Contact resistance Rc (SI unit: ·m2)
to be specified and the following current density condition is used:

 s bnd –  s
– n  i s = ---------------------------
-
Rc

By right-clicking this node you may add a Harmonic Perturbation subnode.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

Electrode Current Density

The Electrode Current Density boundary condition can be applied at an external
boundary of a model that contains either electrode or porous electrode domains.

Note that using this node in 2D or 3D may result in an uneven potential distribution
along the boundary. To mitigate such effects you may use the Electrode Current node
instead.

By right-clicking this node you may enable Harmonic Perturbation. This means that the
node will only be active when solving for Frequency Domain study steps (typically used
in AC Impedance studies). The frequency spectrum is specified in the study node.

Electrode Current
Use the Electrode Current node to set the total current or average current density over
an external electrode or porous electrode boundary — typically at the interface
between the electrode and the current collector or current feeder. The condition sets

158 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

the total inward current without imposing the current density distribution. The
potential along the boundary is calculated in order to satisfy the total value of the
current.

In the Lithium-Ion Battery and the Battery with Binary Electrolyte interfaces, if the
Define cell state of charge (SOC) and initial charge inventory checkbox has been ticked in
the Cell Settings section on the physics interface top node, the Total current can be
specified either explicitly or as a multiple of the C-rate current variable
(xxx.I_1C_cell, where xxx is the physics interface tag) that is defined by the global
SOC and Initial Charge Distribution node.

When using the Total current option in 1D or 2D, the boundary area is based either
on the Cross-sectional area (1D), or the Out-of-Plane thickness (2D) properties, set on
the physics interface top node.

CONTACT RESISTANCE
Enabling Include contact resistance allows for a Contact resistance Rc (SI unit: ·m2)
to be specified.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

• Electrode Current Theory

Electrode Power
The Electrode Power boundary condition sets the power drawn from, or inserted to, an
electrical cell at external electrode boundary.

When using the Total power option in 1D or 2D, the boundary area is based either on
the Cross-sectional area (1D), or the Out-of-Plane thickness (2D) properties, set on the
physics interface top node.

SHARED PHYSICS FEATURES IN THE CURRENT DISTRIBUTION INTERFACES | 159

 CONTACT RESISTANCE
Enabling Include contact resistance allows for a Contact resistance Rc (SI unit: ·m2)
to be specified.

Electrode Power Theory

Harmonic Perturbation
Use the Harmonic Perturbation subnode to specify the voltage amplitude perturbation
in the frequency domain. The harmonic perturbation is only applied when solving for
a Frequency Domain study type, which is typically used in AC Impedance studies.

The Harmonic Perturbation subnode can be added to the Electric Potential, Electrolyte
Potential. The subnode is available from the context menu (right-click the parent
node) or from the Physics toolbar in the Contextual group.

The frequency spectrum is specified in the study node.

You may also specify harmonic perturbations to the Electrode Current Density,
Electrolyte Current Density and Electrode Surface nodes.

Frequency-Domain Perturbation Study Step in the COMSOL

Multiphysics Reference Manual

Electrode Potential
Use the Electrode Potential node to set a boundary condition for the electric potential
with respect to a defined reference potential.

Electric potentials defined by the Electric Reference Potential and Reference Electrode
point nodes can be used as input when specifying the Electric reference potential vs,ref
(SI unit: V).

160 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

Electrode Potentials and Reference Electrodes

External Short
Use the External Short node to connect two electrodes over an external connector with
a given Resistance R (SI unit: ohm).

The boundary selected in the External Short node will be set to a constant potential,
s,here, and the integrated current over the boundary will be computed according to
Ohm’s law:

 s =  s here

 s here –  s there
 s here :   n  Is dd = – ---------------------------------------
R
-
d

where s,there (V) is the potential of the connected electrode. Use the Connected
Potential list to choose among available connection potentials for the value of s,there.

Initial Values for Adsorbing–Desorbing Species

The node sets the initial values for the surface coverage of adsorbing–desorbing species
dependent variables.

This node is available as a subnode for the Electrode Surface, Perforated Electrode
Surface and Internal Electrode Surface. The node is not available if no adsorbing–
desorbing species are present in the parent node.

Initial Values for Dissolving–Depositing Species

The node sets the initial values for the dissolving–depositing surface concentrations
dependent variables.

This node is available as a subnode for the Internal Electrode Surface and
Electrode-Electrolyte Boundary Interface. The node is not available if no dissolving–

SHARED PHYSICS FEATURES IN THE CURRENT DISTRIBUTION INTERFACES | 161

 depositing species are present or if the Solve for dissolving–depositing species
concentrations checkbox is cleared in the parent node.

Nonfaradaic Reactions
Use the Nonfaradaic Reactions node to define the reaction rate for dissolving–
depositing species due to nonfaradaic (not electrochemical) reactions that occur on the
boundary.

This node is available as a subnode to an Electrode Surface if there are dissolving–

depositing species present in the parent node.

Reference Electrode
The Reference Electrode node is a point feature applicable to electrolyte domains. It
defines a global electric reference potential and can be used in the Electrode Potential
node for setting the electric potential of an electrode boundary with respect to the
reference potential.

EQUILIBRIUM POTENTIAL
See Electrode Reaction for information about the settings of this section.

Electrode Potentials and Reference Electrodes

Electric Reference Potential

The Electric Reference Potential node is a point feature applicable to electrode domains.
It defines a global electric reference potential equal to the electric potential in the point
and can be used in the Electrode Potential feature for setting the electric potential of
an electrode boundary with respect to the reference potential.

Electrode Potentials and Reference Electrodes

162 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Circuit Terminal
This feature is only available with an AC/DC Module or a Battery Design Module
license.

Use the Circuit Terminal node to specify a coupling to the External I vs U node in the
Electrical Circuit interface.

SHARED PHYSICS FEATURES IN THE CURRENT DISTRIBUTION INTERFACES | 163

 The Concentrated Electrolyte
Transport Interface
The Concentrated Electrolyte Transport (cet) ( ), interface, found under the
Electrochemistry branch ( ) when adding a physics interface, is used to compute the
electrolyte potential and concentration distribution for an electrolyte mixture
comprising charged (ions) and neutral species. The interface is suitable for modeling
electrolytes featuring large concentration gradients of the charge-carrying ions, where
at the same time the concentrations of the charge-carrying ions are significant in
relation to the total concentration of all species in the electrolyte mixture.

OVERVIEW
The interface defines the individual transport of electrolyte species using the
Onsager-Stefan-Maxwell equations, which account for the binary interactions between
all modeled species in the mixture. In the Species section (see below), the active species
(cations, anions and neutral species) of the electrolyte mixture are defined. A minimum
of one cation and one anion needs to be defined. Internally, various matrix operations
are then performed in order to render an equation system defining an electrolyte phase
potential (default variable name: [Link]) and n  2 (where n is the total number of
electrolyte species) neutral electrolyte component fraction (cet.y_XXX) dependent
variables.

The names (XXX) of the electrolyte component fractions are generated automatically,
and are based on the species names and charges. For instance, the species set consisting
of the cation Fe (set to charge +2), the anion OH (charge -1) and the neutral species
H2O, will render the two electrolyte components Fe_OH_2 and H2O. Note that the main
purpose of defining the neutral electrolyte components is to generate an electroneutral
base for the dependent variables of the equation system, and the resulting chemical
formulas for the neutral salts may have no correspondence to for instance solid salt
compounds in nature.

Due to the constraint of local electroneutrality, a system of n species results in n-1

electrolyte components. Since it is also assumed all major electrolyte species are
included in the formulation, the equation system can be reduced further to include
only n-2 electrolyte component dependent variables.

164 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Throughout the user interface, symbols related to species are generally shown as Pi,
with P being the parameter or variable name, and i the species index, whereas symbols
related to electrolyte components are shown as P˜ i , i.e. with a tilde (~) above the
parameter/variable symbol, with i referring to the electrolyte component index.

The interface adds one Electrolyte node and one Reference Electrode as default
domain nodes. Both nodes are singletons and cannot be removed.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is cet.

DOMAIN SELECTION
If any part of the model geometry should not partake in the electrolyte transport part
of the model, remove that part from the selection list.

SPECIES
Specify the names and molar masses of the species in the Cations, Anions and Neutral
species tables. Only letters and numbers are permitted in the Name field. For the Cations
and Anions tables, also define the species charges. As described above in the Overview
section, this will result in a set of n-1 electrolyte components.

The Electrolyte component from molar-fraction constraint setting defines which

electrolyte component to be eliminated when defining what n-2 electrolyte
component fractions to use as dependent variables.

CONSISTENT STABILIZATION
To display this sections, click the Show More Options button ( ) and select
Stabilization in the Show More Options dialog.

Streamline diffusion (enabled by default) adds weak terms to the equation system that
may improve the quality of results for convection-dominated problems on coarse
meshes. When the Crosswind diffusion checkbox is selected, additional weak terms that
reduce spurious oscillations are added to the transport equations.

THE CONCENTRATED ELECTROLYTE TRANSPORT INTERFACE | 165

 DISCRETIZATION
To display the settings available in this section, click the Show More Options button
( ) and select Advanced Physics Options in the Show More Options dialog.

The same discretization is used for all dependent variables in the interface. Quadratic
elements are used by default. See Overview above regarding what dependent variables
are solved for, and naming conventions.

The Compute boundary fluxes checkbox is activated by default. When this option is
checked, the solver computes variables storing accurate boundary fluxes from each
boundary into the adjacent domain. If the checkbox is cleared, the boundary flux
variables are computed from the dependent variables using extrapolation, which is less
accurate in postprocessing results. This setting does however not create extra
dependent variables on the boundaries for the fluxes.

The flux variables affected in the interface are:

• <name>.nIl, where <name> is the name of the interface. This is the normal
electrolyte current density.
• <name>.ntflux_XXX. This is the normal total flux for each electrolyte component.

Also the Apply smoothing to boundary fluxes checkbox is available if the previous
checkbox is checked. The smoothing can provide a more well-behaved flux value close
to singularities.

For details about the boundary fluxes settings, see Computing Accurate Fluxes in the
COMSOL Multiphysics Reference Manual.

Regarding the Value type when using splitting of complex variables, see Splitting
Complex-Valued Variables in the COMSOL Multiphysics Reference Manual.

• See also Theory for the Concentrated Electrolyte Transport Interface

Electrolyte
The Electrolyte node defines various mixture properties.

In the Convection section, the mixture-averaged velocity of the electrolyte mixture is

defined. The Reference-electrode-stoichiometry based option defines the velocity based
on the electrolyte current vector and the stoichiometry of the Reference Electrode
reaction. This option is applicable when computing steady-state current distributions

166 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

when all charge-transfer reactions in the cell are the same as for the reference electrode.
This option may also serve as a fair approximation in certain time-dependent problems.
The User defined option can be used to provide an arbitrary velocity expression, for
instance velocity variables computed by a Fluid Flow interface.

The Volumetric Equation of State section defines the interrelations between the total
concentration of all species, the density, and the individual electrolyte component
fractions. The equation of state may be based either on a Density expression, or Molar
volumes. The density expression is typically composition dependent. Note that the
species Molar mass settings, defined on the interface top node, are also used when
defining the state equations.

The Chemical Potentials section defines how composition changes impacts the chemical
potentials of the electrolyte components. By default ideal conditions are assumed,
implying that unit chemical activity coefficients are used. Enabling Specify nonideal
thermodynamic factors allows for defining all independent partial derivatives of the
thermodynamic factors with respect to the electrolyte components.

When specifying the nonideal factors, the index k in the formula

˜  ln̃ i
 i j = ̃ i  ------------- (4-2)
 ỹ j  T p ỹ l  j k

equals the index of the component selected as Electrolyte component from

molar-fraction constraint on the interface top node.

Diffusion Coefficients
Use this node to define the transport properties of the electrolyte mixture. The
Maxwell-Stefan diffusivity coefficients define how gradients in the chemical potentials
of the species relate to the difference in species fluxes between all binary species pairs
in the mixture.

This node is a child node of the Electrolyte node.

Separator
This node is contributing to the Electrolyte node and can be used to specify a nonunit
electrolyte volume fraction, and to alter the bulk diffusivity coefficients defined in the
Diffusion Coefficients node as a result of a lowered electrolyte volume fraction and the
topology (tortuosity) of the porous media. The node is typically used to define an inert

THE CONCENTRATED ELECTROLYTE TRANSPORT INTERFACE | 167

 separator, but may also be used to define porous electrodes by combining it by one or
multiple Species Sources nodes.

The Effective diffusivity correction setting defines a correction factor which is used as a
prefactor to the chemical potential gradients when computing the diffusion fluxes on
the Electrolyte node. The correction may be set to either Bruggeman, Tortuosity,
Effective diffusivity factor, or No correction.

The Electrolyte volume fraction setting defines the relative pore volume of electrolyte
in the porous matrix. The setting impacts the mass conservation equation used in
time-dependent simulations, but also the diffusion correction factor used in the
Bruggeman and Tortuosity options.

Initial Values
Use this node to define the initial values of the dependent variables solved for by the
physics interface.

For the electrolyte components, the initial values may be specified as either
Concentrations, Mass proportions, Molar proportions, or Electrolyte component fractions.

Species Sources
Use this node to define species sources, typically stemming from charge transfer,
heterogeneous or homogeneous electrolyte reactions, in an Electrolyte domain.

Reference Electrode
Use this node to define the stoichiometry of a reference electrode reaction, which in
turn defines the electrolyte phase potential dependent variable added by the
Electrolyte node. By defining a reference electrode stoichiometry, the electrolyte phase
potential can be directly related to an experimentally measurable voltage.

By convention the stoichiometry is defined for a reduction reaction (with the electron
on the left-hand side). Nonzero values should be specified for at least one anion or one
cation.

The reference electrode stoichiometry defined on this node is also used when using a
Reference-electrode-stoichiometry based mass-averaged velocity in the Electrolyte node,
as well as when defining the resulting electrolyte component fluxes when imposing an
Electrolyte Potential boundary condition.

168 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

No Flux
This node defines a passive/insulating boundary.

Flux
Use this node to define species fluxes, typically stemming from charge transfer and/or
nonfaradaic reactions at an electrode surface adjacent to an Electrolyte domain.

Inflow
Use this node at an exterior boundary to define the electrolyte mixture of a convective
inlet stream. The electrolyte mixture may be defined either by using the same values as
defined on an adjacent Initial Values domain node, or by specifying the Concentrations
or Electrolyte component fractions on this node.

The boundary condition may be defined using either a Pointwise constraint or a Flux
(Danckwerts) condition. The Pointwise constraint condition can sometimes be a more
stable option, whereas the Flux (Danckwerts) may render more consistent results when
specifying an upstream mixture composition when there are concentration gradients
present in the electrolyte close to the boundary.

The electrolyte current density in the normal direction to boundary is set to zero.

Outflow
This node defines a convective outlet featuring no concentration gradients in the
normal direction to the boundary.

The electrolyte current density in the normal direction to boundary is set to zero.

Electrolyte Potential
Use this boundary node to define an electrode surface operating at a fixed potential in
the absence of kinetic activation losses.

The reaction stoichiometry of the electrode reaction is taken from the Reference
Electrode node, and the species fluxes in the normal direction of the boundary will
balance the electrolyte current density accordingly.

THE CONCENTRATED ELECTROLYTE TRANSPORT INTERFACE | 169

 T he E le c tr o de, S h el l In t erfac e
The Electrode, Shell (els) interface ( ), found under the Electrochemistry
branch ( ) when adding a physics interface, models electric current conduction in
the tangential direction on a boundary.

The physics interface is suitable for modeling thin electrodes where the potential
variation in the normal direction to the electrode is negligible. This assumption allows
for the thin electrode domain to be replaced by a partial differential equation
formulation on the boundary. In this way the problem size can be reduced, and
potential problems with mesh anisotropy in the thin layer can be avoided.

Ohm’s law is used in combination with a charge balance to describe the conduction of
currents in the shell electrode.

When this physics interface is added, these default nodes are also added to the Model
Builder — Electrode, Electric Insulation (the default edge or point condition), and Initial
Values. Then, from the Physics toolbar, add other nodes that implement, for example,
edge or point conditions and current sources. You can also right-click Electrode, Shell
to select physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is els.

THICKNESS

For 2D components, enter a value or expression for the Out-of-plane

thickness d (SI unit: m). The default value is 1 m.

170 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

DEPENDENT VARIABLES
The dependent variable (field variable) is for the Electric potential. The name can be
changed but the names of fields and dependent variables must be unique within a
model.

DISCRETIZATION
To see all settings in this section, click the Show More Options button ( ) and select
Advanced Physics Options in the Show More Options dialog.

• Theory for the Electrode, Shell Interface

• Boundary, Edge, Point, and Pair Nodes for the Electrode, Shell
Interface

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Boundary, Edge, Point, and Pair Nodes for the Electrode, Shell
Interface
The Electrode, Shell Interface has these boundary, edge, point, and pair nodes, listed
in alphabetical order, available from the Physics ribbon toolbar (Windows users),
Physics context menu (Mac or Linux users), or right-click to access the context menu
(all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

THE ELECTRODE, SHELL INTERFACE | 171

 These physics nodes are described in this section:

• Boundary Current Source • Electric Potential

• Current Source • External Current Density
• Electrode • Ground
• Electric Insulation • Initial Values
• Normal Current Density

These nodes are available and described for the Current Distribution interfaces, where
edges (3D components) or points (2D and 2D axisymmetric components) are selected
instead of boundaries.

• Electrode Potential
• Electric Reference Potential

For axisymmetric components, COMSOL Multiphysics takes the axial

symmetry boundaries (at r = 0) into account and automatically adds an
Axial Symmetry node to the component that is valid on the axial symmetry
boundaries only.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Electrode
The Electrode node defines the current conduction in the tangential plane. Use the
node to define the electrode thickness and electric conductivity.

ELECTRODE
The Electrode thickness s (SI unit: m) defaults to 1014 m.

The default Electric conductivity (SI unit: S/m) uses values From material. Or select
User defined. For User defined enter values or expressions for an isotropic or anisotropic
conductivity. Select Isotropic, Diagonal, Symmetric, or Full depending on the properties

172 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

of the electrolyte. To use another type of temperature dependence than a linear
temperature relation, enter any expression for the conductivity as a function of
temperature. Then enter a value or expression in the field or matrix.

Theory for the Electrode, Shell Interface

Initial Values
The Initial Values node adds the electric potential that can serve as an initial guess for
a nonlinear solver. If more than one initial value is needed, add Initial Values nodes from
the Physics toolbar.

INITIAL VALUES
Enter values or expressions for the Electric potential (SI unit: V). The default value
is 0 V.

External Current Density

Use the External Current Density node to define the external current density.

EXTERNAL CURRENT DENSITY

Based on space dimension, enter the vector components (x, y, and z for 3D
components, x and y for 2D, or r and z for 2D axisymmetric components) of the
External current density ie (SI unit: A/m2). The defaults are 0 A/m2.

Current Source
The Current Source node adds a source term to Equation 4-46. Use this node to define
the current source.

ELECTRODE CURRENT SOURCE

Enter a value or expression for the Current source Qi (SI unit: A/m3). The default is
0 A/m3.

THE ELECTRODE, SHELL INTERFACE | 173

 Normal Current Density
The Normal Current Density node is applicable to exterior boundaries that represent
either a source or a sink of current. It sets a condition for the current density according
to:

T  is = in

The node can be used to couple the Electrode, Shell interface to the electrode
reactions in an Electrochemistry interface that describes the electrolyte currents in the
adjacent domain.

NORMAL CURRENT DENSITY

Enter a value for the Electrode current source in (SI unit: A/m2). The default is
0 A/m2.

Electric Insulation
The Electric Insulation node is the default edge (3D components) and point (2D and
2D axisymmetric components) condition and describes the edges of the shell
(boundary) that do not conduct electricity.

Boundary Current Source

The Boundary Current Source node is available on edges (3D components) and points
(2D and 2D axisymmetric components) and sets a condition for the current density
according to:

i s  n = i s 0

BOUNDARY CURRENT SOURCE

Enter a value or expression for the Inward current density is,0 (SI unit: A/m2). The
default is 0 A/m2.

Ground
The Ground node is available on edges (3D components) and points (all components)
and sets the potential according to s  0.

174 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

EDGE OR POINT SELECTION

Beware that constraining the potential on edges or points in 3D or on

points in 2D usually yields a current outflow that is mesh dependent.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

Electric Potential
The Electric Potential node is available on edges (3D components) and points (all
components) and sets the potential according to s  s,0.

EDGE OR POINT SELECTION

Beware that constraining the potential on edges or points in 3D or on

points in 2D usually yields a current outflow that is mesh dependent.

ELECTRIC POTENTIAL
Enter the value or expression for the Electric potential s,0 (SI unit: V).

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

THE ELECTRODE, SHELL INTERFACE | 175

 The Electroanalysis Model Wizard
Entry
The Electroanalysis (tcd) ( ) Model Wizard entry, found under the Electrochemistry
branch ( ) adds a Tertiary Current Distribution, Nernst–Planck interface to the model
with the Electrolyte Charge Conservation Model set to Electroanalysis by default.

The Electroanalysis option is suitable for modeling mass transport of diluted species in
electrolytes using the diffusion-convection equation, solving for electroactive species
concentrations.

The Tertiary Current Distribution, Nernst–Planck Interface also features options for
modeling cyclic voltammetry and electrochemical impedance spectroscopy.

Use this Model Wizard entry to model electroanalytical problems with electrolyte
solutions containing a large quantity of inert “supporting” electrolyte. Ohmic loss is
assumed to be negligible.

The Model Wizard entry is available in 1D, 2D, and 3D as well as for axisymmetric
components in 1D and 2D.

The default dependent variables are the molar concentrations, c1 and c2, of the two
electroactive species in a redox couple and the electric potential, phis, which is solved
for either in the Electrode or Porous Electrode domain feature.

• The Tertiary Current Distribution, Nernst–Planck Interface

• Introduction to Electrochemistry Modeling
• Theory for Electroanalysis

176 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Theory for the Current Distribution
Interfaces
In this section:

• The Nernst–Planck Equations

• Domain Equations for Primary and Secondary Current Distributions
• Electrochemical Reactions and the Difference Between a Primary and a Secondary
Current Distribution
• Domain Equations for Tertiary Current Distributions Using the Nernst–Planck
Equations and Electroneutrality
• Mass Fluxes and Sources Due to Electrochemical Reactions
• Deposition-Dissolution Rates, Growth Velocities, and Thicknesses on an Electrode
Surface
• Stoichiometric Coefficients for Double Layer Capacitive Charging
• Film Resistance
• Equilibrium Potentials and the Nernst Equation
• Electrode Kinetics Expressions
• Theory for Specific Current Distribution Feature Nodes
• Reference

This theory is also relevant to the Current Distribution on Edges, BEM

interface. However, due to the use of BEM, current source terms in the
electrolyte, and porous electrodes, cannot be used in this interface.

The Nernst–Planck Equations

The general mass balance for the diluted species in an electrolyte is described by the
following equations for each species i:

c i
+   N i = R i tot
t

THEORY FOR THE CURRENT DISTRIBUTION INTERFACES | 177

 where Ni is the total flux of species i (SI unit: mol/(m2·s)). The flux in an electrolyte
is described by the Nernst–Planck equations and accounts for the flux of charged solute
species (ions) by diffusion, migration, and convection; these are respectively the first,
second, and third term on right side in the equation below.

N i = – D i c i – z i u m i Fc i  l + c i u = J i + c i u

where

• ci represents the concentration of the ion i (SI unit: mol/m3),

• zi its valence,
• Di the diffusion coefficient (SI unit: m2/s),
• um,i its mobility (SI unit: s·mol/kg),
• F denotes the Faraday constant (SI unit: C/mol),
• l the electrolyte potential,
• u is, the velocity vector (SI unit: m/s), and
• Ji denotes the molar flux relative to the convective transport.

J i = – D i c i – z i u m i Fc i  l

The net current density can be described using the sum of all species fluxes:

il = F  zi Ni
where il denotes the current density vector (SI unit: A/m2) in the electrolyte.

The Nernst-Planck Equations can be written in nonconservative or conservative form.

The nonconservative formulations are applicable to systems that contain
incompressible electrolytes. For these electrolytes, the following continuity equation is
satisfied:

u = 0

The velocity vector in the convective term can therefore be moved out of the
divergence operator which gives the nonconservative form of the Nernst-Planck
equations:

c i
+   J i + u  c i = R i
t

178 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

On the other hand, for compressible electrolytes, the convective term cannot be
simplified and the use of the following conservative form of the Nernst-Planck
equations is applicable:

c i
+    Ji + ci u  = Ri
t

For mass transport in porous systems, the accumulation term is deviating between the
conservative or nonconservative formulations. The electrolyte volume fraction
(porosity), i, is only differentiated in the conservative form giving the following
formulation:

  i ci 
- +    Ji + ci u  = Ri
----------------
t

While it is not for the nonconservative case:

c i
 i ------- +   J i + u  c i = R i
t

Considering a system with no producing or consuming species reactions, these

formulations ensure constant species concentration (mass conservation) over the
model domains together with accumulated mass across in/outflow boundaries over
time.

Domain Equations for Primary and Secondary Current

Distributions
Assuming electroneutrality (which cancels out the convection term) and negligible
concentration gradients of the current-carrying ion (which cancels out the diffusion
term), the following expression is left for the current density vector in an electrolyte:

 zi um i ci l .
2 2
il = –F

Further, assuming approximately constant composition of charge carriers, we can

define a constant electrolyte conductivity as:

 zi um i ci
2 2
l = F

the current density in the electrolyte can be written as

THEORY FOR THE CURRENT DISTRIBUTION INTERFACES | 179

 i l = –  l  l

This equation takes the same form as Ohm’s law; in an electrolyte, charge transport is
ohmic, subject to the above assumptions.

Conservation of charge yields the domain equation usually used for the electrolyte in
the Primary and Secondary Current Distribution interfaces:

  il= 0

In a pore electrolyte, the homogenization used in the theory for porous

electrodes introduces a source or sink term in the pore electrolyte current
balances due to the charge transfer reactions at the electrode-electrolyte
interface within the porous material. In such cases, a source term, Ql, is
introduced on the right-hand side of the equation above.

The Primary and Secondary Current Distribution interfaces define two dependent
variables: one for the potential in the electrolyte and one for the electric potential in
the electrode. The conduction of current in the electrolyte is assumed to take place
through transport of ions as described above, while electrons conduct the current in
the electrode.

Since Ohm’s law is also used for current conduction in the solid electrode phase, the
general equation in these interfaces is according to the following:

  ik = Qk

with

i k = –  k  k

where Qk denotes a general source term, k denotes an index that is l for the electrolyte
or s for the electrode, k denotes the conductivity (SI unit: S/m) and k the potential
(SI unit: V).

Electrochemical Reactions and the Difference Between a Primary and

a Secondary Current Distribution
Both the primary and secondary current distribution models assume an electrolyte that
conducts current according to Ohm’s law, with a constant conductivity. The difference
between a primary and a secondary current distribution lies in the description of the

180 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

electrochemical charge transfer reaction at the interface between an electrolyte and
an electrode. The charge transfer reactions can be defined as boundary conditions, or
as sources or sinks in a domain for the case of porous electrodes.

The rate of the electrochemical reactions can be described by relating the reaction rate
to the activation overpotential. For an electrode reaction, with index m, the activation
overpotential, denoted m, is the following:

 m =  s –  l – E eq m

where Eeq,m denotes the equilibrium potential (also known as a reduction potential)
for reaction m.

In the Primary Current Distribution interface, the electrochemical reactions are

assumed to be fast enough that their kinetics have negligible influence on the cell
voltage. Consequently, the activation overpotential, m, in this physics interface is
equal to zero, and the difference in potential between the electrode and the electrolyte
is always equal to the equilibrium potential, Eeq,m. This condition is implemented as
the following constraint on boundaries between electrodes and electrolyte domains:

 l =  s – E eq m

A good modeling practice is to estimate the influence of finite reaction

kinetics on the cell voltage and to compare the associated activation losses
to the ohmic losses in a cell to check the validity of the primary current
distribution assumption.

In the Secondary Current Distribution interface, the electrochemical reactions are

described as a function of the overpotential. The physics interface uses several relations
for the charge transfer current density and the overpotential, such as Butler–Volmer
and Tafel expressions. The most general expression is of Butler–Volmer type:

 a m F m –  c m F m
i loc,m = i 0 m  exp  ------------------------- – exp  ---------------------------- 
 RT  RT

THEORY FOR THE CURRENT DISTRIBUTION INTERFACES | 181

 where iloc,m denotes the local charge transfer current density for reaction m, i0,m the
exchange current density, a,m the anodic transfer coefficient, c,m the cathodic
charge transfer coefficient, F Faraday’s constant, and R the universal gas constant.

Both the exchange current density and the overpotential are typically
concentration dependent. It is possible include the dependence of
kinetics on concentration in the expression above. It also possible to use
other kinetics expressions.

See Equilibrium Potentials and the Nernst Equation and Electrode

Kinetics Expressions.

The sum of all electrode reaction currents is implemented as a current density

condition on the boundary between an electrode and an electrolyte domain according
to:

il  n =  iloc m
m

is  n = –  iloc m
m

Both the Primary Current Distribution and Secondary Current Density Distribution
interfaces allow for a domain definition for porous electrodes. For the Primary Current
Distribution interface, the same constraint as above is applied.

In porous electrodes for the Secondary Current Distribution interface, the sum of all
reaction currents appears as a source in the domain equations:

  il =  Av m iloc m
m

  is = –  Av m iloc m
m

where Av,m is the surface area.

The only reason to add Porous Electrode descriptions to the Primary

Current Distribution interface is to be able to expand a model to include
charge transfer resistance.

182 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Domain Equations for Tertiary Current Distributions Using the
Nernst–Planck Equations and Electroneutrality
The domain equations can be derived by starting with The Nernst–Planck Equations.
The material balance equation for each species i in the electrolyte is given by the
continuity equation, with a flux given by the Nernst–Planck equation:

c i
+    J i + c i u  = R i tot
t

where

• ci denotes the concentration of species i (SI unit: mol/ m3),

• u is, the velocity vector (SI unit: m/s), and
• Ji denotes the molar flux relative to the convective transport(see Equation )

The material balances give one equation per unknown species concentration.

The concentration gradients are not assumed to be negligible here, and

so the contribution of ion diffusion to overall current density can be
nonzero. (Compare with the Domain Equations for Primary and
Secondary Current Distributions.)

There is one more unknown dependent in the variable, the electrolyte potential, which
requires an additional equation to close the system. This equation is the
electroneutrality condition, which follows from dimensional analysis of Gauss’s law. In
a typical electrolyte solution, it is accurate over lengths greater than a few nanometers:

 zi ci = 0

Using the electroneutrality condition reduces the number of dependent concentration

variables by one.

Further, by combining the electroneutrality condition with the sum of all species flux
vectors and the species mass balance equation, multiplied by the individual species
charges, results in expressions for the current vector and a current balance equation,
respectively.

The expression for the current density in the electrolyte, il, reads:

THEORY FOR THE CURRENT DISTRIBUTION INTERFACES | 183

 n n

il = F  zi  –Di ci – zi um i F ci l  = F  zi Ji

i=1 i=1

whereas current balance in the electrolyte, used for solving for the electrolyte potential,
then becomes:

  il = Ql

where Ql can here be any source or sink. (Ql is typically nonzero for porous
electrodes). These formulations are also valid for the pore electrolyte in porous
electrodes, except for the transport properties that have to be corrected for porosity
and tortuosity. In such cases, the source or sink, Ql, denotes the charge transfer
reactions in the porous electrode and/or the nonfaradaic source or sink due to double
layer charge and discharge.

The charge transfer reaction can be parameterized by arbitrary functions of the

concentrations of the species in the redox couple and the local electric and electrolyte
potentials. The most common way to describe the reaction kinetics is to use a Butler–
Volmer expression for the charge transfer current density. See The Butler–Volmer
Equation.

In the current balance in a porous electrode, the local current density multiplied by the
specific surface area of an electrode gives a contribution to the source or sink, Ql, due
to electrochemical reactions.

Mass Fluxes and Sources Due to Electrochemical Reactions

When relating the species fluxes at an electrode surface to the electric currents, the
general electrochemical reaction formula:

 ox Sox + ne  red Sred

-

ox red

is used to define the stoichiometric coefficients, i, with i being positive (red) for
products and negative (ox) for the reactants in a reduction reaction. The number of
participating electrons, n, is always positive.

The molar species fluxes, Ni (SI unit: mol/(m2·s)), perpendicular to an

electrode-electrolyte interface are calculated by summing all the flux contributions
from the electrode reactions of index m, according to Faraday’s laws:

184 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

  i m i loc m
Ni =  -------------------------
nm F
-
m

where iloc,m is the local current density (SI unit: A/m2) of the electrochemical
reaction, nm the number of participating electrons and F (SI unit: C/mol) is the
Faraday constant. Note that iloc,m denotes the current density perpendicular to the
electrode surface.

The molar species flux, Ni, is obtained from the normal component of the molar
species flux vector over the electrode-electrolyte interface:

Ni = Ni  n

where n is the normal vector of the boundary pointing into the domain.

For a porous electrode, the electrochemical reactions result in species source terms
calculated from:

 i m i loc m
R i molar = –  av m --------------------------
nm F
-
m

where av,m (SI unit: m2/m3) is the specific surface area.

Deposition-Dissolution Rates, Growth Velocities, and Thicknesses on

an Electrode Surface
The rate of dissolution or deposition on due to an electrode reaction is defined based
on the reaction stoichiometry, im, and the local current density, ilocm, as described
in Mass Fluxes and Sources Due to Electrochemical Reactions.

If the reaction rate is known, the total growth vdep, tot (m/s) is defined as the sum of
the velocity contributions for all species and electrode reactions according to:

Mi  i m i loc m
v dep tot = –  ------
i  nm F
- -------------------------- (4-3)
i m

Where Mi (SI unit: kg/mol) is the molar mass and i (SI unit: kg/m3) the density of
the species. (i is the species index, and m the index of the electrode reaction).

This velocity may be used in deforming geometry models as a boundary condition for
the geometry deformation by assuming that dissolution or deposition always occurs in

THEORY FOR THE CURRENT DISTRIBUTION INTERFACES | 185

 the normal direction to an electrode boundary, with the velocity being directed into
the electrolyte domain:

x
 n = v dep tot
t

In a time-dependent simulation one may also introduce a surface concentration

variable, cs,i (mol/m2) on the boundary and calculate the accumulated surface
concentration change by using a local ordinary differential equation (ODE):

dc s i  i m i loc m
dt
= –  -------------------------
nm F
-
m

The total deposited thickness stot(m) can then be defined as

Mi
s tot =  ------
-c .
 i s i
i

Stoichiometric Coefficients for Double Layer Capacitive Charging

A negative capacitive current corresponds to the accumulation of negative charge on
the electrode surface, and a corresponding counterbalanced positive charge
accumulating in the double layer. This can be understood in terms of conservation of
current: cathodic capacitive current at the electrode corresponds to the overall
injection of negative charge (from the double layer) into the electrolyte, just as a
cathodic Faradaic current would add negative charge to the electrolyte through
electron transfer.

Consider an overall reaction as:

- + -
2e (electrical circuit) + A (solution) + X (double layer)

- + -
2e (electrode surface) + A (double layer) + X (solution)

This corresponds to a negative charge of 2F being moved from double layer to

solution. Following the convention that the reaction is written in the order of cathodic
current (as above), then with respect to the free solution species, the cation is a reactant
(negative stoichiometry) and the anion is a product (positive stoichiometry). This
suggests A = 1, X = 1, and n = 2 for an ideal double layer where both anion and
cation have similar contributions to the double layer charge.

186 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Film Resistance
If a resistive film forms on the interface between an electrode and an electrolyte, this
results in additional potential losses. To model a film resistance, an extra dependent
variable for the potential drop over the film, s,film, is introduced. The governing
equation is:

 s film = i tot R film

Where Rfilm (SI unit: ohm·m2) is a generalized film resistance and itot the sum of all
currents over the interface. The activation overpotentials, m, for all occurring
electrode reactions on the electrode with the film receive an extra potential
contribution due to the film resistance according to:

 m =  s –  s film –  l – E 0 m

If the thickness and conductivity of the film are known, the resistance can be written as:

s 0 + s
R film = -----------------
-
 film

where s0 is the reference/initial film thickness, s the electrode thickness change, and
film the conductivity (SI unit: S/m) of the film.

Equilibrium Potentials and the Nernst Equation

Consider an arbitrary, possibly multi-electron step, redox reaction involving a number
of species Si defined as

 
-
 i S i + ne  i Si
i: i  0 i: i  0

where i is the stoichiometric coefficient of the reacting species of index i and n is the
number of participating electrons.

The equilibrium potential of the electrode reaction, Eeq (V), is the electrode potential
(the difference between the electrode phase and electrolyte phase potentials, s  l)
for which the net reaction rate (and the local current density, iloc) is zero.

The equilibrium potential is directly related to the change of Gibbs free energy of the
reacting species, G, as

THEORY FOR THE CURRENT DISTRIBUTION INTERFACES | 187

 E eq = – G
--------
nF

Since the free energy of the reacting species is concentration dependent, also the
equilibrium potential is concentration (activity) dependent. Eeq is usually defined
according to the Nernst equation as

ai i
  ------------
RT
E eq = E eq ref – -------- ln
nF a i ref
i

where Eeq,ref (V) is the equilibrium potential for a reference state for which all species
activities ai (unitless) are equal to a chosen set of reference activities ai,ref (unitless).

For ideal solutions, the activities are replaced by concentrations. Standard conditions
correspond to reference concentrations of 1M for soluble species in the electrolyte,
partial pressures of 1 atm for gaseous species. Constant activities of 1 is used for solid
(metal) species and solvents.

Electrode Kinetics Expressions

A number of different analytical expressions for the current density perpendicular to
the electrode surface, iloc,m, are available. In the following paragraphs, the index m is
dropped. All parameters are understood to refer to a specific reaction.

THE EQUILIBRIUM POTENTIAL AND THE OVERPOTENTIAL

The rate of the electrochemical reactions can be described by relating the reaction rate
to the activation overpotential. For an electrode reaction, the activation overpotential,
denoted  (V), is the following:

 =  s –  l – E eq

where Eeq denotes the equilibrium potential.

THE BUTLER–VOLMER EQUATION

A common expression for the current density as a function of the activation
overpotential, in modeling of electrochemical systems, is the Butler–Volmer equation:

 a F –  c F
i loc = i 0  exp  --------------- – exp  ----------------- 
  RT   RT  

188 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

where c (unitless) denotes the cathodic charge transfer coefficient, a (unitless) the
anodic charge transfer coefficient, and i0 (SI unit: A/m2) is the exchange current
density.

It should be noted that, although used extensively for all sorts of reactions in the
electrochemical community, the Butler–Volmer equation was originally derived for
single-electron transfer reactions (for instance implying that the reaction does not
involve the breaking or creation of a chemical bond).

Concentration Dependence and the Exchange Current Density

The exchange current density is generally concentration dependent. For certain
conditions, it is possible to derive analytical expressions for i0. Assuming a
concentration dependent equilibrium potential defined by the Nernst equation (see
previous section Equilibrium Potentials and the Nernst Equation), the kinetics to
follow the law of mass action, and the condition that a  c  n, then the exchange
current density becomes as follows:

c i –a i
---------
- --------------
ai 
 ------------
n ai 
 ------------
n
i 0 = i 0 ref   a i ref   a i ref
i: i  0 i: i  0

where i0,ref is the exchange current density (SI unit: A/m2) at the reference state. The
above expression can be derived from the mass action law, which gives the following
expression for the local current density:

 ai  i
 ------------  a F ref ai  –i –  c F ref 
- exp  --------------------  ------------ exp  ----------------------- 
i loc = i 0 ref 
   a i ref  RT 
- –
  a i ref
-
 RT  
i: i  0 i: i  0

where the overpotential ref (SI unit: V) is measured using relative to a reference state,
which yields:

 ref =  s –  l – E eq ref .

This latter form of the Butler–Volmer equation, where the reference overpotential and
the exchange current density do not vary with concentration, is usually preferable
numerically.

The law of mass action is usually not suitable for defining complex reactions involving
multiple electron steps. For certain multi-electron reactions, where one electron
transfer step is rate limiting, it is possible to derive a lumped Butler–Volmer expressions
using the following relation for the exchange current density (see Ref. 1):

THEORY FOR THE CURRENT DISTRIBUTION INTERFACES | 189

 ai i
i 0 = i 0 ref   ------------
a i ref
-
i

where i are generic exponential coefficients. For this case, a and c may be
independently defined. By rearranging the Butler–Volmer expression using ref
similarly to what was done for the mass action law case above, the i coefficients relate
to the anodic, ai, or cathodic, ci, reaction orders according to

a i
 a i =  i + -----------
n

and

c i
 c i =  i – ----------
-.
n

Exchange Current Density and Rate Constants

In electroanalysis, one commonly defines electrode kinetics in terms of rate constants
rather than exchange current densities.

For instance, for a one electron redox couple of concentrations co and cr, with the same
reference concentration cref for both species, and i0,ref  k0Fcref, the mass action law
expression above can be rewritten as

 a F –  c F
i loc = k 0 F  c r exp  --------------- – c o exp  ----------------- 
RT RT

where k0 (m/s) is the heterogeneous rate constant.

Generic Concentration Dependent Butler–Volmer Type Kinetics

This expression type gives more freedom for the user to define concentration
dependent Butler–Volmer types of expressions, where the anodic and cathodic terms
of the current density expression, typically depending on the local concentrations of
the electroactive species at the electrode surface, may be individually defined:

 a F –  c F
i loc = i 0  C R exp  --------------- – C O exp  ----------------- 
  RT   RT  

Here CR and CO are dimensionless expressions, describing the dependence on the

reduced and oxidized species in the reaction. Note that if CO  CR when   0, this
kinetics expression results in iloc  0, thus violating the concept of equilibrium. This

190 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

may result in thermodynamical inconsistencies, for instance when coupling an
electrochemical model to heat transfer.

SIMPLIFICATIONS OF THE BUTLER–VOLMER EQUATION

Linearized Butler–Volmer
The charge transfer reaction can be expressed by a linearized Butler–Volmer
expression, which can be used for small overpotentials ( << RT/F) and is usually
referred to as the low-field approximation. This approximation gives the following
linearized equation:

  a +  c F
i loc = i 0  ---------------------------- 
RT

Anodic and Cathodic Tafel Equations

By assuming either high anodic or cathodic overpotentials for a given current (that is,
slow kinetics or low i0), one of the terms in the original Butler–Volmer potentials can
be neglected.

The anodic Tafel equation is implemented as follows:

  Aa
i loc = i 0  10

where Aa (SI unit: V) is the so-called Tafel slope. Aa relates to the corresponding
transfer coefficient as follows

A a = RT ln 10-
---------------------
a F

The cathodic Tafel expression is defined according to:

  Ac
i loc = – i 0  10

where the sign accounts for the negative cathodic charge transfer current. Here, Ac is
required to be negative and relates to the transfer coefficient according to

RT ln 10-
A c = – ---------------------
c F

THEORY FOR THE CURRENT DISTRIBUTION INTERFACES | 191

 LIMITING CURRENT DENSITY
The steady-state rate of electrode reactions can never exceed the rate at which reactants
and products can be transported to and from the electrode surface. When explicitly
including mass transport in a model, this dependence is typically described in
concentration dependence of the equilibrium potential and the exchange current
density as described above.

When not explicitly including mass transfer in the domain equations one can still
include the effect of transport limitations by the assumption of a Nernst diffusion layer
at the electrode surface, and a first-order dependence between the charge transfer
current and the local concentration of a reacting species, resulting in the following
kinetics expression:

i expr
i loc = -------------------------
-
i expr
1 + -----------
i lim

where iexpr (A/m2) is the current density expression in the absence of mass transport
limitations for the species, and ilim (A/m2) is the limiting current density that
corresponds to the maximum transport rate of the species. The derivation of this
expression assumes high overpotentials so that either the anodic or a cathodic term in
the Butler–Volmer equation may be neglected.

LINEARIZE CONCENTRATION DEPENDENCE FOR LOW CONCENTRATIONS

Consider a concentration-based kinetic expression

dc 
= r = – kc
dt

where  and k are positive numbers and the desired behavior is that the rate r and the
concentration c should equal zero in the converged solution at infinite time. However,
if c, due to numerical fluctuations in the solver process, becomes negative during
iterating, issues may arise.

First consider the case when  equals 1 (or any odd positive integer). Negative values
of c will then cause the rate to become positive, resulting in a “self-stabilizing”
situation where c will be approaching 0 with time.

A second case to consider is when  is an even integer larger than 1. The rate then will
become increasingly negative for negative values of c, resulting in an “exploding”
solution, iterating c toward minus infinity. The standard solution for these cases, which

192 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

also works for noninteger  values larger than 1, is to change the expression c in the
rate term to max(c,eps), where eps is a small number. This will avoid the “exploding”
behavior, but result in poor convergence rate for negative c values since the Jacobian
of the rate with respect to c then becomes zero for negative c values.

The third case is when is a noninteger between 0 and 1. Note now that Jacobian with
respect to c of the rate expression now contains c, which will approach minus
infinity when c approaches zero from the positive side. This may result in poor
convergence, and the max() wrapping will not improve convergence in this case.

The solution for the third case is to linearize the concentration dependence for low
concentrations; that is, to use


r = – kc c  c lim

–1
r = – kcc lim c  = c lim

which results in the desired convergence behavior for low and negative concentrations.
Note, however, that the linearization may result in thermodynamic inconsistencies so
that, for instance, relations like the Nernst equation for the equilibrium potential are
no longer fulfilled. The linearization may also improve convergence of the second case
above.

Theory for Specific Current Distribution Feature Nodes

ELECTROLYTE THEORY
The Electrolyte node defines a current balance in the electrolyte. The domain equation
is:

  il = 0

where il denotes the current density vector. In free electrolyte, there is no source or
sink of charge.

The definition of the current density vector depends on the equation formulation of
the electrolyte charge transport, as discussed above in Domain Equations for Primary

THEORY FOR THE CURRENT DISTRIBUTION INTERFACES | 193

 and Secondary Current Distributions and Domain Equations for Tertiary Current
Distributions Using the Nernst–Planck Equations and Electroneutrality.

• For the Primary and Secondary Current Distribution interfaces, see .

• For the Tertiary Current Distribution, Nernst–Planck interface, see .

POROUS ELECTRODE THEORY

A porous electrode is a mixed material with one electrode phase and one electrolyte
phase. (A special case of a porous electrode is the gas diffusion electrode, as found in
a fuel cell. These contain a gas pore phase which is inert to charge transfer.) To model
a porous electrode we define two separate current balances according to the following
equations:

  i l = Q l and   i s = Q s

In these equations, il denotes the current density vector in the electrolyte, as discussed
above in Domain Equations for Primary and Secondary Current Distributions and
Domain Equations for Tertiary Current Distributions Using the Nernst–Planck
Equations and Electroneutrality.

In addition to the current balances, it is necessary to also formulate mass balance

equations for the species in the electrolyte phase for the tertiary case.

It is also common to used corrected conductivities and diffusion parameter values in a

porous electrode to account for the lowered volume fraction of the conducting phase,
and the longer transport distance due to the tortuosity of the pores.

The current balances in the pore electrolyte and in the electrode matrix contain sources
and sinks according to the charge transfer reactions that take place in the electrode
catalyst. For example, if only one charge transfer reaction takes place in the porous
electrode, the domain equations are the following:

  i l = A v i loc
  i s = – A v i loc

where Av denotes the specific surface area (dimension L2/L3), and iloc the local
current density defines the rate of the charge transfer reactions, for instance according
to the Butler–Volmer equation. For various ways of defining iloc, see Electrode Kinetics
Expressions.

194 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

If the porous electrode is a cathode, then the charge transfer reaction is a source for
the current balance in the electrode, because it receives current from the pore
electrolyte. The charge transfer reaction is then a sink for the current balance in the
pore electrolyte, because the current is transferred from the pore electrolyte to the
electrode in a cathodic reaction.

The corresponding sources and sinks in the current balances that are due to the charge
transfer reactions are also coupled to the material balances for the charged species. This
means that the local current density expression above is also included in the material
balances as a reaction term, Ri, by using Faraday’s laws for each of the species that take
part in charge transfer reactions.

• For the Primary and Secondary Current Distribution interfaces, see .

• For the Tertiary Current Distribution, Nernst–Planck interface, see .

ELECTRODE REACTIONS THEORY

Charge transfer reactions occurring at an interface between an electrode and an
electrolyte domain gives rise to a normal current flux that equals the sum of all reaction
currents according to

il  n =  iloc m
m

is  n = –  iloc m
m

where iloc,m (A/m2) is the current density of the charge transfer electrode reaction of
index m, il the current density vector in the electrolyte and is the current density
vector in the electrode.

POROUS ELECTRODE REACTIONS THEORY

For a porous electrode, the electrode reaction current densities are multiplied by the
surface area to yield a source or sink in the current balance domain equation according
to:

THEORY FOR THE CURRENT DISTRIBUTION INTERFACES | 195

   il =  Av iloc m
m

  is = –  Av iloc m
m

where Av is the specific surface area of the electrocatalyst.

ELECTRODE THEORY
Electron conduction in an is modeled using Ohm’s law. The domain equation is the
following:

  is = 0

where is denotes the current density vector according to:

i s = –  s  s

and where s denotes the electric conductivity and s the potential of the electron
conducting (metal) phase.

ELECTROLYTE CURRENT DENSITY THEORY

An applied current density can be defined as its component perpendicular to the
boundary according to:

i l  n = i n l

The current density can also be defined including all its components:

i l = i l bnd

where ilbnd is a given expression for the current density vector.

196 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

ELECTROLYTE CURRENT THEORY
The boundary condition sets the total current at a given position in the electrolyte
without imposing a current density distribution. The conditions yields a constant
electrolyte potential, along the given boundary, that satisfies the total value of the
current. The boundary condition is a good choice in the middle of a cell with planar
electrodes, where the isopotential level can be a plane (or close to a plane in 3D, or line
in 2D) but where the current density distribution is unknown.

The feature adds one unknown variable, the electrolyte potential, lbnd, along the
boundary. It then adds one additional equation for the total current, which is an
integral over the boundary:

 il  n ds = I n l


The average current density condition imposes the same equation but multiplies the
current density by the area of the boundary to obtain the value of the total current Inl.

ELECTRODE CURRENT THEORY

The adds one unknown variable, the electric potential, sbnd, along the boundary. It
then adds one additional equation for the total current, which is an integral over the
boundary:

 is  n ds = I n s


where

i s = –  s  s

and s denotes the electrode conductivity and s the electric potential. The average
current density condition imposes the same equation but multiplies the current density
by the area of the boundary to obtain the value of the total current, In,s.

If a contact resistance Rc is included, the following current density condition is used:

 s bnd –  s
– n  i s = ---------------------------
-
Rc

THEORY FOR THE CURRENT DISTRIBUTION INTERFACES | 197

 SYMMETRY THEORY
The boundary condition, in the Primary Current Distribution and Secondary Current
Distribution interfaces is identical to the Insulation condition and is expressed
according to the equation below.

ik  n = 0

where ik denotes the current density vector and k = l, s is an index for the electrolyte
and electrode, respectively.

The Symmetry boundary condition for the Tertiary Current Distribution, Nernst–
Planck interface imposes a no-flux condition for the molar flux of species at a
boundary. The condition is expressed as follows:

Ji  n = 0

ELECTRODE CURRENT DENSITY THEORY

An applied can be defined as its component perpendicular to the boundary according
to:

i s  n = i n s

where

i s = –  s  s

and s denotes the electrode conductivity and s the electric potential.

The current density can also be defined including all its components:

i s = i s bnd

where isbnd is a given expression for the current density vector.

ELECTRODE POWER THEORY

The Electrode Power boundary condition is used to specify either the total electrode
power or the average electrode power density drawn from or inserted into an
electrochemical cell at an electrode boundary.

For a total power condition, the boundary electric potential of an electrode is set to a
potential s,bnd, defined by the condition for the total power on the boundary 
according to:

198 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

  s =  s bnd on 

  s bnd – s ground   is  n  dS = Ptotal



where s,ground is the ground potential of the cell, and Ptotal (W) is the power to be
drawn.

If a contact resistance Rc is included, the following current density condition is used:

 s bnd –  s
– n  i s = ---------------------------
-
Rc

For an average power condition, Ptotal is calculated by:

P total = P avg A

where Pavg is the average power density on the boundary, and A is the boundary area.

For a galvanic cell, such as a battery during discharge or a fuel cell, there
is a maximum power level, beyond which a further current increase causes
a lowered output power due to increasing voltage losses. A result of this
is that there can be two existing solutions for the same power setting. In
these cases the choice of initial values determines the final solution.

Electrode Power

ION EXCHANGE MEMBRANE THEORY

Ion-exchange membranes typically contain a polymer matrix with a number of fixed
ionic groups.

Since these charges are fixed, there is no need to explicitly model the transport of these
charges, but when calculating the sum of charges, used in the Nernst–Planck (with
electroneutrality) or the Nernst–Planck–Poisson set of equations, one need to add this
fixed space charge.

For Nernst–Planck with electroneutrality, the electroneutrality condition reads

 fix + F  zi ci = 0

THEORY FOR THE CURRENT DISTRIBUTION INTERFACES | 199

 For the Nernst–Planck–Poisson case, the total space charge density becomes

 v =  fix + F  zi ci
ION EXCHANGE MEMBRANE BOUNDARY THEORY
The electrochemical potential i of a charged species of index i is

 i = RT ln a i +  l z i F

where T(K) is the temperature, R (mol/(J·K)) the molar gas constant, ai is the species
activity, l is the electrolyte potential, zi the species charge, and F(C/mol) is Faraday’s
constant.

At equilibrium the electrochemical potentials on each side of the free electrolyte -

ion-exchange membrane interface are equal.

Setting the species activity to equal the concentration and denoting the liquid
electrolyte phase and an ion-exchange membrane phases as 1 and 2, respectively, the
Donnan potential,  (V), describes the relation between the concentration of a
species, ci (mol/m3), at each side of the boundary and the electrolyte potentials:

c i 1
 l =  l 1 –  l 2 = – -------- ln  ---------
RT
z i F  c i 2

which may be rearranged to

z i F   l 1 –  l 2 
c i 2 = c i 2 exp  – ---------------------------------------
RT

The molar flux of each species in the liquid electrolyte is continuous over the
membrane-liquid interface

n  J i 1 = n  J i 2

Since the total current density is the sum of all species fluxes, times the individual
species charges, the current densities Il in the normal direction n of the
membrane-liquid interface boundary is also continuous:

n  I l 1 = n  I l 2

200 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Reference
1. J. O’M. Bockris, A.K.N. Reddy, and M. Gamboa-Aldeco, Modern Electrochemistry,
vol. 2A, 2nd ed., ch. 7, sec. 7.6, Kluwer Academic/Plenum Press, New York, 2000.

THEORY FOR THE CURRENT DISTRIBUTION INTERFACES | 201

 Theory for the Concentrated
Electrolyte Transport Interface
In this chapter:

• Dependent Variables, Species, and Electrolyte Components

• Volumetric Equations of State
• Chemical Potentials and the Stefan–Maxwell–Onsager Relations
• The Reference Electrode
• Reference

This chapter is serves as an introduction of the mathematical framework used for

defining the transport equations in the Concentrated Electrolyte Transport interface. For
a complete description of the theory, including matrix transformations between the
species and electrolyte component basis, see Ref. 1.

Dependent Variables, Species, and Electrolyte Components

The Concentrated Electrolyte Transport interface defines the transport of a total number
of n either positively (cations), negatively (anions), or neutrally charged species in an
electrolyte phase. At least one cation and one anion must be defined.

The mathematical framework makes use of a number of electrolyte components that

are constructed by forming a set of n  1 neutral salts based on the n electrolyte
species. It should be noted that the purpose of the neutral electrolyte components is
to form a basis for the mathematical framework only; they do not necessarily have any
physical correspondence as a real salt in nature. As an example, for an electrolyte
composed of the four species Mg2+, H+, OH-, and H2O, the set Mg-(OH)2, H-OH,
and H2O form the corresponding electrolyte component basis.

The mathematical framework expresses variables either in using the individual

electrolyte species as a basis or the electrolyte components. The index i, j, or k is used
generally to define the index of either a species or an electrolyte component. Symbols
referring to an electrolyte component are denoted by a tilde (~) on top of the greek or
roman character, whereas indexed symbols without the tilde refer to a species.

Transformations between the species and electrolyte component basis are made by the
use of a (n  n) transformation matrix Z, which consists n  1 rows that define the

202 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

stoichiometric coefficients of the species for each electroneutral salt, and with an nth
row defining the individual species charges, divided by the norm of the charge vector.

For the Mg2+, OH-, H+, and H2O example, the Z matrix would look like

1 0 2 0
0 1 1 0
Z = 0 0 0 1 (4-4)
2- 1- – 1-
------ ------ ------ 0
6 6 6

˜ (kg/mol) of the electrolyte component salts can

For instance, the molar weights M i
be computed from the first n  1 rows of the column resulting from the matrix
multiplication

˜ = ZM
M (4-5)

where M  M1 M2 … MnT is a column vector of the defined by the individual species
molar masses Mi (kg/mol).

Also the inverse transpose matrix ZT is used for various variable basis transformations.

The electroneutrality condition, applied locally at all times, defines the sum of all
charges to equal zero

 zi ci = 0 (4-6)

where zi (dimensionless) are the individual species charges, and ci the molar
concentrations (mol/m3).

Electroneutrality in combination with a volumetric equation of state (see next section),

assuming all major constituents of the electrolyte to be included in the electrolyte
model, allows for reducing the number of dependent variables by 2, resulting in an
equation system consisting of n  1 electrolyte component fraction y˜i (dimensionless)
and one electrolyte potential l (V) dependent variables.

Volumetric Equations of State

The interplay between molar masses, molar volumes and concentrations will define the
total concentration cT (mol/m3) of species, and the density  (kg/m3).

THEORY FOR THE CONCENTRATED ELECTROLYTE TRANSPORT INTERFACE | 203

 ˜
If the average molar volumes V 3
m i (m /mol) of the electrolyte components are
known, the total concentration of species may be defined as

1
c T = ------------------------
- (4-7)
˜ ˜
 yi V m i
i

and the density as

 =  Mi ci (4-8)

Alternatively, one may define the density as a function of some of the composition
variables, and define the total concentration as

 -
c T = ------------------- (4-9)
 y˜i M ˜
i

Flux Definitions and Material Balances

The total flux Ni (mol/m2/s) of each individual species is defined as

Ni = ji + ci u (4-10)

where the first term ji represents the diffusive (and migrative) flux and the second term
the convective flux, where u (m/s) is the mass-averaged velocity of the electrolyte.

The electrolyte current density Il (A/m2) is defined as

il =  zi FNi (4-11)

The material balances for the individual species are defined as

 l c i
- +   Ni = Ri
----------- (4-12)
t

where l(dimensionless) is the electrolyte volume fraction and Ri the molar source
term.

Summing all individual material balances and multiplying by the corresponding species
charges results in the charge balance equation for the electrolyte

204 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

   Il =  zi FRi (4-13)

When formulating the material balances in the interface, it is assumed that the
continuity equation for the fluid (which is not solved for by the interface) is fulfilled

 l 
t
- +    u  =
----------  Mi Ri = Qm (4-14)

This allows for reformulating the individual material balances as


c i Q m – u   –  l ------
t-
 l ------- +   j i + u  c i = R i – ------------------------------------------------ (4-15)
t 

which is the actual (nonconservative) form of the material balance equation used
internally by the interface. This equation form avoids accuracy issues related to
evaluating derivatives of the velocity variable.

Chemical Potentials and the Stefan–Maxwell–Onsager Relations

Gradients of the electrochemical potentials of the electrolyte species act as driving
forces for diffusion and migration in the transport model.

The electrochemical potential i (J/mol) of the individual species are defined as

0
 i =  i + RTlnx i + RTln i + Fz i  (4-16)
0
where R is the molar gas constant (8.31 J/mol/K),  i is the electrochemical potential
with respect to some reference state, T is the temperature (K), xi is the molar fraction,
the i the species activity coefficient (dimensionless) and  (V) is the solution
potential.

The molar fractions are defined from the species concentrations as

c
x i = -----i- (4-17)
cT

For ideal conditions, the species activity coefficients equal unity. For nonideal
solutions, two issues experimental arise in relation to the definition of individual
species activity coefficients. The first is that, since charge separation is virtually
impossible to achieve in a solution, it is hard to assess the chemical activity of individual
ions experimentally. The second is that, since all major species in the solution are

THEORY FOR THE CONCENTRATED ELECTROLYTE TRANSPORT INTERFACE | 205

 accounted for, one cannot alter the concentration of a single species (or electrolyte salt)
with out changing the concentration of at least one other compound. To circumvent
these issues, activity coefficients are defined in the form of (n  1)(n  2)2
˜
electroneutral component Darken factors  i j (dimensionless) in an electrolyte
component basis as

˜  ln̃ i
 i j = ̃ i  ------------- (4-18)
 ỹ j  T p ỹ l  j k

where the derivative should be made keeping the temperature, pressure and all
electrolyte components but the ones with index j and k constant. In the Concentrated
Electrolyte Transport interface, the index k equals the index of the electrolyte
component taken from the molar fraction constraint.

Using the Stefan–Maxwell–Onsager relations, the gradient of the chemical potentials

are related to the species fluxes as

n
RT
–  c-------------
T D ij
- jj if i  j

–  i = j=1 (4-19)
n n
RT ck
 --------
c T j  D ik c j
j -------------- if i  j
j=1 ki

where Dij (m2/s) are the binary Maxwell–Stefan diffusivity coefficients. The above
equation system is transformed and inverted in order express the electrolyte
component fluxes in terms of the gradients of the electrolyte component chemical
potentials.

In porous media the binary diffusivities above may be multiplied by an efficiency factor
to accommodate for the lowered effective diffusivity due to the lower available volume
for transport and tortuosity (topology) effects of the porous matrix.

The Reference Electrode

In order to relate the electrolyte phase potential to an experimentally measurable
voltage, a reference electrode, defined by it’s reaction stoichiometric coefficients, is
deployed in the model using the generic reaction formula

 si Xi +  s'j Yj
zi -
= n e ref e (4-20)

206 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

where si and Xi define the stoichiometric coefficients and species of the electrolyte
phase, and s’j and Y’j define the stoichiometric coefficients and species of an adjacent
(electrode) phase, respectively, and ne,ref the number of electrons participating in the
reaction.

The gradient of the electrolyte phase potential, used for defining the chemical potential
is defined as

s̃ i 
 =  --------------------
– n e ref F
- +  (4-21)

The stoichiometry of the reference electrode can also be used to define a steady-state
convective velocity of the electrolyte for the case when all electrode reactions in a cell
have the same stoichiometry as the reference electrode. For this case, the convective
velocity may be calculated as

–Il
u = ---------------------
n e ref F  si Mi (4-22)

Reference
1. A. Van-Bunt, P.E. Farrel, and C.W. Monroe, “Structural electroneutrality in
Onsager–Stefan–Maxwell transport with charged species,” Electrochimica Acta,
vol. 441, p. 141769, 2023, [Link]

THEORY FOR THE CONCENTRATED ELECTROLYTE TRANSPORT INTERFACE | 207

 Theory for Electrochemical Heat
Sources
Irreversible voltage losses in an electrochemical cell can occur due to the following
phenomena:

• Charge transport in the electrolyte (Joule heating)

• Charge transport in the solid conductor materials (Joule heating)
• Activation overpotentials in the electrode reactions
• Heat of mixing

In addition, reversible heat sources and sinks can appear due to the entropy changes in
the electrode reactions.

Most Electrochemistry interfaces define and announce heat source variables that can
be used by the General Source and the Boundary Heat source nodes in the Heat
Transfer interfaces.

An alternative way to couple an electrochemical interface to a heat transfer interface is

by the use of the Multiphysics nodes that become available in the model builder tree if
there is both a Heat Transfer interface and an Electrochemistry interface available in
the same component.

The Electrochemical Heating multiphysics node defines a domain heat source in the
heat transfer interface, based on the sum of irreversible (Joule heating and activation
losses) and reversible heat in the electrochemistry interface.

You can also use the heat source variables defined by the electrochemistry interfaces
when setting up manual heat couplings between different components in a model. For
instance if you are using a 1D electrochemical cell model to calculate an average heat
source in a 3D heat transfer model. The names of the heat source variables are [Link]
(domain, Joule heating and porous electrode reactions) and [Link] (boundary,
electrode surface reactions), where xxx is the electrochemistry interface identifier.

• Electrochemistry Interfaces
• Multiphysics Coupling Nodes

208 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

The First Law of Thermodynamics
As a basis for deriving the heat source expressions needed for modeling electrochemical
cells, we start from the first law of thermodynamics for a small system (volume
element), stating that

U = Q – W (4-23)

where U (J) is differential change of the internal energy, Q (J) is the differential
amount of heat added to the system, and W (J) is the differential work performed by
the system element on the surroundings.

Rearranging we may write

Q = U – W (4-24)

It should be noted that in the absence of pressure or volume changes, which are often
neglected in electrochemical models, the differential internal energy equals the
differential enthalpy H (J). In addition, thermodynamic work is usually not involved
in electrochemical reactions. We may hence simplify the above to

Q = H (4-25)

In order to conserve heat, the above differential heat added to the system equals the
added heat drawn from the surroundings.

Electrochemical Contributions to The Heat Equation

The heat equation solved for by the Heat Transfer interface typically has the following
form

T
C p ------- +   q = Q (4-26)
t

where Q is the domain heat source (W/m3).

The electrochemistry interfaces typically solve for the transport and electrochemical
reactions of a multicomponent system comprising a number of different species with
index i. Adding the additional contributions from the multicomponent system to the
heat transfer equation, and assuming no other additional heat sources (Q = 0), results
in

T c i
C p ------- +    q + H i N i  = – H i ------- (4-27)
t t

THEORY FOR ELECTROCHEMICAL HEAT SOURCES | 209

 where Hi is the partial molar enthalpy (J/mol), Ni is the species flux (mol/m2/s), and
ci is the concentration.

By treating the contributions from the electrochemical multicomponent system as a

heat source defined as

 c i
Q = –
     Hi Ni  + Hi ------
t 
- (4-28)
i

Equation 4-26 can be made to represent the heat transfer equation of a the
multi-component electrochemical system.

Writing the general material balance for each species as

c
-------i +   N i = R i (4-29)
t

where Ri (mol/m3/s) is the local rate production source, the heat source expression
may be simplified to

 
Q = –
  Hi  Ni + Hi Ri (4-30)
i

where the second term may be used to define the reaction enthalpies of non-Faradaic
reactions (for Faradaic reactions, see below).

The derivations below follow a simplified approach along the principles lined out in
Ref. 1. For deepened discussion on thermal effects in electrochemical cells, the reader
is referred to Ref. 1 to Ref. 4.

In the derivations, the partial molar species enthalpies are defined as

H i = H i + z i F (4-31)

where H i is the potential-independent partial molar enthalpy (J/mol) and  the

electrostatic potential (V). Apart for when heat of mixing effects are present, H i is
constant.

Joule Heating Due to Charge Transport

To derive the a heat source for Joule heating we assume a domain where there are no
species sources, nor heat of mixing effects present.

210 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

As charged species travel in an electric field, the enthalpy changes according to the
gradient in the electric potential  (V)

H i = z i F (4-32)

where zi is the species charge.

For a current conductor, the only species carrying charge is the electron, with a
negative unit charge. The electron flux is related to the current density is as

i
N e = – ---s- (4-33)
F

resulting in

Q s = –  i s   s  (4-34)

where the subscript s denotes the electron conducting phase.

For the electrolyte phase, the heat source becomes

Ql = –    zi Fl   Ni = –  l   zi FNi = –  l  i l (4-35)

i i

For a homogenized porous electrode domain, the Joule heating term for both phases
are added together:

Q JH = –  i s   s + i l   l  (4-36)

Heating Due to Electrochemical Reactions

A heat source variable for a boundary defining an electrode-electrolyte interface
(electrode surface) may be derived by considering the net enthalpy flux into the
interface

 H  i H i
q = –  – ------e- +
 F  -----------
nF  loc
- i (4-37)
i

where q is the boundary heat source (W/m2), n and i are the number of electrons
participating in the reaction and the stoichiometric coefficients of the reaction,
respectively, and iloc local current density at the electrode surface. Subscript e above
denotes the electron, and i the electrolyte phase species.

THEORY FOR ELECTROCHEMICAL HEAT SOURCES | 211

 Considering the overall charge neutrality of the electrode reaction

n = –  i zi (4-38)
i

and rearranging, the above relation equates to

H
q =  – --------- –   s –  l  i loc (4-39)
nF

where H is the potential-independent net partial molar enthalpy change, defined as

H =  i Hi – nHe (4-40)
i

Defining the thermoneutral voltage of the reaction as

H-
E therm = – --------
nF

the heat source term can be also be written as

q =  E therm –   s –  l  i loc (4-41)

Alternatively, using the relation

G = H – TS

where G is the Gibbs free energy of the reaction, and where S is the net entropy
change, and defining the equilibrium potential as

E eq = – G
-------- (4-42)
nF

the boundary heat source may also be written as

E eq
q =   + T ------------ i loc (4-43)
 T 

where the overpotential is defined as

 =  s –  l – E eq (4-44)

and the temperature derivative of the equilibrium potential is defined as

212 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

 E eq
------------ = S
-------- (4-45)
T nF

The total heat source due to the electrochemical reactions, qEC, for an electrode
surface is the sum of all individual heat sources of the electrode reactions according to

q EC =  qm
m

where m is the index of the electrode reaction.

For a homogenized formulation of a porous electrode, the domain heat sources due
to electrochemical reactions can be obtained by multiplying by the internal surface area
av,m (m2/m3). For a porous electrode, the electrochemical sources are hence summed
up according to

Q EC =  av m Qm
m

Heating Due to Heat of Mixing

If the partial molar enthalpy varies with the local concentration of the species, there are
also heat of mixing sources associated with the concentration gradients. These heat of
mixing effects are typically small (zero for ideal gases) and are generally not included
in the Electrochemistry interfaces.

Total Overpotential Calculation for Concentration Dependent

Kinetics
When using the Concentration dependent kinetics option in the Porous Electrode
Reaction and Electrode Reaction nodes, the equilibrium potential is typically a
constant referring to a reference state, and the concentration dependence of the
kinetics is accounted for only in the pre-exponential terms according to

 a m F m –  c m F m
i loc m = i 0 m  C R m exp  ------------------------- – C O m exp  ---------------------------- 
RT RT

where

 m =  s –  l – E eq m

THEORY FOR ELECTROCHEMICAL HEAT SOURCES | 213

 A result of this is that, when the pre-exponential factors CO,m and CR,m differ due to
concentration gradients, there can at zero activation overpotential be a net
charge-transfer current density flowing over the electrode–electrolyte interface. This
local current density should obviously give rise to a heat source. The solution is to add
a concentration overpotential term to the total overpotential when calculating the heat
sources.

When using the Concentration dependent kinetics option, the overpotential, used in
the irreversible heat term above, is calculated as follows:

 m tot =  s –  l – E eq m +  m conc

where

RT C O m
 m conc = ------------ ln -------------
-
n m F C R m

If the number of electrons is not available,

n m =  aꞏ  m +  c m

is assumed.

References
1. John Newman, “Thermoelectric Effects in Electrochemical Systems,” Ind. Eng.
Chem. Res., vol. 34, no. 10 pp. 3208–3216, 1995

2. Bernardi and others, “A General Energy Balance for Battery Systems,”

J. Electrochem. Soc., vol. 132, pp. 5–12, 1985.

3. Thomas and others, “Thermal Modeling of Porous Insertion Electrodes,”

J. Electrochem. Soc., vol. 150, pp. A176–A192, 2003.

4. J. Newman and K.E. Thomas-Alyea, Electrochemical Systems, Chapter 13 Thermal

Effects, Wiley, 2012.

214 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Theory for the Electrode, Shell
Interface
The Electrode, Shell Interfacedefines electrode current conduction in a thin shell on a
boundary. The electrolyte current distribution in the adjacent domain is typically
solved for by a separate physics interface.

In this section:

• Governing Equations
• Coupling to Other Physics Interfaces

Governing Equations
The Electrode, Shell interface solves for the electric potential s (SI unit: V) on a
boundary, using the following governing equation:

 T   si s  = – i n (4-46)

where T is the tangential gradient operator, s (SI unit: m) is the electrode layer
thickness, and in (SI unit: A/m2) are the sum of all currents flowing out from the
electrode (in the normal direction to the boundary). Furthermore, is (SI unit: A/m2)
is the tangential current density vector along the electrode boundary, defined as

i s = – T  s  s (4-47)

where s is the electric conductivity (SI unit: S/m). The next section discusses
Coupling to Other Physics Interfaces.

Coupling to Other Physics Interfaces

ELECTRODE POTENTIAL IN OTHER PHYSICS INTERFACES

Typically the boundary electric potential in the Electrode Surface node is set to the
electric potential in the Electrode, Shell interface.

ELECTRODE REACTION CURRENTS

The sum of currents can be coupled to other electrochemistry interfaces, calculating
in as

THEORY FOR THE ELECTRODE, SHELL INTERFACE | 215

 i n = i tot

where itot (SI unit: A/m2) is the sum of all electrode currents in the coupled
Electrochemistry interface.

ELECTRODE LAYER THICKNESS

The electrode thickness, s, can typically be coupled to an External Depositing
Electrode (External Corroding Electrode) node in the Electrodeposition, Moving
Mesh (Corrosion, Moving Mesh) interface, in that case s is calculated as

s = s 0 + s tot (4-48)

where s0 is the initial electrode layer thickness, and stot is the electrode thickness
change, calculated by the coupled Electrochemistry interface.

HEAT SOURCE
The electron conduction gives rise to a Joule heating source QH (SI unit: W/m2)
according to

Q H = – si s   t  s

216 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Theory for Electroanalysis
The theory related to the Electroanalysis charge conservation model in The Tertiary
Current Distribution, Nernst–Planck Interface (tcd) is described in this section:

• Electroanalytical Methods
• Supporting Electrolyte
• Domain Equations for the Electroanalysis Case
• Electrode Boundary Conditions in the Electroanalysis Model
• The Electroanalytical Butler–Volmer Equation
• Counter Electrodes and Overall Charge Balance

Electroanalytical Methods
Electroanalysis is the science of quantitative electrochemical measurement of the
composition or properties of a chemical system. Common electroanalytical methods
include: (cyclic) voltammetry, (chrono)amperometry, potentiometry, coulometry, and
electrochemical impedance spectroscopy (EIS). These methods are experiments
performed either in a static electrolyte solution or in an electrolyte solution subject to
a forced fluid flow. The results sought in electroanalysis include:

• Measurement of the concentration of an analyte — for example, glucose, or a

gas-phase species that dissolves in the electrolyte solution. This technique is often
referred to as electrochemical sensing. The most common sensing method is by
recording the concentration-dependent current due to an analyte-specific reaction:
this is called amperometric sensing.
• Measurement of the kinetic and/or thermodynamic parameters describing an
electrochemical reaction on a given electrode surface.
• Measurement of the transport properties of an electroactive chemical species.
• Investigation of the mechanism of an electrochemical reaction and the further
reactivity of the electrochemically generated chemical species.

Supporting Electrolyte
When performing electroanalytical experiments, it is conventional to add a large
quantity of inert salt to the solution — this artificially added salt is called supporting

THEORY FOR ELECTROANALYSIS | 217

 electrolyte. The purpose of the supporting electrolyte is to increase the conductivity of
the solution, and hence to eliminate the electric field from the electrolyte.

A negligible electric field provides two advantages for electroanalysis:

• The voltage due to the resistance of the electrolyte when the cell draws current
(“ohmic drop”) is minimal. Therefore, the potential difference applied across the
electrochemical cell is localized at the electrode–electrolyte interfaces, and so the
activation overpotential perceived by the redox couple at this interface is almost
exactly proportional to the applied cell voltage. The kinetic behavior of the
electrochemical cell then has no explicit dependence on the magnitude of the drawn
current.
• The contribution of migration to the transport of charged chemical species is
negligible compared to the contribution of diffusion (and of convection, in a forced
flow). Therefore the transport properties of the system are linearized, and they do
not depend on the magnitude of the drawn current.

These properties greatly simplify the design and analysis of electroanalytical

experiments. Therefore, the use of a supporting electrolyte is very common in
electrochemical sensing and electroanalysis.

The amount of supporting electrolyte required for the assumption of zero

electric field to be sound is dependent on the system. Typically it becomes
valid when the concentration of inert salt exceeds the concentration of
analyte by a factor of 10–100, but this is not always the case.

Even for the conductivities of electrolyte solutions in the presence of excess supporting
electrolyte, the electric field is not negligible if significant current density is drawn.
Electroanalysis typically draws small currents because the purpose is measurement. In
processes where an electrochemical reaction is driven — such as electrolysis,
electrodeposition, batteries, and fuel cells — current densities are typically much larger,
so that the desired extent of reaction is achieved in a reasonable time. Under these
conditions, significant electric fields are likely and other charge conservation models
should be used instead of the Electroanalysis option.

Domain Equations for the Electroanalysis Case

The Electroanalysis option assumes that the electric field is zero, and so the electrolyte
potential is constant. Since you can arbitrarily choose to ground the electrolyte
potential at any point, set it to l  0.

218 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

This is equivalent to the assumption of infinite electrolyte conductivity. Therefore the
Electroanalysis option does not solve for charge transfer within domains, because
current density is not meaningfully defined within the electrolyte.

The transport of chemical species in an electrolyte solution obeys the Nernst–Planck

equation for the flux of species i:

N i = – D i c i – z i u m i Fc i  l + c i u = J i + c i u

Here Ji denotes the molar flux relative to the convective transport.

At zero electric field, this reduces to:

N i = – D i c i + c i u (4-49)

where the only contributions to the flux of a chemical species are from diffusion and
convection respectively. In the absence of convection (no fluid flow, u  0), this is also
known as Fick’s first law of diffusion:

N i = J i = – D i c i

A mass balance also applies to each chemical species:

c i
+   N i = R i tot (4-50)
t

The Electroanalysis charge conservation model solves Equation 4-49 and

Equation 4-50 for the unknown concentrations of each chemical species under
analysis.

This combination is often written as a single equation for the unknown ci. For zero
convection, zero reaction, and a constant diffusion coefficient, the domain equation is:

c i 2
= Di  ci (4-51)
t

THEORY FOR ELECTROANALYSIS | 219

 Equation 4-51 is Fick’s second law of diffusion.

The Electroanalysis charge conservation model is not suitable to explicitly

model the transport of the supporting electrolyte, since the migration of
the supporting electrolyte is always its dominant mode of mass transport.
For a coupled model including the chemical species of all charge-carrying
species, use the electroneutrality model in Tertiary Current Distribution,
Nernst–Planck interface.

You can also include additional chemical species and reactions that are not
involved in the electrochemical reaction.

• Electrode Potentials and Reference Electrodes

• The Tertiary Current Distribution, Nernst–Planck Interface

Electrode Boundary Conditions in the Electroanalysis Model

The activation overpotential at an electrode–electrolyte interface with respect to the
electrode reaction m is defined as:

 m =  s –  l – E eq m

This is the potential difference perceived by a redox couple, measured against the
equilibrium potential of the couple; it provides the thermodynamic driving force for
an electrochemical reaction by faradaic charge transfer between the electrode and the
electrolyte domains.

Subject to the approximation of constant electrolyte potential (l  0), this equation
reduces to:

 m =  s – E eq m

The activation overpotential is independent of the properties of the adjacent

electrolyte. It only depends on the applied electrode potential s and the equilibrium
potential of the redox couple, where both are measured against a common reference
potential.

220 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

The current density due to an electrode reaction at a point on an electrode surface is
computed using an electrochemical rate expression. For most practical electroanalytical
applications, the rate depends on the local concentration which varies during the study,
and so the most relevant expression is The Electroanalytical Butler–Volmer Equation.

Electrode Kinetics Expressions

The flux Ni of the chemical species i (SI unit: mol/m2) across an electrode surface
depends on the current densities im associated with the electrode reactions m
according to Faraday’s laws of electrolysis. These can be written as:

 i m i m
Ni =  -----------------
nm F
- (4-52)
m

where i,m is the stoichiometric coefficient of species i with respect to reaction m (in
the reductive direction), and nm is the number of transferred electrons. F is the
Faraday constant, which is the charge on a mole of electrons (96485.3365 C/mol).

The surface flux is applied as a boundary condition:

Ni = Ni  n (4-53)

Equation 4-52 and Equation 4-53 constitute the coupling between charge balance
and mass balance. This coupling only applies at the electrode–electrolyte interface,
which is a boundary to the domain where the electroanalysis charge conservation
model solves for chemical species transport.

The total current density is the sum of faradaic (electrode reaction) components and
nonfaradaic components (inf) such as current due to Electrode Kinetics Expressions:

i total =  iloc m + inf

m

The experimentally measurable total current I (SI unit: A) drawn at an electrode can
be computed by integration of the local current density (SI unit: A/m2) across the
electrode area:

THEORY FOR ELECTROANALYSIS | 221

 I total =  itotal ds


Domain Equations for the Electroanalysis Case

The Electroanalytical Butler–Volmer Equation

A one-electron electrochemical reaction between two solution-phase species can be
–
written as a reduction Ox + e  Red .

Ox and Red represent the oxidized and reduced forms of the chemical species,
respectively.

The most general equation to describe the rate of this reaction as it proceeds at an
electrode surface is the electroanalytical Butler–Volmer equation:

 a F –  c F
i loc = k 0 F  c Red exp  --------------- – c Ox exp  -----------------  (4-54)
  RT   RT  

where k0 is the heterogeneous rate constant (SI unit: m/s) and c is the (cathodic)
transfer coefficient (dimensionless). For a one-electron reduction, the anodic and
cathodic transfer coefficients are related as a  c  1.

When the current is zero, the electroanalytical Butler–Volmer equation can be

rearranged to the thermodynamic Nernst equation relating the equilibrium
concentrations of the reacting species:

– FE eq
c Red = c Ox exp  ----------------
RT

Where the flux of the reacting species is negligible compared to the concentration of
these species, the concentrations are roughly constant (cRed ~ cOx ~ c). This converts
Equation 4-54 into the Butler–Volmer equation written in terms of an exchange
current density i0 (SI unit: A/m2):

 a F –  c F
i loc = i 0  exp  --------------- – exp  ----------------- 
RT RT

The exchange current density i0 (SI unit: A/m2) is then related to the heterogeneous
rate constant as i0  k0Fc.

222 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

Counter Electrodes and Overall Charge Balance
The Electroanalysis charge conservation model does not solve explicitly for the
electrolyte potential in the domains. This has the implication that an overall charge
balance of a modeled cell does not follow naturally as a result of the domain potential
equation (as is the case for instance when solving for Ohm’s law in the Secondary
Current Distribution interface).

In order to model a charge balance of a cell with an arbitrary number of working

electrodes and counter electrodes, the solution is to add an extra degree of freedom
for the counter electrode potential, s, CE, and then solve for this potential to fulfill the
following current balance over all electrode surfaces:

 s CE :  i tot dS = 0 (4-55)
electrodes

The counter electrode potential is used in the overpotential, , in the kinetics

expression(s) on the counter electrode.

 m =  s CE – E eq m (4-56)

Note that only one counter electrode potential degree of freedom is added in the
model, regardless of the number of counter electrodes that are active.

Domain Equations for the Electroanalysis Case

THEORY FOR ELECTROANALYSIS | 223

 Electrode Potentials and Reference
Electrodes
In all the Electrochemistry branch interfaces, the dependent potential variables are s
(SI unit: V), the electric potential of the electrode phase (the electron conductor, such
as metal), andl (SI unit: V), the potential of the electrolyte phase (ion conductor).

Typically the kinetics of the electrochemical reactions are defined using the
overpotential,  (SI unit: V), defined as

 =  s –  l – E eq (4-57)

where Eeq (SI unit: V) is the equilibrium potential. If it is to apply for all
overpotentials, a general kinetic expression for an electrode reaction must be set up so
that the charge-transfer current over the electrolyte-electrode interface is zero for zero
overpotential (equilibrium conditions).

An implication of Equation 4-57 is that it is the potential difference, s  l, that

governs the kinetics, not the absolute individual values of s andl. A global change
in the reference for both potentials has no impact on the electrode kinetics. As a result
of this, the potentials have to be “boot-strapped” in a model in some way, typically by
making an arbitrary choice of electric ground — for example, on an external boundary
— in order to ensure that there is a unique solution to the problem.

In this section:

• Reference Electrodes
• Boundary Conditions Using Reference Electrode Potentials
• Nodes for Handling Electrode Potentials and Reference Electrodes

Reference Electrodes
In experimental electrochemistry, it is common to use reference electrodes when
controlling current or voltage with a potentiostat. Potential differences in the system
are recorded with respect to the equilibrium potential of the redox couple at the
reference electrode. A good reference electrode is designed so that no net charge
transfer takes place at its electrode-electrolyte interface. Then the overpotential of the
reference is zero, so:

224 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

  s ref =  l + E eq ref (4-58)

where s,ref (SI unit: V) is the electric potential of the reference electrode and Eeq,ref
(SI unit: V) is the equilibrium potential of the reference electrode reaction.

The electric potentials of the electrodes in the electrochemical cell can then be defined
with respect to the reference electrode according to:

E vs ref =  s –  s ref (4-59)

where Evs,ref (SI unit: V) is the electrode potential versus the reference potential.

It is important to realize that the presence of an ideal reference electrode in the system
has no impact on the physics; the only purpose of the reference electrode is to define
a stable reference point for the potential levels.

Boundary Conditions Using Reference Electrode Potentials

The set of boundary conditions defined in a model of an electrochemical cell needs to
ensure that no current flows through the reference electrode. A constraint for the
potential on the actual reference electrode is not suitable in general since it can create
unwanted local sources or sinks of current.

When modeling the control of an electrochemical cell by a potentiostat, which

monitors and controls the potential of an electrodes versus a reference electrode, the
electric potential conditions should be applied to the boundaries over which the
current flows in the cell. One electrode (the counter electrode) is typically grounded,
whereas an electric potential condition is set on all other (working) electrode
boundaries according to

 s bnd = E vs ref +  s ref (4-60)

where s,bnd (SI unit: V) is the applied electric boundary potential on the electrode.

Nodes for Handling Electrode Potentials and Reference Electrodes

The node is a point feature applicable to electrolyte domains (in all the
Electrochemistry interfaces). It defines a global electric reference potential s,ref (SI
unit: V) according to Equation 4-58. The s,ref can be used in the Electrode Potential
feature for setting the electric potential of an electrode boundary with respect to the
reference potential. The variable Evs,ref (SI unit: V), according to Equation 4-59, is
also defined in all electrode domains.

ELECTRODE POTENTIALS AND REFERENCE ELECTRODES | 225

 The node is a point feature applicable to electrode domains (in all the
Electrochemistry interfaces). It defines a global electric reference potential s,ref (SI
unit: V) equal to the dependent variable s (SI unit: V) in the point. The s,ref can be
used in the Electrode Potential feature for setting the electric potential of an electrode
boundary with respect to the reference potential.

Whenever a s,ref is defined, the variable Evs,ref (SI unit: V), according to
Equation 4-59, is also defined in all electrode domains.

The is a boundary condition feature, applicable to external boundaries of electrode

domains (in all the Electrochemistry interfaces). It sets a boundary condition for the
electric potential with respect to a defined reference potential according to
Equation 4-60. Electric potentials defined by the Electric Reference Potential and
Reference Electrode point feature can be used as input when specifying the Electrode
Potential feature.

An Electrode Potential alternative in the Boundary Condition settings list is also

available for the Electrode Surface node.

The variable Evs,ref (SI unit: V), according to Equation 4-59, is also defined on these
features.

226 | CHAPTER 4: ELECTROCHEMISTRY INTERFACES

 5

Chemical Species Transport Interfaces

This chapter describes the physics interfaces found under the Chemical Species
Transport branch ( ).

In this chapter:

• The Transport of Diluted Species Interface

• The Transport of Diluted Species in Porous Media Interface
• The Transport of Concentrated Species Interface
• The Transport of Concentrated Species in Porous Media Interface
• The Chemistry Interface
• The Nernst–Planck–Poisson Equations Interface
• The Electrophoretic Transport Interface
• The Surface Reactions Interface
• The Reacting Flow Interfaces
• The Reacting Flow in Porous Media Interfaces
• The Nonisothermal Reacting Flow Multiphysics Interfaces

The theory for all the interfaces is also discussed at the end of the chapter and links
are included with the interfaces.

227
 T he T r a ns po r t of D i l u t ed S p ec i es
Interface
The Transport of Diluted Species (tds) interface ( ), found under the Chemical Species
Transport branch ( ), is used to calculate the concentration field of a dilute solute in
a solvent. Transport and reactions of the species dissolved in a gas, liquid, or solid can
be handled with this interface. The driving forces for transport can be diffusion by
Fick’s law, convection when coupled to a flow field, and migration, when coupled to
an electric field.

The interface supports simulation of transport by convection and diffusion in 1D, 2D,
and 3D as well as for axisymmetric components in 1D and 2D. The dependent variable
is the molar concentration, c. Modeling multiple species transport is possible, whereby
the physics interface solves for the molar concentration, ci, of each species i.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is tds.

DOMAIN SELECTION
If any part of the model geometry should not partake in the mass transfer model,
remove that part from the selection list.

OUT-OF-PLANE THICKNESS
For 2D components, the Thickness field (default value: 1 m) defines a parameter for the
thickness of the geometry perpendicular to the two-dimensional cross-section. Both
constant and varying thicknesses are supported. The value of this parameter is used,
among other things, to automatically calculate molar flow rates from the total molar
flux.

228 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

CROSS-SECTIONAL AREA
For 1D components, enter a Cross-sectional area Ac (SI unit: m2) to define a parameter
for the area of the geometry perpendicular to the 1D component. Both constant and
varying areas are supported. The value of this parameter is used, among other things,
to automatically calculate molar flow rates from the total molar flux. The default is 1
m2.

TRANSPORT MECHANISMS
Mass transport due to diffusion is always included. Use the checkboxes available under
Additional transport mechanisms to control other transport mechanisms.

Note: Some of the additional transport mechanisms listed below are only available in
certain products. For details see [Link]

• By default, the Convection checkbox is selected. Clear the checkbox to disable

convective transport.
• Select the Migration in electric field checkbox to activate transport of ionic species in
an electric field. See further the theory section Adding Transport Through
Migration.

Transport of ionic species in an electric field is available in a limited set of

add-on products. See [Link]
specifications/ for more details on availability.

Mass Transport in Porous Media

The Mass transport in porous media checkbox activates functionality specific to species
transport in porous media:

• Porous Medium
• Unsaturated Porous Medium
• Porous Electrode Coupling
• Volatilization
• Species Source

Note: Mass transport in porous media is only available in a limited set of

add-on products. See [Link]
specifications/ for more details on availability.

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 229

 CONSISTENT STABILIZATION
To display this sections, click the Show button ( ) and select Stabilization.

• When the Crosswind diffusion checkbox is selected, a weak term that reduces spurious
oscillations is added to the transport equation. The resulting equation system is
always nonlinear. There are two options for the Crosswind diffusion type:
- Do Carmo and Galeão — the default option. This type of crosswind diffusion
reduces undershoots and overshoots to a minimum but can in rare cases give
equation systems that are difficult to fully converge.
- Codina — this option is less diffusive compared to the Do Carmo and Galeão
option but can result in more undershoots and overshoots. It is also less effective
for anisotropic meshes. The Codina option activates a text field for the Lower
gradient limit glim. Its defaults to 0.1[mol/m^3]/[Link], where [Link]
is the local element size.
• For both consistent stabilization methods, select an Equation residual. Approximate
residual is the default and means that derivatives of the diffusion tensor components
are neglected. This setting is usually accurate enough and is computationally faster.
If required, select Full residual instead.

INCONSISTENT STABILIZATION
To display this section, click the Show button ( ) and select Stabilization. By default,
the Isotropic diffusion checkbox is not selected, because this type of stabilization adds
artificial diffusion and affects the accuracy of the original problem. However, this
option can be used to get a good initial guess for under resolved problems.

ADVANCED SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.
Normally these settings do not need to be changed. Select a Material Balance Form —
Nonconservative (the default) or Conservative. The conservative formulation should be
used for compressible flow. See Material Balance Formulation for more information.

DISCRETIZATION
To display all settings available in this section, click the Show button ( ) and select
Advanced Physics Options.

The Compute boundary fluxes checkbox is activated by default so that COMSOL

Multiphysics computes predefined accurate boundary flux variables. When this option
is selected, the solver computes variables storing accurate boundary fluxes from each
boundary into the adjacent domain.

230 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

If the checkbox is cleared, the COMSOL Multiphysics software instead computes the
flux variables from the dependent variables using extrapolation, which is less accurate
in postprocessing results but does not create extra dependent variables on the
boundaries for the fluxes.

The flux variables affected in the interface are:

• ndflux_c (where c is the dependent variable for the concentration). This is the
normal diffusive flux and corresponds to the boundary flux when diffusion is the
only contribution to the flux term.
• ntflux_c (where c is the dependent variable for the concentration). This is the
normal total flux and corresponds to the boundary flux plus additional transport
terms, for example, the convective flux when you use the nonconservative form.

Also the Apply smoothing to boundary fluxes checkbox is available if the previous
checkbox is selected. The smoothing can provide a more well-behaved flux value close
to singularities.

For details about the boundary fluxes settings, see Computing Accurate Fluxes in the
COMSOL Multiphysics Reference Manual.

The Value type when using splitting of complex variables setting should in most pure
mass transfer problems be set to Real, which is the default. It makes sure that the
dependent variable does not get affected by small imaginary contributions, which can
occur, for example, when combining a Time Dependent or Stationary study with a
frequency-domain study. For more information, see Splitting Complex-Valued
Variables in the COMSOL Multiphysics Reference Manual.

DEPENDENT VARIABLES
The dependent variable name is the Concentration c by default. The names must be
unique with respect to all other dependent variables in the component.

Add or remove species variables in the model and also change the names of the
dependent variables that represent the species concentrations.

Enter the Number of species. Use the Add concentration ( ) and Remove
concentration ( ) buttons as needed.

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 231

 FURTHER READING

• Theory for the Transport of Diluted Species Interface

• Numerical Stabilization in the COMSOL Multiphysics Reference
Manual.
• In the COMSOL Multiphysics Reference Manual, see Table 2-4 for
links to common sections and Table 2-5 for common feature nodes.
You can also search for information: press F1 to open the Help window
or Ctrl+F1 to open the Documentation window.

• Effective Diffusivity in Porous Materials: Application Library path

COMSOL_Multiphysics/Diffusion/effective_diffusivity
• Micromixer: Application Library path COMSOL_Multiphysics/
Fluid_Dynamics/micromixer

The Transport of Diluted Species in Porous Media Interface

This interface ( ), found under the Chemical Species Transport branch ( ), is used
to calculate the species concentration and transport in free and porous media. The
interface is the same as the Transport of Diluted Species interface but it uses other
defaults: The Mass Transport in Porous Media property is selected, and a Porous
Medium node is added by default. The interface includes reaction rate expressions and
solute sources for modeling of species transport and reaction in porous media.

This interface is dedicated to modeling transport in porous media, including immobile

and mobile phases, where the chemical species may be subjected to diffusion,
convection, migration, dispersion, adsorption, and volatilization in porous media. It
supports cases where either the solid phase substrate is exclusively immobile, or when
a gas-filling medium is also assumed to be immobile.

It applies to one or more diluted species or solutes that move primarily within a fluid
that fills (saturated) or partially fills (unsaturated) the voids in a solid porous medium.
The pore space not filled with fluid contains an immobile gas phase. Models including
a combination of porous media types can be studied.

The main feature nodes are the Porous Medium and Unsaturated Porous Medium nodes,
which add the equations for the species concentrations and provide an interface for
defining the properties of the porous media, as well as additional properties governing

232 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

adsorption, volatilization, dispersion and diffusion, migration, and the velocity field to
model convection.

The physics interface can be used for stationary and time-dependent analysis.

When this physics interface is added, these default nodes are also added to the Model
Builder — Porous Medium, No Flux (the default boundary condition), and Initial Values.
Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions, reaction rate expressions, and species sources. You can also
right-click Transport of Diluted Species in Porous Media to select physics features from
the context menu.

SETTINGS
The rest of the settings are the same as The Transport of Diluted Species Interface.

FURTHER READING

• Mass Balance Equation for Transport of Diluted Species in Porous

Media
• Domain, Boundary, and Pair Nodes for the Transport of Diluted
Species Interface

• Variably Saturated Flow and Transport — Sorbing Solute:

Application Library path Subsurface_Flow_Module/Solute_Transport/
sorbing_solute
Web link: [Link]
variably-saturated-flow-and-transport-sorbing-solute-490

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 233

 Domain, Boundary, and Pair Nodes for the Transport of Diluted
Species Interface
has the following domain, boundary, point, and pair nodes, listed in alphabetical
order, available from the Physics ribbon toolbar (Windows users), Physics context menu
(Mac or Linux users), or by right-clicking to access the context menu (all users).

• To add a node, go to the Physics toolbar, no matter what operating

system you are using.
• Contextual subnodes (attributes) are available by clicking the parent
node in the Model Builder, and then selecting the subnode from the
Attributes menu.

234 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

• Adsorption • Periodic Condition
• Concentration • Point Mass Source
• Continuity on Interior Boundaries • Porous Matrix
• Dispersion • Porous Medium
• Electrode Surface Coupling • Porous Electrode Coupling
• Equilibrium Reaction • Reaction Coefficients
• Fast Irreversible Surface Reaction • Reactions
• Fluid • Solid
• Fluid (Porous Medium) • Species Properties
• Flux • Species Source
• Flux Discontinuity • Surface Reactions
• Fracture • Surface Equilibrium Reaction
• Gas • Symmetry
• Inflow • Thin Diffusion Barrier
• Initial Values • Thin Impermeable Barrier
• Line Mass Source • Unsaturated Porous Medium
• Liquid • Volatilization
• Mass-Based Concentrations
• No Flux
• Open Boundary
• Out-of-Plane Flux
• Outflow
• Partition Condition

Some features require certain add-on modules. For details see https://
[Link]/products/specifications/

For axisymmetric components, COMSOL Multiphysics takes the axial

symmetry boundaries (at r  0) into account and automatically adds an
Axial Symmetry node that is valid on boundaries representing the
symmetry axis.

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 235

 In the COMSOL Multiphysics Reference Manual, see Table 2-4 for links
to common sections and Table 2-5 for common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Prescribing Conditions on Fluid-Solid Interfaces

Defining conditions on interfaces between different phases, typically a wall in contact
with a fluid, it is worthwhile to be careful with how variables in the conditions are
evaluated. COMSOL Multiphysics supports that variables are defined differently on
adjacent domains (or any other geometry dimension). Evaluating a variable with
differing definitions, the mean value is returned on a boundary separating the different
domains. The density, pressure, or temperature are examples of variables that may
differ in a fluid and a solid. In order to specify on which side a variable is evaluated, the
up and down operators (described in the COMSOL Multiphysics Reference Manual)
can be used.

Species Properties
Use this node to define parameters specific to each of the species. It is available when
the Migration in electric field checkbox is selected in the Transport Mechanisms section
of the interface.

The node will change its label depending on what inputs are available in the node. If
only Charge is available, it will be labeled “Species charges”.

CHARGE
Enter the charge number zc (dimensionless, but requires a plus or minus sign) for each
species.

ACTIVITY
This section is available when the Species Activity combo box is set to Debye–Hückel or
User defined.

If Debye–Hückel is selected, enter the ion size number, a0, for each species. If User
defined is selected, enter the activity coefficient, f, for each species.

236 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Fluid
The settings in this node are dependent on the checkboxes selected under Transport
Mechanisms in the Settings window for the Transport of Diluted Species interface. It
includes only the sections required by the activated transport mechanisms. It has all the
equations defining transport of diluted species as well as inputs for the material
properties.

MODEL INPUTS
The temperature model input is always available. Select the source of the Temperature.
For User defined, enter a value or expression for the temperature in the input field. This
input option is always available.

You can also select the temperature solved for by a Heat Transfer interface added to
the model component. These physics interfaces are available for selection in the
Temperature list.

CONVECTION
If transport by convection is active, the velocity field of the solvent needs to be
specified. Select the source of the Velocity field. For User defined, enter values or
expressions for the velocity components in the input fields. This input option is always
available.

You can also select the velocity field solved for by a Fluid Flow interface added to the
model component. These physics interfaces are available for selection in the Velocity
field list.

DIFFUSION
Use this section to specify diffusion coefficients describing the diffusion of each species
in the solvent fluid.

Use the Source list to select diffusion coefficients defined in a material or a Chemistry
interface.

Select Material to use a diffusion coefficient in a material available in the model.

User-defined property groups including a diffusion coefficient output property are
available for selection in the Fluid diffusion coefficient list.

Select Chemistry to use a diffusion coefficient from a Chemistry interface added to

model component. Diffusion coefficients are automatically defined when Calculate
transport properties is selected on the interface level of a Chemistry interface. All

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 237

 defined diffusion coefficients are available for selection in the Fluid diffusion coefficient
list.

For User defined, enter a value or expression for the Fluid diffusion coefficient Dc of each
species in the corresponding input field. This can be a scalar value for isotropic
diffusion or a tensor describing anisotropic diffusion. Select the appropriate tensor
type — Isotropic, Diagonal, Symmetric, or Full that describes the diffusion transport, and
then enter the values in the corresponding element (one value for each species).

Note that multiple species, as well as Migration in Electric fields (described below) is
only available for certain COMSOL Multiphysics add-on products. See details: https:/
/[Link]/products/specifications/.

MIGRATION IN ELECTRIC FIELD

This section is available when the Migration in electric field checkbox is selected. From
the Electric potential list, select the source of the electric field.

• Enter a value or expression for the Electric potential V, which is User defined; this
input option is always available.
• Select the electric potential solved by an AC/DC-based interface that has also been
added to the model.
• Select the electric potential defined or solved by Electrochemistry interface that has
been added to the component.

By default the Mobility is set to be calculated based on the species diffusivity and the
temperature using the Nernst-Einstein relation. For User defined, and under Mobility,
select the appropriate scalar or tensor type — Isotropic, Diagonal, Symmetric, or Full —
and type in the value of expression of the mobility um,c.

Enter the Charge number zc (dimensionless, but requires a plus or minus sign) for each
species.

The temperature (if you are using mobilities based on the Nernst–Einstein relation) is
taken from Model Inputs section.

Note that the migration in electric fields feature is only available in some COMSOL
products. See details: [Link]

238 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

EXAMPLE MODELS

• Separation Through Dialysis: Application Library path

Chemical_Reaction_Engineering_Module/Mixing_and_Separation/dialysis
Web link: [Link]
separation-through-dialysis-258

• Transport in an Electrokinetic Valve: Application Library path

Microfluidics_Module/Fluid_Flow/electrokinetic_valve
Web link: [Link]
transport-in-an-electrokinetic-valve-603

Solid
Use this node to describe diffusion in a solid phase domain. In this feature, the
activities of the species are assumed to be ideal.

DIFFUSION
Specify the diffusion coefficients describing the diffusion of each species in the solid.

Enter a value or expression for the Diffusion coefficient Dc of each species in the
corresponding input field. This can be a scalar value for isotropic diffusion or a tensor
describing anisotropic diffusion. Select the appropriate tensor type — Isotropic,
Diagonal, or Symmetric that describes the diffusion transport, and then enter the values
in the corresponding element (one value for each species).

ADVANCED
This section is only available with the Advanced Physics Options selected and when the
model contains a split between the material frame and the spatial frame. A split of these
frames occurs when using a Moving Mesh node or a Structural Mechanics interface like
Solid Mechanics. In such a model, the reference geometry (the material frame) differs
from the deformed geometry (the spatial frame). The diffusive transport of species is
solved for on the material frame, but the equation is adjusted for geometry
deformations so that species conservation is achieved on the spatial frame. Use the
Diffusion model to define the frame of reference for the diffusion tensor.

Select Reference geometry based (Lagrangian) to specify the diffusion tensor in the
undeformed geometry, implying that the diffusion is defined in the material frame

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 239

 (indicated by coordinates X, Y, Z). This is the preferred setting for an anisotropic
diffusion in combination with solid body translation or rotation. Also in general when
the diffusion tensor components have been measured in an undeformed geometry.

Select Deformed geometry based (Eulerian) to specify the diffusion tensor in the
deformed geometry. This means that the diffusion is defined in the spatial frame
(indicated by coordinates x, y, z). Use this setting when the diffusion tensor
components have been measured in a deformed geometry.

The Advanced Physics Options is found in the Show More Options dialog, accessed by
clicking the Show More Options button ( ).

Note that multiple species are only available for certain COMSOL Multiphysics
add-on products. See details: [Link]

Initial Values
The Initial Values node specifies the initial values for the concentration of each species.
These serve as an initial guess for a stationary solver or as initial conditions for a
transient simulation.

DOMAIN SELECTION
If there are several types of domains with different initial values defined, it might be
necessary to remove some domains from the selection. These are then defined in an
additional Initial Values node.

INITIAL VALUES
Enter a value or expression for the initial value of the Concentration or concentrations,
ci. This also serves as a starting guess for stationary problems.

Mass-Based Concentrations
Use the Mass-Based Concentrations node to add postprocessing variables for mass-based
concentrations (SI unit: kg/m3) and mass fractions (dimensionless) for all species.

MIXTURE PROPERTIES
The default Solvent density solvent is taken From material. For User defined, enter a
value or expression manually. Define the Molar mass of each species, which is needed
to calculate the mass-based concentration.

240 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Reactions
Use the Reactions node to account for the consumption or production of species
through chemical reactions. Define the rate expressions as required.

DOMAIN SELECTION
From the Selection list, choose the domains on which to define rate expression or
expressions that govern the source term in the transport equations.

Several reaction nodes can be used to account for different reactions in different parts
for the modeling geometry.

REACTION RATES
Add a rate expression Ri (SI unit: mol/(m3·s)) for species i. Enter a value or expression
in the field. Note that if you have the Chemistry interface available, provided with the
Chemical Reaction Engineering Module, the reaction rate expressions can be
automatically generated and picked up using the drop-down lists. For an example, see
the application Fine Chemical Production in a Plate Reactor as linked below.

REACTING VOLUME
This section is only available when the Mass Transport in Porous Media property is
available and selected. See [Link] for
more details on availability.

When specifying reaction rates for a species in porous media, the specified reaction rate
may have the basis of the total volume, the pore volume, or in the case of using the
volume of a particular phase (applicable when modeling an unsaturated porous
medium).

• For Total volume, the reaction expressions, in mol/(m3·s), are specified per unit
volume of the model domain (multiplied by unity).
• For Pore volume, the reaction expressions, in mol/(m3·s), are specified per unit
volume of total pore space. The reaction expressions will be multiplied by the
domain porosity, p. (p equals unity for nonporous domains).
• For Liquid phase, the reaction expressions, in mol/(m3·s), are specified per unit
volume of liquid in the pore space. The expressions will be multiplied by the liquid
volume fraction l. (l equals the porosity (p) on Porous Medium domains).
• For Gas phase, the reaction expressions, in mol/(m3·s), are multiplied by the gas
volume fraction g  p  lg equals 0 for Porous Medium domains.

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 241

 FURTHER READING
See the theory chapter on chemical species transport, starting with the section Mass
Balance Equation.

• Fine Chemical Production in a Plate Reactor: Application Library

path Chemical_Reaction_Engineering_Module/
Reactors_with_Mass_and_Heat_Transfer/plate_reactor
Web link: [Link]
fine-chemical-production-in-a-plate-reactor-8589

No Flux
This node is the default boundary condition on exterior boundaries. It should be used
on boundaries across which there is no mass flux, typically solid walls where no surface
reactions occur. The condition applied for each species corresponds to

– n   – Dc  = 0

where n denotes the outward pointing normal of the boundary. When the mass
transport includes migration of ionic species, the no flux condition is:

– n   – Dc – zu m Fc  = 0

CONVECTION
By default, the feature prescribes a vanishing flux due to diffusion and migration in an
electric field. This is the appropriate no flux condition when the relative convective
velocity at the boundary is zero. When the fluid velocity at the boundary is not equal
to that of the boundary, it is often convenient to prescribe the total flux including the
convection. To do this select Include in the Convection section.

When including the convection, the no flux condition prescribed is

– n   – Dc + uc  = 0

or

– n   – Dc – zu m Fc + uc  = 0

when migration of ionic species is included.

242 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Inflow
Use this node to specify all species concentrations at an inlet boundary.

If you want to specify the concentration of a subset of the partaking species, this can
be done by using the Concentration node instead.

For the Electroanalysis interface, this node is available when you select the Convection
checkbox on the physics interface Settings window.

CONCENTRATION
For the concentration of each species c0,c (SI unit: mol/m3), enter a value or
expression.

BOUNDARY CONDITION TYPE

The option Concentration constraint constrains the concentration values on the
boundary by the use of pointwise constraints.

The other option, Flux (Danckwerts) can be used when the concentration at the
boundary is not known, or when it varies in a nontrivial manner. This may, for example,
be useful when reactions with high reaction rates occur in the vicinity of the inlet. In
this case the concentration far upstream of the boundary is instead prescribed. The Flux
(Danckwerts) condition prescribes the total flux defined by the upstream concentration
and the fluid velocity at the boundary.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options. You can find details about the different constraint settings in the
section Constraint Reaction Terms in the COMSOL Multiphysics Reference Manual.

FURTHER READING
See the theory chapter in the section .

Outflow
Apply this condition at outlets boundaries where species should be transported out of
the model domain by fluid motion or by an electric field (in the case of ions). It is
assumed that convection and migration (in an electric field) are the dominating
transport mechanisms across the boundary, and therefore that the diffusive transport
can be ignored, that is:

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 243

 n   – D c  = 0

Note that the Convection or the Migration in electric field transport mechanisms needs
to be included for this node to be available.

Concentration
This condition node adds a boundary condition for the species concentration. For
example, a c = c0 condition specifies the concentration of species c.

CONCENTRATION
Individually specify the concentration for each species. Select the checkbox for the
Species to specify the concentration, and then enter a value or expression in the
corresponding field. To use another boundary condition for a specific species, click to
clear the checkbox for the concentration of that species.

When defining conditions on boundaries between different phases, see the section
Prescribing Conditions on Fluid-Solid Interfaces.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.
You can find details about the different constraint settings in the section Constraint
Reaction Terms in the COMSOL Multiphysics Reference Manual.

Flux
This node can be used to specify the species molar flux across a boundary. The flux can
for example occur due to chemical reactions or a phase change at the boundary. The
flux can also represent the transport to or from a surrounding environment currently
not included model.

The prescribed flux of a species c is by default defined as

– n   – Dc  = J 0

where n denotes the outward pointing normal of the boundary. When the mass
transport includes migration of ionic species, the flux is defined as:

– n   – Dc – zu m Fc  = J 0

244 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The flux prescribed, J0, can include any arbitrary user-specified expression. It can be
constant or a function of a dependent variable or independent variable. Common
examples are a flux dependent of the concentration, temperature, pressure, or the
electric potential .

When defining conditions on boundaries between different phases, see the section
Prescribing Conditions on Fluid-Solid Interfaces.

CONVECTION
By default, the flux due to diffusion and migration in an electric field is prescribed. This
is the appropriate flux condition when the relative velocity at the boundary is zero.
When the fluid velocity is not equal to that of the boundary, it is often convenient to
prescribe the total flux, including the convection. To do this select Include in the
Convection section.

When including the convection, the prescribed flux is defined as:

– n   – Dc + uc  = J 0

or

– n   – Dc – zu m Fc + uc  = J 0

when migration of ionic species is included.

INWARD FLUX
Select the Species checkbox for the species for which to specify the flux, and enter a
value or expression for the inward flux in the corresponding field. Use a minus sign
when specifying a flux directed out of the system. To use another boundary condition
for a specific species, click to clear the checkbox for that species.

External convection
Set Flux type to External convection to prescribe a flux to or from an exterior domain
(not modeled) assumed to include convection. The exterior can for example include a
forced convection to control the temperature or to increase the mass transport. In this
case the prescribed mass flux corresponds to

J0 = kc  cb – c 

where kc is a mass transfer coefficient and cb is the bulk concentration, the typical
concentration far into the surrounding exterior domain.

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 245

 Symmetry
The Symmetry node can be used to represent boundaries where the species
concentration is symmetric, that is, where there is no mass flux across the boundary.

This boundary condition is identical to that of the No Flux node.

Flux Discontinuity
This node represents a discontinuity in the mass flux across an interior boundary:

where the value N0 (SI unit: mol/(m2·s)) specifies the jump in total flux at the
boundary. This can be used to model a boundary source, for example a surface
reaction, adsorption, or desorption.

FLUX DISCONTINUITY
In this section the jump in species flux (or surface source) is specified.

Select the Species checkbox for the species to specify and enter a value or expression
for the material flux jump in the corresponding field. To use a different boundary
condition for a specific species, click to clear the checkbox for the flux discontinuity of
that species.

Partition Condition
The Partition Condition node can be used to prescribe the relation between the
concentration of a solute species in two adjoining immiscible phases. It can for example
be used on interior boundaries separating two liquid phases, a gas-liquid interface, or
on a boundary separating a liquid phase and a solid or porous media.

Select Partition coefficient from the Relation list to prescribe a linear relationship
between the concentration on either side of a boundary. For a species concentration
ci, the ratio between the concentration on the upside and on the downside of the
boundary (ci,u and ci,d respectively) is then defined in terms of a partition coefficient
Ki in the manner of

c i u
K i = --------
-
c i d

Use the associated input field to prescribe the partition coefficient for each species.

246 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Select User defined from the Relation list to set up a nonlinear relation between the up
and downside concentrations. For such relations f(ci)u  f(ci)d, where f(ci)u is the
expression enforced on the upside. Use the associated input fields to enter the upside
and downside expressions for each species.

The up and downside of the selected boundary is indicated with a red arrow in the
Graphics window. The arrow points from the downside into the upside. Select the
Reverse direction checkbox to reverse the direction of the arrow on the selected
boundary, and update the definition of the up and downside concentrations
accordingly.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

FURTHER READING
For an example of using a partition condition, see this application example:

Separation Through Dialysis: Application Library path

Chemical_Reaction_Engineering_Module/Mixing_and_Separation/dialysis

Periodic Condition
The Periodic Condition node can be used to define periodicity for the mass transport
between two sets of boundaries. The node prescribes continuity in the concentration
and the mass flux between the “source” and the “destination” side, respectively. Note
that these names are arbitrary and does not influence the direction in which mass is
transported. It is dictated by mass transfer equations in the adjacent domains.

BOUNDARY SELECTION
The node can be activated on more than two boundaries, in which case the feature tries
to identify two separate surfaces that each consist of one or several connected
boundaries.

DESTINATION SELECTION
For more complex geometries, it might be necessary to specify the destination
selection manually. To do so, right-click the Periodic Condition node and choose Manual

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 247

 Destination Selection. You can then specify the boundaries that constitute the
destination surfaces in the Destination Selection section.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog. This section contains settings for
specifying the type of constraint and whether to use a pointwise or weak constraint. See
Constraint Settings for more information.

ORIENTATION OF SOURCE
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog. For information about the Orientation
of Source section, see Orientation of Source and Destination.

ORIENTATION OF DESTINATION
This section appears if the setting for Transform to intermediate map in the Orientation
of Source section is changed from the default value, Automatic, and Advanced Physics
Options is selected in the Show More Options dialog. For information about the
Orientation of Destination section, see Orientation of Source and Destination.

MAPPING BETWEEN SOURCE AND DESTINATION

To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog. For information about the Mapping
Between Source and Destination section, see Mapping Between Source and Destination.

FURTHER READING
For an example of using a periodic condition, see this application example:

The KdV Equation and Solitons: Application Library path

COMSOL_Multiphysics/Equation_Based/kdv_equation

248 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Line Mass Source
The Line Mass Source feature models mass flow originating from a tube or line region
with an infinitely small radius.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

SELECTION
The Line Mass Source feature is available for all dimensions, but the applicable selection
differs between the dimensions.

MODEL DIMENSION APPLICABLE GEOMETRICAL ENTITY

2D Points
2D Axisymmetry Points not on the symmetry axis and the symmetry axis
3D Edges

SPECIES SOURCE
Enter the source strength, qꞏ l,c , for each species (SI unit: molm·s)). A positive value
results in species injection from the line into the computational domain, and a negative
value means that the species is removed from the computational domain.

Line sources located on a boundary affect the adjacent computational domains. This
effect makes the physical strength of a line source located in a symmetry plane twice
the given strength.

Point Mass Source

The Point Mass Source feature models mass flow originating from an infinitely small
domain around a point.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 249

 SPECIES SOURCE
Enter the source strength, qꞏ p,c , for each species (SI unit: mols). A positive value
results in species injection from the point into the computational domain, and a
negative value means that the species is removed from the computational domain.

Point sources located on a boundary or on an edge affect the adjacent computational

domains. This has the effect, for example, that the physical strength of a point source
located in a symmetry plane is twice the given strength.

Open Boundary
Use this node to set up mass transport across boundaries where both convective inflow
and outflow can occur. On the parts of the boundary where fluid flows into the
domain, an exterior species concentration is prescribed. On the remaining parts, where
fluid flows out of the domain, a condition equivalent to the Outflow node is instead
prescribed.

The direction of the flow across the boundary is typically calculated by a fluid flow
interface and is provided as a model input to the Transport of Diluted Species
interface.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

EXTERIOR CONCENTRATION
Enter a value or expression for the Exterior concentration.

Thin Diffusion Barrier

Use this boundary condition to model a thin layer through which mass is transported
by diffusion only. The node is applicable on interior boundaries and can be used to
avoid meshing thin structures.

THIN DIFFUSION BARRIER

Specify the Layer thickness, ds, and input a Diffusion coefficient, Ds,c, for each of the
species included.

250 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Thin Impermeable Barrier
This feature models a thin mass transfer barrier. It is available on interior boundaries
and introduces a discontinuity in the concentration across the boundary. On each side,
a no-flux condition is prescribed for the mass transport implying that it acts as a barrier.
The feature can be used to avoid meshing thin structures.

Solving a model involving coupled fluid flow and mass transfer, the Thin Impermeable
Barrier feature can be combined with an Interior Wall feature in order to model a thin
solid wall.

Equilibrium Reaction
Use this node to model a reaction where the kinetics is so fast that the equilibrium
condition is fulfilled at all times. The node solves for an additional degree of freedom
(the reaction rate Req) to fulfill the equilibrium condition at all times in all space
coordinates.

If the Apply equilibrium condition on inflow boundaries checkbox is selected, the

specified inflow concentration values in all active Inflow boundary nodes for the physics
interface are modified to comply with the equilibrium condition.

• A necessary requirement for this is feature to be available is that two or

more species are solved for by the interface.
• This feature is only available in a limited set of add-on products. See
[Link] for more details
on availability.

EQUILIBRIUM CONDITION
The list defaults to Equilibrium constant or select User defined. For either option, the
Apply equilibrium condition on inflow boundaries checkbox is selected by default.

For Equilibrium constant, enter an Equilibrium constant Keq (dimensionless). Also enter
a value or expression for the Unit activity concentration Ca0 (SI unit: mol/m3).
Selecting Equilibrium constant defines an equilibrium condition based on the
stoichiometric coefficients, the species activities, and the law of mass action.

For User defined, enter an Equilibrium expression Eeq (dimensionless).

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 251

 STOICHIOMETRIC COEFFICIENTS
Enter a value for the stoichiometric coefficientc (dimensionless). The default is 0. Use
negative values for reactants and positive values for products in the modeled reaction.

Species with a stoichiometric coefficient value of 0 are not affected by the Equilibrium
Reaction node.

Surface Reactions
The Surface Reactions node can be used to account for the species boundary flux due
to chemical reactions occurring on a surface (heterogeneous reactions). For a domain
species participating in a surface reaction, the boundary flux corresponds to the
reaction rate at the surface.

SURFACE REACTION RATE

Specify the surface reaction rate J0 of each species resulting from the reactions. Note
that if you have the Chemistry interface available, provided with the Chemical
Reaction Engineering Module, the reaction rate expressions can be automatically
generated and picked up using the drop-down list.

When defining conditions on boundaries between different phases, see the section
Prescribing Conditions on Fluid-Solid Interfaces.

FURTHER READING
For an example of using the Surface Reactions node, see this application example:

Chemical Vapor Deposition of GaAs: Application Library path

Chemical_Reaction_Engineering_Module/
Reactors_with_Mass_and_Heat_Transfer/gaas_cvd

Surface Equilibrium Reaction

Use this node to model an equilibrium reaction on a boundary (surface). The settings
for this node are similar to Equilibrium Reaction. Note that a necessary requirement
for this is feature to be available is that two or more species are solved for by the
interface.

252 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

When defining conditions on boundaries between different phases, see the section
Prescribing Conditions on Fluid-Solid Interfaces.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

Fast Irreversible Surface Reaction

This boundary node defines an irreversible reaction where the kinetics is so fast that
the only factor limiting the reaction rate is the transport of a species to the reacting
surface.

The node will set the Rate limiting species concentration to zero at the boundary, and
balance the fluxes of the species participating in the reaction and the current densities
according to the Stoichiometric Coefficients settings.

When defining conditions on boundaries between different phases, see the section
Prescribing Conditions on Fluid-Solid Interfaces.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

Porous Electrode Coupling

Use this node to add a molar source in a domain that is coupled to one or multiple
nodes of an Electrochemistry Interface.

The molar source is calculated from the number of electrons, stoichiometric

coefficients, and volumetric current densities of the coupled porous electrode reactions
specified in the Reaction Coefficients subnodes.

In the Transport of Concentrated Species interface, the molar sources (or sinks) are
multiplied by the species molar masses to obtain the corresponding mass sources.

Additional Reaction Coefficients subnodes are available from the context menu
(right-click the parent node) as well as from the Physics toolbar, Attributes menu.

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 253

 Note that if you are also modeling the momentum transport and expect a
nonnegligible total mass source or sink, which is often the case in gas diffusion
electrodes, you need to also add a corresponding mass source/sink in the Fluid Flow
interface. (A Reacting Flow multiphysics node adds the associated source/sink term in
a Fluid Flow interface automatically.)

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

Reaction Coefficients
Add this node to the Electrode Surface Coupling and Porous Electrode Coupling
features to define molar fluxes and sources based on electrode current densities in an
Electrochemistry interface.

The molar flux or source is proportional to the stoichiometric coefficients and the
current density according to Faraday’s law.

Current densities from Electrode Reaction (iloc, SI unit: A/m2) or Porous Electrode
Reaction nodes (iv, SI unit: A/m3) of any Electrochemistry interface in the model are
available for selection as the Coupled reaction, and user-defined expressions are also
supported.

Enter the Number of participating electrons nm (dimensionless) and the Stoichiometric

coefficient vc (dimensionless) as explained in the theory section linked below.

Use multiple subnodes to couple to multiple reactions.

Electrode Surface Coupling

Use this node to define a flux boundary condition based on current densities of one or
multiple nodes in an Electrochemistry interface.

The flux is proportional to the current densities and the stoichiometric coefficients
according to Faraday’s law as defined by summation over the Reaction Coefficients
subnodes.

Note that if you are also modeling the momentum transport and expect a
nonnegligible total mass flux over the boundary, which is often the case for gas

254 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

diffusion electrodes, you need to also add a corresponding Electrode Surface Coupling
node in the Fluid Flow interface.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

Porous Medium
Use this node to model the concentration of diluted species transported by a solvent
(mobile fluid phase) through interstices in a solid porous medium. In addition to
transport due to convection and diffusion, the node contains functionality to include
species evolution through adsorption and dispersion.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

Species transport through a porous medium is affected both by properties of the fluid
phase, and as well as properties of the solid matrix. These properties are defined in the
Fluid (Porous Medium) and Porous Matrix subnodes respectively. The Porous
Medium node supports material properties using a Porous Material node, where
properties are defined per phase in a similar manner.

Add an Adsorption or a Dispersion subfeature to the Porous Medium to account for the
corresponding transport mechanism.

Fluid (Porous Medium)

Use this node to specify the mass transfer in the mobile fluid solvent filling the pores
of a porous medium. It is used as a subnode to Porous Medium.

MODEL INPUTS
The temperature model input is always available. By default, the Temperature model
input is set to Common model input, and the temperature is controlled from Default
Model Inputs under Global Definitions or by a locally defined Model Input. If a Heat
Transfer interface is included in the component, it controls the temperature Common
model input. Alternatively, the temperature field can be selected from another physics

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 255

 interface. All physics interfaces have their own tags (Name). For example, if a Heat
Transfer in Fluids interface is included in the component, the Temperature (ht) option
is available for T.

You can also select User defined from the Temperature model input in order to manually
prescribe T.

CONVECTION
If transport by convection is active, the velocity field of the solvent needs to be
specified. For User defined, enter values or expressions for the velocity components in
the input fields. This input option is always available. You can also select the velocity
field solved for by a Fluid Flow interface added to the model component. These
physics interfaces are available for selection in the Velocity field list.

DIFFUSION
Use this section to specify diffusion coefficients describing the diffusion of each species
in the solvent fluid.

Use the Source list to select to pick up diffusion coefficients defined in a material or a
Chemistry interface.

Select Material, and a Fluid material to use a diffusion coefficient in a material available
in the model. User-defined property groups including a diffusion coefficient output
property are available for selection in the Fluid diffusion coefficient list.

Select Chemistry to use a diffusion coefficient from a Chemistry interface added to

model component. Diffusion coefficients are automatically defined when Calculate
transport properties is selected on the interface level of a Chemistry interface. All
defined diffusion coefficients are available for selection in the Fluid diffusion coefficient
list.

For User defined, enter a value or expression for the Fluid diffusion coefficient DF,i of
each species in the corresponding input field.

In a porous medium the diffusivity is reduced due to the fact that the solid grains
impede Brownian motion. Select an Effective diffusivity model to account for the
reduced diffusivity. The available models are Millington and Quirk model (the default),
Bruggeman model, Tortuosity model, or No correction. For Tortuosity model, enter a
value for the tortuosity F,i (dimensionless).

256 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Note that multiple species, as well as Migration in Electric fields (described below) is
only available for certain COMSOL Multiphysics add-on [Link] details, see
[Link]

MIGRATION IN ELECTRIC FIELD

This section is available when the Migration in electric field checkbox is selected. From
the Electric potential list, select the source of the electric field.

• For User defined, enter a value or expression for the Electric potential V. This input
option is always available.
• Select the electric potential solved by an AC/DC-based interface that has added to
the component.
• Select the electric potential defined or solved for by an Electrochemistry interface
added to the component.

By default the Mobility is set to be calculated based on the species effective diffusivity
and the temperature using the Nernst-Einstein relation. For User defined, select the
appropriate scalar or tensor type — Isotropic, Diagonal, Symmetric, or Full — and type
in the value or expression of the effective mobility ume,i.

Enter the Charge number zc for each species.

Porous Matrix
This node sets the porosity when modeling transport of diluted species in a Porous
Medium or an Unsaturated Porous Medium. The porosity corresponds to the volume
fraction occupied by the species containing fluid.

• By default the porosity of the solid matrix, p, is taken From material. The Porous
Matrix node supports the use of a Porous Material added to the Materials branch. In
this case the porosity is defined in the manner of

p = 1 –  s i – imf i
i i

where s,i and imf,i are the volume fractions given in Solid and Immobile Fluids
subnodes under the Porous Material node.
• Select From pellet bed densities to compute the porosity using the (dry bulk) Bed
density b and the (single phase) Pellet density pe. The porosity is then defined from

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 257

 b
 p = 1 – -------
-
 pe

• For User defined, enter a value or expression for the porosity.

Dispersion
Local variations in the velocity as the fluid flows around solid particles lead to
mechanical mixing, referred to as dispersion. Use this feature to account for dispersion
in a Porous Medium or a Unsaturated Porous Medium

This subfeature is available when both the Mass transfer in porous media mass transfer
checkbox and the Convection checkbox are selected in the Settings window for the
physics interface.

Select the Specify dispersion for each species individually checkbox to specify the
dispersion tensor DD (SI unit: m2/s) for each species separately. When not selected the
same dispersion tensor DD is used for all species.

Select an option from the Dispersion tensor list — Dispersivity or User defined.

Select Dispersivity to specify the dispersion in terms of dispersivities (SI unit: m). Select
an option from the Dispersivity model list: Isotropic (the default) or Transverse isotropic
based on the properties of the porous media. For isotropic porous media, specify the
longitudinal and transverse dispersivities. For transverse isotropic porous media,
specify the longitudinal, horizontal transverse, and vertical transverse dispersivities.

For User defined, specify the dispersion components in terms of constants or

expressions. Select Isotropic, Diagonal, Symmetric, or Full to enable the appropriate
tensor components.

Unsaturated Porous Medium

Use this node to model the concentration of diluted species transported by a solvent
(mobile fluid phase) through interstices in a solid porous medium. In addition to
transport due to convection and diffusion, the node contains functionality to include
species evolution through adsorption and dispersion.

Use this node to model the concentration of diluted species transported by a liquid
(mobile fluid phase) in a partially filled solid porous medium. The interstices of the
porous medium contains the liquid carrier phase and gas pockets. Apart from

258 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

convection and diffusion, the node contains functionality to include species evolution
through adsorption, dispersion, and volatilization.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

The properties of each phase present are defined using the Liquid, the Gas, and the
Porous Matrix subnodes respectively. The Unsaturated Porous Medium node supports
material properties using a Porous Material node, where properties are defined per
phase in a similar manner.

Add an Adsorption or a Dispersion subfeature to the Unsaturated Porous Medium to

account for the corresponding transport mechanism.

Liquid
Use this node to specify the mass transfer in the mobile liquid solvent present in the
pores of the Unsaturated Porous Medium.

MODEL INPUTS
The temperature model input is always available. By default, the Temperature model
input is set to Common model input, and the temperature is controlled from Default
Model Inputs under Global Definitions or by a locally defined Model Input. If a Heat
Transfer interface is included in the component, it controls the temperature Common
model input. Alternatively, the temperature field can be selected from another physics
interface. All physics interfaces have their own tags (Name). For example, if a Heat
Transfer in Fluids interface is included in the component, the Temperature (ht) option
is available for T.

You can also select User defined from the Temperature model input in order to manually
prescribe T.

SATURATION
Select Saturation or Liquid volume fraction from the list.

For Saturation, enter a value for s (dimensionless) between 0 and 1. The liquid volume
fraction is then computed from the saturation and porosity as l = ps.

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 259

 For Liquid volume fraction, enter a value for l (dimensionless) between 0 and the value
of the porosity.

Select a Liquid fraction time change: Liquid fraction constant in time (the default), Time
change in liquid fraction, or Time change in pressure head.

• For Time change in fluid fraction, enter d/dt (SI unit: 1/s).
• For Time change in pressure head, enter dHp/dt (SI unit: m/s) and a Specific
moisture capacity Cm (SI unit: 1/m). If a Darcy’s Law interface is included in the
component, the time change in pressure head solved for can be selected.

CONVECTION
If transport by convection is active, the velocity field of the solvent needs to be
specified. For User defined, enter values or expressions for the velocity components in
the input fields. This input option is always available. You can also select the velocity
field solved for by a Fluid Flow interface added to the model component. These
physics interfaces are available for selection in the Velocity field list.

DIFFUSION
Use this section to specify diffusion coefficients describing the diffusion of each species
in the solvent liquid.

Use the Source list to select to pick up diffusion coefficients defined in a material or a
Chemistry interface.

Select Material, and a Liquid material to use a diffusion coefficient in a material available
in the model. User-defined property groups including a diffusion coefficient output
property are available for selection in the Liquid diffusion coefficient list.

Select Chemistry to use a diffusion coefficient from a Chemistry interface added to

model component. Diffusion coefficients are automatically defined when Calculate
transport properties is selected on the interface level of a Chemistry interface. All
defined diffusion coefficients are available for selection in the Liquid diffusion coefficient
list.

For User defined, enter a value or expression for the Liquid diffusion coefficient DL,i of
each species in the corresponding input field.

In a porous medium the diffusivity is reduced due to the fact that the solid grains
impede Brownian motion. Select an Effective diffusivity model, liquid to account for the
reduced diffusivity in the liquid. The available models are Millington and Quirk model

260 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

(the default), Bruggeman model, Tortuosity model, or No correction. For Tortuosity
model, enter a value for the tortuosity L,i (dimensionless).

Note that multiple species, as well as Migration in Electric fields (described below) is
only available for certain COMSOL Multiphysics add-on products. For details see:
[Link]

MIGRATION IN ELECTRIC FIELD

This section is available when the Migration in electric field checkbox is selected. From
the Electric potential list, select the source of the electric field.

• For User defined, enter a value or expression for the Electric potential V. This input
option is always available.
• Select the electric potential solved by an AC/DC-based interface that has added to
the component.
• Select the electric potential defined or solved for by an Electrochemistry interface
added to the component.

By default the Mobility is set to be calculated based on the species effective diffusivity
and the temperature using the Nernst-Einstein relation. For User defined, select the
appropriate scalar or tensor type — Isotropic, Diagonal, Symmetric, or Full — and type
in the value or expression of the effective mobility ume,i.

Enter the Charge number zc for each species.

Gas
Use this node to specify the mass transfer in the gas phase present in the pores of the
Unsaturated Porous Medium.

DIFFUSION
The settings for the diffusion of each species in the gas phase are identical to those in
the Liquid subnode.

VOLATILIZATION
Enter a value for the volatilization isotherm kG,c (dimensionless) for each species.

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 261

 Adsorption
Use this node to model adsorption of the (fluid phase) solute species onto the surface
of the porous matrix. It is available as a subnode to the Porous Medium and the
Unsaturated Porous Medium nodes.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

MATRIX PROPERTIES
The density of the porous media is needed when modeling adsorption to the surface
of the porous matrix. Choose to input either the Dry bulk density , or the Solid phase
density s. The former is the density of the porous matrix including empty pores, while
the latter corresponds to the density of the pure solid phase. The density can be defined
from the domain material by selecting From material, or from a user defined expression.
When a Porous Material is used on the selection, the density will be requested from a
Solid subfeature. Several Solid features can be used to model a homogeneous mixture
of several solid components.

ADSORPTION
Select an Adsorption isotherm — Langmuir, Freundlich, Toth, BET, or User defined to
specify how to compute cP, the amount of species sorbed to the solid phase (moles per
unit dry weight of the solid):

• For Langmuir:

KL c c K L c Pmax
c P = c Pmax --------------------  K P = -------P- = ---------------------------
1 + KL c c  1 + KL c 
2

Enter a Langmuir constant kL,c (SI unit: m3/mol) and an Adsorption maximum
cp,max,c (SI unit: mol/kg).

• For Freundlich:

262 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

 c N c cP
c P = K F  --------  K P = -------P- = N -----
c ref c c

Enter a Freundlich constant kF,c (SI unit: mol/kg), a Freundlich exponent NF,c
(dimensionless), and a Reference concentration cref,c (SI unit: mol/m3).
• For Toth:
bT c
c P = c Pmax ----------------------------------------------
N 1  NT
 1 +  bT c  T 

and

–  1 + ------
1-
c  N T
N
KP = -------P- = c Pmax b T  1 +  b T c  T 
c

Enter a Toth constant bT,c (SI unit: m3/mol), a Toth exponent NT,c (dimensionless),
and an Adsorption maximum cp,max,c (SI unit: mol/kg).
• For BET (Brunauer-Emmett-Teller):
KB c0 c
c P = -------------------------------------------------------------
-
 c S – c   1 +  K B – 1  -----
c
 c S

and

2 2
c K B c 0 c S   K B – 1 c + c S 
K P = -------P- = -----------------------------------------------------------------
-
c 2
 c – c S   c S +  K B – 1 c 
2

Enter a BET constant KB,c (dimensionless), a Monolayer adsorption capacity c0,c

(SI unit: mol/kg), and an Saturation concentration cS,c (SI unit: mol/m3).
• For User defined enter an Adsorption isotherm cP,c (SI unit: mol/kg):

cP = f  c 

For more information, see in the theory section.

FURTHER READING
See the theory chapter in the section .

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 263

 Volatilization
This feature is available when the Mass transfer in porous media checkbox is selected in
the Settings window for the physics interface.

Use this feature to model mass transfer at the boundary due to volatilization. The
species dissolved in the liquid are assumed to be vaporized at the boundary, and
transported into the surrounding bulk region due to convection and diffusion. The
mass transfer at the boundary is defined as

– n  J c = – h c  k G,c c – c Gatm,c 

where hc is the mass transfer coefficient, kG,c the volatilization coefficient, and cGatm,c
the concentration in the surrounding atmosphere.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

VOLATILIZATION
Enter a Mass transfer coefficient hc defining the transfer into the surrounding media.
This can be given by boundary layer theory. When assuming that no convective flow
is present in the surrounding, the mass transfer coefficient can be defined from the gas
diffusion coefficient DGc and the thickness of the diffusion layer ds in the manner of

D Gc
h c = ----------
ds

Also give the atmospheric concentration for each species, cGatm,c. The Volatilization
coefficient kG,c for each species are taken from the adjacent Unsaturated Porous
Medium domain.

Species Source
In order to account for consumption or production of species in porous domains, the
Species Source node adds source terms expressions Si to the right-hand side of the
species transport equations.

264 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

DOMAIN SELECTION
From the Selection list, choose the domains on which to define rate expression or
expressions that govern the source term in the transport equations.

If there are several types of domains, with subsequent and different reactions occurring
within them, it might be necessary to remove some domains from the selection. These
are then defined in an additional Species Source node.

SPECIES SOURCE
Add a source term Si (SI unit: mol/(m3·s)) for each of the species solved for. Enter a
value or expression in the field of the corresponding species.

Hygroscopic Swelling
The Hygroscopic Swelling multiphysics coupling node ( ) is used for moisture
concentration coupling between the Solid Mechanics interface and either the
Transport of Diluted Species or Transport of Diluted Species in Porous Media
interfaces.

Hygroscopic swelling is an effect of internal strain caused by changes in moisture

content. This volumetric strain can be written as

 hs =  h M m  c mo – c mo,ref 

where h is the coefficient of hygroscopic swelling, Mm is the molar mass, cmo is the
moisture concentration, and cmo,ref is the strain-free reference concentration.

This feature requires a license of either the MEMS Module or the Structural Mechanics
Module. The multiphysics feature will appear automatically if both the Transport of
Diluted Species and the Solid Mechanics interfaces are added to the same component.
For the most current information about licensing, please see https://
[Link]/products/specifications/.

FURTHER READING
More information about how to use hygroscopic swelling can be found in the
Hygroscopic Swelling section in the Structural Mechanics Module User’s Guide.

More information about multiphysics coupling nodes can be found in the section The
Multiphysics Branch.

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 265

 Out-of-Plane Flux
The out-of-plane flux node is used to prescribe a flux in an out-of-plane domain, and
it can be used to reduce a model geometry to 2D, or even 1D, when the concentration
variation is small in one or more directions. This could be the case for example when
the object to model is thin or slender. Figure 5-1 shows examples of possible situations
in which this type of geometry reduction can be applied.

Jz

Jup

Jdown

Figure 5-1: Geometry reduction from 3D to 1D (top) and from 3D to 2D (bottom).

For a 1D component this node adds a single out-of-plane molar flux J0,z,i for species
i. For a 2D component two fluxes can be prescribed for each species; J0,u,i for the
upside of the domain, and J0,d,i for the downside of the domain.

The fluxes are added as a source contribution to the right-hand side of the transport
equation. For example, in 2D components with diffusion, convection, and
out-of-plane flux present, the mass transfer of species i is defined by the mass
conservation equation

c i
------- +   J i + u  c i = R i + S opf i , (5-1)
t

where Sopf,i is the out-of-plane source for species i

J 0 i J 0 u i + J 0 d i
- = --------------------------------------
S opf i = --------- .
dz dz

For external convection on the upside and the downside of the domain, the
out-of-plane flux is

266 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

 J 0 i = k c u i  c b u i – c i  + k c d i  c b d i – c i 

where kc is a mass transfer coefficient and cb is the bulk concentration, the typical
concentration far into the surrounding exterior domain. The prescribed flux, J0, can
include any arbitrary user-specified expressions. It can be a constant or a function of a
dependent variable or independent variables.

UPSIDE INWARD FLUX

For 1D components, enter the cross-sectional perimeter Pc to get the out-of-plane flux

J 0 i = P c J 0 z i .

The default value of Pc is the circumference. Either keep the default value, for a circular
cross-section shape, or edit the value to get a user-defined shape of the out-of-plane
cross-section.

The available flux type options are General inward flux and External convection. Select
the Species checkbox for the species for which to specify the flux, and enter a value or
expression for the inward flux in the corresponding field. Use a minus sign when
specifying a flux directed out of the system. To use another boundary condition for a
specific species, click to clear the checkbox for that species.

Set Flux type to External convection to prescribe a flux to or from an exterior domain
(not modeled) assumed to include convection. The exterior can for example include a
forced convection to control the temperature or to increase the mass transport. In this
case the prescribed mass flux corresponds to

J0 = kc  cb – c 

where kc is a mass transfer coefficient and cb is the bulk concentration, the typical
concentration far into the surrounding exterior domain.

DOWNSIDE INWARD FLUX

The available options are General inward flux and External convection. The settings are
the same as for the Upside Inward Flux section.

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 267

 FURTHER READING
For an example of using the Out-of-Plane Flux node, see this application example:

Cross-Flow Mass Transfer in a Thin Domain: Application Library path

Chemical_Reaction_Engineering_Module/Tutorials/thin_domain

Fracture
Use this node to model mass transport along thin fractures in porous media. The node
assumes that the transport in the tangential direction along the fracture is dominant as
a result of lower flow resistance.

The Fracture feature has two default subfeatures — the Fluid (Fracture) subfeature
where convection and diffusion are specified and the Fracture Material subfeature
where the porosity p is defined.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

FRACTURE PROPERTIES
Specify a value for the Fracture thickness dfr.

For the Fracture boundary condition, there are a number of Attributes available which
can be accessed either via the Physics Ribbon toolbar (for Windows) or by
right-clicking on the Fracture node. These attributes are identical to the Boundary,
Edge, Point, and Pair Nodes for the Transport of Diluted Species in Fractures
Interface.

Fluid (Fracture)
Use this node to specify the mass transfer in the mobile fluid solvent filling the pores
of a fracture.

It uses the same input options as the Fluid (Porous Medium) subfeature under the
Porous Medium feature. There are further details about the settings.

268 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

MODEL INPUTS
The temperature model input is always available. By default, the Temperature model
input is set to Common model input, and the temperature is controlled from Default
Model Inputs under Global Definitions or by a locally defined Model Input. If a Heat
Transfer interface is included in the component, it controls the temperature Common
model input. Alternatively, the temperature field can be selected from another physics
interface. All physics interfaces have their own tags (Name). For example, if a Heat
Transfer in Fluids interface is included in the component, the Temperature (ht) option
is available for T.

CONVECTION
If transport by convection is active, the velocity field of the solvent needs to be
specified. Select an option from the Velocity field list to specify the convective velocity
along the fracture. For a consistent model, use a Fracture feature in a Darcy’s Law
interface to compute the fluid flow velocity in the fracture.

For User defined, enter values or expressions for the velocity components in the table
shown

DIFFUSION
Use this section to specify diffusion coefficients describing the diffusion of each species
in the solvent fluid.

The Diffusion settings are similar to those for the diffusion coefficients in
the Fluid (Porous Medium) node under Porous Medium.

Fracture Material
Specify the Porosity p (dimensionless) of the porous matrix. This is by default taken
From material. Select From pellet bed densities to compute the porosity from the (dry
bulk) Bed density b and the (single phase) Pellet density pe through the relation

b
 p = 1 – -------
-
 pe

For User defined, enter a value or expression for the porosity.

THE TRANSPORT OF DILUTED SPECIES INTERFACE | 269

 The Transport of Diluted Species in
Fractures Interface
The Transport of Diluted Species in Fractures (dsf) interface ( ), found under the
Chemical Species Transport branch ( ), is used to model the transport of a solute
species along thin porous fractures, taking into account diffusion, dispersion,
convection, and chemical reactions. The fractures are defined by boundaries and the
solute species is assumed to be present in a solvent.

The interface supports simulation of species transport along boundaries in 2D and 3D,
and axisymmetric components in 2D. The dependent variable is the molar
concentration, c. Modeling multiple species transport is possible, whereby the physics
interface solves for the molar concentration, ci, of each species i.

This interface is only available in a limited set of add-on products. For a

detailed overview of which interfaces are available in each product, visit
[Link]

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is dsf.

BOUNDARY SELECTION
If the model geometry includes boundaries that should not be included in the mass
transfer simulation, remove those from the selection list.

OUT-OF-PLANE THICKNESS
For 2D components, the Thickness field (default value: 1 m) defines a parameter for the
thickness of the geometry perpendicular to the two-dimensional cross-section. Both
constant and varying thicknesses are supported. The value of this parameter is used,

270 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

among other things, to automatically calculate molar flow rates from the total molar
flux.

TRANSPORT MECHANISMS
Use the Convection checkbox, available under Additional transport mechanisms, to
control whether to also include convective transport.

CONSISTENT STABILIZATION
To display this sections, click the Show button ( ) and select Stabilization. Use this
section to control the application of the available consistent stabilization methods;
Streamline diffusion and Crosswind diffusion.

• When the Crosswind diffusion checkbox is selected, a weak term that reduces spurious
oscillations is added to the transport equation. The resulting equation system is
always nonlinear. There are two options for the Crosswind diffusion type:
- Do Carmo and Galeão — the default option. This type of crosswind diffusion
reduces undershoots and overshoots to a minimum but can in rare cases give
equation systems that are difficult to fully converge.
- Codina. This option is less diffusive compared to the Do Carmo and Galeão
option but can result in more undershoots and overshoots. It is also less effective
for anisotropic meshes. The Codina option activates a text field for the Lower
gradient limit glim. Its default value is 0.1[mol/m^3)/[Link], where
[Link] is the local element size.
• For both consistent stabilization methods select an Equation residual. Approximate
residual is the default and means that derivatives of the diffusion tensor components
are neglected. This setting is usually accurate enough and is computationally faster.
If required, select Full residual instead.

INCONSISTENT STABILIZATION
To display this section, click the Show button ( ) and select Stabilization. By default,
the Isotropic diffusion checkbox is not selected, because this type of stabilization adds
artificial diffusion and affects the accuracy of the original problem. However, this
option can be used to get a good initial guess for under resolved problems.

DISCRETIZATION
To display all settings available in this section, click the Show button ( ) and select
Advanced Physics Options.

THE TRANSPORT OF DILUTED SPECIES IN FRACTURES INTERFACE | 271

 The Value type when using splitting of complex variables setting should in most pure
mass transfer problems be set to Real, which is the default. It makes sure that the
dependent variable does not get affected by small imaginary contributions, which can
occur, for example, when combining a Time Dependent or Stationary study with a
frequency-domain study. For more information, see Splitting Complex-Valued
Variables in the COMSOL Multiphysics Reference Manual.

DEPENDENT VARIABLES
The dependent variable name is Concentration c by default. A dependent variable name
must be unique with respect to all other dependent variables in the component.

Add or remove species variables in the model and also change the names of the
dependent variables that represent the species concentrations.

Enter the Number of species. Use the Add concentration ( ) and Remove
concentration ( ) buttons as needed.

FURTHER READING

• Mass Transport in Fractures in the theory section.

• Numerical Stabilization in the COMSOL Multiphysics Reference
Manual.
• Domain, Boundary, and Pair Nodes for the Transport of Diluted
Species Interface
• In the COMSOL Multiphysics Reference Manual, see Table 2-4 for
links to common sections and Table 2-5 for common feature nodes.
You can also search for information: press F1 to open the Help window
or Ctrl+F1 to open the Documentation window.

Boundary, Edge, Point, and Pair Nodes for the Transport of Diluted
Species in Fractures Interface
The Transport of Diluted Species in Fractures Interface has the following boundary,
edge, point, and pair nodes, listed in alphabetical order, available from the Physics

272 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

ribbon toolbar (Windows users), Physics context menu (Mac or Linux users), or by
right-clicking to access the context menu (all users).

• Concentration • No Flux
• Flux • Outflow
• Fracture • Reactions
• Inflow • Species Source
• Initial Values

Furthermore, there are the following attributes to the Fracture node

• Adsorption
• Dispersion

Adsorption
Use this node to model adsorption of the fluid phase species onto the porous media
surface of the fracture.

MATRIX PROPERTIES
Use the Porous material list to define a material specifying the matrix properties on the
current selection. By default the Boundary material is used as The Transport of Diluted
Species in Fractures Interface is active on boundaries only. The Density  of the porous
media is needed when modeling adsorption to the surface of the porous matrix. By
default the option From material is selected.

ADSORPTION
Select an Adsorption isotherm — Langmuir (the default), Freundlich, Toth, BET, or User
defined to specify how to compute cP, the amount of species sorbed to the solid phase
(moles per unit dry weight of the solid). Also activate the checkbox for the species ci
for which adsorption takes place.

• For Langmuir:

THE TRANSPORT OF DILUTED SPECIES IN FRACTURES INTERFACE | 273

 KL c c P K L c Pmax
c P = c Pmax -------------------- -------- = ---------------------------
1 + KL c c  1 + KL c 
2

Enter a Langmuir constant kL,c (SI unit: m3/mol) and an Adsorption maximum
cp,max,c (SI unit: mol/kg):

• For Freundlich:

c N c P cP
c P = K F  -------- -------- = N ----- Freundlich
c ref c c

Enter a Freundlich constant kF,c (SI unit: mol/kg), a Freundlich exponent NF,c
(dimensionless), and a Reference concentration cref,c (SI unit: mol/m3).

• For Toth:
bT c
c P = c Pmax ----------------------------------------------
N 1  NT
 1 +  bT c  T 

and

–  1 + ------
1-
c P N N T
-------- = c Pmax b T  1 +  b T c  T 
c

Enter a Toth constant bT,c (SI unit: m3/mol), a Toth exponent NT,c (dimensionless),
and an Adsorption maximum cp,max,c (SI unit: mol/kg).
• For BET (Brunauer-Emmett-Teller):
KB c0 c
c P = -------------------------------------------------------------
-
 c S – c   1 +  K B – 1  -----
c
cS

and

2 2
K B c 0 c S   K B – 1 c + c S 
-----------------------------------------------------------------
-
2 2
 c – c S   c S +  K B – 1 c 

Enter a BET constant KB,c (dimensionless), a Monolayer adsorption capacity c0,c

(SI unit: mol/kg), and a Saturation concentration cS,c (SI unit: mol/m3).
• For User defined:

274 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

 cP = f  c 

Enter an adsorption isotherm cP,c (SI unit: mol/kg).

For more information, see Adsorption.

FURTHER READING
See the theory chapter in the section Mass Balance Equation for Transport of Diluted
Species in Porous Media.

Concentration
Use this node to specify the species concentration on a fracture boundary (applied in
points in 2D and along edges in 3D). For example, a c = c0 condition specifies the
concentration of species c.

CONCENTRATION
Individually specify the concentration for each species. Select the checkbox for the
Species to specify the concentration, and then enter a value or expression in the
corresponding field. To use another boundary condition for a specific species, click to
clear the checkbox for the concentration of that species.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.
You can find details about the different constraint settings in the section Constraint
Reaction Terms in the COMSOL Multiphysics Reference Manual.

Dispersion
Local variations in the velocity as the fluid flows around solid particles lead to
mechanical mixing, referred to as dispersion. Use this feature to account for dispersion
in a Fracture.

This subfeature is available when the Convection checkbox is selected in the Settings
window for the physics interface.

Select the Specify dispersion for each species individually checkbox to specify the
dispersion tensor DD (SI unit: m2/s) for each species separately. When not selected the
same dispersion tensor DD is used for all species.

Select an option from the Dispersion tensor list — Dispersivity or User defined.

THE TRANSPORT OF DILUTED SPECIES IN FRACTURES INTERFACE | 275

 Select Dispersivity to specify the dispersion in terms of dispersivities (SI unit: m). Select
an option from the Dispersivity model list: Isotropic (the default) or Transverse isotropic
based on the properties of the porous media. For isotropic porous media, specify the
longitudinal and transverse dispersivities. For transverse isotropic porous media,
specify the longitudinal, horizontal transverse, and vertical transverse dispersivities.

For User defined, specify the dispersion components in terms of constants or

expressions. Select Isotropic, Diagonal, Symmetric, or Full to enable the appropriate
tensor components.

Flux
This node can be used to specify the species flux across a boundary of a porous fracture
(applied in points in 2D and along edges in 3D). The flux of species c is defined as

n  d fr  D e c  = d fr N 0

where N0 is an arbitrary user-specified flux expression. For example, N0 can represent

a flux due to chemical reactions, or a phase change. A positive N0 implies that the
concentration inside the fracture increases.

INWARD FLUX
The available options are General inward flux and External convection. If the latter is
chosen, define a Mass transfer coefficient and a Bulk concentration. Specify the flux of
each species individually. To use another boundary condition for a specific species,
make sure that the checkbox for the mass fraction of that species is unchecked.

Inflow
Use this node to specify all species concentrations at a fracture inlet. The condition is
applied in points in 2D and along edges in 3D.

If you want to specify the concentration of a subset of the partaking species, this can
be done by using the Concentration node instead.

CONCENTRATION
For the concentration of each species c0,c (SI unit: mol/m3), enter a value or
expression.

276 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

BOUNDARY CONDITION TYPE
This section in the settings is only available for some products. Search for “Inflow” on
the page: [Link] for more details on
availability.

The option Concentration constraint constrains the concentration values on the

boundary by the use of pointwise constraints. The other option, Flux (Danckwerts) can
be more stable and fast to solve when high reaction rates are anticipated in the vicinity
of the inlet. Oscillations on the solutions can also be avoided in such cases. The latter
condition uses a flux boundary condition based on the velocity across the boundary
and the concentration values. See further details in the theory section.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.
You can find details about the different constraint settings in the section Constraint
Reaction Terms in the COMSOL Multiphysics Reference Manual.

FURTHER READING
See the theory chapter in the section Danckwerts Inflow Boundary Condition.

No Flux
This node can be used to specify that the species flux across a boundary of a porous
fracture is zero. The condition is applied in points in 2D and along edges in 3D.

Outflow
Set this condition at fracture outlets where species are transported out of the model
domain by fluid motion. The condition is applied in points in 2D and along edges in
3D. It is assumed that convection is the dominating transport mechanism across
outflow boundaries, and therefore that diffusive transport can be ignored, that is:

n   – D e c  = 0

Reactions
Use the Reactions node to account for the consumption or production of species
through chemical reactions in the fracture. Define the rate expressions as required.

THE TRANSPORT OF DILUTED SPECIES IN FRACTURES INTERFACE | 277

 BOUNDARY SELECTION
From the Selection list, choose the boundaries on which to define rate expression or
expressions that govern the source term in the transport equations.

Several reaction nodes can be used to account for different reactions in different parts
of the fracture.

REACTION RATES
Add a rate expression Ri for species i. Enter a value or expression in the field. Note that
if you have the Chemistry interface available, included with the Chemical Reaction
Engineering Module, the reaction rate expressions can be automatically generated and
picked up using the drop-down list.

REACTING VOLUME
When specifying reaction rates for a species in a fracture, the specified reaction rate may
have the basis of the pore volume of the fracture, or the total volume.

• For Total volume, the reaction expressions in are specified per unit volume of the
fracture. The reaction expressions will be multiplied by the fracture thickness dfr.
• For Pore volume, the reaction expressions in mol/(m3·s) are specified per unit
volume of total pore space in the fracture. The reaction expressions will be
multiplied by the fracture thickness dfr and the fracture porosity, p.

Species Source
In order to account for consumption or production of species in a fracture, the Species
Source node adds source terms expressions Si to the right-hand side of the species
transport equations.

BOUNDARY SELECTION
From the Selection list, choose the boundaries on which to define expressions that
govern the source term in the transport equations.

If there are several different parts of the fracture, with subsequent and different sources
occurring within them, it might be necessary to remove some boundaries from the
selection. The sources in these can then be defined using an additional Species Source
node.

278 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

SPECIES SOURCE
Add a source term Si for each of the species solved for. Enter a value or expression in
the field of the corresponding species.

THE TRANSPORT OF DILUTED SPECIES IN FRACTURES INTERFACE | 279

 T he T r a ns po r t of Con cen t rat ed
Species Interface
The Transport of Concentrated Species (tcs) interface ( ), found under the Chemical
Species Transport branch ( ) when adding a physics interface, is used to study
gaseous and liquid mixtures where the species concentrations are of the same order of
magnitude and none of the species can be identified as a solvent. In this case, properties
of the mixture depend on the composition, and the molecular and ionic interactions
between all species need to be considered. The physics interface includes models for
multicomponent diffusion, where the diffusive driving force of each species depends
on the mixture composition, temperature, and pressure.

The physics interface solves for the mass fractions of all participating species. Transport
through convection, diffusion, and migration in an electric field can be included.

It supports simulations of transport by convection, migration, and diffusion in 1D,

2D, and 3D as well as for axisymmetric components in 1D and 2D. The physics
interface defines the equations for the species mass fractions, including a diffusion
model (Mixture-averaged, Maxwell–Stefan, or Fick’s law).

The available transport mechanisms and diffusion models differs between various
COMSOL products (see [Link]

Some examples of what can be studied with this physics interface include:

• The evolution of a chemical species transported by convection and diffusion.

• The migration in an electric field in the case of ionic species, in mixtures and
solutions that cannot be deemed as being diluted.
• Concentrated solutions or gas mixtures, where the concentration of all participating
species are of the same order of magnitude, and their molecular and ionic interaction
with each other therefore must be considered. This implies that the diffusive
transport of a single species is dependent on the mixture composition, and possibly
on the temperature, the electric potential, the pressure, or any combination.

When this physics interface is added, the following default nodes are also added in the
Model Builder — Fluid, No Flux, and Initial Values. Then, from the Physics toolbar, add
other nodes that implement, for example, boundary conditions and reactions. You can
also right-click Transport of Concentrated Species to select physics features from the
context menu.

280 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is tcs.

EQUATION
The basic equation for the conservation of mass of a species i is:

 
 i  +     i u  = –   j i + R i (5-2)
t

Using the continuity equation, the sum of all species equations, the equation can be
recast in its nonconservative form:


 ( i) +   u    i = –   j i + R i (5-3)
t

This form of the equation is the one used in the Transport of Concentrated Species
interface. The equation displayed in the interface changes depending on the active
transport mechanisms, the selected diffusion model, and the type of reactions
modeled.

OUT-OF-PLANE THICKNESS
For 2D components, the Thickness field (default value: 1 m) defines a parameter for the
thickness of the geometry perpendicular to the two-dimensional cross-section. Both
constant and varying thicknesses are supported. The value of this parameter is used,
among other things, to automatically calculate mass flow rates from the total mass flux.

CROSS-SECTIONAL AREA
For 1D components, enter a Cross-sectional area Ac (SI unit: m2) to define a parameter
for the area of the geometry perpendicular to the 1D component. Both constant and
varying areas are [Link] value of this parameter is used, among other things,
to automatically calculate mass flow rates from the total mass flux. The default is 1 m2.

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 281

 TRANSPORT MECHANISMS
The Transport of Concentrated Species interface always accounts for transport due to
convection and diffusion.

The available diffusion models and the additional transport mechanisms differs
between various COMSOL products (see [Link]
specifications/).

Diffusion Model
• The Maxwell–Stefan option employs the most detailed diffusion model, but is also
the most computationally expensive. The model is intended for diffusion dominated
models, and requires that the multicomponent Maxwell–Stefan diffusivities of all
component pairs are known.
• The Mixture-averaged option is less computationally expensive than the Maxwell–
Stefan model. It is a simpler model that can be used when variations in the partial
pressures and temperature can be assumed to not affect the multicomponent
diffusion. The model includes stabilization but requires the multicomponent
Maxwell–Stefan diffusivities of all component pairs.
• The Fick’s law model is a general model that should be used when the diffusion is
assumed Fickian, or when no multicomponent diffusivities are available. Also, when
molecular diffusion is not the dominating transport mechanism and a robust but
low order model is wanted, the Fick’s law options should be used. The model
includes stabilization.

Additional Transport Mechanisms

Under Additional transport mechanisms, click to select or clear any combination of
checkboxes as needed.

Migration in Electric Field

Select the Migration in electric field checkbox to activate migration of ionic species due
to an electric field. The resulting migration term is part of the relative mass flux vector.

Mass Transport in Porous Media

The Mass transport in porous media checkbox activates functionality specific to species
transport in porous media. When selected the following domain features are enabled:

• Porous Medium
• Porous Electrode Coupling

Maxwell–Stefan Diffusion Model

When using the Maxwell–Stefan diffusion model the relative mass flux vector is

282 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

 Q T
Di
j i = –  i  D̃ ik d k – -------- T
T
k=1

where D ˜ 2
ik (SI unit: m /s) are the multicomponent Fick diffusivities, dk (SI unit: 1/
T
m) is the diffusional driving force, T (SI unit: K) is the temperature, and D i (SI
unit: kg/(m·s)) is the thermal diffusion coefficient.

The diffusional driving force is defined as

Q
1
d k = x k + ---  x k –  k  p –  k g k +  k
p  l gl (5-4)
l=1

where gk is an external force (per unit mass) acting on species k. In the case of an ionic
species, the external force due to the electric field, which is added by selecting the
Migration in electric field checkbox, is

zk F
g k = – ---------  (5-5)
Mk

where zk is the species charge number, F (SI unit: A·s/mol) is Faraday’s constant and
 (SI unit: V) is the electric potential.

Mixture-Averaged Diffusion Model

When using the Mixture-averaged diffusion model, the diffusive flux is formulated in
terms of a mixture-averaged diffusion coefficient representing the diffusion of each
species into the resulting mixture. The diffusion coefficient is based on the
multicomponent Maxwell–Stefan diffusivities Dik. The Mixture-averaged diffusion
model is computationally less expensive, and significantly more robust than the
Maxwell–Stefan Diffusion Model, but constitutes an approximation of the
multicomponent flux. For information on the flux formulation in this case see
Multicomponent Diffusion: Mixture-Averaged Approximation.

Fick’s Law Diffusion Model

When using the Fick’s law diffusion model, the diffusive flux is formulated in terms of
a Fickian diffusion coefficient. The Fick’s law diffusion model is computationally less
expensive and significantly more robust than the Maxwell–Stefan Diffusion Model, but
constitutes an approximation of the multicomponent flux. For information on the flux
formulation in this case see Multispecies Diffusion: Fick’s Law Approximation.

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 283

 SPECIES
Select the species that this physics interface solves for using the mass constraint in
Equation 5-40 (that is, its value comes from the fact that the sum of all mass fractions
must equal 1). In the From mass constraint list, select the preferred species. To
minimize the impact of any numerical and model introduced errors, use the species
with the highest concentration. By default, the first species is used.

1 = 1 –  i (5-6)
i=2

CONSISTENT AND INCONSISTENT STABILIZATION

To display this section, click the Show button ( ) and select Stabilization.

• Two consistent stabilization methods are available — Streamline diffusion and

Crosswind diffusion. Both are active by default.
The Residual setting applies to both the consistent stabilization methods.
Approximate residual is the default setting and it means that derivatives of the
diffusion tensor components are neglected. This setting is usually accurate enough
and computationally faster. If required, select Full residual instead.
• There is one inconsistent stabilization method, Isotropic diffusion, which is available
when using the Mixture-Averaged Diffusion Model or Fick’s Law Diffusion Model.

ADVANCED SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.
Normally these settings do not need to be changed.

Regularization
From the Regularization list, select On (the default) or Off. When turned On, regularized
mass fractions are calculated such that

0  w i reg  1

Regularized mass fractions are used for the calculation of composition-dependent

material properties, such as the density.

Diffusion
The Diffusion settings are available for the approximate diffusion models
Mixture-averaged and Fick’s law.

284 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

When the Mixture diffusion correction is enabled, a flux correction is added to ensure
that the net diffusive flux is zero. This typically also mean that the solution becomes
less sensitive to the species selected to be computed from the mass constraint in the
Species section. More information on this correction is available in the theory section
Multicomponent Diffusion: Mixture-Averaged Approximation.

The Diffusion flux type list controls the whether the molecular flux is assumed
proportional to the mole fraction or the mass fraction. See Multicomponent Diffusion:
Mixture-Averaged Approximation or Multispecies Diffusion: Fick’s Law
Approximation for information on the diffusive flux formulation.

Pseudo Time Stepping

The Use pseudo time stepping for stationary equation form option adds pseudo time
derivatives to the equation when the Stationary equation form is used in order to speed
up convergence. When selected, a CFL number expression should also be defined. For
the default Automatic option, the local CFL number (from the Courant–Friedrichs–
Lewy condition) is determined by a PID regulator. For more information, see Pseudo
Time Stepping for Mass Transport.

DISCRETIZATION
To display all settings available in this section, click the Show button ( ) and select
Advanced Physics Options.

For more information about these settings, see the Discretization section under The
Transport of Diluted Species Interface.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

• Domain, Boundary, and Pair Nodes for the Transport of Concentrated

Species Interface
• Theory for the Transport of Concentrated Species Interface

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 285

 • Species Transport in the Gas Diffusion Layers of a PEM: Application
Library path Fuel_Cell_and_Electrolyzer_Module/Fuel_Cells/
pem_gdl_species_transport_2d
Web link: [Link]
species-transport-in-the-gas-diffusion-layers-of-a-pem-260

DEPENDENT VARIABLES
Add or remove species in the model and also change the names of the dependent
variables that represent the species concentrations.

Specify the Number of species. There must be at least two species. To add a single
species, click the Add concentration button ( ) under the table. To remove a species,
select it in the list and click the Remove concentration button ( ) under the table.
Edit the names of the species directly in the table.

The species are dependent variables, and their names must be unique with
respect to all other dependent variables in the component.

The Transport of Concentrated Species in Porous Media Interface

This interface ( ), found under the Chemical Species Transport branch ( ), is used
to calculate the chemical composition in a fluid mixture transported through the
interstices of a porous media. The interface is the same as the Transport of
Concentrated Species interface but it has a Porous Medium node added by default. All
other features for porous media transport is always available.

This interface is dedicated to analyzing mass transport in porous media where the
chemical species may be subjected to convection, diffusion, and migration in an
electric field. The interface also includes reaction rate expressions and source terms for
modeling chemical reactions in porous media. It is also possible to include regions with
free flow.

The main feature of the interface is the Porous Medium node which adds the governing
equations for the mass fractions of all present species. It has two subnodes, Fluid and
Porous Matrix, which in turn are used to define the physical properties corresponding
to each phase, and to prescribe the transport properties. The latter are typically the
flow field, and when applicable, the electric field. The Porous Medium node supports

286 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

the use of a Porous Material for defining phase-specific properties, and for sharing
them among different physics interfaces.

When this physics interface is added, the following default nodes are also added to the
Model Builder — Porous Medium, No Flux (the default boundary condition), and Initial
Values. Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions, reaction rate expressions, and species sources. You can also
right-click Transport of Concentrated Species in Porous Media to select physics features
from the context menu.

SETTINGS
The rest of the settings are the same as The Transport of Diluted Species Interface.

Domain, Boundary, and Pair Nodes for the Transport of

Concentrated Species Interface
The Transport of Concentrated Species Interface has these domain, boundary, and pair
nodes, listed in alphabetical order, available from the Physics ribbon toolbar (Windows
users), Physics context menu (Mac or Linux users), or right-click to access the context
menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

For axisymmetric components, COMSOL Multiphysics takes the axial

symmetry boundaries (at r = 0) into account and automatically adds an
Axial Symmetry node to the component that is valid on the axial symmetry
boundaries only.

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 287

 These nodes are described in this section or as indicated:

• Electrode Surface Coupling • Porous Matrix

• Equilibrium Reaction • Porous Medium
• Fluid (Porous Medium) • Porous Electrode Coupling1
• Flux • Reaction
• Flux Discontinuity • Reaction Coefficients1
• Inflow • Reaction Sources
• Initial Values • Species Properties
• Mass Fraction • Surface Equilibrium Reaction
• No Flux • Symmetry
• Open Boundary • Thin Impermeable Barrier1
• Out-of-Plane Flux • Fluid
• Outflow

Some features require certain add-on modules. See details https://

[Link]/products/specifications/
1 The node is described for the Transport of Diluted Species interface.

Prescribing Conditions on Fluid-Solid Interfaces

Defining conditions on interfaces between different phases, typically a wall in contact
with a fluid, it is worthwhile to be careful with how variables in the conditions are
evaluated. COMSOL Multiphysics supports that variables are defined differently on
adjacent domains (or any other geometry dimension). Evaluating a variable with
differing definitions, the mean value is returned on a boundary separating the different
domains. The density, pressure, or temperature are examples of variables that may
differ in a fluid and a solid. In order to specify on which side a variable is evaluated, the
up and down operators (described in the COMSOL Multiphysics Reference Manual)
can be used.

Species Properties
Use this node to define parameters specific to each of the species.

288 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The node will change its label depending on what inputs are available in the node. If
only Molar Mass is available, it will be labeled “Species Molar Masses”.

MOLAR MASS
Enter a value or expression for the Molar mass Mw for each species. The default value
is 0.032 kg/mol, which is the molar mass of O2 gas.

CHARGE
This section is available when the Migration in electric field checkbox is selected in the
Transport Mechanisms section of the interface. Enter the charge number zw
(dimensionless, but requires a plus or minus sign) for each species.

Fluid
The Fluid node is the main node used to model mass transfer in a fluid mixture with
the Transport of Concentrates species interface. The node adds the equations
governing the mass fractions of all present species, and provides inputs for the
transport mechanisms and for the material properties of the fluid mixture.

The settings in this node are dependent on the checkboxes selected under Transport
Mechanisms in the Settings window of the Transport of Concentrated Species
interface.

The options available in this feature differs between COMSOL products. (See https:/
/[Link]/products/specifications/).

MODEL INPUTS
Specify the temperature and pressure to be used in the physics interface. The
temperature model input is used when calculating the density from the ideal gas law,
but also when thermal diffusion is accounted for by supplying thermal diffusion
coefficients. The pressure model input is used in the diffusional driving force in
Equation 5-4 (that is, when a Maxwell–Stefan Diffusion Model is used) and when
calculating the density from the ideal gas law.

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 289

 Temperature
Select the source of the Temperature field T:

• Select User defined to enter a value or an expression for the temperature (SI unit: K).
This input is always available.
• If required, select a temperature defined by a Heat Transfer interface present in the
model (if any). For example, select Temperature (ht) to use the temperature defined
by the Heat Transfer in Fluids interface with the ht name.

Absolute Pressure
Select the source of the Absolute pressure p:

• Select User defined to enter a value or an expression for the absolute pressure
(SI unit: Pa). This input is always available.
• In addition, select a pressure defined by a Fluid Flow interface present in the model
(if any). For example, select Absolute pressure (spf) to use the pressure defined in a
Laminar Flow interface with spf as the name.

DENSITY
Define the density of the mixture and the molar masses of the participating species.

Mixture Density
Select a way to define the density from the Mixture density list — Ideal gas or User
defined:

• For Ideal gas, the density is computed from the ideal gas law in the manner of:
pM-
 = ----------
Rg T

Here M is the mean molar mass of the mixture and Rg is the universal gas constant.
The absolute pressure, p, and temperature, T, used corresponds to the ones defined
in the Model Inputs section.
• For User defined enter a value or expression for the Mixture density .

CONVECTION
Select the source of the Velocity field u:

• Select User defined to enter manually defined values or expressions for the velocity
components. This input is always available.
• Select a velocity field defined by a Fluid Flow interface present in the model (if any).
For example, select Velocity field (spf) to use the velocity field defined by the Fluid

290 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

 Properties node fp1 in a Single-Phase Flow, Laminar Flow interface with spf as the
Name.

DIFFUSION
Specify the molecular and thermal diffusivities of the present species based on the
selected Diffusion model.

When using a Maxwell–Stefan Diffusion Model or a Mixture-Averaged Diffusion

Model, select the Binary diffusion input type (Table or Matrix) and specify the Maxwell–
Stefan diffusivities in the table or matrix, then enter the Thermal diffusion coefficients
T
Dw .
F
When using a Fick’s Law Diffusion Model, specify the Diffusion coefficient D w and the
T
Thermal diffusion coefficients D w for each of the species.

Maxwell–Stefan Diffusivity Matrix

Using a Maxwell–Stefan Diffusion Model or a Mixture-Averaged Diffusion Model, the
Maxwell–Stefan diffusivity matrix Dik (SI unit: m2/s) can be specified by a table or
matrix. For a simulation involving Q species the Maxwell–Stefan diffusivity matrix is a
Q-by-Q symmetric matrix, where the diagonal components are 1. Enter values for the
upper triangular components, Dij, which describe the interdiffusion between species i
and j. For the table input type, only upper triangular components (Dij) are listed. The
name of species pair consists of species in the first and second column. For the matrix
input type, the numbering of the species corresponds to the order, from top to
bottom, used for all the input fields for species properties (see for example the molar
mass fields in the Density section). The Maxwell–Stefan diffusivity matrix is used to
compute the multicomponent Fick diffusivities as described in Multicomponent
Diffusivities.

Diffusion Coefficient
Using a Fick’s Law Diffusion Model, the diffusion is by default assumed to be isotropic
F
and governed by one Diffusion coefficient D w (SI unit: m2/s) for each species. To allow
for a general representation, it is also possible to use diffusion matrices (diagonal,
symmetric, or anisotropic).

Thermal Diffusion Coefficient

T
To model thermal diffusion, prescribe the Thermal diffusion coefficients D i
(SI unit: kg/(m·s)), by entering one thermal diffusion coefficient for each species in
the corresponding field. In a multicomponent mixture, the sum of the thermal
diffusion coefficients is zero. The default value for all thermal diffusion coefficients
is 0.

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 291

 Specify the molecular and thermal diffusivities of the present species based on the
selected Diffusion model.

MIGRATION IN ELECTRIC FIELD

This section is available when the Migration in electric field checkbox is selected for the
Transport of Concentrated Species interface.

Electric Potential
• Select User defined to enter a value or expression for the electric potential. This input
is always available.
• If required, select an electric potential defined by an AC/DC interface that is
present in the model (if any). For example, select Electric potential (ec) to use the
electric field defined by the Current Conservation node cucn1 in an Electric
Currents interface ec.

Settings for the mobilities are used for the Mixture-averaged and Fick’s law transport
models. By default the mobility is set to be calculated based on the species diffusivities
and the temperature using the Nernst-Einstein relation. To manually specify the
mobilities, select User defined for the mobility um,c (SI unit: s·mol/kg) and enter one
value for each species.

The temperature (if you are using mobilities based on the Nernst–Einstein relation) is
taken from the Model Inputs section.

Porous Medium
Use this node to model the composition in a gas or liquid mixture as it is transported
though the interstices of solid porous medium. In addition to transport due to
convection and diffusion, the node contains functionality to include species evolution
through reactions in the fluid phase.

Species transport in a porous medium is affected both by properties of the fluid phase
as well as properties of the solid matrix. These properties are defined in the Fluid
(Porous Medium) node and the Porous Matrix node respectively. The Porous Medium
node supports material properties using a Porous Material node, where properties are
defined per phase in a similar manner.

Fluid (Porous Medium)

Use this node to specify the mass transport in a fluid phase filling the pores of a porous
medium. It is used as a subnode to Porous Medium.

292 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

MODEL INPUTS
Specify the temperature and pressure in the fluid. The temperature model input is used
when calculating the density from the ideal gas law, but also when thermal diffusion is
accounted for by supplying thermal diffusion coefficients. The pressure model input is
used in the diffusional driving force in Equation 5-4 (that is, when a Maxwell–Stefan
Diffusion Model is used) and when calculating the density from the ideal gas law.

Temperature
Select the source of the Temperature field T:

• Select User defined to enter a value or an expression for the temperature.

• When present, select a temperature defined by a Heat Transfer interface in the
model. For example, select Temperature (ht) to use the temperature defined by the
Heat Transfer in Fluids interface with the ht name.

Absolute Pressure
Select the source of the Absolute pressure p:

• Select User defined to enter a value or an expression for the absolute pressure.
• When present, select a pressure defined by a Fluid Flow interface present in the
model. For example, select Absolute pressure (spf) to use the pressure defined in a
Laminar Flow interface with spf as the name.

DENSITY
Use this section to define the density of the fluid phase, and to specify the molar masses
of the participating species.

Mixture Density
Select a way to define the density from the Mixture density list — Ideal gas or User
defined:

• For Ideal gas, the density is computed from the ideal gas law in the manner of:
pM-
 = ----------
Rg T

Here M is the mean molar mass of the mixture and Rg is the universal gas constant.
The absolute pressure, p, and temperature, T, used corresponds to the ones defined
in the Model Inputs section.
• For User defined enter a value or expression for the Mixture density .

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 293

 CONVECTION
Select the source of the Velocity field u:

• Select User defined to enter values or expressions for the velocity components. This
input is always available.
• Select a velocity field defined by a Fluid Flow interface that solves for the velocity of
the fluid. For example, select Velocity field (spf) to use the velocity field defined by
in a Single-Phase Flow, Laminar Flow interface with spf as the Name.

When the interface is used in a reacting flow multiphysics coupling, the velocity is
automatically defined and the input is disabled.

DIFFUSION
Specify the species molecular and thermal diffusivities in fluid phase in the manner
described for the Fluid node.

To account for the effect of porosity in the diffusivities, select an Effective diffusivity
model — Millington and Quirk model, Bruggeman model, Tortuosity model, or No
correction. Using one of the first four models, the effective transport factor, fe, is
defined from the porosity and the fluid tortuosity factor in the manner of:


f e = ----p- (5-7)
F

For No correction, the effective transport factor is set to one.

–1  3
• For the Millington and Quirk model, the effective transport factor is  F =  p .
–1  2
• For the Bruggeman model, the effective transport factor is  F = p .
• For the Tortuosity model, specify the tortuosity factor F. Select either Isotropic to
define a scalar value, or Diagonal or Symmetric to define anisotropic tensor values.

The species diffusivities and mobilities are automatically adjusted for porous media
transport using the effective transport factor.

294 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

MIGRATION IN ELECTRIC FIELD
This section is available when the Migration in electric field checkbox is selected in the
Transport Mechanisms section of the interface. Select the source of the Electric
potential V:

• Select User defined to enter a value or expression for the electric potential.
• When present, select an electric potential defined by an AC/DC interface that is
present in the model. For example, select Electric potential (ec) to use the electric
field defined an Electric Currents interface ec.

Settings for the mobilities are needed for the Mixture-averaged and Fick’s law diffusion
models. By default the mobility is set to be calculated based on the species diffusivities
(adjusted by the Effective diffusivity model in the Diffusion section) using the
Nernst-Einstein relation. To manually specify the mobilities, select User defined for the
mobility um,w and enter one value for each species.

The temperature (if you are using mobilities based or the Nernst–Einstein relation) is
taken from the Model Inputs section.

PORE WALL INTERACTIONS

In porous media, it is possible to include Pore Wall Interactions. This accounts for
species collisions with the surrounding media, that is, the pore walls.
M
For the Fick’s law and Mixture-averaged diffusion models, the diffusion coefficient D i
W
is corrected with the Wall diffusion coefficient D i in the following way

MW  1 1 -
-1
Di =  --------
- + -------- 
 D M D W
i i

For the Maxwell-Stefan diffusion model, the following term is added to the species force
balance

xi  ji
- --------- + u W = d i
 + ------------
W   
D i eff i

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 295

 For weakly adsorbing gases, the Knudsen formula can be applied for computing the wall
diffusivities, based on the Pore diameter dpore (SI unit: m). For other cases, choose User
defined.

For isobaric systems featuring concentration gradients in porous media,

concentration gradients may give rise to nonzero fluid velocity as a result
of the wall interactions. This effect can be modeled using the
Maxwell-Stefan diffusion model in combination Pore Wall Interactions. For
this case the wall velocity variable uW ([Link]) can be used to contribute
to the fluid momentum balance (either as a Contributing Velocity in Darcy’s
law, or as a Volume Force in either the Brinkman Equations or Free and
Porous Media Flow, Brinkman interfaces. The multiphysics Reacting Flow
nodes incorporates this contribution from uW automatically.

Porous Matrix
This node defines the porosity, when modeling transport in a Porous Medium. The
porosity corresponds to the volume fraction occupied by the species containing fluid.

• By default the porosity of the solid matrix, p, is taken From material. The Porous
Matrix node supports the use of a Porous Material added to the Materials branch. In
this case the porosity is defined in the manner of

p = 1 –  s i – imf i
i i

where s,i and imf,i are the volume fractions given in Solid and Immobile Fluids
subnodes under the Porous Material node.
• Select From pellet bed densities to compute the porosity using the (dry bulk) Bed
density b and the (single phase) Pellet density pe. The porosity is then defined from

b
 p = 1 – -------
-
 pe

• For User defined, enter a value or expression for the porosity.

Electrode Surface Coupling

Use this node to define a flux boundary condition based on current densities of one or
multiple Electrode Reaction nodes in an Electrochemistry interface.

296 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The flux is proportional to the current densities and the stoichiometric coefficients
according to Faraday’s law as defined by summation over the Reaction Coefficients
subnodes. The molar fluxes are multiplied by the species molar masses to obtain the
corresponding mass fluxes.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

MASS TRANSFER TO OTHER PHASES

When using the Reacting Flow coupling feature to model coupled momentum and
mass transport, boundary conditions that results in a nonnegligible mass flux can be
consistently handled by accounting for The Stefan Velocity.

Select Account for Stefan velocity to update the Stefan velocity in accordance with the
mass flux from the electrode reactions. One example that may benefit from this is when
modeling gas diffusion electrodes.

Reaction
Use the Reaction node to specify the reaction kinetics for a single homogeneous
chemical reaction. The resulting mass source terms for the species involved in the
reaction are automatically defined and added on the selected domains. For turbulent
flow, the Reaction node includes the Eddy-dissipation turbulent-reactions model.

REACTION RATE
Select a Reaction rate — Automatic (the default), or User defined. Selecting Automatic
the laminar flow reaction rate is computed using the mass action law

For User defined, input a custom expression or constants for the Reaction rate r.

Specify the reaction stoichiometry by entering values for the stoichiometric coefficients
(dimensionless) of each species. Enter negative values for reactants and positive values
for products.

RATE CONSTANTS
When the Use Arrhenius expressions checkbox is not selected, input custom expressions
or constants for the Forward rate constant kf and Reverse rate constant kr.

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 297

 When the Use Arrhenius expressions checkbox is selected, enter values for the following
parameters of the forward and reverse reactions:

• Forward frequency factor Af and Reverse frequency factor Ar (dimensionless)

• Forward activation energy Ef and Reverse activation energy Er
• Forward temperature exponent nf and Reverse temperature exponent nr
(dimensionless)

TURBULENT FLOW
Note this section is only available when then licensed to the CFD Module (see https:/
/[Link]/products/specifications/).

When the Turbulent-reaction model is set to None, laminar flow is assumed and the
reaction source terms are defined from the reaction stoichiometry and reaction rates
prescribed.

When the Turbulent-reaction model is set to Eddy-dissipation, turbulent flow will be

accounted for in the reaction mass sources. In this case, enter values for the Turbulent
reaction model parameters ED and  ED (dimensionless).

The Eddy-dissipation model also requires an estimation of the turbulent mixing time of
the fluid flow turbulence. When a Fluid Flow interface defining it is present in the
model, it can be selected from the Turbulence time scale list. For example, select
Turbulence time scale (spf/fp1) to use the time scale defined by the Fluid Properties node
fp1 in a Turbulent Flow, k- interface with the Name set to spf.

REGULARIZATION
Select Rate expression in order to regularize the individual rate expressions that are
added to each species. If the mass fraction for a reactant species i becomes smaller
than its damping limit, idl, the rate expression added to species i is reduced linearly.
If i  0 for a reactant species, the reaction rate contribution to that species is
completely removed. Similarly, if the mass fraction for a product species j becomes
larger than 1  jdl, the rate expression added to that species is damped linearly. If j 
1 for a product species, the reaction rate contribution to that species is completely
removed.

The default value for the damping limit, idl, is 106, which is appropriate for most
applications, but can require adjustment when working with for example catalytic trace
species.

298 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Regularization of the rate expressions adds considerable stability to the reaction
expressions, but should optimally only be used as a mean to reach convergence. If the
regularization is active in too large parts of the domain, the mass balance can become
affected since the regularization acts on the contributions to each individual species,
not the reaction as a whole.

Reaction Sources
Use the Reaction Sources node to define mass source terms from one or more chemical
reactions. Both homogeneous reactions as well as heterogeneous reactions occurring
in a porous media can be studied. The node also includes the possibility to pick up
automatically defined source term definitions from a Chemistry interface.

REACTIONS
Add an expression for the reaction mass source, Ri, for each individual species present,
except for the one computed from the mass constraint (see Species). Enter a value or
expression in the field for the corresponding species.

Select the Mass transport to other phases checkbox if mass is leaving or entering the
fluid as a result of the reactions, for instance due to condensation or vaporization in a
porous matrix. In this case the mass source for the species calculated from the mass
constraint can also be specified. The net mass transfer corresponds to the sum of the
mass sources for all species.

REACTING VOLUME
When specifying reaction sources for a species in porous media, the specified mass
source may have the basis of the total volume, or the pore volume. For nonporous
domains, the Reacting Volume setting has no impact.

• For Total volume, the reaction mass source expressions are specified per unit volume
of the model domain.
• For Pore volume, the reaction mass source expressions are specified per unit volume
pore space. In this case the reaction mass sources will be multiplied by the domain
porosity p (p equals unity for nonporous domains).

Initial Values
The Initial Values node adds initial values for the mass fractions that can serve as an
initial condition for a transient simulation, or as an initial guess for a nonlinear solver.
If required, add additional Initial Values nodes from the Physics toolbar.

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 299

 INITIAL VALUES
The initial mass fractions can be specified using a number of quantities. Select the type
of input from the Mixture specification list. Select:

• Mass fractions (the default) to enter mass fractions (01 for example)
• Mole fractions to enter mole fractions (x01 for example)
• Molar concentrations (SI unit: mol/m3) to enter molar concentrations (c01 for
example)
• Number densities (SI unit: 1/m3) to enter number densities (n01 for example)
• Densities (SI unit: kg/m3) to enter densities (01 for example)

Enter a value or expression in the field for each species except for the one computed
from the mass constraint.

INITIAL MIXTURE DENSITY

When the selecting Molar concentrations, Number densities, or Densities are selected
from the Mixture specification list, the should also be specified. Select Ideal gas or User
defined from the Initial mixture density list.

• For Ideal gas, also specify the Initial pressure p0 and the Initial Temperature T0. Note
that dependent variables solved for are evaluated to zero for initial values. When
solving for pressure or temperature together with the mass fractions, apply the initial
values from the corresponding interfaces here as well.
• For User defined, input a custom for the Initial mixture density m0.

Mass Fraction
The Mass Fraction node adds boundary conditions for the species mass fractions. For
example, the following condition specifies the mass fraction of species i: i = i,0.

Set the mass fractions of all species except the one computed from the mass constraint.
This ensures that the sum of the mass fractions is equal to one (see Species). This node
is available for exterior and interior boundaries.

When defining conditions on boundaries between different phases, see the section
Prescribing Conditions on Fluid-Solid Interfaces.

MASS FRACTION
Specify the mass fraction for each species individually. Select the checkbox for the
species to specify the mass fraction, and enter a value or expression in the

300 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

corresponding field. To use another boundary condition for a specific species, click to
clear the checkbox for the mass fraction of that species.

MASS TRANSFER TO OTHER PHASES

When using the Reacting Flow coupling feature to model coupled momentum and
mass transport, boundary conditions that results in a nonnegligible mass flux can be
consistently handled by accounting for The Stefan Velocity.

Select Account for Stefan velocity to update the Stefan velocity in accordance with the
prescribed mass fractions. Examples of cases that may benefit from this are, for
example, when modeling surface reactions or phase change on an exterior boundary.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

Flux
The Flux node is available on exterior boundaries and can be used to specify the mass
flux. The boundary mass flux for each species is defined in the manner of

– n  j i = j 0 i (5-8)

where n denotes the outward pointing normal of the boundary. The prescribed flux
j0,i can contain an arbitrary flux expression. It can, for example, be used to represent
a heterogeneous reaction or a separation process occurring at the boundary. The flux
can, for example, be a function of i, the temperature, the pressure or even the electric
potential.

When defining conditions on boundaries between different phases, see the section
Prescribing Conditions on Fluid-Solid Interfaces.

CONVECTION
By default, the flux due to diffusion and migration in an electric field is prescribed. This
is the appropriate flux condition when the relative velocity at the boundary is zero.
When the fluid velocity is not equal to that of the boundary, for example due to a fluid
injection, it is often convenient to prescribe the total flux including convection. To do
so select Include in the Convection section. In this case the prescribed flux is defined as:

– n   j i + u i  = j 0

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 301

 MASS TRANSFER TO OTHER PHASES
When using the Reacting Flow coupling feature to model coupled momentum and
mass transport, boundary conditions that results in a nonnegligible mass flux can be
consistently handled by accounting for The Stefan Velocity.

Select Account for Stefan velocity to update the Stefan velocity in accordance with the
prescribed flux. Examples of cases that benefit from this are for example when
modeling surface reactions or phase change on an exterior boundary. The Stefan
velocity represents the mixture velocity resulting from mass fluxes at the wall. This
section is not available when Include is selected in the Convection section.

INWARD FLUX
Specify the Inward flux for each species individually. Select the checkbox for the species
to prescribe a flux for and enter a value or expression in the corresponding field. To
use another boundary condition for a specific species, click to clear the checkbox for
the flux of that species. Use a positive value for an inward flux.

External convection
Set Flux type to External convection to prescribe a mass flux to or from an exterior
domain (not modeled) assumed to include convection. The exterior can for example
include a forced convection to control the temperature or to increase the mass
transport. In this case the prescribed mass flux corresponds to

j 0 i = k  i   b i –  i  (5-9)

where ki is a mass transfer coefficient and b,i is the bulk mass fraction, the typical
mass fraction far into the surrounding exterior domain.

Inflow
The Inflow node adds a boundary condition for an inflow boundary, defining the
composition of the mixture. The node is available for exterior boundaries. The mixture
composition can be specified using the following quantities:

• Mass fractions: i  0,i

• The mole fraction: xi  x0,i
• The molar concentration: ci  c0,i
• The number density, which describes the number of particles per volume: ni  n0,i

302 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

• The density: i  0,i
• The mass flow rates.

The node requires input for all species. (The Mass Fraction node can be used to specify
boundary mass fractions for a subset of the active species in the interface.)

INFLOW
Select a Mixture specification:

• Mass fractions (the default) to enter mass fractions (01, for example)
• Mole fractions to enter mole fractions (x01 for example)
• Molar concentrations (SI unit: mol/m3) to enter molar concentrations (c01, for
example)
• Number densities (SI unit: 1/m3) to enter number densities (n01, for example)
• Densities (SI unit: kg/m3) to enter densities (01, for example)
• Mass flow rates (SI unit: kg/s) to enter the total mass flow over the boundary
(Jin1, for example)

Enter a value or expression in the field for each species except for the one computed
from the mass constraint.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

No Flux
The No Flux node is the default boundary condition available for exterior boundaries.
It should be used on boundaries across which there is no mass flux, typically exterior
solid walls where no surface reactions occur. The condition applied for each species
corresponds to

–n  ji = 0

where n denotes the outward pointing normal of the boundary.

Out-of-Plane Flux
The out-of-plane flux node is used to prescribe a flux in an out-of-plane domain, and
it can be used to reduce a model geometry to 2D, or even 1D, when the variation of
mass fraction is small in one or more directions. This could be the case for example

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 303

 when the object to model is thin or slender. Figure 5-1shows examples of possible
situations in which this type of geometry reduction can be applied.

jz

jup

jdown

Figure 5-2: Geometry reduction from 3D to 1D (top) and from 3D to 2D (bottom).

For a 1D component this node adds a single out-of-plane mass flux j0,z,i for species i.
For a 2D component two fluxes can be prescribed for each species; j0,u,i for the upside
of the domain, and j0,d,i for the downside of the domain.

The fluxes are added as a source contribution to the right-hand side of the transport
equation. For example, in 2D components with diffusion, convection, and
out-of-plane flux present, the mass transfer of species i is defined by the mass
conservation equation


 ( i) +   u    i +   j i = R i + S opf i , (5-10)
t

where Sopf,i is the out-of-plane source for species i

j 0 i j 0 u i + j 0 d i
- = ----------------------------------- .
S opf i = -------
dz dz

For external mass transfer on the upside and the downside of the domain, the
out-of-plane flux is

j 0 i = k w u i  w b u i – w i  + k w d i  w b d i – w i 

where ki is a mass transfer coefficient and b,i is the bulk mass fraction, the typical
mass fraction far into the surrounding exterior domain. The prescribed flux, j0, can
include any arbitrary user-specified expressions. It can be a constant or a function of a
dependent variable or independent variables.

304 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

UPSIDE INWARD FLUX
For 1D components, enter the cross-sectional perimeter Pc to get the out-of-plane flux

j 0 i = P c j 0 z i .

The default value of Pc is the circumference. Either keep the default value, for a circular
cross-section shape, or edit the value to get a user-defined shape of the out-of-plane
cross-section.

The available options are General inward flux and External convection. Specify the Inward
flux for each species individually. Select the Species checkbox for the species for which
to specify the flux, and enter a value or expression in the corresponding field. To use
another boundary condition for a specific species, click to clear the checkbox for the
flux of that species. Use a positive value for an inward flux.

Set Flux type to External convection to prescribe a mass flux to or from an exterior
domain (not modeled) assumed to include convection. The exterior can for example
include a forced convection to control the temperature or to increase the mass
transport. In this case the prescribed mass flux corresponds to

j 0 i = k  i   b i –  i 

where ki is a mass transfer coefficient and b,i is the bulk mass fraction, the typical
mass fraction far into the surrounding exterior domain.

DOWNSIDE INWARD FLUX

The available options are General inward flux and External convection. The settings are
the same as for the Upside Inward Flux section.

FURTHER READING
For an example of using the Out-of-Plane Flux node, see this application example:

Cross-Flow Mass Transfer in a Thin Domain: Application Library path

Chemical_Reaction_Engineering_Module/Tutorials/thin_domain

Outflow
The Outflow node is the preferred boundary condition at outlets where the species are
to be transported out of the model domain. It is useful, for example, in mass transport
models where it is assumed that convection is the dominating effect driving the mass

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 305

 flow through the outflow boundary. This node is available for exterior boundaries. The
boundary condition is applied to all species and corresponds to one of the following
equations depending on the selected diffusion model:

• For the Mixture-Averaged Diffusion Model:

m
– n  D i  i = 0

• For the Fick’s Law Diffusion Model:

f
– n  D i  i = 0

Symmetry
The Symmetry node can be used to represent boundaries where the species
concentration is symmetric; that is, there is no mass flux in the normal direction across
the boundary:

– n  N = – n    i u + j i  = 0

This boundary condition is identical to the No Flux node, but applies to all species and
cannot be applied to individual species. The Symmetry node is available for exterior
boundaries.

Flux Discontinuity
The Flux Discontinuity node represents a discontinuity in the mass flux across an interior
boundary:

–n   Nd – Nu  = N0 N =   i u + j i 

where the value of N0 specifies the size of the flux jump evaluated from the down to
the upside of the boundary.

FLUX DISCONTINUITY
Specify the jump in species mass flux. Use a positive value for increasing flux when
going from the downside to the upside of the boundary. The boundary normal points
in the direction from the downside to the upside of an interior boundary and can be
plotted for visualization.

Select the Species checkboxes to specify a flux discontinuity, and enter a value or
expression for the Flux discontinuity N0 (SI unit: kg/(m2·s)) in the corresponding field,

306 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

N0, w1, for example. To use a different boundary condition for a specific species, click
to clear the checkbox for the flux discontinuity of that species.

Open Boundary
Use the Open Boundary node to set up mass transport across boundaries where both
convective inflow and outflow can occur. Use the node to specify an exterior species
composition on parts of the boundary where fluid flows into the domain. A condition
equivalent to the Outflow node applies to the parts of the boundary where fluid flows
out of the domain. The direction of the flow across the boundary is typically calculated
by a Fluid Flow interface and is entered as Model Inputs.

EXTERIOR COMPOSITION
Enter a value or expression for the species composition. Select:

• Mass fractions (the default) to enter mass fractions (0,1, for example)
• Mole fractions to enter mole fractions (x0,1, for example)
• Molar concentrations (SI unit: mol/m3) to enter molar concentrations (c0,1, for
example)
• Number densities (SI unit: 1/m3) to enter number densities (n0,1, for example)
and to describe the number of particles per volume n  n0
• Densities (SI unit: kg/m3) to enter densities (0,1, for example)

A concentration quantity other than the mass fractions can only be used
when all species are defined.

Equilibrium Reaction
Use this node to model a reaction where the kinetics is assumed so fast that the
equilibrium condition is fulfilled at all times. The node solves for an additional degree
of freedom (the reaction rate Req) to fulfill the equilibrium condition at all times in all
space coordinates.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 307

 EQUILIBRIUM CONDITION
Selecting Equilibrium constant in the list, the following equilibrium condition based on
the species activities and the law of mass action is used


 ai i
i  products
K eq = ----------------------------------
–
 ai i
i  reactants

where i are the stoichiometric coefficients and the species activities are defined from
the concentration, ci, and the unit activity concentration ca0.

c
a i = ------i-
c a0

Enter a value or expression for the dimensionless Equilibrium constant Keq, and the Unit
activity concentration Ca0.

Select User defined from the list to instead enter a manually defined Equilibrium
expression Eeq.

STOICHIOMETRIC COEFFICIENTS
Enter a value for the stoichiometric coefficient for all participating species. Use
negative values for reactants and positive values for products in the modeled reaction.
Species with a stoichiometric coefficient value of 0 are not affected by the Equilibrium
Reaction node.

Surface Equilibrium Reaction

Use this node to model an equilibrium reaction on a boundary (surface). The kinetics
of the reaction is assumed so fast that the equilibrium condition is fulfilled at all times.
The node solves for an additional degree of freedom (the mass flux Jeq) to fulfill the
equilibrium condition at all times in all space coordinates along the boundary.

This feature is only available in a limited set of add-on products. See

[Link] for more details on
availability.

The settings for this node are the same as for Equilibrium Reaction except for the
setting in the section below.

308 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

MASS TRANSFER TO OTHER PHASES
When using the Reacting Flow coupling feature to model coupled momentum and
mass transport, boundary conditions that results in a nonnegligible mass flux can be
consistently handled by accounting for The Stefan Velocity. Select Account for Stefan
velocity to update the Stefan velocity in accordance with the prescribed mass fractions.
Examples of cases that may benefit from this are for example when modeling surface
reactions or phase change on an exterior boundary.

THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 309

 The Chemistry Interface
The Chemistry (chem) interface ( ) is found under the Chemical Species Transport
branch ( ) when adding a physics interface.

This physics interface is a tool for generating a set of variables to be used for modeling
chemical species and reactions systems. The variables are generated from species and
reaction properties and can be divided in two categories:

• Rate expressions and heat sources for use in mass and heat balances.
• Material property variables (mixture density, diffusivities, viscosity, and so on) for
use in space-dependent transport equations.

Many of the fields and nodes described in this section are only made available when
either a Reaction or a Species (or both) subnode is added to the Model Builder. All
predefined constants and expressions can be overwritten by user-defined expressions.
This makes it possible to go beyond the modeling assumptions set as defaults in this
physics interface.

The following is a description of the features and fields available in the Settings window
for the Chemistry interface.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is chem.

MODEL INPUT
This section sets the Temperature, Pressure, and Electrode potential (only available with
a Battery Design Module, Corrosion Module, Electrochemistry Module,
Electrodeposition Module, or Fuel Cell & Electrolyzer Module license) to be used by
the current interface. Toggle the Enable electrode reactions checkbox to enable the
Electrode Potential input in the Model Input section, as well as the Electrode Reaction
and Electrode Reaction Group features. Use the lists to select a temperature, pressure,

310 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

or electrode potential defined and announced by another interface in the model. For
example, when a heat transfer interface is also included, the temperature solved for is
available in the Temperature list.

For Temperature or Pressure, you can also select Common model input to use a globally
available common model input. In all three cases, select User defined to manually define
the variable in question.

MIXTURE PROPERTIES

Type
Use this setting to define specify what kind of mixture assumption is used. For Diluted
Species the mixture is composed of low concentration solutes present in a solvent. For
Concentrated Species a mixture where no single species is considered to be in excess is
assumed.

Select Diluted species from the Type list to use the concentration variables from a
Transport of Diluted Species interface in the Species Matching section. The same
setting should be used for any other interface solving for molar concentrations using a
diluted species assumption. Select Concentrated species from the Type list to use the
mass fractions from a Transport of Concentrated Species interface.

Thermodynamics
All property parameters and property functions required by the interface can be
automatically created by coupling to a system added to the Thermodynamics node. To
do so, click the Thermodynamics checkbox and select an existing Thermodynamic
System.

The Thermodynamics checkbox is enabled when the Thermodynamics node, including

one or more systems, is available under Global Definitions. Also, the Chemistry interface
needs to include at least one species.

Phase
Use the Phase list to specify the state of aggregation of the mixture.

SPECIES MATCHING
This section is used to match species solved for in space dependent physics interfaces
to the species defined in Chemistry. By doing so the variables for the reaction kinetics
and the mixture properties defined by Chemistry can be applied to study chemically
reacting or composition dependent systems.

THE CHEMISTRY INTERFACE | 311

 Species Solved For
Use this list to populate the inputs in the Bulk species table with dependent variables
from a certain mass transfer interface. The Species solved for list contains all present
interfaces of the Type selected in the Mixture Properties section. For example, when the
Type is Diluted species, all Transport of Diluted Species interfaces or variants thereof,
such as Transport of Diluted Species in Porous Media interface, are available in the list.

Bulk Species
Use the Bulk species table to specify the concentrations to be used as arguments in
reaction kinetics variables, for example the reaction rate. Reaction kinetics variables are
generated by the species features (Species and Species Group) and by the reaction
features (Reaction, Electrode Reaction, Reversible Reaction Group, Electrode
Reaction Group, and Equilibrium Reaction Group).

The bulk species concentrations are also used in the mixture properties, both for
transport properties like the density and viscosity, and thermodynamic properties like
the enthalpy or heat capacity.

When a mass transfer interface has been selected in the Species solved for list, use the
Molar concentration column to select one of the dependent variables in the selected
interface to the corresponding Species in Chemistry. When using a concentrated
species interface (defined by the Type list in the Mixture Properties section), instead
use the Mass fraction column to select a dependent variable for each species in
Chemistry. For a consistent set up, the species molar mass in the Transport of
Concentrated Species interface is updated to the one defined in the Chemistry
interface.

The Molar concentration and Mass fraction columns also includes a User defined option,
in which case a constant, parameter, or variable expression can be entered in the Value
column.

Thermodynamics Coupling
When the Thermodynamics is enabled, the species in the Chemistry interface can be
coupled to the species in the Thermodynamic System. This is needed to ensure that
arguments for the thermodynamic functions are correctly defined. Use the lists in the
From Thermodynamics column to match each species in the Chemistry to a species in
the coupled thermodynamic system. For each thermodynamics-coupled species, the
required property parameters and functions are added under the corresponding
thermodynamic system. When all species are matched, the Chemistry is considered
fully coupled and functions representing mixture properties, such as the density, are
also added automatically under the same thermodynamic system.

312 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Surface Species
The Surface Species table is available when there is at least one surface species (with
“(ads)” as suffix).

Similar to the table for Bulk Species, the surface species can be matched (or coupled)
directly to the surface species dependent variables if the Chemistry is coupled to a mass
transfer interface where the surface concentration is solved. There are three cases
where mass transfer interfaces solve surface species concentration (s).

• Surface Reactions physics interface;

• Packed Bed feature (with surface species transport in pellet) available in the Transport
of Diluted Species or Transport of Concentrated Species;
• Porous Catalyst feature (with surface species transport in catalyst) available in the
Transport of Diluted Species or Transport of Concentrated Species.

The last column Value(mol/m^2) will be automatically set to Solved for when a surface
species dependent variable is selected under the column Surface concentration.
Otherwise for a User defined being for Surface concentration column, specify the
concentration of surface species Value(mol/m^2) column.

In addition, if a surface species is present at the same time as an Electrode Reaction,

the Density of sites input field becomes visible. The Density of sites input value will be
used as the default reference concentration for surface species reacting in an Electrode
Reaction.

The species present in this section will not be used to define mixture properties.

Solid Species
The Solid Species table is available when there is at least one solid species (with “(s)” as
suffix).

Specify the concentration of solid species in the table in this section.

The species present in this section will not be used to define mixture properties.

Aqueous Species
The Aqueous Species table is available when there is at least one aqueous species (with
“(aq)” as suffix).

Specify the concentration of solid species in the table in this section.

The species present in this section will not be used to define mixture properties.

THE CHEMISTRY INTERFACE | 313

 Equilibrium Reaction Rate Table
This table is only available when equilibrium reactions are present in the interface. Use
the Reaction rate column to specify the rate for each equilibrium reaction. By doing so
the postprocessing variable for the total reaction rate for each species, of form
chem.Rsum_species, will be updated correctly. For a mass transport interface, say
Transport of Diluted Species, the reaction rate needed for an equilibrium reaction is
typically a dependent variable. In that case, the name of the variable can be found in
the Shape Function section of the Equation View of the node.

When the Chemistry interface is created using from the Generate Space-Dependent
Model the table is automatically set up in accordance with the added equilibrium
reaction nodes.

CALCULATE TRANSPORT PROPERTIES

Select the Calculate mixture properties checkbox (selected as default) to calculate
mixture properties that can be picked up in the space-dependent model interfaces. The
properties that can be calculated are shown beneath the checkbox. There are four
properties, heat capacity, ratio of specific heats, thermal conductivity and dynamic
viscosity. Foe there properties, the Automatic and User defined are always available, the
Thermodynamics is available when all bulk species are coupled to a Thermodynamic
System (fully coupled). Consider also if the built-in Automatic expressions fit the model
or if User defined expressions are more suitable.

The density and diffusion coefficient are always calculated.

It could be costly to calculate mixture properties by Automatic or Thermodynamics if

there are a lot of species. The calculations of mixture properties are set to the User
defined when the number of species from an imported file (chemkin file or a reaction
file) exceeds the maximum number (default 200). The maximum number can be
changed under the Advanced Settings section (activated by Advanced Physics Options).

Transport Properties

ACTIVITY
This section is available when the Thermodynamics checkbox is cleared.

Select the Use activity checkbox to solve for species activities instead of species
concentrations, which is a common approach when nonideal fluids are modeled.

314 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

An activity coefficient other than 1 can be set for each species for the Species node in
the Species Concentration/Activity section.

CHEMKIN IMPORT FOR SPECIES PROPERTIES

It is available when the Thermodynamics checkbox is cleared.

This section enables CHEMKIN® import to simulate complex chemical reactions in

gas phase.

Two types of CHEMKIN input files can be imported here: Thermo and Transport, for
thermodynamic properties and transport properties, respectively. Properties for either
volumetric or surface species are supported. Click Browse to locate the CHEMKIN file
to be imported, then click Import. For Thermo, the imported data is directly entered in
the NASA format fields in the Species node’s Thermodynamic Expressions section; for
Transport, the imported data is entered in its Transport Expressions section.

Feature Nodes Available for the Chemistry Interface

The following feature nodes are available from the Chemistry ribbon toolbar (Windows
users), Chemistry context menu (Mac or Linux users), or right-click to access the
context menu (all users).

• Equilibrium Reaction Group • Species

• Reaction • Species Activity
1
• Reaction Thermodynamics • Species Group
• Electrode Reaction • Species Thermodynamics1
• Electrode Reaction Group
• Reversible Reaction Group
1 This is a subnode and it is only available together with a parent node.

Reaction
To add a Reaction node ( ) either right-click the Chemistry node or in the Chemistry
toolbar click Reaction.

THE CHEMISTRY INTERFACE | 315

 REACTION FORMULA

Formula
Enter a chemical equation in the Formula field. The chemical equation should be of the
format “A + B arrow C + D”. Here, A and B are reactants, C and D are product species,
and arrow denotes a reaction arrow. The participating species should be written with
Valid Species Names. Valid reaction arrows are “<=>” for reversible reaction, “=>” for
irreversible reaction, and “=” for equilibrium reaction. Species can be given trivial
names, or their chemical formulas can be used as names. Examples of valid formulas
are; “carbon+oxygen=>carbondioxide”, and “C(s)+O2(g)=>CO2(g)”.

Click Apply to make the interface examine the species taking part in the chemical
equation, and automatically add the associated Species features to the Model Builder.

Balance ( )
Use the Balance button to automatically calculate the stoichiometric coefficients such
that the number of atoms of each kind are the same on both sides of the reaction. This
turns, for example, the formula “H2+O2=>H2O” into “2H2+O2=>2H2O”. The
balancing involves parsing all participating species for elements in the periodic table. It
therefore requires that all species in the Formula field are written either using their
chemical formula, say “H2O”, or that each species in the reaction have an enabled
Chemical Formula field in their Species node. For example, it is possible to balance the
formula “H2+O2=>water” as long as there is a Species node with the name “water”
already present, and an enabled Chemical Formula.

For automatic reaction balancing to be successful, requires that the problem is well
posed. One example of a problem that is not well posed is “C+H2=>CH4+C2H6”, in
which case any ratio of CH4/C2H6 could be obtained. Another example is
“H2=>O2” where not all elements are present on both sides.

Reaction Type
Select the Reaction type — Reversible, Irreversible, or Equilibrium — or edit the
expression directly in the Formula field. In the latter case, specify the reaction type with
a delimiter separating the two sides of the equation:

• <=> denotes a Reversible reaction

• => denotes an Irreversible reaction
• = denotes a reaction at chemical Equilibrium

316 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Each Reaction type has its own set of reaction kinetics:

• If the reaction is Reversible or Irreversible, the reaction rate for reaction i contributes
to the change in species i as follows:

Ri =  ij rj (5-11)
j

where ij is the stoichiometric coefficient.

• If it is an Equilibrium reaction, the equilibrium expression is equal to the equilibrium

constant:
K eq = K eq0

REACTION RATE
This section is available when the Reaction type is either Reversible or Irreversible.

When Mass action law is selected (default), the rate expression is automatically derived
from the stoichiometric coefficients in the reaction formula:

• For an Irreversible reaction type, the reaction rate becomes:

–  ij

f
rj = kj ci (5-12)
i  react

• For a Reversible reaction type, the expression instead becomes:

–  ij 
 
f r
rj = kj ci – kj c i ij (5-13)
i  react i  prod

The deduced overall reaction order is shown in text below the respective equation in
the Reaction Rate section.

Arbitrary Rate Expression

If the reaction order differs from the stoichiometric coefficients, or if an arbitrary rate
expressions is applicable, change Reaction Rate to User defined. An expression field r
appears with the default expression being that from the mass action law. Below this
there are fields to set the reaction order. For a reversible reaction the reverse reaction
order may be specified in addition to the forward one. The unit of the rate constant k
(or frequency factor A in the case of Arrhenius behavior), is derived from the reaction
order, in SI units: (m/mol)  1/s, where  equals the order with respect to
volumetric species. When surface species are present — identified by their “(ads)”

THE CHEMISTRY INTERFACE | 317

 suffix — the unit is instead given by m  2/mol/s, where is the order with
respect to surface species.

Consider for example the reaction:

2 S 2 O 42 – + H 2 O  S 2 O 32 – + 2 HSO 3–

The automatically deduced reaction order is three, however, in the case that water is
the solvent, the order should probably be two. This is specified as follows:

Change to User Defined

Adjust rate expression

Override reaction order

Note effect on unit

Figure 5-3: The reaction order can be manually specified for a user defined reaction rate.

RATE CONSTANTS
This section applies for Reversible or Irreversible reactions and defines the reaction rate
constants used in the reaction rates.

Forward Rate Constant and Reverse Rate Constant

The Forward rate constant kf is used for both Reversible and Irreversible reactions. The
Reverse rate constant kr is only used for Reversible reactions (Equation 5-12).

The SI units of the rate constants are automatically based on the order of the reaction
with respect to the concentrations, as defined in the Reaction formula.

318 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Specify Equilibrium Constant
The Specify equilibrium constant checkbox is available for Reversible reactions. If the
checkbox is selected the rate constants are defined in a different manner with the
reverse rate constant being computed from the following expression:

r f
k = k  K eq0

Thus, in this case, the forward rate constant and equilibrium constant for the reaction
are needed. The Equilibrium constant is edited in the Equilibrium Settings section.

Use Arrhenius Expressions

When the Use Arrhenius expressions checkbox is selected the Arrhenius parameters are
automatically used in predefined expressions for the forward and reverse rate constants
kfand kr, respectively.
f
f f n f
k = A  T  T ref  exp  – E   R g T  

r
r r n r
k = A  T  T ref  exp  – E   R g T  

Specify the activation energy and the frequency factor in the Arrhenius expressions to
account for temperature variations. The reference temperature, Tref equals 1 K. The
available fields are based on the Reaction type chosen in the Reaction node. Enter
values or expressions for each of the following (reverse expressions are only available
for reversible reactions):

• Forward frequency factor Af and Reverse frequency factor Ar (unit depends on

reaction order)
• Forward temperature exponent nf and Reverse temperature exponent nr
• Forward activation energy Ef and Reverse activation energy Er (SI unit: J/mol)

EQUILIBRIUM SETTINGS
This section is available for equilibrium reactions, and for reversible reactions when the
Specify equilibrium constant checkbox has been selected.

Equilibrium Expression
For an equilibrium reaction, specify the Equilibrium expression. When the Equilibrium
expression is set to Automatic the following expression is used:

THE CHEMISTRY INTERFACE | 319

 
 c i ij
i  prod
K eqj = ---------------------------
–
 c i ij
i  react

Select User defined from the Equilibrium expression list to instead enter a manually
defined equilibrium expression.

Equilibrium Constant
Specify the Equilibrium constant Keq0 for an equilibrium reaction, or for a reversible
reaction when the Specify equilibrium constant checkbox has been selected (in the Rate
Constants section).

The Equilibrium constant can either be User defined, or automatically defined when set
to Automatic or Thermodynamics.

Use the Automatic option to compute the equilibrium constant for an ideal system.

The Thermodynamics option is available when all reactions in the interface are
equilibrium reactions, and the interface is fully coupled to a Thermodynamic System
(see Species Matching). Use this setting to automatically compute the equilibrium
constant for an ideal or nonideal system, dependent on the thermodynamic model
applied for the coupled system.

Using Automatic or Thermodynamics, Keq0 is calculated from the Gibbs free energy of
the reaction. For more details see The Equilibrium Constant and the Automatically
Defined Equilibrium Constants section therein.

REACTION THERMODYNAMIC PROPERTIES

This section contains information about thermodynamic properties that relate to a
selected reaction. Several Automatic definitions are available here.

Enthalpy of Reaction
The Enthalpy of reaction H (SI unit: J/mol) is calculated by the interface from species
properties and the related stoichiometric coefficients:

Hj =   ij h i –   –  ij h i (5-14)
i  prod i  react

Entropy of Reaction
The Entropy of reaction S (SI unit: J/(mol·K)) comes from a similar expression:

320 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

 Sj =   ij s i –   –  ij s i (5-15)
i  prod i  react

In Equation 5-14 and Equation 5-15, hi and si are the species’ molar enthalpy and
molar entropy, respectively.

Enter these quantities in the Species Thermodynamic Expressions section for the Species
node either by using the predefined polynomial or by providing a custom expression
or constants.

The stoichiometric coefficients, ij, are defined as being negative for reactants and
positive for products. Using Equation 5-14 and Equation 5-15 to equate the Gibbs
free energy of reaction enables the equilibrium constant to be expressed according to
Equation 5-14.

Heat Source of Reaction

The Heat source of reaction (SI unit: W/m3) is automatically computed from the heat
of each reaction j, given by
Qj = –Hj rj

TURBULENT FLOW
This section is available if the following conditions are met:

1 Type under Mixture Properties is Concentrated species

2 Reversible or irreversible bulk reaction
3 Activity not selected
4 Pellet Chemistry not selected

When this section is activated, the Reaction feature can be coupled to turbulent flow in
the Transport of Concentrated Species interface.

In a turbulent flow, there are many small turbulent eddies (known as fine structures),
where actual dissipation of turbulent kinetic energy into heat and simultaneous mixing
occur. The chemical reactions can take place inside these isolated regions. According
to the ratio between the time scales of chemical reactions and mixing (the Damköhler
number, Da), the reaction rate can be modeled in the following three ways:

• Da << 1 — the mixing process is very fast, the limiting rate step is the chemical
reaction

THE CHEMISTRY INTERFACE | 321

 • Da >> 1 — the chemical reaction rate is very fast, the limiting rate step is the
turbulent mixing rate
• Da being not far away from 1 — the reaction rate is controlled by both chemical
reaction and turbulent mixing rates

The equations for the reaction rate controlled by both chemical reaction and turbulent
mixing are

c for for rev rev

r j =  min  r MV j r ED j  – min  r MV j r ED j  

–  i r
  ci 
for f
r MV j = k
r

–  i p
  ci 
rev r
r MV j = k
p

 p
 --------------
for
r ED j = ------ min min 
T p Mp
p

 p r
= ------ min min  --------------    ------------
rev
r ED j -
T   p M p r Mr
r

For very fast mixing process (Da << 1), the reaction rate equations are the same as that
for normal reaction (without turbulence).

For very fast chemical reaction (Da >> 1), the equation for reaction rate is,

c for rev
r j = r ED j –r ED j

Turbulence-Reaction Model
Select a turbulent model from the combo box, for example, Eddy dissipation.

Turbulence-Reaction Model Parameters

This subsection is available when a valid turbulence model is selected in the
Turbulence-reaction model list. There are two dimensionless parameters, ED and ED,
with default values 4 and 0.5, respectively.

Turbulence Time Scale

The time scale feature input (T) is available when a valid turbulence model is selected.
It contains all available turbulence time scales from turbulence flow interfaces.

322 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Assume Infinitely Fast Reaction
This checkbox is available when a valid turbulence model is selected. With the
checkbox selected, the reaction rate is controlled completely by the turbulence time
scale, and the Reaction rate list and the sections Rate Constants and Reaction Orders will
be not available.

REGULARIZATION
This section is available under the same conditions as those for the Turbulent Flow
section.

With the checkbox Reaction rate regularization selected, a table for Reaction rate
damping limits will be available. For each species in the reaction, the default damping
limit is 1e-6.

Using the Reaction Node

Species
When a Reaction is defined, a Species node ( ) is automatically generated for the
participating reactants and products. This feature enables you to review and enter
species specific information regarding chemical kinetics, thermodynamics, and
transport properties.

It is also possible to add and define an individual Species node: in the Chemistry toolbar
click Species or right-click the Chemistry node and select it from the context menu.

NAME
When a Species node is automatically generated using the Formula text field for the
Reaction node, the Name is also automatically generated.

For a Species node added individually, enter a Name in the field and click Apply. By
entering a name consisting of elements from the periodic table, the molar mass is
calculated and added in the Chemical Formula section. The species charge, and type, are
also deduced from the species name.

Valid Species Names

THE CHEMISTRY INTERFACE | 323

 TYPE
Select a species type — Bulk species, Surface species, or Solvent. The latter is only
available when a diluted solution is assumed. That is when Type is set to Diluted Species
in the Mixture Properties section (in the interface level).

Bulk species and Solvent are solved for volumetric concentrations (SI unit: mol/m3),
while Surface species are solved for surface concentration (SI unit: mol/m2). The
compositions for Bulk species and Solvent use the syntax c_speciesname, while Surface
species uses csurf_speciesname_surf.

When Surface species is selected, the corresponding reaction formula introduces (ads)
after the species notation and changes the species’ name to speciesname_surf.
Additionally, the Species node name is updated in a similar fashion.

CHEMICAL FORMULA
The Chemical Formula section contains the species chemical formula, Molar mass, M,
and Charge, z.

If the name entered for the species consists of only elements from the periodic table,
and optionally charge and phase indication, then the Chemical Formula field will be
populated with the name entered in the Name section, and the molar mass and charge
will be added to this section. The species molar mass is computed from the mass of
occurring individual elements1. The rules for writing chemical formulas are the same
as those for Species names.

Valid Species Names

When the species name contains parts not in the periodic table, for example when a
descriptive name such as water is used, the molar mass is set to 0.0 kg/mol and the
charge to zero. In this case, the molar mass needs to be defined in order to achieve
mass balanced reactions and correct definitions of mass basis properties. By entering a
chemical formula consisting only of elements from the periodic table (and optionally
charge and phase indication), the molar mass, and charge are derived automatically.

When needed, the Molar mass can be edited in the corresponding text field. Editing
the Chemical formula and pressing apply will override the molar mass and charge
[Link] Charge field cannot be edited as long as a chemical formula is enabled.

1. Values are averages of upper and lower bounds for each atom as published by CIAAW. Source: CIAAW.
Atomic weights of the elements 2019. Available online at [Link].

324 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

With an enabled Chemical Formula it is possible to use the trivial name of a species and
balance reactions. Having added the Species water, oxygen, and hydrogen, and filled
in H2O, O, and H in their respective chemical formula sections, the Formula
oxygen+hydrogen=water is balanced into oxygen+2hydrogen=water by clicking
Balance in the Reaction node.

It is possible to specify the species density  when the fluid Mixture is specified as Liquid.
The default value is that of water at 293 K.

REACTION RATE
Change the Automatic default setting to User defined to use a species reaction rate other
than the one set up in the associated Reaction node. For individual species, use the User
defined option to set a reaction rate other than zero (that is, nonreactive).

Edit either the Rate expression (SI unit: mol/(m3·s)), the Surface rate expression (SI
unit: mol/(m2·s)), or both. For a bulk species, both expressions appear if surface
reactions are present since the reaction of the species can depend both on bulk reaction
R and surface reaction Rads rates. For a surface species, only the surface reaction rate
Rads appears.

ADDITIONAL SOURCE
The Additional Source section is available in order to include additional rate
contribution for the species to the reaction kinetics. When the Additional source
checkbox is selected, add an Additional rate expression in the text field (SI unit: mol/
m3).

CONSTANT CONCENTRATION/ACTIVITY
To account for nonideality in the fluid mixture, adjust the activity coefficient in the
Activity coefficient input field. This input field is only shown if activity instead of
concentration has been chosen in the interface, that is, the Use activity checkbox is
selected on the Chemistry interface Settings window

Click to select the Keep concentration/activity constant checkbox if the species

concentration or activity should be treated as constant.

THERMODYNAMIC EXPRESSIONS
The parameters utilized for calculation of thermodynamic mixture and reaction
properties are set in this section.

Choose the User defined alternative to specify Cp, h, and s directly.

THE CHEMISTRY INTERFACE | 325

 Electrode Reaction
To add an Electrode Reaction node ( ) either right-click the Chemistry node or in the
Chemistry toolbar click Electrode Reaction.

The Electrode Reaction subnode defines the electrode kinetics for a charge transfer
reaction that occurs on an electrolyte-electrode interface boundary. Use multiple
nodes to model multiple reactions, for instance in mixed potential problems.

This node is available with a Battery Design Module, Corrosion Module,

Electrochemistry, Module, Electrodeposition Module, or Fuel Cell & Electrolyzer
Module license.

REACTION FORMULA

Formula
Enter an electrode reaction Formula, as a reduction reaction. Use the single letter “e”
to symbolize an electron. The delimiter between reactants and products has to be
written as “<=>” (without quotation marks). The phase of the species, that is (aq) for
aqueous dilute ions, (s) for solids, and (g) for gases can be indicated in the name. The
phase is used to select an automatic reference state (see Reference Concentrations) for
computing the equilibrium potential using the Nernst equation. Click Apply to make
the interface examine the species taking part in the model’s reactions and automatically
add the associated Species features to the Model Builder.

EQUILIBRIUM POTENTIAL
The Equilibrium potential, Eeq (SI unit: V), is used in the electrode kinetics
expressions in the following section (via the definition of the overpotential), or for
setting up primary current distribution potential constraints.

Nernst Equation
After providing the reference electrode potential, the equilibrium potential is
calculated using the Nernst equation. The concentrations entered in the Species
Matching section of the Chemistry interface will be used to calculate the equilibrium
potential. The reference concentrations, defined in the Reference Concentrations
section, will be used to compute the activities in the reaction quotient expression, Qr.

Automatic
This option is only available if the Calculate mixture properties checkbox in the
Calculate Transport Properties section is checked.

326 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The Species enthalpies and entropies are used to calculate the equilibrium potential of
the electrode reaction at the temperature and composition of the system.

If only species enthalpies and entropies at standard state are available, use the Nernst
Equation option instead.

REFERENCE CONCENTRATIONS
The reference concentrations for the species in the reaction can be changed here.

Automatic
The phase indicated when first naming a species will be used to select a suitable
reference concentration. The species must indicate the phase in the name when the
species is first created.

For aqueous dilute species, denoted by (aq) in their names, the activity is calculated as
their concentration divided by 1 M. For gaseous species, denoted by (g) in their
names, the activity is calculated as their concentration divided by the standard
reference pressure of 1 atm, converted into concentration using the ideal gas law. For
solid species, denoted by (s) in their names, the reference concentration 1. For surface
species, denoted by (ads) in their names, the reference surface concentration will be set
to the site density configured in the Species Matching section. Any species not
containing a phase in their name, will use a reference state depending on the phase
selected under Mixture Properties. If the phase is Gas, the gaseous species reference
concentration described above will be used, while if the phase is Liquid, the aqueous
species 1 M reference concentration will be used.

User Defined
The reference concentrations for each type of species can be set manually, for aqueous,
gaseous, solid, surface, and any other species, respectively.

ELECTRODE KINETICS
See the Electrode Reaction node in Shared Physics Features in the Current
Distribution Interfaces for a general description of the Electrode Kinetics section.

Electrode reactions in the Chemistry interface support User-defined, Butler-Volmer, and

Linearized Butler-Volmer kinetics. For the latter two options, User-defined exchange
current densities and exchange current densities according to the Mass Action Law are
supported.

If the equilibrium potential is calculated using Nernst Equation, the exponents in the
local current density expression are evaluated using the reference overpotential. If

THE CHEMISTRY INTERFACE | 327

 instead the equilibrium potential is calculated using the Automatic option, the
exponents are evaluated using the overpotential.

HEAT OF REACTION
See the Electrode Reaction node in Shared Physics Features in the Current
Distribution Interfaces for a general description of the Heat of Reaction section.
Electrode Reactions in Chemistry do not support the Temperature Derivative option.

By choosing Automatic, the Thermoneutral voltage parameter, Etherm (SI unit: V), will
be calculated according to

r H
E therm = – ----------
- (5-16)
nF

using the Species enthalpies.

• Electrode Reactions Theory

• Electrode Kinetics Expressions

Electrode Reaction Group

To add an Electrode Reaction Group node ( ) either right-click the Chemistry node or
in the Chemistry toolbar click Electrode Reaction Group.

An Electrode Reaction Group is typically used to represents all electrode reactions

occurring on one electrode.

Multiple Electrode Reaction nodes may be added to the Electrode Reaction Group.
Variables for the sum of all electrode reactions in one group can be used as input in
Electrode Reaction features in current distribution interfaces.

This node is available with a Battery Design Module, Corrosion Module,

Electrochemistry Module, Electrodeposition Module, or Fuel Cell & Electrolyzer
Module license.

Reversible Reaction Group

The Reversible Reaction Group node ( ) allows input of a large number of reversible
reactions in the same table. All reactions are treated in the same way. More than one
Reversible Reaction Group can be added to a component.

328 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Add the node from the Chemistry toolbar or right-click Chemistry and add it from the
context menu.

For the case when reaction kinetics data are entered manually into the Reaction table
and temperature dependent reaction kinetics apply, right-click to add a Reaction
Thermodynamics subnode or select it from the Chemistry toolbar, Attributes menu. In
it, the Enthalpy of Reaction (J/mol) for each reaction can be specified.

REACTION TABLE
The reversible reactions in the Reaction table are numbered and contain reactants,
products, and kinetic parameters describing the reaction. Use the buttons under the
Reaction table to add and sort the reaction details.

• In general, use the Move Up ( ), Move Down ( ), and Delete ( ) buttons and
the fields under tables to edit the table contents. Or right-click a table cell and select
Move Up, Move Down, or Delete.
• The Add button ( ) adds default reactant, A, and product, B, with a default. Click
the corresponding field to edit the reactant, product, or parameters. After editing
the Species Group node is also updated. It is created together with the reaction
group.
• You can save the parameters to a text file to reuse in other models. Click the Save to
File button ( ) and enter a File name in the Save to File dialog, including the
extension .txt. Click Save to store the parameters in a text file or in a Microsoft
Excel Workbook spreadsheet if the license includes LiveLink™ for Excel®. The
information is saved in space-separated columns in the same order as displayed on
screen. When saving to Excel, an Excel Save dialog appears where you can specify the
sheet and range and whether to overwrite existing data or include a header.
• You can import or load data in files from a spreadsheet program, for example, with
the Load from File button ( ) and the Load from File dialog that appears. Data must
be separated by spaces or tabs. If there is already data in the table, imported
parameters are added after the last row. Move or edit rows as needed. If the license
includes LiveLink™ for Excel® you can also load parameters from a Microsoft Excel
Workbook spreadsheet. Then an Excel Load dialog appears where you can specify the
sheet and range and whether to overwrite existing data. It is also possible to import
from a spreadsheet containing a separate column for units.

CREATE REACTION AND SPECIES

Enter a reaction number in the Moving reaction (with the number) from table text field
and click Create Reaction to remove the reaction from the Reaction table to the model

THE CHEMISTRY INTERFACE | 329

 tree. This introduces a reaction node named after the reaction number and species
names and simultaneously creates corresponding species nodes. This action is
powerful, since it enables all the editing possibilities available for Reaction and Species
nodes.

Equilibrium Reaction Group

The Equilibrium Reaction Group node ( ) allows for input of a large number of
equilibrium reactions in the same table. This node functions in the same way as the
Reversible Reaction Group for reversible reactions.

Add the node from the Chemistry toolbar or right-click Chemistry and add it from the
context menu.

For the case when reaction kinetics data are entered manually into the Reaction table
and temperature dependent reaction kinetics apply, right-click to add a Reaction
Thermodynamics subnode or select it from the Chemistry toolbar, Attributes menu. In
it, the Enthalpy of Reaction (J/mol) for each reaction can be specified.

REACTION TABLE
The equilibrium reactions in the table are numbered and contain reactants, products,
and kinetic parameters describing the reaction. Use the buttons under the Reaction
table to add and sort the reaction details.

• In general, use the Move Up ( ), Move Down ( ), and Delete ( ) buttons and
the fields under tables to edit the table contents. Or right-click a table cell and select
Move Up, Move Down, or Delete.
• The Add button ( ) adds default reactant, A, and product, B, with a default. Click
the corresponding field to edit the reactant, product, or parameters. After editing
the Species Group node is also updated. It is created together with the reaction
group.
• You can save the parameters to a text file to reuse in other models. Click the Save to
File button ( ) and enter a File name in the Save to File dialog, including the
extension .txt. Click Save to store the parameters in a text file or in a Microsoft
Excel Workbook spreadsheet if the license includes LiveLink™ for Excel®. The
information is saved in space-separated columns in the same order as displayed on

330 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

 screen. When saving to Excel, an Excel Save dialog appears where you can specify the
sheet and range and whether to overwrite existing data or include a header.
• You can import or load data in files from a spreadsheet program, for example, with
the Load from File button ( ) and the Load from File dialog that appears. Data must
be separated by spaces or tabs. If there is already data in the table, imported
parameters are added after the last row. Move or edit rows as needed. If the license
includes LiveLink™ for Excel® you can also load parameters from a Microsoft Excel
Workbook spreadsheet. Then an Excel Load dialog appears where you can specify the
sheet and range and whether to overwrite existing data. It is also possible to import
from a spreadsheet containing a separate column for units.

Species Group
The Species Group node ( ) contains information on a molecular level about the
volumetric species and the surface species present in the model. The Property for
Volumetric Species or Property for Surface Species tables typically collect parameters from
when importing CHEMKIN transport files.

This node is automatically added when either the Reversible Reaction Group or the
Equilibrium Reaction Group is used.

For the case of temperature dependent reaction kinetics, a Species Thermodynamics

subnode is automatically created in which the thermodynamic properties of the species
can be specified.

Reaction Thermodynamics
The Reaction Thermodynamics subnode ( ), the Enthalpy of Reaction (J/mol) of
each reaction can be specified. This node overrides all the automatically calculated
reaction enthalpies as defined in the Species Thermodynamics subnode.

Add the Reaction Thermodynamics node from the Chemistry toolbar, Attributes menu.
Alternatively, right-click a Reversible Reaction Group or Equilibrium Reaction Group
to add the Reaction Thermodynamics subnode.

Species Activity
The Species Activity node ( ) creates variables for the activities of all the species/
surface species present in the Species Group parent feature. Edit the Activity coefficient
field in the Species Activity or Surface Species Activity tables by clicking in these.

THE CHEMISTRY INTERFACE | 331

 Species Activity is a subnode to the Species Group node. It is automatically generated
when nonideality in the fluid mixture is accounted for; when the Use activity checkbox
is selected in the Activity section.

Species Thermodynamics
The Species Thermodynamics node ( ) creates variables for the enthalpies, entropies,
and heat capacities for all the species/surface species present in the Species Group
parent feature. The purpose is to compute thermodynamic mixture properties and the
heat of reactions.

This node is a subnode to the Species Group node.

332 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The Nernst–Planck–Poisson
Equations Interface
The Nernst-Planck-Poisson Equations interface ( ), found under the Chemical Species
Transport branch ( ) when adding a physics interface, is a predefined multiphysics
interface for modeling transport of electrolyte species without the otherwise common
assumption of local electroneutrality. This allows for simulating charge separation that
typically arises close to an electrode surface, where ions in the electrolyte are attracted
and repelled by unscreened excess charge on the electrode. The charge separation
region, also called the diffuse double layer, normally extends a few nanometers away
from the electrode surface into the electrolyte. The study of the charge separation is
important to applications that consider very thin layers of electrolyte, such as
electrochemical capacitors, atmospheric corrosion problems, ion-selective field effect
transistors (ISFETs), and nanoelectrochemistry.

The Nernst–Planck–Poisson Equations predefined multiphysics interface adds ( )

(with Migration enabled), which solves for the Nernst–Planck equations without
charge neutrality, and The Electrostatics Interface( ), that solves for the Poisson’s
equation.

A ( ) multiphysics node is also added. This node computes the local space charge,
based on the local concentrations and species charges in the Transport of Diluted
Species interface, and adds it to Poisson’s Equation in the Electrostatics interface.

THE NERNST–PLANCK–POISSON EQUATIONS INTERFACE | 333

 Finally, a ( ) multiphysics node is added that applies the potential dependent
variable in the Electrostatics interface into the migration term of the Transport of
Diluted Species interface.

• The Multiphysics Branch in the COMSOL Multiphysics

Reference Manual.
• The Transport of Diluted Species Interface
• The Electrostatics Interface
• Space Charge Density Coupling
• Potential Coupling

Diffuse Double Layer: Application Library path

Fuel_Cell_and_Electrolyzer_Module/General_Electrochemistry/
diffuse_double_layer

334 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The Electrophoretic Transport
Interface
The Electrophoretic Transport (el) interface ( ), found under the Chemical Species
Transport branch ( ), is used to solve for the electrophoretic transport of an
arbitrarily number of species in water-based system, subject to potential gradients. The
species transported can be any combination of weak and strong acids and bases,
ampholytes, and uncharged species. The transport of masses and charge is based on the
Nernst–Planck equations for molecular transport, in combination with
electroneutrality, dissociation equilibria for weak acids, bases, and ampholytes as well
as the water autoionization reaction.

The physics interface can simulate most forms of electrophoresis modes, such as zone
electrophoresis, isotachophoresis, isoelectric focusing, and moving boundary
electrophoresis.

Gel electrophoresis can be simulated by the inclusion of immobile charged species.

The interface supports simulation in 1D, 2D, and 3D as well as for axisymmetric
components in 1D and 2D.

The dependent variables are the electrolyte potential, and the molar concentrations of
the included species, added individual by each species node in the model tree.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is el.

DOMAIN SELECTION
If any part of the model geometry should not partake in the mass transfer model,
remove that part from the selection list.

THE ELECTROPHORETIC TRANSPORT INTERFACE | 335

 OUT-OF-PLANE THICKNESS

For 2D components, enter a value or expression for the out-of-plane

Thickness d (SI unit: m). The value of d determines the size of the domain
perpendicular to the modeled 2D cross section. This value yields, for
example, the correct total current when the current density is obtained
from a 2D simulation.

CROSS SECTIONAL AREA

For 1D components, enter a Cross sectional area Ac (SI unit: m2) to define
a parameter for the area of the geometry perpendicular to the 1D
component. The value of this parameter is used, among other things, to
automatically calculate the total current from the current density vector.
The analogy is valid for other fluxes. The default is 1 m2.

TRANSPORT MECHANISMS
Mass transport due to diffusion and migration is always included. Use the checkboxes
available under Additional transport mechanisms to control other transport
mechanisms.

• By default, the Convection checkbox is selected. Clear the checkbox to disable

convective transport.
• The Mass transfer in porous media checkbox activates functionality specific to species
transport in porous media. When selected, the Porous Matrix Properties node can
be added to a domain to specify the electrolyte volume fraction and tortuosity, and
the Effective Transport Parameter Correction sections are enabled in the species
nodes.

CONSISTENT STABILIZATION
To display this sections, click the Show More Options button ( ) and select
Stabilization in the Show More Options dialog.

• When the Crosswind diffusion checkbox is selected, a weak term that reduces spurious
oscillations is added to the transport equation. The resulting equation system is
always nonlinear.
• For both Streamline diffusion and Crosswind diffusion, select an Equation residual.
Approximate residual is the default and means that derivatives of the diffusivity are

336 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

 neglected. This setting is usually accurate enough and is computationally faster. If
required, select Full residual instead.

INCONSISTENT STABILIZATION
To display this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog. By default, the Isotropic diffusion checkbox is not
selected because this type of stabilization adds artificial diffusion and affects the
accuracy of the original problem. However, this option can be used to get a good initial
guess for under-resolved problems.

ADVANCED SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog. Normally these settings do not need
to be changed. Select a Convective term — Nonconservative form (the default) or
Conservative form. Use the conservative formulation for compressible flow.

DISCRETIZATION
To display all settings available in this section, click the Show More Options button ( )
and select Advanced Physics Options in the Show More Options dialog.

The concentration variables are set to use Linear elements by default.

The potential variable is set to use Quadratic elements by default.

The Compute boundary fluxes checkbox is activated by default so that COMSOL

Multiphysics computes predefined accurate boundary flux variables. When this option
is checked, the solver computes variables storing accurate boundary fluxes from each
boundary into the adjacent domain.

If the checkbox is cleared, the COMSOL Multiphysics software instead computes the
flux variables from the dependent variables using extrapolation, which is less accurate
in postprocessing results but does not create extra dependent variables on the
boundaries for the fluxes.

The flux variables affected in the interface are:

• <name>.nIl, where <name> is the name of the interface (default is el), set on the
interface top node. This is the normal electrolyte current density.
• <name>.ntflux_<species_name> is the Species name (see Common Settings for
the Species Nodes in the Electrophoretic Transport Interface below). This is the
normal total flux for each species.

THE ELECTROPHORETIC TRANSPORT INTERFACE | 337

 Also the Apply smoothing to boundary fluxes checkbox is available if the previous
checkbox is checked. The smoothing can provide a more well-behaved flux value close
to singularities.

For details about the boundary fluxes settings, see Computing Accurate Fluxes in the
COMSOL Multiphysics Reference Manual.

Regarding the Value type when using splitting of complex variables, see Splitting
Complex-Valued Variables in the COMSOL Multiphysics Reference Manual.

DEPENDENT VARIABLES
The dependent variable name for the electrolyte potential variable is phil by default.

The name of the concentration dependent variables are named as [Link], where el is
the name of the interface as set above, and the xxx string is controlled by the Species
name setting on the individual species nodes.

FURTHER READING

• Theory for the Electrophoretic Transport Interface

• Numerical Stabilization in the COMSOL Multiphysics Reference
Manual.
• In the COMSOL Multiphysics Reference Manual, see Table 2-4 for
links to common sections and Table 2-5 for common feature nodes.
You can also search for information: press F1 to open the Help window
or Ctrl+F1 to open the Documentation window.

Common Settings for the Species Nodes in the Electrophoretic

Transport Interface
The interface features the following species nodes, applicable on the domain level:

• Ampholyte
• Fully Dissociated Species
• Uncharged Species
• Weak Acid
• Weak Base

338 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Each species node add a dependent variable for the concentration. The initial and
boundary condition, as well as adding additional source reaction terms, for each
species concentration is controlled by adding subnodes to the species nodes:

• Concentration
• Flux
• Inflow
• Initial Concentration
• No Flux
• Outflow
• Species Source

The Ampholyte, Weak Acid, and Weak Base nodes are dissociation species and may
define an arbitrary number of dissociation steps. Each dissociation step is defined by
its pKa (the acid equilibrium constant) parameter. For the weak bases the pKa refers
to the acid constant of the conjugate acid. Each dissociation step adds one additional
subspecies concentration variable so that the concentration dependent variable
represents the sum of all subspecies, and initial and boundary conditions are defined
with respect to this total concentration.

All species node have a setting for the Species name, which needs to be unique. The
species name is used for naming of all related variables of the species. For species nodes
not defining any subspecies, the concentration variables are named as
<name>.c_<species_name> where <name> is the name of the interface (default is el),
set on the interface top node, and <species_name> is the Species name. For
dissociation species nodes defining multiple subspecies, the concentration nodes are
named as <name>.c<X>_<species_name> where <X> is the integer from 1 up to the
total number of subspecies. Note that the Solvent node automatically defines the
concentration variables for protons (<name>.cH) and hydroxide ions (<name>.cOH).

All species except the Uncharged Species carry charge and contribute to the total
electrolyte current which is used in the equation for solving the electrolyte potential.

The Immobile Species checkbox can be used to lock the concentration of a species, to,
for instance, define the immobile charges in an ion-selective membrane or a gel. When
the checkbox is enabled the concentration of the species is not added as a dependent
variable to the model; instead the concentration will be set to the value provided in the
Concentration field. The contribution to the electrolyte current for immobile species is
zero.

THE ELECTROPHORETIC TRANSPORT INTERFACE | 339

 Diffusion and Migration Settings
All species, except when the Immobile Species checkbox has been enabled, may be
transported by diffusion, and charged species are also transported by migration in the
electric field.

For dissociation species you may choose to set the transport parameters to be the Same
for all species appearing in the different dissociation steps, or you may use Individual
settings for each subspecies.

Typically the mobilities and diffusivities for small species are related by the
Nernst-Einstein relation, and when this relation is enabled you can choose whether to
specify either the Diffusivity (SI unit: m2/s) or the Mobility (SI-unit: s·mol/kg). The
Debye-Hückel-Henry relation is commonly used for larger molecules, such as proteins.

Note: There are other definitions of the migration transport equations in the
literature which use mobilities expressed in m2/(V·s), whereas COMSOL
Multiphysics uses s·mol/kg. To convert mobilities expressed in m2/(V·s) to the
corresponding values in s·mol/kg, you typically divide by the Faraday constant,
F_const (about 96,485 C/mol).

EFFECTIVE TRANSPORT PARAMETER CORRECTION

If Mass transfer in porous media is enabled on the interface top node, you may use the
settings of this section in the species nodes to account for the changed mass transport
in a porous domain due to the lowered porosity and the increased tortuosity.

The default correction model is Bruggeman, which multiplies the diffusivity and
mobility values by the porosity to the power of 1.5. The porosity of a domain is set by
the Porous Matrix Properties node.

IONIC STRENGTH CONTRIBUTION

To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

The Debye-Hückel-Henry relation makes use of the ionic strength for calculating the
species mobility from the diffusivity. All charged species contribute to the ionic
strength, either assuming the species contributing to an Ideal solution or by using the
Linderstrøm-Lang assumption. The latter is usually used for macromolecules.

340 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Domain, Boundary, and Pair Nodes for the Electrophoretic Transport
Interface
The interface has the following domain, boundary, and pair nodes, listed in
alphabetical order, available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or by right-clicking to access the context menu
(all users).

To add a node, go to the Physics toolbar, no matter what operating system

you are using. Subnodes are available by clicking the parent node and
selecting it from the Attributes menu.

• Ampholyte • No Flux
• Concentration • Outflow
• Current • Porous Matrix Properties
• Current Density • Potential
• Current Source • Protein
• Flux • Species Source
• Fully Dissociated Species • Solvent
• Inflow • Uncharged Species
• Initial Concentration • Weak Acid
• Initial Potential • Weak Base
• Insulation

For axisymmetric components, COMSOL Multiphysics takes the axial

symmetry boundaries (at r  0) into account and automatically adds an
Axial Symmetry node that is valid on boundaries representing the
symmetry axis.

In the COMSOL Multiphysics Reference Manual, see Table 2-4 for links
to common sections and Table 2-5 for common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

THE ELECTROPHORETIC TRANSPORT INTERFACE | 341

 Solvent
The settings of this node are used to define the properties of the aqueous solvent.

If Convection is enabled on the interface top node, you can specify the Velocity field
(m/s) as user defined input using analytical expressions or the velocity field variables
solved for by a separate physics interface.

The Solvent node automatically defines the concentration variables and for protons
(<name>.cH) and hydroxide ions (<name>.cOH), and the corresponding flux
expressions. See Diffusion and Migration Settings for how to set up the transport
parameters for the proton and hydroxide ions.

In the Solvent Properties section you can modify the Dynamic viscosity (Pa·s) and
Relative permittivity (unitless) values. The Built in and default values are applicable to
water. These parameters are used when calculating mobilities according to the
Debye-Hückel-Henry relation in the species nodes.

In the Water Self-Ionization section you can change the default Built in expression for
the Water self-ionization constant, pKw (unitless), to any user-defined expression.

Porous Matrix Properties

Use this node, available if Mass transfer in porous media has been enabled on the
interface top-node, to define the Porosity (electrolyte volume fraction) and Tortuosity
of a domain.

The porosity should be a number between 0 and 1.

Fully Dissociated Species

Use this node to define a fully dissociated charged species, such as strong bases and
acids.

The Charge number, Z0 (unitless), defines the species charge.

See also Common Settings for the Species Nodes in the Electrophoretic Transport
Interface and Diffusion and Migration Settings.

Uncharged Species
Use this node to define a species that does not carry any charge, nor is impacted by the
electric field.

342 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

See also Common Settings for the Species Nodes in the Electrophoretic Transport
Interface and Diffusion and Migration Settings.

Weak Acid
The Weak acid node supports multiple dissociation steps, where the acid of the first
dissociation step is uncharged.

The species may be either Monoprotic, subject to one dissociation step only, or
Polyprotic. For the latter case any Number of dissociation steps larger than one may be
used.

See also Common Settings for the Species Nodes in the Electrophoretic Transport
Interface and Diffusion and Migration Settings.

Weak Base
The Weak base node supports multiple dissociation steps, where the base of the last
dissociation step is uncharged.

The species may be either Monoprotic, subject to one dissociation step only, or
Polyprotic. For the latter case any Number of dissociation steps larger than one may be
used.

Note that the pKa refers to the acid constant of the conjugate acid of the weak base.

See also Common Settings for the Species Nodes in the Electrophoretic Transport
Interface and Diffusion and Migration Settings.

Ampholyte
Use the Ampholyte node generically to define any species where the average charge
depends on the pH of the solution. The Dissociation model may be based either on a
set of Equilibrium constants or an Average charge.

When using the Equilibrium constants the Base charge in last dissociation step, Z0
(unitless), needs to be set.

The Average charge, Z (unitless) parameter is typically a function of pH and may be

added as an analytical function or an interpolation polynomial under Definitions. The
average squared charge, used in the electrolyte potential equation, is calculated
automatically based on Z .

THE ELECTROPHORETIC TRANSPORT INTERFACE | 343

 See also Common Settings for the Species Nodes in the Electrophoretic Transport
Interface and Diffusion and Migration Settings.

Protein
Use the Protein node to define macromolecules. The features of the Protein node are
similar to the Ampholyte node, but with the default settings applicable for larger
molecules.

See also Common Settings for the Species Nodes in the Electrophoretic Transport
Interface and Diffusion and Migration Settings.

Current Source
To make this node available, click the Show More Options button ( ) and select
Advanced Physics Options in the Show More Options dialog.

Use this node to add a current source in a domain. A current source may appear in a
domain in homogenized porous electrode modeling, but should normally not be used.

Initial Potential
Use this node to specify the Initial Value of the electrolyte potential for the solver.

Current
The Current boundary condition sets the total current or average current density over
a boundary. It will set a constant electrolyte potential along the given boundary, which
satisfies the current value setting.

This node is typically used to model electrode surfaces or boundaries facing an

electrolyte reservoir (containing an electrode).

When using the Total current option in 1D or 2D, the boundary area is based either
on the Cross sectional area (1D) or the Out-of-Plane thickness (2D) properties, set on
the physics interface top node.

Current Density
Use the Current Density node to specify the current density distribution along a
boundary.

344 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

This node is typically used to model electrode surfaces where the electrode kinetics
depends on the electrolyte potential.

Note that using this node in 2D or 3D may result in an uneven potential distribution
along the boundary. To mitigate such effects you may use the Current node instead.

Insulation
The Insulation boundary condition describes the walls of a cell or the boundaries of the
cell that do not face an electrode (or a reservoir containing an electrode). The
boundary condition imposes the following equation:

il  n = 0

Potential
Add the Potential node on a boundary to set a fixed potential. This node is typically
used to model electrode surfaces or boundaries facing an electrolyte reservoir.

The node sets the potential in the electrolyte, l, to be equal to the Boundary electrolyte
potential, l, bnd (SI unit: V).

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

Species Source
To make this node available, click the Show More Options button ( ) and select
Advanced Physics Options in the Show More Options dialog.

This node may be added as a subnode to any species node. See also Common Settings
for the Species Nodes in the Electrophoretic Transport Interface.

In order to account for consumption or production of species, for example in porous

electrodes or as a result of reversible reactions, the Species Source node adds source
terms expressions S (SI unit: mol/(m3·s) to the right-hand side of the transport
equation of the parent species.

THE ELECTROPHORETIC TRANSPORT INTERFACE | 345

 Initial Concentration
This node may be added as a subnode to any species node. See also Common Settings
for the Species Nodes in the Electrophoretic Transport Interface.

This node specifies the initial value for the Concentration, c (mol/m3), of the parent
species. This value serve as the initial condition for a transient simulation. The value
also serves as a start guess for stationary problems.

You can use spatially dependent functions (such as smoothed step functions) available
under Definitions when defining the Concentration expression to specify different
concentrations in different parts of the geometry. You can also use additional Initial
Values node and modify the Selection to set different values for different domains.

Concentration
This node may be added as a subnode to any species node. See also Common Settings
for the Species Nodes in the Electrophoretic Transport Interface

This condition node adds a boundary condition for the parent species concentration.
Use the node to, for instance, specify the inlet concentration at the boundary facing
an electrolyte reservoir.

No Flux
This node may be added as a subnode to any species node. See also Common Settings
for the Species Nodes in the Electrophoretic Transport Interface.

This node is the default boundary condition on exterior boundaries. It represents

boundaries where no mass flows in or out of the boundaries. Hence, the total flux is
zero.

Flux
This node may be added as a subnode to any species node. See also Common Settings
for the Species Nodes in the Electrophoretic Transport Interface.

This node can be used to specify the species inward flux across a boundary. The flux
can represent a flux from or into a much larger surrounding environment, a phase
change, or a flux due to chemical reactions.

346 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

INWARD FLUX
Enter a value or expression for the species mass flux J0. Use a minus sign when
specifying a flux directed out of the system.

External convection
Set Flux type to External convection to prescribe a flux to or from an exterior domain
(not modeled) assumed to include convection. The exterior can for example include a
forced convection to control the temperature or to increase the mass transport. In this
case the prescribed mass flux corresponds to

J0 = kc  cb – c 

where kc is a mass transfer coefficient and cb is the bulk concentration, the typical
concentration far into the surrounding exterior domain.

Inflow
This node is available when you select the Convection checkbox on the physics interface
Settings window.

Use this node to specify the species concentrations at an inlet boundary.

BOUNDARY CONDITION TYPE

The option Concentration constraint constrains the concentration values on the
boundary by the use of pointwise constraints. When using this option the boundary
condition is identical to the Concentration node.

The other option, Flux (Danckwerts) can be more stable and fast to solve when high
reaction rates are anticipated in the vicinity of the inlet. Oscillations on the solutions
can also be avoided in such cases. The latter condition uses a flux boundary condition
based on the velocity across the boundary and the concentration values.

Outflow
Set this condition at outlets where species are transported out of the model domain by
migration or fluid motion. It is assumed that migration and convection are the
dominating transport mechanisms across outflow boundaries, and therefore that
diffusive transport can be ignored; that is:

n   – D c  = 0

THE ELECTROPHORETIC TRANSPORT INTERFACE | 347

 The Surface Reactions Interface
The Surface Reactions (sr) interface ( ), found under the Chemical Species Transport
branch ( ) when adding a physics interface, is used to model the chemical reactions
of surface and bulk species on a boundary. Surface species can be transported in the
tangential direction of the surface by Fick’s law whereas bulk species are assumed to be
immobile on the surface.

The physics interface supports simulation of surface reactions on boundaries in 1D,

2D, and 3D as well as for axisymmetric components in 1D and 2D. The dependent
variables are the surface concentrations, cs,i (SI unit: mol/m2) and the bulk
concentrations, cb,i (SI unit: mol/m2).

When this physics interface is added, these default nodes are also added to the Model
Builder — Surface Properties, No Flux, and Initial Values. Then, from the Physics toolbar,
add other nodes that implement, for example, boundary conditions. You can also
right-click Surface Reactions to select physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is sr.

DEPENDENT VARIABLES
Add or remove species and also change the names of the dependent variables that
represent the species concentrations. Note that the names can be changed but the
names of fields and dependent variables must be unique within a model.

Enter the Number of surface species. Use the Add surface concentration ( ) and
Remove surface concentration ( ) buttons as needed. The same number of Surface
concentrations cs, cs2, cs3, … are then listed in the table.

348 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Enter the Number of bulk species. Use the Add bulk concentration ( ) and Remove bulk
concentration ( ) buttons as needed. The same number of Bulk concentrations cb,
cb2, cb3, … are then listed in the table.

DISCRETIZATION
To display all settings available in this section, click the Show More Options button ( )
and select Advanced Physics Options in the Show More Options dialog.

CONSISTENT AND INCONSISTENT STABILIZATION

To display these sections, click the Show More Options button ( ) and select
Stabilization in the Show More Options dialog.

By default the Compensate for boundary stretching checkbox is selected for the Surface
Properties node. This section is then used to stabilize the tangential mesh velocity
term.

When the Compensate for boundary stretching checkbox is cleared (not selected), and
for fixed geometries or moving geometries, the stabilization has no effect.

See Surface Reaction Equations on Deforming Geometries for more information.

• Boundary, Edge, Point, and Pair Nodes for the Surface Reactions
Interface
• Theory for the Surface Reactions Interface

Boundary, Edge, Point, and Pair Nodes for the Surface Reactions
Interface
The Surface Reactions Interface has these boundary, edge, point, and pair nodes, listed
in alphabetical order, available from the Physics ribbon toolbar (Windows users),
Physics context menu (Mac or Linux users), or right-click to access the context menu
(all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are [Link] are available by clicking the
parent node and selecting it from the Attributes menu.

THE SURFACE REACTIONS INTERFACE | 349

 These nodes are described in this section:

• Initial Values
• Reactions
• Surface Concentration
• Surface Properties

All other available nodes are described for the Transport of Diluted Species interface.
See Domain, Boundary, and Pair Nodes for the Transport of Diluted Species Interface.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Surface Properties
Use the Surface Properties node to define the density of sites, the site occupancy
number, and the surface diffusion.

SPECIES CONSERVATION ON DEFORMING GEOMETRY

The Compensate for boundary stretching checkbox is selected by default. Click to clear
the checkbox if required.

SITES
Enter a value or expression for the Density of sites s (SI unit: mol/m2). The default
is 2 x 10-5 mol/m2.

Enter a Site occupancy number i (dimensionless), indicating how many surface sites a
surface species block upon adsorption.

This section is only present if the number of surface species is 1 or higher.

SURFACE DIFFUSION
Select a Surface material from the list. The default is None.

For each surface concentration species, the default Diffusion coefficient Di (SI unit: m2/
s) is User defined. Select Isotropic, Diagonal, Symmetric, or Full depending on the
characteristics of the surface, and then enter values or expressions in the field or matrix.

This section is only present if the number of surface species is 1 or higher.

350 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

BULK SPECIES
For each bulk species enter the Molar mass Mi (SI unit: kg/mol) and the Density i
(SI unit: kg/m3). The default molar mass is 0.144 kg/mol and the default density is
5320 kg/m3).

This section is only present if the number of bulk species is 1 or higher.

Surface Reaction Equations on Deforming Geometries

Initial Values
The Initial Values node allows the initial value or guess for the surface and bulk
concentrations.

INITIAL VALUES
Based on the number of surface species and number of bulk species entered for the
physics interface under Dependent Variables section, enter values for the same number
of Surface concentration cs, cs2, cs3, … (SI unit: mol/m2) and Bulk concentration cb,
cb2, cb3, … (SI unit: mol/m2) in each field.

Reactions
The Reactions node adds rate expression terms to the species transport equations in
order to account for consumption or production of species due to reactions.

REACTION RATE FOR SURFACE SPECIES

Add a reaction rate expression, Rs,cs1,cs2… (SI unit: mol/(m2·s)), for each surface
species taking part in a surface reaction.

This section is only present if the number of surface species is 1 or higher.

REACTION RATE FOR BULK SPECIES

Add a reaction rate expression, Rb,cb1,cb2… (SI unit: mol/(m2·s)), for each bulk
species taking part in a surface reaction.

This section is only present if the number of bulk species is 1 or higher.

THE SURFACE REACTIONS INTERFACE | 351

 Surface Concentration
Use the Surface Concentration node to set the surface concentrations for one or more
species on an edge (3D components) or a point (2D and 2D axisymmetric
components).

SURFACE CONCENTRATION
Select each species checkbox as needed and enter a value or expression for each species
concentration, cs,0,cs1,cs2… (SI unit: mol/(m2·s)).

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

352 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The Reacting Flow Interfaces
The Fuel Cell & Electrolyzer Module includes predefined multiphysics interfaces that
couple fluid with mass transport and reactions.

Selecting any of these interfaces under the Chemical Species Transport > Reacting Flow
or Chemical Species Transport > Reacting Flow in Porous Media branches of the Model
Wizard or Add Physics windows, adds a fluid flow interface and either a Transport of
Diluted Species or Transport of Concentrated Species interface to the Model Builder.

In addition, the Multiphysics node Reacting Flow is added. The Reacting Flow node
predefines and controls the couplings between the mass and fluid transport in order to
facilitate easy set up of models.

In this section:

• The Reacting Laminar Flow Interface

• The Reacting Flow Coupling Feature
• Physics Interface Features

The Reacting Laminar Flow Interface

The Reacting Laminar Flow ( ) multiphysics interface is used to simulate laminar flow
coupled to species transport in a gas or liquid.

It combines the Laminar Flow, and Transport of Concentrated Species interfaces. The
Reacting Flow multiphysics coupling, which is added automatically, couples fluid flow
and mass transport. The fluid flow can either be free flow or flow in a porous medium.
The species transport supports both a mixture, where the concentrations are of
comparable order of magnitude, and low-concentration solutes in a solvent.

The interface can be used for stationary and time-dependent analysis in 2D, 2D axial
symmetry, and 3D.

ON THE CONSTITUENT PHYSICS INTERFACES

The equations solved by the Laminar Flow interface are the Navier–Stokes equations for
conservation of momentum and the continuity equation for conservation of mass. A
Fluid Properties feature is active by default on the entire interface selection. A Porous
Medium feature can be added in order to model flow in porous media by solving the
Brinkman equations.

THE REACTING FLOW INTERFACES | 353

 The Transport of Concentrated Species interface solves for an arbitrary number of mass
fractions. The species equations include transport by convection, diffusion and,
optionally, migration in an electric field.

The Reacting Laminar Flow interface triggers pseudo time stepping for the
flow equations when Use pseudo time stepping for stationary equation form
in the Fluid Flow interface is set to Automatic from physics.

The Reacting Flow Coupling Feature

Use the Reacting Flow ( ) multiphysics coupling to simulate mass transport and
reactions in a gas or liquid mixture where the fluid flow can be dependent on the
mixture composition. When a Heat Transfer and a Chemistry interface are selected, use
this coupling to simulate heat transfer additionally to mass transport and reactions.

The coupling adds the heat source of reaction when a heat transfer interface is selected.
No additional heat source needs to be defined to account for it in the heat transfer
interface. It also accounts for the multiphysics stabilization terms, for work due to
pressure forces and viscous dissipation.

Select a Fluid Flow interface and a Species transport interface to couple fluid flow with
mass transport. Chemistry and Heat Transfer are optional. They can be set to None when
the coupling is used to simulate isothermal mixtures.

When a Chemistry interface is selected and Heat Transfer is set to None, fluid properties
are taken from the Chemistry interface. Set the Temperature to evaluate the fluid
properties synchronized with all the physics interfaces at the given temperature.

Select a Chemistry interface and Heat Transfer interface in order to account for the heat
of reaction, enthalpy diffusion, viscous heating and mass fluxes contributing to the
heat and energy balance.

When Chemistry is set to None and a Heat Transfer interface is selected, the
coupling is solved in the same way as when no Heat Transfer interface is
selected. Thermodynamic properties are required by the Heat Transfer
interface.

354 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

 The pressure, velocity, and temperature variables of the Reacting Flow
coupling node are set to the Common Model Input values of the Default
Model Inputs node on the complementary selection, that is, all domains
except those from the Selection list. It allows to couple multiple fluid flow
or transport of species interfaces with a single heat transfer interface. See
Default Model Inputs in the COMSOL Multiphysics Reference Manual
for details.

DOMAIN LEVEL SYNCHRONIZATION

The Reacting Flow coupling synchronizes the features from a Chemistry interface, Heat
Transfer interface, Single-Phase Flow, or Brinkman Equations, interface and a
Transport of Concentrated Species interface. When the Chemistry interface is not
selected, the density in the Single-Phase Flow interface is automatically synchronized
to the one defined by the Transport of Concentrated Species interface.

The velocity field used by the Transport of Concentrated Species interface and Heat
Transfer interface is synchronized to the one computed in the Single-Phase Flow
interface.

When a Chemistry interface is selected, the Reacting Flow coupling synchronizes the
definition of the thermal conductivity, density, heat capacity, enthalpy, and dynamic
viscosity with the other coupled physics interfaces. The reference temperature is taken
from the Heat Transfer interface.

THE STEFAN VELOCITY

The Reacting Flow coupling feature automatically couples mass transfer on boundaries
and applies a corresponding velocity contribution for the flow. Prescribing a net mass
boundary flux in the Transport of Concentrated Species interface, either using a Flux or
Mass Fraction feature, the Reacting Flow feature computes The Stefan Velocity and
applies this in Wall features using the same selection.

MASS TRANSFER TO OTHER PHASES IN POROUS MEDIA

When coupled to the Brinkman Equations interface, the Reacting Flow node
automatically computes the net mass source or sink in a Reactions (when Mass transfer
to other phases is enabled) node in the Transport of Concentrated Species interface and
adds the corresponding source/sink to the momentum equations of the Fluid and
Matrix Properties domains.

THE REACTING FLOW INTERFACES | 355

 TURBULENT MASS TRANSFER
When a turbulence model is used, the Reacting Flow coupling applies turbulence
modeling for the mass transport in the following manners:

• Turbulent heat transfer and mass transport are added, defined from the turbulent
viscosity and a turbulent Schmidt number (for more information, see Turbulent
Mass Transport Models).
• Temperature and mass transport wall functions. When a turbulence model using
wall functions is used for the fluid flow, the Reacting Flow coupling automatically
adds wall functions for the temperature and mass transport on the same boundaries
(for more information, see Mass Transport Wall Functions).

SETTINGS
The Label is the default multiphysics coupling feature name.

The Name is used primarily as a scope prefix for variables defined by the coupling node.
Refer to such variables in expressions using the pattern <name>.<variable_name>. In
order to distinguish between variables belonging to different coupling nodes or physics
interfaces, the name string must be unique. Only letters, numbers, and underscores (_)
are permitted in the Name field. The first character must be a letter.

The default Name (for the first multiphysics coupling feature in the model) is nirf1.

DOMAIN SELECTION
The Reacting Flow coupling is automatically defined on the intersection of the
selections for the coupled interfaces. When modeling porous media transport a
Porous Medium feature, applied on the same domains, is needed in all coupled transport
interfaces. Reacting Flow coupling supports porous media transport in Fluid flow and
Species transport interfaces.

The Selection list displays the domains where the coupling feature is active.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Fluid flow,
Species transport, Chemistry, Heat Transfer lists include all applicable physics interfaces.

356 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The default values depend on how this coupling node is created.

• If it is added from the Physics ribbon (Windows users), Physics contextual toolbar
(Mac and Linux users), or context menu (all users), then the first physics interface
of each type in the component is selected as the default.
• If it is added automatically when a multiphysics interface is chosen in the Model
Wizard or Add Physics window, then the two participating physics interfaces are
selected.

You can also select None from a list to uncouple the node from a physics interface.

Click the Go to Source buttons ( ) to move to the main physics interface node for
the selected physics interface.

Click the Show or Hide Physics Properties Settings button ( ) to toggle the display of
physics properties settings affecting the coupling feature. When a turbulence model is
used, turbulent heat and mass transfer is automatically accounted for (see the settings
in the Turbulence section below). Using Reacting Flow, the heat and mass transfer
treatment at walls follows that applied for the fluid flow. Therefore the Wall treatment
setting is also displayed when using a turbulence model. For more information on
turbulent mass transfer at walls, see the section Mass Transport Wall Functions in the
CFD Module User’s Guide.

If a physics interface is deleted and then added to the model again, then
in order to reestablish the coupling, you need to choose the physics
interface again from the Fluid flow, Species transport, Chemistry and Heat
Transfer lists. This is applicable to all multiphysics coupling nodes that
would normally default to the once present interface. See Multiphysics
Modeling Workflow in the COMSOL Multiphysics Reference Manual.

MASS TRANSPORT TURBULENCE MODEL

When the fluid flow interface uses a turbulence model, select an option from the Mass
transport turbulence model list — Kays-Crawford, High Schmidt Number, or User-defined
turbulent Schmidt number.

For User-defined turbulent Schmidt number, enter a Turbulent Schmidt number ScT
(dimensionless).

The turbulent mass transfer added to the mass fraction equations is defined as

THE REACTING FLOW INTERFACES | 357

 T
-  i
N i T = – --------
Sc T

where T is the turbulent viscosity defined by the flow interface, and the turbulent
Schmidt number, ScT, depends on the Mass transport turbulence model used.

HEAT TRANSFER TURBULENCE MODEL

This section is available when a Heat Transfer interface is selected and the fluid flow
interface uses a turbulence model. Select an option from the Heat transport turbulence
model list: Kays-Crawford (the default), Extended Kays-Crawford, or User-defined
turbulent Prandtl number.

For Extended Kays-Crawford, enter a Reynolds number at infinity Reinf (dimensionless).

For User-defined turbulent Prandtl number, enter a Turbulent Prandtl number PrT
(dimensionless).

When the flow interface uses a RANS turbulence model, the conductive heat flux is
defined as

q = –  k + k T  T

with the turbulent thermal conductivity defined as

T Cp
k T = -------------
-
Pr T

where T is defined by the flow interface, and PrT depends on the Heat transport
turbulence model. See Turbulent Conductivity for details.

The Turbulence model type used by the fluid flow interface can be displayed by selecting
the Show or Hide Physics Property Settings button at the right of the Fluid flow list.

Physics Interface Features

Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using.

358 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

LAMINAR FLOW
The available physics features for The Laminar Flow interface are listed in the section
Domain, Boundary, Pair, and Point Nodes for Single-Phase Flow in the COMSOL
Multiphysics Reference Manual.

TRANSPORT OF CONCENTRATED SPECIES

The available physics features for The Transport of Concentrated Species interface are
listed in the section Domain, Boundary, and Pair Nodes for the Transport of
Concentrated Species Interface.

THE REACTING FLOW INTERFACES | 359

 The Reacting Flow in Porous Media
Interfaces
The Fuel Cell & Electrolyzer Module includes two predefined multiphysics interface
that couples fluid flow in porous media with mass transport and reactions in porous
media.

Selecting a multiphysics interface under the Chemical Species Transport > Reacting Flow
in Porous Media branch of the Model Wizard or Add Physics windows, a Brinkman
Equations interface combined with either the Transport of Diluted Species in Porous
Media interface or the Transport of Concentrated Species interfaces are added to the
Model Builder.

In addition, the Multiphysics node is added, which includes the multiphysics coupling
feature. The multiphysics coupling feature controls the coupling between the separate
interfaces in order to facilitate easy setup of models.

In this section:

• The Reacting Flow in Porous Media, Transport of Diluted Species Interface

• The Reacting Flow in Porous Media, Transport of Concentrated Species Interface
• The Reacting Flow, Diluted Species Coupling Feature
• The Reacting Flow Coupling Feature
• Physics Interface Features

The Reacting Flow in Porous Media, Transport of Diluted Species

Interface
The Reacting Flow in Porous Media, Transport of Diluted Species interface ( ) is used
to study the flow and chemical composition of a gas or liquid moving through the
interstices of a porous medium.

It combines the Brinkman Equations, and Transport of Diluted Species in Porous Media
interfaces. The Reacting Flow, Diluted Species multiphysics coupling feature, which is
added automatically, couples the fluid flow and mass transport. A Porous Material node,
including a Fluid and a Solid subnode, is also added to the component.

360 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The interface can be used for stationary and time-dependent analysis in 2D, 2Daxi, and
3D.

On the constituent physics interfaces:

The Brinkman Equations interface computes the fluid velocity and pressure fields of
single-phase flow in porous media in the laminar flow regime. A Porous Medium feature
is active by default on the entire interface selection.

The Transport of Diluted Species in Porous Media interface computes the species
concentration in free and porous media, assuming that the species are solutes,
dissolved in a solvent of significantly higher concentration. A Porous Medium feature is
active by default on the entire interface selection.

The Reacting Flow in Porous Media, Transport of Concentrated

Species Interface
The Reacting Flow in Porous Media, Transport of Concentrated Species interface ( ) is
used to study the flow and chemical composition of a gas or liquid moving through
the interstices of a porous medium. The fluid can consist of a mixture of species where
the individual concentrations are of comparable order of magnitude.

It combines the Brinkman Equations, and Transport of Concentrated Species interfaces.

The Reacting Flow multiphysics coupling feature, which is added automatically, couples
the fluid flow and mass transport. A Porous Material node is also added to the
component.

The interface can be used for stationary and time-dependent analysis in 2D, 2Daxi, and
3D.

On the constituent physics interfaces:

The Brinkman Equations interface computes the fluid velocity and pressure fields of
single-phase flow in porous media in the laminar flow regime. A Porous Medium feature
is active by default on the entire interface selection.

The Transport of Concentrated Species in Porous Media interface solves for an arbitrary
number of mass fractions in free and porous media. In the current multiphysics
interface a Porous Medium feature is active by default on the entire interface selection.

THE REACTING FLOW IN POROUS MEDIA INTERFACES | 361

 The Reacting Flow, Diluted Species Coupling Feature
For the settings of this feature see The Reacting Flow, Diluted Species Coupling
Feature in the Chemical Reaction Engineering Module User’s Guide.

The Reacting Flow Coupling Feature

For the settings of this feature see The Reacting Flow Coupling Feature in the
Chemical Reaction Engineering Module User’s Guide.

Physics Interface Features

Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

BRINKMAN EQUATIONS
The available physics features for The Brinkman Equations interface are listed in the
Domain, Boundary, Point, and Pair Nodes for the Brinkman Equations Interface
section in the CFD Module User’s Guide.

TRANSPORT OF DILUTED SPECIES

The available physics features for The Transport of Diluted Species interface are listed
in the section.

TRANSPORT OF CONCENTRATED SPECIES

The available physics features for The Transport of Concentrated Species interface are
listed in the section.

362 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The Nonisothermal Reacting Flow
Multiphysics Interfaces
Nonisothermal Reacting Flow is a predefined multiphysics interface, available in the
Fuel Cell & Electrolyzer Module, that couples fluid flow and heat transfer with species
transport and reactions.

Selecting an entry in the Chemical Species Transport > Nonisothermal Reacting Flow
branch of the Model Wizard or Add Physics windows, the following four physics
interfaces are added to the Model Builder:

• Chemistry
• Transport of Concentrated Species
• Laminar Flow, or a turbulent flow equivalent like Turbulent Flow, k-.
• Heat Transfer in Fluids

In addition, the Multiphysics node is added, which includes the multiphysics coupling
feature Reacting Flow. The Reacting Flow feature predefines and controls the couplings
between the separate interfaces in order to facilitate set up of models.

In this section:

• The Nonisothermal Reacting Laminar Flow Interface

• The Nonisothermal Reacting Turbulent Flow, k-e Interface
• The Nonisothermal Reacting Turbulent Flow, k-w Interface
• The Nonisothermal Reacting Turbulent Flow, SST Interface
• The Nonisothermal Reacting Turbulent Flow, Low Re k-e Interface
• The Reacting Flow Coupling Feature
• Physics Interface Features

The Nonisothermal Reacting Laminar Flow Interface

The Nonisothermal Reacting Laminar Flow ( ) multiphysics interface is used to
simulate laminar flow and heat transfer coupled to species transport in a gas or liquid
moving either in a free flow or through the interstices of a porous medium.

THE NONISOTHERMAL REACTING FLOW MULTIPHYSICS INTERFACES | 363

 It combines the Chemistry, Transport of Concentrated Species, Laminar Flow, and Heat
Transfer in Fluids interfaces. The Reacting Flow multiphysics coupling, which is added
automatically, couples fluid flow, heat transfer, and mass transfer. The species transport
supports both a mixture, where the concentrations are of comparable order of
magnitude, and low-concentration solutes in a solvent.

The multiphysics coupling takes into account the heat of reaction in the fluid and at
the surface of a porous catalyst, enthalpy diffusion, and contributing mass fluxes. In
addition, the temperature dependency of the chemical properties and reactions are
accounted for.

The interface can be used for stationary and time-dependent analysis in 2D, 2D axial
symmetry, and 3D.

On the constituent physics interfaces:

The Chemistry interface defines thermodynamic properties and transport properties of

the fluid. Provided that properties of each species have been defined, composition
dependent mixture properties such as the heat capacity, the density, and the heat
conduction are defined. The Chemistry interface also defines reaction rates for species
involved in the chemical reactions added to the system.

The Transport of Concentrated Species interface solves for an arbitrary number of mass
fractions. The species equations include transport by convection, diffusion and,
optionally, migration in an electric field. A Transport Properties feature is active by
default on the entire interface selection

The equations solved by the Laminar Flow interface are the Navier–Stokes equations for
conservation of momentum and the continuity equation for conservation of mass. A
Fluid Properties feature is active by default on the entire interface selection.

The Heat Transfer interface solves for conservation of energy. A Fluid feature is active
by default on the entire interface selection.

The Nonisothermal Reacting Turbulent Flow, k- Interface

The Nonisothermal Reacting Turbulent Flow, k- ( ) multiphysics interface is used to
simulate flow in the turbulent regime, heat transfer, and species transport and in a gas
or liquid moving either in a free flow or through the interstices of a porous medium.
This interface requires a license for either the CFD Module or the Heat Transfer
Module.

364 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

It combines the Chemistry, Transport of Concentrated Species, Turbulent Flow, k-, and
Heat Transfer in Fluids interfaces. The Reacting Flow multiphysics coupling, which is
added automatically, couples fluid flow, heat transfer and mass transfer. The species
transport supports both a mixture, where the concentrations are of comparable order
of magnitude, and low-concentration solutes in a solvent.

The multiphysics coupling takes into account the heat of reaction in the fluid and at
the surface of a porous catalyst, enthalpy diffusion, and contributing mass fluxes. In
addition, the temperature dependency of the chemical properties and reactions are
accounted for.

The interface can be used for stationary and time-dependent analysis in 2D, 2D axial
symmetry, and 3D.

On the constituent physics interfaces:

The Chemistry interface defines thermodynamic properties and transport properties of

the fluid. Provided that properties of each species have been defined, composition
dependent mixture properties such as the heat capacity, the density, and the heat
conduction are defined. The Chemistry interface also defines reaction rates for species
involved in the chemical reactions added to the system.

The Transport of Concentrated Species interface solves for an arbitrary number of mass
fractions. The species equations include transport by convection, diffusion and,
optionally, migration in an electric field. Mass transfer close to walls is modeled using
wall functions

The equations solved by the Turbulent Flow, k- interface are the Navier–Stokes
equations for conservation of momentum and the continuity equation for
conservation of mass. Turbulence effects are modeled using the standard two-equation
k- model with realizability constraints. Flow close to walls is modeled using wall
functions.

The Heat Transfer interface solves for conservation of energy. A Fluid feature is active
by default on the entire interface selection. Heat transfer close to walls is modeled
using wall functions.

The Nonisothermal Reacting Turbulent Flow, k- Interface

The Nonisothermal Reacting Turbulent Flow, k- ( ) multiphysics interface is used to
simulate flow in the turbulent regime, heat transfer, and species transport and in a gas

THE NONISOTHERMAL REACTING FLOW MULTIPHYSICS INTERFACES | 365

 or liquid moving either in a free flow or through the interstices of a porous medium.
This interface requires a license for the CFD Module.

It combines the Chemistry, Transport of Concentrated Species, Turbulent Flow, k-, and
Heat Transfer in Fluids interfaces. The Reacting Flow multiphysics coupling, which is
added automatically, couples fluid flow, heat transfer and mass transfer. The species
transport supports both a mixture, where the concentrations are of comparable order
of magnitude, and low-concentration solutes in a solvent.

The multiphysics coupling takes into account the heat of reaction in the fluid and at
the surface of a porous catalyst, enthalpy diffusion, and contributing mass fluxes. In
addition, the temperature dependency of the chemical properties and reactions are
accounted for.

The interface can be used for stationary and time-dependent analysis in 2D, 2D axial
symmetry, and 3D.

On the constituent physics interfaces:

The Chemistry interface defines thermodynamic properties and transport properties of

the fluid. Provided that properties of each species have been defined, composition
dependent mixture properties such as the heat capacity, the density, and the heat
conduction are defined. The Chemistry interface also defines reaction rates for species
involved in the chemical reactions added to the system.

The Transport of Concentrated Species interface solves for an arbitrary number of mass
fractions. The species equations include transport by convection, diffusion and,
optionally, migration in an electric field. Mass transfer close to walls is modeled using
wall functions

The equations solved by the Turbulent Flow, k- interface are the Navier–Stokes
equations for conservation of momentum and the continuity equation for
conservation of mass. The fluid flow turbulence is modeled using the Wilcox revised
k- model with realizability constraints. Flow close to walls is modeled using wall
functions.

The Heat Transfer interface solves for conservation of energy. A Fluid feature is active
by default on the entire interface selection. Heat transfer close to walls is modeled
using wall functions.

366 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The Nonisothermal Reacting Turbulent Flow, SST Interface
The Nonisothermal Reacting Turbulent Flow, SST ( ) multiphysics interface is used to
simulate flow in the turbulent regime, heat transfer, and species transport and in a gas
or liquid moving either in a free flow or through the interstices of a porous medium.
This interface requires a license for either the CFD Module or the Heat Transfer
Module.

It combines the Chemistry, Transport of Concentrated Species, Turbulent Flow, SST, and
Heat Transfer in Fluids interfaces. The Reacting Flow multiphysics coupling, which is
added automatically, couples fluid flow, heat transfer and mass transfer. The species
transport supports both a mixture, where the concentrations are of comparable order
of magnitude, and low-concentration solutes in a solvent.

The multiphysics coupling takes into account the heat of reaction in the fluid and at
the surface of a porous catalyst, enthalpy diffusion, and contributing mass fluxes. In
addition, the temperature dependency of the chemical properties and reactions are
accounted for.

The interface can be used for stationary and time-dependent analysis in 2D, 2D axial
symmetry, and 3D.

On the constituent physics interfaces:

The Chemistry interface defines thermodynamic properties and transport properties of

the fluid. Provided that properties of each species have been defined, composition
dependent mixture properties such as the heat capacity, the density, and the heat
conduction are defined. The Chemistry interface also defines reaction rates for species
involved in the chemical reactions added to the system.

The Transport of Concentrated Species interface solves for an arbitrary number of mass
fractions. The species equations include transport by convection, diffusion and,
optionally, migration in an electric field. Mass transfer close to walls is modeled using
wall functions

The equations solved by the Turbulent Flow, SST interface are the Navier–Stokes
equations for conservation of momentum and the continuity equation for
conservation of mass. The fluid flow turbulence is modeled using the SST model. The
SST model is a low-Reynolds number model which means that it resolves the velocity,
pressure, and concentrations all the way down to the wall. For that reason this physics
interface is suited for studying mass transfer at high Schmidt numbers. The SST model

THE NONISOTHERMAL REACTING FLOW MULTIPHYSICS INTERFACES | 367

 depends on the distance to the closest wall, and the interface therefore includes a wall
distance equation.

The Heat Transfer interface solves for conservation of energy. A Fluid feature is active
by default on the entire interface selection. Heat transfer close to walls is modeled
using wall functions.

The Nonisothermal Reacting Turbulent Flow, Low Re k- Interface

The Nonisothermal Reacting Turbulent Flow, Low Re k- ( ) multiphysics interface is
used to simulate flow in the turbulent regime, heat transfer, and species transport and
in a gas or liquid moving either in a free flow or through the interstices of a porous
medium. This interface requires a license for either the CFD Module or the Heat
Transfer Module.

It combines the Chemistry, Transport of Concentrated Species,

Turbulent Flow, Low Re k-, and Heat Transfer in Fluids interfaces. The Reacting Flow
multiphysics coupling, which is added automatically, couples fluid flow, heat transfer
and mass transfer. The species transport supports both a mixture, where the
concentrations are of comparable order of magnitude, and low-concentration solutes
in a solvent.

The multiphysics coupling takes into account the heat of reaction in the fluid and at
the surface of a porous catalyst, enthalpy diffusion, and contributing mass fluxes. In
addition, the temperature dependency of the chemical properties and reactions are
accounted for.

The interface can be used for stationary and time-dependent analysis in 2D, 2D axial
symmetry, and 3D.

On the constituent physics interfaces:

The Chemistry interface defines thermodynamic properties and transport properties of

the fluid. Provided that properties of each species have been defined, composition
dependent mixture properties such as the heat capacity, the density, and the heat
conduction are defined. The Chemistry interface also defines reaction rates for species
involved in the chemical reactions added to the system.

The Transport of Concentrated Species interface solves for an arbitrary number of mass
fractions. The species equations include transport by convection, diffusion and,
optionally, migration in an electric field. Mass transfer close to walls is modeled using
wall functions

368 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The equations solved by the Turbulent Flow, Low Re k- interface are the Navier–Stokes
equations for conservation of momentum and the continuity equation for
conservation of mass. The fluid flow turbulence is modeled using the AKN
low-Reynolds number k-e model. The low-Reynolds number model resolves the
velocity, pressure, and concentrations all the way down to the wall. For that reason this
physics interface is suited for studying mass transfer at high Schmidt numbers. The
AKN model depends on the distance to the closest wall, and the interface therefore
includes a wall distance equation.

The Heat Transfer interface solves for conservation of energy. A Fluid feature is active
by default on the entire interface selection. Heat transfer close to walls is modeled
using wall functions.

The Reacting Flow Coupling Feature

For the settings of this feature see The Reacting Flow Coupling Feature.

Physics Interface Features

Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using.

CHEMISTRY
The available physics features for the Chemistry interface are listed in the section
Feature Nodes Available for the Chemistry Interface.

LAMINAR FLOW
The available physics features for The Laminar Flow interface are listed in the section
Domain, Boundary, Pair, and Point Nodes for Single-Phase Flow in the CFD Module
User’s Guide.

TURBULENT FLOW
The available physics features for a single-phase flow interface using a turbulence
model, for example the Turbulent Flow, k- interfaceis listed in the section Domain,

THE NONISOTHERMAL REACTING FLOW MULTIPHYSICS INTERFACES | 369

 Boundary, Pair, and Point Nodes for Single-Phase Flow in the CFD Module User’s
Guide.

TRANSPORT OF CONCENTRATED SPECIES

The available physics features for the Transport of Concentrated Species interface are
listed in the section Domain, Boundary, and Pair Nodes for the Transport of
Concentrated Species Interface.

HEAT TRANSFER
The available physics features for the Heat Transfer interface are listed in the section
The Heat Transfer Features in the COMSOL Multiphysics Reference Manual.

370 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

T he o r y f o r th e Tran sp ort of D i l u t ed
S pe c i e s I nte r fac e
provides a predefined modeling environment for studying the evolution of chemical
species transported by diffusion and convection. The physics interface assumes that all
species present are dilute; that is, that their concentration is low compared to a solvent
fluid or solid. As a rule of thumb, a mixture containing several species can be
considered dilute when the concentration of the solvent is more than 90 mol%. Due
to the dilution, mixture properties such as density and viscosity can be assumed to
correspond to those of the solvent.

When studying mixtures that are not dilute, the mixture and transport properties
depend on the composition, and a different physics interface is recommended. See The
Transport of Concentrated Species Interface in the for more information.

Fick’s law governs the diffusion of the solutes, dilute mixtures, or solutions, while the
phenomenon of ionic migration is sometimes referred to as electrokinetic flow. The
Transport of Diluted Species interface supports the simulations of chemical species
transport by convection, migration, and diffusion in 1D, 2D, and 3D as well as for
axisymmetric components in 1D and 2D.

In this section:

• Adding Transport Through • Mass Balance Equation

Migration • Mass Sources for Species Transport
• Material Balance Formulation • Solving a Diffusion Equation Only
• Crosswind Diffusion • Supporting Electrolytes
• Danckwerts Inflow Boundary • References
Condition
• Equilibrium Reaction Theory

Note: Some features explained in this section require certain add-on modules. For
details see [Link]

THEORY FOR THE TRANSPORT OF DILUTED SPECIES INTERFACE | 371

 The section also includes the theory for :

• Adsorption • Mass Balance Equation for

• Convection in Porous Media Transport of Diluted Species in
Porous Media
• Diffusion in Porous Media
• Mass Transport in Fractures
• Dispersion
• Reactions

Mass Balance Equation

The default node attributed to the Transport of Diluted Species interface models
chemical species transport through diffusion and convection and solves the mass
conservation equation for one or more chemical species i:

c i
------- +   J i + u  c i = R i (5-17)
t

Equation 5-17 in its form above includes the transport mechanisms diffusion and
convection. If Migration in Electric Field is activated (only available in some add-on
products), the migration mechanism will be added to the equation as well. See more
details in the section Adding Transport Through Migration.

• ci is the concentration of the species (SI unit: mol/m3)

• Di denotes the diffusion coefficient (SI unit: m2/s)
• Ri is a reaction rate expression for the species (SI unit: mol/(m3·s))
• u is the mass averaged velocity vector (SI unit: m/s)
• Ji is the mass flux diffusive flux vector (SI unit: mol/(m2·s))

The mass flux relative to the mass averaged velocity, Ji (SI unit: mol/(m2·s)), is
associated with the mass balance equation above and used in boundary conditions and
flux computations. The Transport of Diluted Species interface always includes mass
transport due to molecular diffusion. In this case the mass flux Ji defines the diffusive
flux vector

J i = – D c i (5-18)

An input field for the diffusion coefficient is available.

372 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

When Migration in Electric Fields is activated, the migration term is also added to
the diffusive flux vector as shown in the section Adding Transport Through Migration.

The third term on the left side of Equation 5-17 describes the convective transport due
to a velocity field u. This field can be expressed analytically or obtained from coupling
the physics interface to one that solves for fluid flow, such as Laminar Flow. Note that
all fluid flow interfaces solve for the mass averaged velocity.

On the right-hand side of the mass balance equation (Equation 5-17), Ri represents a
source or sink term, typically due to a chemical reaction or desorption on a porous
matrix. To specify Ri, another node must be added to the Transport of Diluted Species
interface — the Reaction node for example, which includes an input field for specifying
a reaction expression using the variable names of all participating species.

Equilibrium Reaction Theory

The feature Equilibrium Reaction is described in this section. A chemical equilibrium
reaction system is defined by the stoichiometry of the reaction and the relation
between the chemical activities of the chemical species participating in the reaction (the
equilibrium condition).

The kinetics of the reaction is so fast that the equilibrium condition is fulfilled at all
times in all space coordinates.

The equilibrium condition is commonly based on the stoichiometric coefficients,

i (dimensionless), of the reaction; the species activities of the reacting species
ai (dimensionless); and an equilibrium constant, Keq (1) according to


 ai i
i  products
K eq = ----------------------------------
–
 ai i
i  reactants

where the species activities are defined as

ci
a i =  c i -------
c a0

where ca0 (SI unit: mol/m3) is the standard molarity, and c,i (dimensionless) an
activity coefficient.

THEORY FOR THE TRANSPORT OF DILUTED SPECIES INTERFACE | 373

 Defining the stoichiometric coefficients positive for products and negative for
reactants, the above equilibrium condition can also be written

i
K eq =  ai
i

The Equilibrium Reaction node solves for a reaction rate so that the equilibrium
condition is always fulfilled in the domain. It is available for the Chemical Reaction
Engineering Module; Battery Design Module; Corrosion Module; Electrochemistry
Module; Electrodeposition Module; and Fuel Cell & Electrolyzer Module.

c,i is set to unity when the Equilibrium constant is selected in the Settings
window. For nonunity activity coefficients, a user-defined equilibrium
condition can be used.

EQUILIBRIUM REACTIONS AND INFLOW BOUNDARY CONDITIONS

Contradictory constraints arise if the boundary conditions for concentrations or
activities are set so that the domain equilibrium condition is not fulfilled. Special
treatment is therefore needed at Inflow boundaries, where the concentrations are set
for all species in the mass transport interfaces.

One way of avoiding competing constraints on an inflow boundary is to add an

additional reaction coordinate degree of freedom, solved for to create a set of modified
inflow concentrations that fulfill the domain equilibrium condition. The reaction
coordinate gives rise to a concentration shift, which is the offset to the inflow
concentrations provided by the user. The shift for each species obeys the stoichiometry
of the reaction and the equilibrium expression. The modified inflow concentrations are
then used in the boundary conditions for the domain mass transport equations. The
resulting modified inflow concentrations can be seen as the stationary solution for a
batch reactor with the user inflow concentrations as initial concentrations. In addition,
the domain reaction rate degree of freedom of the equilibrium reaction is constrained
to zero on all Inflow boundaries.

EQUILIBRIUM REACTIONS AND CONCENTRATION BOUNDARY

CONDITIONS
No special treatment is made with regards to input concentration values of the
Concentration boundary node. Using this feature, you can explicitly set one or a set of
concentrations, and the equilibrium condition acts on the rest of the concentrations.

374 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

However, there is no solution to the problem if more concentrations than the number
of species minus the number of equilibrium reactions are set using this feature.

EQUILIBRIUM REACTIONS AND TIME-DEPENDENT SIMULATIONS

Spurious oscillations may occur in a time-dependent problem if the initial conditions
do not fulfill the equilibrium condition. Since equilibrium reactions are assumed to be
infinitely fast, the solution is to initialize the problem using an additional study step,
solving for a stationary problem with all nonequilibrium reaction rates set to zero.
Manual scaling of the reaction rate dependent variables is needed in this study step.

Material Balance Formulation

The default node attributed to assumes chemical species transport through diffusion
and convection (depending on the modules licensed, a checkbox to activate migration
is available) and implements the mass balance equation in Equation 5-17.

Two formulations of the mass balance are available, these are referred to as the
nonconservative and conservative form:

c
nonconservative: ----- +   J i + u  c = R (5-19)
t

conservative: ----c- +   J i +    cu  = R (5-20)

t

In these equations Ji is the diffusive flux vector, u is the solvent velocity field, and R
is a production or consumption rate expression. The nonconservative is often more
numerically robust, and therefore chosen by default for the interface. This formulation
however disregards density changes in the fluid. This can be seen by expanding the
conservative convection term using the chain rule. One of the resulting terms, c·u,
equals zero for an incompressible fluid which results in the nonconservative form. To
switch between the two formulations, click the Show button ( ) and select Advanced
Physics Options.

Solving a Diffusion Equation Only

Remove the convection term from Equation 5-19 and Equation 5-20 by clearing the
Convection checkbox in the Transport Mechanisms section for . The equation then
becomes

THEORY FOR THE TRANSPORT OF DILUTED SPECIES INTERFACE | 375

 ----c- =   J + R
t i

Mass Sources for Species Transport

Note: The features below are only available in a limited set of add-on products. For a
detailed overview of which features are available in each product, visit
[Link]

There are two types of mass sources in the Transport of Diluted Species interface: point
sources and line sources.

POINT SOURCE
A point source is theoretically formed by assuming a mass injection/ejection, Qꞏ c (SI
unit: mol/(m3·s)), in a small volume V and then letting the size of the volume tend
to zero while keeping the total mass flux constant. Given a point source strength, qꞏ p,c
(SI unit: mol/s), this can be expressed as

ꞏ
lim
V  0  Qc = qꞏ p,c (5-21)
V

An alternative way to form a point source is to assume that mass is injected/extracted

through the surface of a small object. Letting the object surface area tend to zero while
keeping the mass flux constant results in the same point source. For this alternative
approach, effects resulting from the physical object’s volume need to be neglected.

The weak contribution

qꞏ p,c test  c 

is added at a point in the geometry. As can be seen from Equation 5-21, Qꞏ c must tend
to plus or minus infinity as V tends to zero. This means that in theory the
concentration also tends to plus or minus infinity.

Observe that “point” refers to the physical representation of the source. A point source
can therefore only be added to points in 3D components and to points on the
symmetry axis in 2D axisymmetry components. Other geometrical points in 2D
components represent physical lines.

376 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The finite element representation of Equation 5-21 corresponds to a finite
concentration at a point with the effect of the point source spread out over a region
around the point. The size of the region depends on the mesh and on the strength of
the source. A finer mesh gives a smaller affected region but also a more extreme
concentration value. It is important not to mesh too finely around a point source since
this can result in unphysical concentration values. It can also have a negative effect on
the condition number for the equation system.

LINE SOURCE
A line source can theoretically be formed by assuming a source of strength Qꞏ l,c (SI
unit: mol/(m3·s)), located within a tube with cross section S and then letting S tend
to zero while keeping the total mass flux per unit length constant. Given a line source
strength, qꞏ l,c (SI unit: mol/(m·s)), this can be expressed as

lim
S  0  Qꞏ l,c = qꞏ l,c (5-22)
S

As in the point source case, an alternative approach is to assume that mass is injected/
extracted through the surface of a small object. This results in the same mass source,
but requires that effects resulting from the physical object’s volume are neglected.

The weak contribution

qꞏ l,c test  c 

is added on lines in 3D or at points in 2D (which represent cut-through views of lines).

Line sources can also be added on the axisymmetry line in 2D axisymmetry
components. It cannot, however, be added on geometrical lines in 2D since those
represent physical planes.

As with a point source, it is important not to mesh too finely around the line source.

For feature node information, see Line Mass Source and Point Mass
Source.

THEORY FOR THE TRANSPORT OF DILUTED SPECIES INTERFACE | 377

 Adding Transport Through Migration

Note: Migration is only available in a limited set of add-on products. For a detailed
overview of which features are available in each product, visit
[Link]

In addition to transport due to convection and diffusion, the Transport of Diluted

Species interface supports ionic species transport by migration. This is done by
selecting the Migration in Electric Field checkbox under the Transport Mechanisms
section for the physics interface. The mass balance then becomes:

c i
+    – D i c i – z i u m i F c i V + c i u  = R i (5-23)
t

where

• ci (SI unit: mol/ m3) denotes the concentration of species i

• Di (SI unit: m2/s) is the diffusion coefficient of species i
• u (SI unit: m/s) is the fluid velocity
• F (SI unit: A·s/mol) refers to Faraday’s constant
• V (SI unit: V) denotes the electric potential
• zi (dimensionless) is the charge number of the ionic species, and
• um,i (SI unit: mol·s/kg) is its ionic mobility

In this case the diffusive flux vector is

J i = – D i c i – z i u m i Fc i V

The velocity, u, can be a computed fluid velocity field from a Fluid Flow interface or
a specified function of the spatial variables x, y, and z. The potential can be provided
by an expression or by coupling the system of equations to a current balance, such as
the Electrostatics interface. Sometimes it is assumed to be a supporting electrolyte
present, which simplifies the transport equations. In that case, the modeled charged
species concentration is very low compared to other ions dissolved in the solution.
Thus, the species concentration does not influence the solution’s conductivity and the
net charge within the fluid.

378 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The Nernst–Einstein relation can in many cases be used for relating the species
mobility to the species diffusivity according to

D
u m i = --------i
RT

where R (SI unit: J/(mol·K)) is the molar gas constant and T (SI unit: K) is the
temperature.

Note: In the Nernst–Planck Equations interface, the ionic species contribute to the
charge transfer in the solution. It includes an electroneutrality condition and also
computes the electric potential field in the electrolyte. For more information, see
Theory for the Nernst–Planck Equations Interface. This interface is included in the
Chemical Reaction Engineering Module.

Supporting Electrolytes
In electrolyte solutions, a salt can be added to provide a high electrolyte conductivity
and decrease the ohmic losses in a cell. These solutions are often called supporting
electrolytes, buffer solutions, or carrier electrolytes. The added species, a negative and
a positive ion pair, predominates over all other species. Therefore, the supporting
electrolyte species can be assumed to dominate the current transport in the solution.
In addition, the predominant supporting ions are usually selected so that they do not
react at the electrode surfaces since the high conductivity should be kept through the
process, that is, they should not be electroactive species. This also means that the
concentration gradients of the predominant species in a supporting electrolyte are
usually negligible.

Modeling and solving for a supporting electrolyte in the Electrostatics or Secondary

Current Distribution interfaces will give a potential distribution that drives the
migration in the Transport of Diluted Species Interface.

The current density vector is proportional to the sum of all species fluxes as expressed
by Faraday’s law:

i = F  zi Ni
i

THEORY FOR THE TRANSPORT OF DILUTED SPECIES INTERFACE | 379

 The electroneutrality condition ensures that there is always a zero net charge at any
position in a dilute solution. Intuitively, this means that it is impossible to create a
current by manually pumping positive ions in one direction and negative ions in the
other. Therefore, the convective term is canceled out to yield the following expression
for the electrolyte current density, where j denotes the supporting species:

 – z j u m  j F c j 
2
i = F (5-24)
j

Equation 5-24 is simply Ohm’s law for ionic current transport and can be simplified to

i = –   (5-25)

where  is the conductivity of the supporting electrolyte. A current balance gives the
current and potential density in the cell

i = 0

which, in combination with Equation 5-25, yields:

   –    = 0 (5-26)

Equation 5-26 can be easily solved using the Electrostatics or Secondary Current
Distribution interface and, when coupled to the Transport in Diluted Species interface,
the potential distribution shows up in the migration term.

Crosswind Diffusion
Transport of diluted species applications can often result in models with a very high
cell Péclet number — that is, systems where convection or migration dominates over
diffusion. Streamline diffusion and crosswind diffusion are of paramount importance
to obtain physically reasonable results. The Transport of Diluted Species interface
provides two crosswind diffusion options using different formulations. Observe that
crosswind diffusion makes the equation system nonlinear even if the transport
equation is linear.

DO CARMO AND GALEÃO

This is the formulation described in Numerical Stabilization in the COMSOL
Multiphysics Reference Manual. The method reduces over- and undershoots to a
minimum, even for anisotropic meshes.

380 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

In some cases, the resulting nonlinear equation system can be difficult to converge.
This can happen when the cell Péclet number is very high and the model contains
many thin layers, such as contact discontinuities. You then have three options:

• Refine the mesh, especially in regions with thin layers.

• Use a nonlinear solver with a constant damping factor less than one.
• Switch to the Codina crosswind formulation.

CODINA
The Codina formulation is described in Ref. 1. It adds diffusion strictly in the direction
orthogonal to the streamline direction. Compared to the do Carmo and Galeão
formulation, the Codina formulation adds less diffusion but is not as efficient at
reducing over- and undershoots. It also does not work as well for anisotropic meshes.
The advantage is that the resulting nonlinear system is easier to converge and that
underresolved gradients are less smeared out.

Danckwerts Inflow Boundary Condition

Constraining the composition to fixed values at an inlet to a reactor may sometimes
result in issues with unreasonably high reaction rates or singularities at the inlet
boundary. These problems may many times be mitigated by using a flux boundary
condition instead, based on the upstream concentrations and the fluid velocity at the
boundary. In chemical engineering, this type of flux boundary condition is also known
as a Danckwerts condition.

Use the Danckwerts condition to specify inlet conditions for domains where high
reaction rates are anticipated in the vicinity to the inlet (Ref. 2).

Given an upstream concentration ci,0, the Danckwerts inflow boundary condition

prescribed the total flux as

n   J i + uc i  = n   uc i 0  (5-27)

THEORY FOR THE TRANSPORT OF DILUTED SPECIES INTERFACE | 381

 Mass Balance Equation for Transport of Diluted Species in Porous
Media

VARIABLY SATURATED POROUS MEDIA

The following equations for the molar concentrations, ci, describe the transport of
solutes in a variably saturated porous medium for the most general case, when the pore
space is primarily filled with liquid but also contain pockets or immobile gas:

 
 c  +   c P i  +  ( g c G i) + u  c i =
t l i t t (5-28)
    D D  i + D e  i  c i  + R i + S i

On the left-hand side of Equation 5-28, the first three terms correspond to the
accumulation of species within the liquid, solid, and gas phases, while the last term
describes the convection due to the velocity field u (SI unit: m/s).

In Equation 5-28 ci denotes the concentration of species i in the liquid (SI unit: mol/
m3), cP, i the amount adsorbed to solid particles (moles per unit dry weight of the
solid), and cG, i the concentration of species i in the gas phase.

The equation balances the mass transport throughout the porous medium using the
porosity p, the liquid volume fraction l; the dry bulk density,  = (1  p)s, and the
solid phase density s.

For saturated porous media, the liquid volume fraction l is equal to the porosity p,
but for unsaturated porous media, they are related by the saturation s as l = psThe
resulting gas volume fraction in the case of an unsaturated porous medium is

 g =  p –  l =  1 – s  p

On the right-hand side of Equation 5-28, the first term introduces the spreading of
species due to mechanical mixing resulting from the porous media (dispersion), as well
as from diffusion and volatilization to the gas phase. The dispersion tensor is denoted
DD (SI unit: m2/s) and the effective diffusion by De (SI unit: m2/s).

The last two terms on the right-hand side of Equation 5-28 describe production or
consumption of the species; Ri is a reaction rate expression which can account for
reactions in the liquid, solid, or gas phase, and Si is an arbitrary source term, for
example due to a fluid flow source or sink.

382 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Adsorption
The time evolution of the adsorption, the solute transport to or from the solid phase,
is defined by assuming that the amount of solute adsorbed to the solid, cP,i, is a
function of the concentration in the fluid ci. This implies that the solute concentration
in the liquid and solid phase are in instant equilibrium. The adsorption term can be
expanded to give

c 
  c  =  c P i c i – c   p = K P,i i – c P i  s p (5-29)
t P i  ci  t P i s  t t t

where KP,i  cP,ici is the adsorption isotherm.

Volatilization
Volatilization is the process where a solute species in the liquid is transported to the
gas phase due to vaporization. Assuming that the amount of solute in the gas phase,
cG,i, is a linear function of the liquid phase concentration, the volatilization term is
defined as

c G i c i  g c i  g
 =  + k c =  g k G i + k G i c i (5-30)
c
 t g G i
g  ci  t G  i i t t t

where kG,i  cG,ici is the linear volatilization.

SATURATED POROUS MEDIA

In the case of transport in a saturated porous medium,l = p, and the governing
equations are

 
 c  +   c P i  + u  c i =     D D i + D e i  c i  + R i + S i (5-31)
t p i t

Convection in Porous Media

Convection (also called advection) describes the movement of a species, such as a
pollutant, with the bulk fluid velocity. The velocity field u corresponds to a superficial
volume average over a unit volume of the porous medium, including both pores and
matrix. This velocity is sometimes called Darcy velocity, and defined as volume flow

THEORY FOR THE TRANSPORT OF DILUTED SPECIES INTERFACE | 383

 rates per unit cross section of the medium. This definition makes the velocity field
continuous across the boundaries between porous regions and regions with free flow.

The velocity field to be used in the Model Inputs section on the physics
interface can, for example, be prescribed using the velocity field from a
Darcy’s Law or a Brinkman Equations interface.

The average linear fluid velocities ua, provides an estimate of the fluid velocity within
the pores:

u-
u a = ---- Saturated
p
u-
u a = --- Unsaturated
l

where p is the porosity and l = sp the liquid volume fraction, and s the saturation,
a dimensionless number between 0 and 1.

Figure 5-4: A block of a porous medium consisting of solids and the pore space between the
solid grains. The average linear velocity describes how fast the fluid moves within the pores.
The Darcy velocity attributes this flow over the entire fluid-solid face.

MATERIAL BALANCE FORMULATION

The Transport of Diluted Species interface in porous media includes two formulations
of the material balance. The conservative formulation of the species equations in
Equation 5-28 is written as:

384 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

  
 c  +   c P i  +  ( g c G i) +   uc i =
t l i t t (5-32)
    D D i + D e i  c i  + R i + S i

If the conservative formulation is expanded using the chain rule, then one of the terms
from the convection part, ci·u, would equal zero for an incompressible fluid and
would result in the nonconservative formulation described in Equation 5-28.

When using the nonconservative formulation, which is the default, the fluid is assumed
incompressible and divergence free:  u = . The nonconservative formulation
improves the stability of systems coupled to a momentum equation (fluid flow
equation).

To switch between the two formulations, click the Show button ( ) and
select Advanced Physics Options. In the section Advanced Settings select
either Nonconservative (the default) or Conservative. The conservative
formulation should be used for compressible flow.

Diffusion in Porous Media

The effective diffusion in porous media, De, depends on the structure of the porous
material and the phases involved. Dependent on whether the transport of diluted
species occurs in free flow or in saturated or unsaturated porous media, the effective
diffusivity is defined as:

De = DF Free Flow
p
D e = ----- D L Saturated Porous Media
L
l
D e = ----- D L Unsaturated Porous Media
L
l g
D e = ----- D L + ------ k G D G Unsaturated with Volatilization
L G

Here DF, DL, and DG are the single-phase diffusion coefficients for the species diluted
in a fluid, a pure liquid, and a gas phase, respectively (SI unit: m2/s), and F, L, and
G are the corresponding tortuosity factors (dimensionless).

The tortuosity factor accounts for the reduced diffusivity due to the fact that the solid
grains impede Brownian motion. The interface provides predefined expressions to

THEORY FOR THE TRANSPORT OF DILUTED SPECIES INTERFACE | 385

 compute the tortuosity factors in partially saturated porous media according to the
Millington and Quirk model (Ref. 3):

–7  3 2 –7  3 2
L = l p , G = g p

and Bruggeman model

–5  2 2 –5  2 2
L = l p , G = g p

For saturated porous media l = p. The fluid tortuosity for the Millington and Quirk
model is

–1  3
L = p

and for the Bruggeman model the tortuosity is defined as

–1  2
L = p

User defined expressions for the tortuosity factor can also be applied.

Dispersion
The contribution of dispersion to the mixing of species typically overshadows the
contribution from molecular diffusion, except when the fluid velocity is very low.

The spreading of mass, as a fluid travel through a porous medium is caused by several
contributing effects. Local variations in fluid velocity lead to mechanical mixing
referred to as dispersion occurs because the fluid in the pore space flows around solid
particles, so the velocity field varies within pore channels. The spreading in the
direction parallel to the flow, or longitudinal dispersivity, typically exceeds the
transverse dispersivity from up to an order of magnitude. Being driven by the
concentration gradient alone, molecular diffusion rate is low relative to the mechanical
dispersion, except at very low fluid velocities.

386 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Figure 5-5: Spreading of fluid around solid particles in a porous medium.

The dispersion is controlled through the dispersion tensor DD. The tensor
components can either be given by user-defined values or expressions or derived from
the directional dispersivities.

Using the longitudinal and transverse dispersivities in 2D, the dispersivity tensor
components are (Ref. 5):

2 2
ui uj
D Dii =  L ------ +  T ------
u u

ui uj
D Dij = D Dji =   L –  T  -----------
u

In these equations, DDii (SI unit: m2/s) are the principal components of the
dispersivity tensor, and DDji and DDji are the cross terms. The parameters L and T
(SI unit: m) specify the longitudinal and transverse dispersivities; and ui (SI unit: m/
s) stands for the velocity field components.

In order to facilitate modeling of stratified porous media in 3D, the tensor formulation
by Burnett and Frind (Ref. 4) can be used. Consider a transverse isotropic media,
where the strata are piled up in the z direction, the dispersivity tensor components are:

THEORY FOR THE TRANSPORT OF DILUTED SPECIES INTERFACE | 387

 2 2 2
u v w
D Lxx =  1 ------ +  2 ------ +  3 -------
u u u
2 2 2
v u w
D Lyy =  1 ------ +  2 ------ +  3 -------
u u u
2 2 2
w u v
D Lzz =  1 ------- +  3 ------ +  3 ------
u u u (5-33)
uv
D Lxy = D Lyx =   1 –  2  -------
u
uw
D Lxz = D Lzx =   1 –  3  --------
u
vw
D Lyz = D Lzy =   1 –  3  --------
u

In Equation 5-33 the fluid velocities u, v, and w correspond to the components of the
velocity field u in the x, y, and z directions, respectively, and 1 (SI unit: m) is the
longitudinal dispersivity. If z is the vertical axis, 2 and 3 are the dispersivities in the
transverse horizontal and transverse vertical directions, respectively (SI unit: m).
Setting 2 = 3 gives the expressions for isotropic media shown in Bear (Ref. 5 and
Ref. 6).

Adsorption
As species travel through a porous medium they typically attach to (adsorb), and
detach (desorb) from the solid phase, which slows chemical transport through the
porous medium. Adsorption and desorption respectively reduces or increases species
concentrations in the fluid. The adsorption properties vary between chemicals, so a
plume containing multiple species can separate into components (Ref. 7). The
Adsorption feature includes four predefined and one user-defined relationships to
predict the solid concentrations, cPi from the concentration in the liquid phase, ci:

388 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

 KL c
c P = c Pmax -------------------- Langmuir
1 + KL c
c N
c P = K F  -------- Freundlich
 c ref
bT c (5-34)
c P = c Pmax ---------------------------------------------- Toth
N 1  NT
 1 +  bT c  T 
KB c0 c
c P = -------------------------------------------------------------
- BET
 c S – c   1 +  K B – 1  -----
c
cS

The above equations contains the following parameters:

• Freundlich (Ref. 8)
- Freundlich constant KF (SI unit:·mol/kg), Freundlich exponent NF
(dimensionless), and reference concentration cref (SI unit: mol/m3).
• Langmuir (Ref. 9)
- Langmuir constant KL (SI unit: m3/mol), and adsorption maximum cPmax
(SI unit: mol/kg).
• Toth (Ref. 10)
- Toth constant bT (SI unit: m3/mol), Toth exponent NT (dimensionless), and
adsorption maximum cPmax (SI unit: mol/kg).
• BET (Brunauer–Emmett–Teller, Ref. 11)
- BET constant KB (dimensionless), the monolayer adsorption capacity c0
(SI unit: mol/kg) and the saturation concentration, cS (SI unit: mol/m3).

These predefined expressions are adsorption isotherms that describe the amount of
species sorbed to the solid. Applying the isotherms, as defined in the Adsorption
section above, assumes that adsorption is at equilibrium locally. This implies that the
species phase transfer, between the fluid and the solid phase (within the porous
medium), is much faster than the transport within the fluid phase.

Using a Species Source feature, arbitrary expressions can be entered to define, for
example, nonequilibrium and temperature-dependent adsorption laws, in the manner
of Fetter (Ref. 12), or Bear and Verruijt (Ref. 13).

The retardation factor, RF, describes how adsorption slows the solute velocity, uc,
relative to the average linear velocity of the fluid, ua, as in

THEORY FOR THE TRANSPORT OF DILUTED SPECIES INTERFACE | 389

  b c P u
RF = 1 + ----- -------- = ----a-
 c uc

If the contaminant moves at the average linear velocity of the fluid for RF = 1. For
RF > 1, the contaminant velocity is lower than the fluid velocity owing to residence
time on solids.

Reactions
Chemical reactions of all types influence species transport in porous media. Examples
include biodegradation, radioactive decay, transformation to tracked products,
temperature- and pressure-dependent functions, exothermic reactions, and
endothermic reactions. The reactions represent change in species concentration per
unit volume porous medium per time. Reaction terms are used on the right-hand side
of the governing equation to represent these processes. For reactions in a fluid phase,
multiply the expression by the fluid volume fraction . Similarly, solid phase reaction
expressions include the bulk density, b, and gas phase reactions include the gas
volume fraction, av.

The following expressions define some common types of reactions:

ln 2
R Li = –  --------- c i Radioactive decay — liquid
 Li
ln 2 c Pi
R Pi = –  b ---------  ---------- c i Radioactive decay — solid
 Pi  c i 
c Gi
R Gi = – --------- a v  ----------- c i Radioactive decay — gas
ln 2
 Gi c
R Lk =  Li c i Creation from parent c Li — liquid
c Pi
R Pk =  b  Pi  ---------- c i Creation from sorbed parent c Pi — solid
 c i 
c Gi
R Gk = – a v  ----------- c i Reaction — gas
c

where  is the chemical half life,  is a reaction rate, and the subscripts L, P, and G
denote liquid, solid, and gas phases, respectively. In the equations, the reactions either
depend on liquid concentration ci or solid phase concentrations cPi obtained using the
sorption derivative with ci or gas phase concentration cGi depending on the gas volume
fraction, the volatilization, and the liquid concentration.

390 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Reaction rates can vary with results from other equations in your model, such as
temperature. For example, enter the Arrhenius rate law given in Ref. 14:

Ea  T – TR 
 T =  R exp ------------------------------ (5-35)
R u TT R

In Equation 5-35, T denotes the current absolute temperature, TR denotes the

reference absolute temperature, Ea is the activation energy, and Ru is the universal gas
constant.

Mass Transport in Fractures

When thin fractures occur in porous media, fluid flow tends to move faster along the
fracture than in the surrounding media. The transport of chemical species therefore
also occur also faster in the direction of the fractures.

The fluid flow in a fracture can be modeled using Darcy’s law formulated in a thin
sheet of porous medium (a fracture):


u = ---  t p


Here u is the tangential Darcy velocity,  is the fracture permeability, the fluid’s
dynamic viscosity, and tp is the tangential gradient of the fluid pressure.

The equation to solve for mass transport of species ci in a thin fracture, embedded in
a porous media, is derived from Equation 5-28. The resulting equation is:

 b c P i  p c i
d fr  ------------------ + ------------- +  t   D e i  t c i  + u   t c i = d fr R i + d fr S i + n 0 (5-36)
t t

Here dfr is the fracture thickness, cP,i the amount of species adsorbed to (or desorbed
from) the porous matrix (moles per unit dry weight of the solid), p is the fracture
porosity, and De is the effective diffusivity. The first two terms on the right hand side
represent source terms from reactions, and n0 corresponds to out-of plane flux from
the adjacent porous domain.

In order to arrive at the tangential differential equation, the gradient is split into the
contributions normal and tangential to the fracture:

c i =  n c i +  t c i

THEORY FOR THE TRANSPORT OF DILUTED SPECIES INTERFACE | 391

 The normal gradient is defined in the direction normal to the boundary representing
the fracture and the tangential gradient is defined along the boundary. Assuming that
the variations in the normal (thin) direction of the fracture are negligible compared to
those in the tangential direction, the gradient is simplified as:

c i =  t c i

Using , the transport along fracture boundaries alone is solved for. In this case the
transport in the surrounding porous media neglected and the out-of plane flux n0
vanishes.

See Fracture for more information about the boundary feature solving
Equation 5-36. See for more information about the physics interface
solving the equation on boundaries only.

References
1. R. Codina, “A discontinuity-capturing crosswind-dissipation for the finite element
solution of the convection-diffusion equation”, Computer Methods in Applied
Mechanics and Engineering, vol. 110, pp. 325–342, 1993.

2. P.V. Danckwerts, “Continuous flow systems: Distribution of residence times”,

Chem. Eng. Sci., vol. 2, no. 1, 1953.

3. R.J. Millington and J.M. Quirk, “Permeability of Porous Solids”, Trans. Faraday
Soc., vol. 57, pp. 1200–1207, 1961.

4. R.D. Burnett and E.O. Frind, “An Alternating Direction Galerkin Technique for
Simulation of Groundwater Contaminant Transport in Three Dimensions: 2.
Dimensionality Effects”, Water Resour. Res., vol. 23, no. 4, pp. 695–705, 1987.

5. J. Bear, Hydraulics of Groundwater, McGraw Hill, 1979.

6. J. Bear, Dynamics of Fluids in Porous Media, Elsevier Scientific Publishing, 1972.

7. D.M. Mackay, D.L. Freyberg, P.V. Roberts, and J.A. Cherry, “A Natural Gradient
Experiment on Solute Transport in a Sand Aquifer: 1. Approach and Overview of
Plume Movement”, Water Resour. Res., vol. 22, no. 13, pp. 2017–2030, 1986.

8. H.M.F. Freundlich, “Über die Adsorption in Lösungen” (in German), Z. Phys.

Chem., vol 57, pp. 385–470, 1906

392 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

9. I. Langmuir, “The Constitution and Fundamental Properties of Solids and Liquids,
Part I Solids”, J. Amer. Chem. Soc., vol. 38, pp. 2221–2295, 1916.

10. J. Toth, “State equation of the solid gas interface layer”, Acta Chem. Acad. Hung.
vol. 69, pp. 311–317, 1971.

11. S. Brunauer, P.H. Emmett, E. Teller, “Adsorption of gases in multimolecular

layers”, J. Am. Chem. Soc., vol. 60 (2), pp. 309–319, 1938.

12. C.W. Fetter, Contaminant Hydrogeology, Prentice Hall, 1999.

13. J. Bear and A. Verruijt, Modeling Groundwater Flow and Pollution, D. Reidel
Publishing, 1994.

14. I. Langmuir, “Chemical Reactions at Low Temperatures”, J. Amer. Chem. Soc.,

vol. 37, 1915.

15. J.M. Coulson and J.F. Richardson, Chemical Engineering, vol. 2, 4th ed.,
Pergamon Press, Oxford, U.K., 1991.

16. J.M. Coulson and J.F. Richardson, Chemical Engineering, vol. 1, 4th ed.,
Pergamon Press, Oxford, U.K., 1991.

17. D.E Rosner, Transport Processes in Chemically Reacting Flow Systems, ISBN-13:
978-1483130262, Butterworth–Heinemann, 1986.

18. R.B. Bird, W.E. Stewart, and E.N. Lightfoot, Transport Phenomena, 2nd ed.,
John Wiley & Sons, Inc., 2007.

THEORY FOR THE TRANSPORT OF DILUTED SPECIES INTERFACE | 393

 Theory for the Transport of
Concentrated Species Interface
The Transport of Concentrated Species Interface theory is described in this section:

• Multicomponent Mass Transport

• Multicomponent Gas Diffusion: Maxwell–Stefan Description
• Multicomponent Diffusivities
• Multicomponent Diffusion: Mixture-Averaged Approximation
• Multispecies Diffusion: Fick’s Law Approximation
• Multicomponent Thermal Diffusion
• Regularization of Reaction Rate Expression
• References for the Transport of Concentrated Species Interface

Multicomponent Mass Transport

Suppose a reacting flow consists of a mixture with i1, …, Q species and j = 1, …, N
reactions. Equation 5-37 then describes the mass transport for an individual species:

 
 i  +     i u  = –   j i + R i (5-37)
t

where,  (SI unit: kg/m3) denotes the mixture density and u (SI unit: m/s) the mass
averaged velocity of the mixture. The remaining variables are specific for each of the
species, i, being described by the mass transfer equation:

• i is the mass fraction (1)

• ji (SI unit: kg/(m2·s)) is the mass flux relative to the mass averaged velocity, and
• Ri (SI unit: kg/ (m3·s)) is the rate expression describing its production or
consumption.

The relative mass flux vector ji can include contributions due to molecular diffusion,
mass flux due to migration in an electric field, and thermal diffusion.

Summation of the transport equations over all present species gives Equation 5-38 for
the conservation of mass

394 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

  +    u  = 0 (5-38)
t

assuming that

Q Q Q

 i = 1,  ji = 0,  Ri = 0
i=1 i=1 i=1

Using the mass conservation equation, the species transport for an individual species,
i, is given by:

    i  +   u    i = –   j i + R i (5-39)
t

Q  1 of the species equations are independent and possible to solve for using
Equation 5-39. To compute the mass fraction of the remaining species, COMSOL
Multiphysics uses the fact that the sum of the mass fractions is equal to 1:

1 = 1 –  i (5-40)
i=2

Multicomponent Gas Diffusion: Maxwell–Stefan Description

The Maxwell–Stefan formulation for a multicomponent mixture defines N  1 (with N
being the number of active species in the model) force balances according to

T T
xi xk  ji jk  xi xk  Dk Di 
–  ---------- --------- – ---------
D ik   i   k 
- = dk –  ----------  ---------- – ---------  ln T
D ik   k   i 
(5-41)
ki ki

where

• Dik (SI unit: m2/s) are the binary Maxwell–Stefan diffusivities

• T (SI unit: K) is the temperature
T
• D i (SI unit: kg/ m·s)) are the thermal diffusion coefficients, and
• dk (SI unit: 1/m) is the diffusional driving force acting on species k.

By inverting the above set of equations, the mass flux relative to the mass average
velocity, ji, can be defined by the generalized Fick equations (Ref. 1):

THEORY FOR THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 395

 Q
˜
 Dik dk – Di  ln T
T
j i = –  i (5-42)
k=1
˜
where D ik (SI unit: m2/s) are the multicomponent Fick diffusivities.

For ideal gas mixtures the diffusional driving force is (Ref. 1)

Q
1
d k = -------------- p k –  k p –  k g k +  k
cR g T  l gl (5-43)
l=1

where

• c (SI unit: mol/m3) is the total molar concentration

• Rg is the universal gas constant 8.314 J/(mol·K)
• p (SI unit: Pa) is the total pressure
• pk (SI unit: Pa) is the partial pressure, and
• k (SI unit: kg/m3) is the density of species k, and
• gk (SI unit: m/s2) is an external force (per unit mass) acting on species k. In the
case of an ionic species the external force arises due to the electric field.
As can be seen in Equation 5-42 and Equation 5-43, the total diffusive flux for the
species depends on the gradients of all species concentrations, temperature, and
pressure as well as any external force on the individual species.

Using the ideal gas law, p  c·Rg·T, and the definition of the partial pressures, pk  xkp,
the equation can be written as

Q
1
d k = x k + ---  x k –  k  p –  k g k +  k
p  l gl (5-44)
l=1

The mole fraction xk is given by

k
x k = -------- M (5-45)
Mk

and the mean molar mass M (SI unit: kg/mol) by

396 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

 Q
1- = i
----
M  ------
Mi
-
i=1

When using the Maxwell–Stefan diffusion model, the transport equations for the
species’ mass are

Q
 
T T
    i  +   u    i =     i D̃ ik d k + D i -------- + R i

t  T
 k=1 
(5-46)
Q
1
d k = x k + ---  x k –  k  p –  k g k +  k
p  l gl
l=1

Multicomponent Diffusivities
˜
The multicomponent Fick diffusivities, D ik , are needed to solve Equation 5-46. The
diffusivities are symmetric

˜
D ˜
ik = D ki

and are related to the multicomponent Maxwell–Stefan diffusivities, Dik, through the
following relation (Ref. 2)

xi xk
  adjB i  jk
ji
---------- = –  i  k ---------------------------------------- ,  B i  = D̃ kj – D̃ ij , ij (5-47)
D ik kj
 D̃ ij  adjB i  jk
ji

where (adjBi)jk is the jkth component of the adjoint of the matrix Bi.

For low-density gas mixtures, the multicomponent Maxwell–Stefan diffusivities, Dij,

can be replaced with the binary diffusivities for the species pairs that are present.

Solving for Equation 5-47 leads to a number of algebraic expressions for each of the
components in the multicomponent Fick diffusivity matrix. For two- and
three-component systems, these are implemented and solved directly by COMSOL
˜
Multiphysics. For instance, the component D 12 in a ternary system is given by:

THEORY FOR THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 397

 1  2 + 3  2  1 + 3   32
– ------------------------------- - – -------------------------------- + --------------- -
˜ x 1 D 23 x 2 D 13 x 3 D 12
D 12 = -------------------------------------------------------------------------------------------------
-
x1 x2 x3
-------------------
- + ------------------- - + ------------------- -
D 12 D 13 D 12 D 23 D 23 D 13

For four components or more, COMSOL Multiphysics obtains the multicomponent

Fick diffusivities numerically through matrix inversion derived from the matrix
properties defined in Ref. 2. The program starts with the multicomponent Maxwell–
Stefan diffusivity matrix Dik to compute the multicomponent Fick diffusivity matrix,
D˜
ik , using the following equation:

D̃ ij = N ij – q (5-48)
˜
where ij, the indices in the matrices D and N, range from 1 to the number of species,
Q.

The elements of the matrix N in Equation 5-48 are defined as

–1
N ij =  P  ij (5-49)

where P1 is the inverse of a matrix P defined as

i j
P ij = -----------
- – C̃ ij
q
˜
The matrix C in turn is defined as

 xi xj
 --------- ij
 D ij
C̃ ij = 

 –  C̃ik i=j
 kj

The term q in Equation 5-48 is a scalar value that provides numerical stability and
should be of the same order of magnitude as the multicomponent Maxwell–Stefan
diffusion coefficients. The physics interface therefore defines q as the sum of the
multicomponent Maxwell–Stefan diffusion coefficients:

n–1
 n 
q =    Dij 

i = 1  j = i+1 

398 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Because the multicomponent Fick diffusivity matrix is symmetric, it is sufficient to
˜
apply Equation 5-49 to the upper triangle elements of D . The remaining elements are
obtained by swapping the indices in the matrix.

Multicomponent Diffusion: Mixture-Averaged Approximation

The mixture-averaged diffusion model assumes that the relative mass flux due to
molecular diffusion is governed by a Fick’s law type approximation.

Assuming that the diffusive flux, relative to the mass averaged velocity, is proportional
to the mole fraction gradient, the mass flux is defined as

m x i
j md i = –  i D i --------- (5-50)
xi

Here i is the density, and xi the mole fraction of species i. Using the definition of the
species density and mole fraction

i
 i =  i , x i = ------- M
Mi

Equation 5-50 can be expressed in terms of the mass fractions (i) in the manner of

m M
j md i = –  D i  i +  i D i ---------
m
M

Using Equation 5-50 together with the Maxwell–Stefan equations, where isobaric and
isothermal conditions have been assumed, the following expression for the
mixture-averaged diffusion coefficients can be derived (Ref. 3):

m 1 – i
D i = -------------------------
- (5-51)
xk

N
--------
-
kiD
ik

If instead the diffusive flux (relative to the mass averaged velocity) is assumed
proportional to the mass fraction gradient, the mass flux is defined as

m*  i m*
j md i = –  i D i ---------- = – D i  i (5-52)
i

THEORY FOR THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 399

 For this assumption, using Equation 5-52 together with the Maxwell–Stefan
equations, also assuming isobaric and isothermal conditions, the following expression
for the mixture-averaged diffusion coefficients can be derived (Ref. 3):

1 = xk xi k
k  i --------
D ik 1 –  i k  i D ik
N N
----------- - + --------------- --------- (5-53)
m*
Di

THE MIXTURE-AVERAGED DIFFUSION COEFFICIENT

When using the mixture-averaged model, the diffusion is proportional to a single
diffusion coefficient. The coefficient describes the diffusion of species i relative to the
remaining mixture and is referred to as the mixture-averaged diffusion coefficient. The
coefficient is explicitly given in terms of the multicomponent Maxwell–Stefan
diffusivities Dik. As a consequence, no matrix inversion operation is required as for the
Maxwell–Stefan diffusion model (when using four or more species). For low-density
gas mixtures, the Dik components can be replaced by the binary diffusivities for the
species pairs present.

MIXTURE DIFFUSION CORRECTION

It should be noted that models for the multicomponent diffusivities based on Fick’s
law are approximations. One drawback of this is that the net diffusive mass flux, the
sum of all mass fluxes relative to the mass averaged velocity, is not constrained to zero.
To account for this we add a flux, defined by a correction velocity uc, to enforce a zero
net diffusive flux in the manner of:

i = 1 jmd i = i = 1 i  ud,i + uc 

N N
= 0 (5-54)

Hence the correction velocity is:

i = 1 i ud,i
N
uc = – (5-55)

Here ud,i is the diffusion velocity resulting from the flux assumption in Equation 5-50
or Equation 5-52. Note that the correction velocity is a constant correction (same for
all species), but varies in space.

Using the correction velocity together with Equation 5-50 (flux proportional to the
mole fraction gradient), the resulting diffusive flux is

m M Mi
i = 1 ------
m N m
j md i = – D i  i –  i D i --------- +  i - D x i (5-56)
M M i

400 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

If instead Equation 5-52 is used (flux proportional to the mass fraction gradient), the
resulting diffusive flux is

i = 1 Di
m* N m*
j md i = – D i  i +  i  i (5-57)

ADDITIONAL TRANSPORT MECHANISMS

Apart from molecular diffusion, transport due to thermal diffusion and migration of
charged species in an electric field can be accounted for by adding the following terms
to the diffusive flux:

T T
j i =    D i -------- +  i z i u m i F 
 T 

where
T
• D i (SI unit: kg/(m·s)) is the thermal diffusion coefficient
• zi (dimensionless) is the charge number
• um,i the mobility of the ith species, and
•  (SI unit: V) is the electric potential.

Multispecies Diffusion: Fick’s Law Approximation

Using a Fick’s law approximation, the relative mass flux due to molecular diffusion is
governed by

F x i
j md i = –  i D i kl --------- (5-58)
xi

when assuming that the diffusive flux is proportional to the mole fraction gradient. If
instead assuming that it is proportional to the mass fraction it becomes

F
j md i = – D i kl  i (5-59)
F
In the equations above D i kl represents a general diffusion matrix (SI unit: m2/s)
describing the diffusion of species i into the mixture. This form makes it possible to
use any diffusion coefficient, matrix, or empirical model based on Fick’s law. For
example, in situations when the mass transport is not dominated by diffusion, an
alternative is to use the diffusion coefficients at infinite dilution,

THEORY FOR THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 401

 F 0
D i kk = D i

These coefficients are typically more readily available compared to the binary diffusion
diffusivities, especially for liquid mixtures.

The mixture diffusion correction described above for the mixture-averaged diffusion
can also be applied in this case. Correspondingly, the resulting diffusive flux is

F M Mi
i = 1 ------
F N F
j md i = – D i kl  i –  i D i kl --------- +  i - D x (5-60)
M M i kl i

when using Equation 5-58 and

i = 1 Di kl i
F N F
j md i = – D i kl  i +  i (5-61)

when instead using Equation 5-59.

When using the Fick’s Law approximation, Additional Transport Mechanisms can be
accounted for in the same manner as described above for the mixture-averaged
approximation.

Multicomponent Thermal Diffusion

Mass diffusion in multicomponent mixtures due to temperature gradients is referred
to as the Soret effect. This occurs in mixtures with high temperature gradients and
large variations in molecular weight (or size) of the species. Typically species with high
molecular weight accumulate in lower temperature regions while the diffusion due to
the Soret effect transports species with low molecular weight to higher temperature
regions. In COMSOL Multiphysics, thermal diffusion is included by prescribing the
T
thermal diffusion coefficients D i . In a multicomponent mixture, the sum of the
thermal diffusion coefficients is zero:

 Di
T
= 0
i=1

Pore Wall Interactions

In porous media, the diffusion models may be extended to account for species
collisions with the pore walls.

402 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

For the Fick’s law or the Mixture-averaged diffusion models, the diffusion coefficient
M W
D i is corrected with the Wall diffusion coefficient D i in the following way

MW  1 1 -
-1
Di =  --------
- + -------- 
D M
D 
W
i i

For the Maxwell-Stefan diffusion model (see Ref. 5), the following term is added to the
species force balance (Equation 5-41)

xi  ji
- --------- + u W = d i + 
 + ------------
W  
D i eff  i 

The wall velocity uW, representing the velocity of the fluid in a representative region
adjacent to the pore wall, is here defined as

p A x k j k'
----------- +
pA  --------------
D
W  
----------
k
k k  eff
uW = – ---------------------------------------------------- (5-62)
xk
 --------------
W
k D k eff

where the modified fluxes j k' have been calculated using the modified driving forces

1
d k' = x k + --- x k p (5-63)
p

It should be noted that uW can become nonzero also in the absence of pressure
gradients. For a proper description of the fluid momentum balance, uW needs to be
added either as a contributing velocity (Darcy’s law) or as a corresponding volume
force F  1uW (Brinkman equations).

The Maxwell–Stefan model formulated in this way correctly captures the Klinkenberg
effect, and has large similarities with other models in literature incorporating Knudsen
diffusion such as the Dusty Gas Model and the (Modified) Binary Friction Model. In
addition, for the limit of vanishing permeabilities (  0), the formulation reduces to
the Lightfoot model for membrane transport.

Regularization of Reaction Rate Expression

The flexibility of COMSOL means that a reaction can result in nearly any reaction rate
contribution Ric to a mass fraction i. But a reactant, r, can physically only be

THEORY FOR THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 403

 consumed if r  0 and a product, k, can only be formed if k  1. The Reaction
feature in Transport of Concentrated Species Interface contains a way to regularize
make the rate expression contributions comply to these restrictions.

When selected the “core”, or unregularized reaction rate contribution, Ric, to a mass
fraction i is replaced by

c c c c
1 Ri – Ri 1 Ri + Ri
R i = --- -------------------------------- max   i 0  + --- -----------------------------------------
- max  1 –  i 0  (5-64)
2 max      dl 2 max  1 –    dl 
i i i i

The first term on the right hand side of Equation 5-64 is active if Ric  0, that is if i
is a reactant. The reaction rate contribution, Ri, is equal to the “core” reaction rate,
Ric, as long as i  idl. As i approaches zero, the regularization damps out negative
Ric and for i  0, Ri for reactant i is equal to zero.

The second term on the right hand side of Equation 5-64 is active if Ric  0, that is if
i is a product. The reaction rate contribution, Ri, is equal to the “core” reaction rate,
Ric, as long as i  1  idl. As i approaches one, the regularization damps out
positive Ric and for i  1, Ri for product i is equal to zero.

The damping limits, idl, should be in an order of magnitude that can be considered
numerical noise for species i. The damping limits are per default set to 1e6, which is
appropriate for most applications. It can be advantageous to lower some limits when
working with for example catalytic trace species and the limits can sometimes be raised
to gain additional robustness.

References for the Transport of Concentrated Species Interface

1. C.F. Curtiss and R.B. Bird, “Multicomponent Diffusion”, Ind. Chem. Res., vol. 38
pp. 2515–2522, 1999.

2. R.B. Bird, W.E. Stewart, and E.N. Lightfoot, Transport Phenomena, 2nd ed., John
Wiley & Sons, 2005.

3. R.J. Kee, M.E. Coltrin, and P. Glarborg, Chemically Reacting Flow, John Wiley &
Sons, 2003.

4.

5. H Ekström, “Combining wall interactions, fluid momentum balances and the

Maxwell-Stefan equations for gas transport in porous media: An alternative approach”,
International Journal of Thermofluids, vol21, 100534, 2024

404 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

THEORY FOR THE TRANSPORT OF CONCENTRATED SPECIES INTERFACE | 405
 Theory for the Electrophoretic
Transport Interface
models transport of charged and uncharged species, and in addition sets up a charge
balance equation for the electrolyte potential.

The species concentrations are denoted, ci (SI unit: mol/m3), and the potential, l (SI
unit: V).

The species are transported by diffusion, migration, and (optionally) convection

according the Nernst–Planck set of equations. The total flux of species i is denoted Ni
(SI unit: mol/(m2·s)) according to

N i = – D i c i – z i u m i Fc i  l + uc i = J i + uc i (5-65)

where Di (SI unit: m2/s) is the diffusion coefficient, zi (1) the corresponding charge,
um,i (SI unit: s·mol/kg) the mobility and u (SI unit: m/s) the velocity vector. Ji
denotes the molar flux relative to the convective transport (SI unit: mol/(m2·s)). For
a detailed description of the theory of these equations and the different boundary
conditions, see .

CHARGE BALANCE EQUATION

The current vector, il (A/m2), is defined as

il =  zi Ni (5-66)
i

where the summation is made over all species in the electrolyte.

The governing equation for the electrolyte potential is

  il = Ql (5-67)

where Ql (SI unit: A/m3) is the electrolyte current source stemming from, for
example, porous electrode reactions. For nonporous electrode domains this source
term is usually zero.

406 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

PROTON AND HYDROXIDE CONCENTRATIONS AND ELECTRONEUTRALITY
Assuming the total number of species to be N  2, the assumption of electroneutrality
is

N+2

 zi ci = 0 (5-68)
i=1

In water-based systems the species H+ and OH- are always present. The auto
ionization reaction for water is

+ -
H 2 O  H + OH (5-69)

This reaction is fast, following the equilibrium relation

2 –6
c H + c OH - = K w  1 mol dm (5-70)

where Kw  1014.

Now, the electroneutrality condition, including the two additional species H+ and
OH-, reads

c H + – c OH - +  zi ci = 0 (5-71)
i=1

Combining these two equations results in the following algebraic expressions for the
concentrations of H+ and OH-.

2
cH+ = – 
--- + 
2 –6
------ + K w  1 mol dm (5-72)
2 4

and

2
c OH - = 
--- + 
2 –6
------ + K w  1 mol dm (5-73)
2 4

where

 =  zi ci (5-74)
i=1

THEORY FOR THE ELECTROPHORETIC TRANSPORT INTERFACE | 407

 MASS BALANCE EQUATIONS FOR DISSOCIATION SPECIES
Assume a set of species Si describing k dissociation steps from

+ z +k–1
z +k z +k–1 +  H   S k0– 1 
S k0  S k0– 1 +H K a 1 = ---------------------------------------- (5-75)
z0 + k
 Sk 

to

+ z
z +1 z +  H   S 00 
S k0  S 00 + H K a k = -------------------------
- (5-76)
z +1
 S 10 

where z0 is the charge (valence) of species S0 (which has no dissociable protons) and
Ka,j is the acid (equilibrium) constant of the jth dissociation reaction. The brackets
“[ ]” here represent the species activity. The charge of each species is always deductible
from the index i according to z0+i and will be dropped from now on.

If the proton activity is known, any species Sm may be expressed using any other
species Sl according to

m–l
H  Sl 
 S m  = ------------------------------------- (5-77)
k–l

 K a j
j = k–m+1

if m  l and

k–m


m–l
 Sm  =  H   Sl  K a j
j = k–l+1
(5-78)

if l  m.

Setting m  i and denoting the flux of species i by Ni using equation Equation 5-65,
the mass balance equation for the concentration ci of each subspecies i in the
dissociation chain is

c
-------i +   N i = R eq i k – i – R eq i k – i + 1 + R i (5-79)
t

where Req,i,j is the reaction source stemming from the jth dissociation step (with
Req,i,k+1  0), and Ri any additional reaction sources.

408 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

The reaction source contributions from the dissociation steps are generally not known,
but may be canceled by taking the sum of all mass balance equations, resulting in

k k
c i
 ------
 - +   N i =  Ri (5-80)
 t 
i=0 i=0

AVERAGE IONIZATION FORMULATION FOR LARGE MOLECULES

Large protein molecules are typically subject to a large number of dissociation steps.

As an alternative way to a long chain of dissociation steps for describing protein

transport, you can instead formulate the protein transport based on the average
number of protons  (1) removed from the molecule.  is typically provided as a
function of pH, based on experimental data. The average charge of the species then is
z = z 0 –  , so that the flux of the species is written as:

N i = – Dc i – zu m Fc i  l + uc i (5-81)

and similarly the addition to the charge neutrality condition is zc i .

When considering the contribution to the current and the charge balance equation you
2 2
need to take into account that the squared average charge, z =  z 0 –   , is not equal
2 2 2 2 2 2
to the “average squared charge”, z = z 0 + 2z 0  +  = z –  +  (Ref. 1).

The addition to the current density vector is

2
i l =  – F  zD i c i + z u m c i  l  (5-82)

The average number of protons removed from the proton typically depends on the
pH. If the average number of removed protons depend only on the pH, the averaged
squared number of protons removed can be written as

2 d  2
 = – c H+ + (5-83)
d cH+

And from this you can derive the average squared charge according to

2 d  z2 d z z2
z = – cH+ + = – cH+ + (5-84)
d c H+ d cH+

DIFFUSIVITY-MOBILITY RELATIONS
The Stokes radius r of a molecule is related to the diffusivity according to

THEORY FOR THE ELECTROPHORETIC TRANSPORT INTERFACE | 409

 kT
r = --------------- (5-85)
6D

where is the  (SI unit: Pa·s) is the dynamic viscosity and k the Boltzmann constant.

For small molecules, there is the frequently used Nernst–Einstein relation between the
diffusivity and the mobility

D
u m = -------- (5-86)
RT

For larger molecules, such as proteins, the mobility may instead be calculated based on
the Debye–Hückel–Henry expression (Ref. 2 ) according to

ef  r  - = -----------------------------
Df  r  -
u m = ----------------------------------- (5-87)
6F  1 + r  RT  1 + r 

where  (1/m) is the Debye parameter, which depends on the ionic strength of the
solution, is defined for ideal solutions as

2 N
2e N A
 zi ci
2 2
 = ----------------- (5-88)
 0 kT
i=1

where  is the dielectric constant of the fluid and 0 the permittivity of free space.
2
( z should be used if available in the formula above when calculating the ionic
strength).

The function f above is based on a sigmoidal function so that it ranges from 1 for r  0
to 1.5 for r  . Note that the Debye–Hückel–Henry expression approaches the
Nernst–Einstein mobility as r  0.

For larger molecules (macro ions), where the distance between the charges is large
compared to 1/, the Linderstrøm–Lang approximation postulates a smaller
contribution of to the ionic strength so that the z-valent ion behaves as a monovalent
ion with a z-fold concentration. For an assemble of N  M smaller molecules and M
macro ions, the Debye parameter then is defined as

N–M N
2e N A  
2
 = -----------------    abs(z i)c i
2 2
zi ci + (5-89)
 0 kT  
 i=1 i = N–M+1 

410 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

REFERENCES
1. R.A. Mosher, D.A. Saville, and W. Thormann, The Dynamics of Electrophoresis,
VCH Verlagsgesellschaft mbH, Weinheim, Germany, 1992.

2. R.A. Mosher, P. Gebauer, J. Caslavska, and W. Thormann, “Computer Simulation

and Experimental Validation of the Electrophoretic Behavior of Proteins. 2. Model
Improvement and Application to Isotachophoresis,” Anal. Chem., vol. 64, pp. 2991–
2997, 1992.

THEORY FOR THE ELECTROPHORETIC TRANSPORT INTERFACE | 411

 Theory for the Surface Reactions
Interface
The Surface Reactions Interface solves for an arbitrary number of surface
concentrations, cs,i (SI unit: mol/m2), referring to species adsorbed on a reactive
surface, and an arbitrary number of bulk concentrations cb,k (SI unit: mol/m3),
referring to the species in the solid material that constitutes the bulk of the reactive
surface.

In this section:

• Governing Equations for the Surface Concentrations

• Governing Equations for the Bulk Concentrations
• ODE Formulations for Surface Concentrations
• Surface Reaction Equations on Deforming Geometries
• Reference for the Surface Reactions Interface

Governing Equations for the Surface Concentrations

Transport of adsorbed species occurs in the tangential direction along the surface. The
Surface Reactions interface models the tangential flux in the surface dimension, the
surface molar flux, Nt,i (SI unit: mol/(m·s)) (the subscript t refers to the tangential
direction), as governed by diffusion according to Fick’s law:

N t i = – D s  i  t c s  i

where Ds,i (SI unit: m2/s) is the surface diffusion coefficient for species i.

The governing equation for the surface concentrations is written as

c s i
= –  t  N t i + R s  i (5-90)
t

where Rs,i (SI unit: mol/(m2·s)) is the sum of all sources due to surface reactions and
adsorption/desorption phenomena.

Of frequent interest for surface reaction kinetics are the fractional surface coverages, i
(dimensionless), of the species (with index i).

412 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

Writing the density of sites of the surface as s (SI unit: mol/m2) and the site
occupancy number for each species as i (dimensionless), the surface coverages can be
calculated from

 i c s i
 i = -------------
-
s

(The site occupancy number accounts for the situation when a large species covers
more than one site on the surface.)

For the case of monolayer adsorption, the sum of all fractional coverages of free and
adsorbed sites is unity, and hence the fraction of free sites on the surface, *, can be
calculated from

* = 1 –  i
i

Governing Equations for the Bulk Concentrations

BULK SPECIES RATES AND GROWTH VELOCITY

The reaction rate for a bulk species of index k, being added to the bulk due to surface
reactions is defined as Rb,k (SI unit: mol/(s·m2)).

The reaction rate in mass basis, rb,k (SI unit: kg/(s·m2)) for species k, is given by

r k = M k R b k

with Mk (SI unit: kg/mol) being the molar mass of the species.

Based on this, the species contribution to the bulk growth velocity, vk (SI unit: m/s),
is given by

r b k
v b k = ---------
-
k

where k is the species density (SI unit: kg/m3).

TOTAL RATES AND VELOCITY

To get the corresponding total values, a summation for all bulk species is made
according to

THEORY FOR THE SURFACE REACTIONS INTERFACE | 413

 R b tot =  R b k
k

r b tot =  r b k
k

v b tot =  vb k
k

FRACTIONAL RATES AND VELOCITY

The fractional deposition rates are calculated by

R b k
R b frac k = --------------
-
R b tot

r b k
r b frac k = ------------
-
r b tot

v b k
v b frac k = -------------
-
v b tot

The fractional deposition rates can be used as measures of the local

fractional composition in a deposited layer at a certain thickness or time.

ACCUMULATED BULK QUANTITIES

By solving for the bulk species concentrations for each time step, the accumulated mass
and thickness changes of the bulk material can be calculated.

The bulk concentration, cb,k (SI unit: mol/m2), for species k is governed by the
equation

c b k
= R b k (5-91)
t

The bulk concentration in mass basis, mb,k (SI unit: kg/m2) for a species k, can be
derived from

m b k = M k c b k

414 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

leading to the bulk thickness sb,k (SI unit: m) of species k according to

m b k
s b k = ------------
-
k

TOTAL BULK QUANTITIES

The total bulk concentration is calculated according to

c b tot =  c b k
k

The deposited total mass, mb,tot (SI unit: kg/m2) is

m b tot =  mb k
k

and the total bulk thickness, stot (SI unit: m) is

s b tot =  sk
k

FRACTIONAL QUANTITIES
The fractional bulk concentration, bulk mass, and thickness (all dimensionless) are
calculated according to

c b k
c b frac k = ------------
-
c b tot

m b k
m b frac k = ---------------
-
m b tot

s b k
s b frac k = ------------
-
s b tot

ODE Formulations for Surface Concentrations

The surface of a 1D geometry has a 0 dimension, hence no surface transport can be
modeled and the governing equation is reduced to an ODE:

THEORY FOR THE SURFACE REACTIONS INTERFACE | 415

 c s i
= R s i
t

Surface Reaction Equations on Deforming Geometries

If a Surface Reaction interface is used in conjunction with a Deformed Geometry or a
Moving Mesh interface, the boundary concentration can either be assumed to be
transported with the deforming geometry (moving mesh), with no compensation for
the stretching (the Compensate for boundary stretching checkbox is not selected), or
the species can be assumed to “float” on the mesh (the Compensate for boundary
stretching checkbox is selected, which is the default). In the latter case the following is
assumed in regard to the coupling between the surface species and bulk species and the
mesh movement:

• Expansion or contraction of the boundary dilutes or increases concentration of the

species, respectively, so that the surface integral (in spatial coordinates) of the species
is kept constant.
• Tangential mesh movement has no impact on the local concentration in spatial
coordinates, that is, the tangential transport of surface and bulk species does not
move with the mesh in the tangential direction.

In order to comply with the additional contributions to the mass balance, equations
are added. First, the following terms are added to the right-hand side of Equation 5-90
and Equation 5-91, respectively:

– c s i  ln  A
t

– c b k  ln  A
t

where A is the infinitesimal mesh area segment (area scale factor). The above terms
account for the concentration change due to a fractional area change.

Second, the resulting unwanted tangential convection, imposed by the mesh

movement, is compensated for by the adding following terms to the right-hand side of
Equation 5-90 and Equation 5-91, respectively:

 t   c s i v t mesh 

 t   c b k v t mesh 

416 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

where vt,mesh is the mesh velocity in the tangential direction.

This convectional term needs often to be stabilized using methods such as streamline
diffusion or isotropic diffusion.

Reference for the Surface Reactions Interface

1. R.J. Kee, M.E. Coltrin, and P. Glarborg, Chemically Reacting Flow, Chapter 11,
John Wiley & Sons, 2003.

THEORY FOR THE SURFACE REACTIONS INTERFACE | 417

 Theory for the Coupling of Mass
Transport to Electrochemical
Reactions
When relating the species fluxes at an electrode surface to the electric currents, the
general chemical reaction formula

 ox Sox + ne  red Sred

-

ox red

is used to define the stoichiometric coefficients, i, with i being positive (red) for
products and negative (ox) for the reactants in a reduction reaction. The number of
participating electrons, n, is always positive.

In this section:

• Molar Sources and Sinks

• Mass Sources and Sinks

Molar Sources and Sinks

The molar species fluxes, Ni (SI unit: mol/(m2·s)), perpendicular to an
electrode-electrolyte interface are calculated by summing all the flux contributions
from the electrode reactions, of index m, according to Faraday’s law:

 i m i m
Ni =  -----------------
nm F
-
m

where im is the local current density (SI unit: A/m2) of the electrochemical reaction,
nm the number of participating electrons and F (SI unit: C/mol) is Faraday’s
constant.

The molar species flux, Ni, is obtained from normal component of the molar species
flux vector over the electrode-electrolyte interface:

Ni = Ni  n

where n is the normal vector of the boundary pointing into the domain.

418 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

For a porous electrode, the electrochemical reactions result in species source terms
calculated from:

 i m i m
R i molar = –  av m -----------------
nm F
-
m

where av,m, (SI unit: m2/m3) is the specific surface area.

Mass Sources and Sinks

To compute the mass flux of a species, ni (SI unit: kg/(m2·s)), the molar flux is
multiplied by the molar mass, Mi (SI unit: kg/mol), resulting in

ni  n = Mi Ni

for a mass flux boundary, and

R i mass = M i R i molar

for a porous source term.

THEORY FOR THE COUPLING OF MASS TRANSPORT TO ELECTROCHEMICAL REACTIONS | 419

 Theory for the Reacting Flow
Interface
The following sections describe theory applicable for the Reacting Flow interface:

• Pseudo Time Stepping for Mass Transport

• The Stefan Velocity
• The Chemical Reaction Rate
• Nonisothermal Reacting Flow
• References for the Reacting Flow Interfaces

Pseudo Time Stepping for Mass Transport

In order to improve the solution robustness, pseudo time stepping can be used for the
Transport of Concentrated Species interface when solving a stationary model. Using
pseudo time stepping, a fictitious time derivative term:

 i – nojac   i 
 ------------------------------------
t̃

is added to the left-hand side of the mass fraction equations. Here  is the fluid mixture
density, i is the mass fraction (dimensionless) of species i, and t̃ is the pseudo time
step. Since inojac(i) is always zero, this term does not affect the final solution. It
does, however, affect the discrete equation system and effectively transforms a
nonlinear iteration into a time step of size t̃ .

For a description of the pseudo time step term for the Navier–Stokes
equations and the pseudo time step see and Pseudo Time Stepping in the
COMSOL Multiphysics Reference Manual.

The Stefan Velocity

Heterogeneous reactions on fluid-solid surfaces can affect the mass, momentum, and
energy balances at the interface separating the fluid and the solid. On the reacting
surface, the production or destruction rate, rs,i (SI unit: mol/(m2·s)), of a fluid phase

420 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

species is balanced by the total mass flux of the species. The mass balance for species i
on a boundary representing a fluid-solid interface is given by

n   j i +  i u  = r s,i M i (5-92)

Here, n is the unit normal pointing out of the fluid domain, u is the mass averaged
velocity of the fluid mixture (SI unit: m/s), ji denotes the mass flux of species i relative
to the mixture (typically due to diffusion), and Mi is the species molar mass (SI unit:
kg/mol). Summing the mass balances at the surface, over all species, results in an
effective mixture velocity:

n  u s =  rs,i Mi (5-93)
i=1

referred to as the Stefan velocity, here denoted us. To reach Equation 5-93 the fact
that the sum of all mass fractions is one, and that the sum of all relative diffusive fluxes
is zero, was used.

Equation 5-93 implies that surface reactions result in a net flux between the surface
and the domain. A net flux in turn corresponds to an effective convective velocity at
the domain boundary; the Stefan velocity. It should be noted here that when solving
for mass transport inside a fluid domain, an outer boundary of the domain corresponds
to a position just outside of the actual physical wall (on the fluid side). The domain
boundary does not coincide with the physical wall.

In most reacting flow models, the species mass fractions in the fluid domain are solved
for without including the surface concentrations (mol per area) on exterior walls. One
reason for this is that the surface reaction rates are often not known. In this case,
surface reactions can be modeled either by applying a mass flux or prescribing the mass
fraction, or a combination of both, on fluid boundaries adjacent to the reacting
surface. The Stefan velocity on a fluid domain boundary is then defined as the net mass
flux resulting from the boundary conditions applied:

n  u s =  n   j0 i + 0 i u  (5-94)

i=1

Here, the first term contains contributions from boundary conditions prescribing the
mass flux, while the second contains contributions from boundary conditions
prescribing the mass fractions. Contributions to the Stefan velocity can be added by

THEORY FOR THE REACTING FLOW INTERFACE | 421

 selecting Account for Stefan velocity in the Flux or Mass Fraction features in The
Transport of Concentrated Species interface.

The resulting Stefan velocity based on mass transport boundary conditions is

computed as

 n  j 0 i
n  u s = ------------------------------------
i=1 - (5-95)
N
 
  1 –   i 0
 i 

Using a Reacting Flow interface, the Stefan velocity, defined in the manner of
Equation 5-95, is automatically computed and applied on boundaries corresponding
to walls in the coupled fluid flow interface. The Stefan velocity is prescribed in the wall
normal direction on the wall selection.

The Chemical Reaction Rate

For laminar flow, or when the Turbulent-reaction model is set to None in a feature (in
Transport of Concentrated Species), the default (Automatic) reaction rate used by the
feature is based on the mass action law. Consider a general reaction belonging to a set
of j reactions and involving i species:

f (5-96)
kj
aA + bB + ... r
xX + yY + ...
kj

For such a reaction set, the reaction rates rj (SI unit: mol/(m3·s)), can be described by
the mass action law:

– 
 
f r
rj = kj c i ij – k j c i ij (5-97)
i  react i  prod
f r
Here, kj and kj denote the forward and reverse rate constants, respectively. The
concentration of species i is denoted ci (SI unit: mol/m3). The stoichiometric
coefficients are denoted ij, and are defined to be negative for reactants and positive
for products. In practice, a reaction seldom involves more than two species colliding
in a reacting step, which means that a kinetic expression is usually of order 2 or less
(with respect to the involved concentrations).

422 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

In addition to the concentration dependence, the temperature dependence can be
included by using the predefined Arrhenius expressions for the rate constants:

k = AT exp  – ----------
n E -
R g T

Here, A denotes the frequency factor, n the temperature exponent, E the activation
energy (SI unit: J/mol) and Rg the gas constant, 8.314 J/(mol·K). The
pre-exponential factor, including the frequency factor A and the temperature factor
Tn, is given the units (m3/mol)  1/s, where  is the order of the reaction (with
respect to the concentrations).

Nonisothermal Reacting Flow

When a Chemistry interface and a Heat Transfer interface are used in the
Nonisothermal Reacting Flow multiphysics coupling, the heat equation is modified to
account for reactions and diffusion.

In presence of chemical reactions, the enthalpy (SI unit: J/kg) is a function of

temperature, pressure and the mixture composition. For a mixture of N components,
the mass fractions verify the condition 1  …  N  1. Therefore the enthalpy is a
function of the temperature, pressure, and only N  1 mass fractions:

ˆ = H
ˆ  T p    
H 1 N – 1

Assuming the same velocity for the convection term of all components, the mass
fractions follow the equation:


  -----   i  +  u    i = –   j i + R i
t

On the other hand, using the chain rule, the enthalpy variation is written:

N–1
ˆ Ĥ
--------- +  Ĥ
DH DT Dp Ĥ D i
 ---------- =  --------
Dt T p 
Dt
--------
p T 
-------- + 
Dt  --------
 i
- ----------
Dt
j j
i=1 T  p  j  i

which can be rewritten using the partial molar enthalpies:

Ĥ
H i = --------
n i
T p n j  i

THEORY FOR THE REACTING FLOW INTERFACE | 423

 N
DH ˆ Ĥ DT Ĥ- Dp Hi
 ---------- =  --------
Dt T p 
--------- +  -------
Dt
j
p T 
-------- +
j
Dt  ------
Mi
-  – j i + R i 
i=1

Additionally, the heat balance equation in terms of internal energy is written:

DE
 --------- = –   q + Q vd + Q
Dt

where Qvd is the heat source from viscous dissipation, and Q represents additional
sources. The heat flux q consists in a term of conduction and a term of diffusion:

N
Hi
q = q cond + q diff = – kT +  ------
Mi i
-j
i=1

ˆ –p
Enthalpy and internal energy are linked through the formula E = H --- , we can thus

rewrite the heat balance equation in terms of enthalpy:

N
Ĥ   Dp
--------- +   Ĥ
DT Hi
 --------
T p 
Dt  p
j
--------
T  j
– 1 -------- +
 Dt  ------
Mi
-  – j i + R i  = –   q + Q vd + Q
i=1

Introducing the heat capacity at constant pressure Cp (SI unit: J/(kg·m3), and the
coefficient of thermal expansion p (SI unit: 1/K) gives:

N
DT Dp Hi
C p --------- –  p T -------- +
Dt Dt  ------
Mi
-  – j i + R i  = –   q + Q vd + Q
i=1

The term with the pressure derivative is the work of pressure forces Qp. The enthalpy
variation due to the gradient of species mass flux can be partly simplified with the
diffusion part of the heat flux. The heat source of reaction can be written in terms of
enthalpies of reaction Hi instead of partial enthalpies (see Ref. 1). The temperature
derivative term can be expanded to make the convective term appear. We finally obtain
the temperature equation for a free flow with N species and M reactions:

T- +  u   T +    – kT  = Q
C p ------ diff + Q react + Q vd + Q p + Q
t

424 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

 N
 Hi 
Q diff = –  ji    ------
M i
-
i=1

Q react = –  Hi ri
i=1

Q vd =  – pI +  :u

Q p =  p T p
------ +  u   p
t

For reacting flows moving through the interstices of a porous medium, the heat
sources of reaction and diffusion are only accounted for the temperature of the fluid
phase. While the porosity is already accounted for in the diffusion term through the
mass flux, the heat source of reaction has to be multiplied by the porosity. Also, in
presence of a porous catalyst, both reactions in the bulk and reactions at the surface are
accounted for in the reaction source term:

M bulk M surf

 
s s
Q react = –  p H i r i – S area Hi ri
i=1 i=1

where p is the porosity of the medium (SI unit: 1), Sarea is the specific surface area
s
(SI unit: 1/m), and r i is the surface reaction rate (SI unit: mol/(m2·s)).

An additional source term is added to model convective heat transfer between the fluid
and the porous matrix to get the temperature equation for nonisothermal reacting
flows in porous media:

T
 f C p f ---------f +  u   T f +    – k f T f  = Q diff + Q react + Q vd + Q p + Q s f + Q
t

This equation is the one that is solved when the Porous medium type of the Porous
Medium feature is set to Local thermal nonequilibrium, since both fluid and porous
matrix temperatures are solved separately. With the Local thermal equilibrium approach,
the equation is written for the common temperature, and in terms of effective material
properties for the time derivative term and conduction term:

THEORY FOR THE REACTING FLOW INTERFACE | 425

 T
 C p  eff ------- +  f C p f  u   T +    – k eff T  = Q diff + Q react + Q vd + Q p + Q
t

References for the Reacting Flow Interfaces

1. H.A. Jakobsen, Chemical Reactor Modeling: Multiphase Reactive Flows,
Springer, 2008.

426 | CHAPTER 5: CHEMICAL SPECIES TRANSPORT INTERFACES

 6

Fluid Flow Interfaces

This chapter describes the physics interfaces found under the Fluid Flow branch
( ).

In this chapter:

• The Laminar Flow and Creeping Flow Interfaces

• The Darcy’s Law Interface
• The Free and Porous Media Flow, Brinkman Interface
• The Free and Porous Media Flow, Darcy Interface
• The Brinkman Equations Interface
• The Multiphase Flow in Porous Media Interface
• The Multiphase Free and Porous Media Flow Interface
• The Bubbly Flow Interface
• The Mixture Model Interface
• The Euler-Euler Model Interface
• Theory for the Laminar Flow and Creeping Flow Interfaces
• Theory for the Darcy’s Law Interface
• Theory for the Free and Porous Media Flow, Brinkman Interface

| 427
 • Theory for the Free and Porous Media Flow, Darcy Interface
• Theory for the Brinkman Equations Interface
• Theory for the Multiphase Flow in Porous Media Interface
• Theory for the Bubbly Flow Interface
• Theory for the Mixture Model Interface
• Theory for the Euler-Euler Interface
• Theory for the Coupling of Fluid Flow to Electrochemical Reactions

428 | CHAPTER 6: FLUID FLOW INTERFACES

The Laminar Flow and Creeping Flow
Interfaces
In this section:

• The Creeping Flow Interface

• The Laminar Flow Interface
• The Bubbly Flow Interface

• Theory for the Coupling of Fluid Flow to Electrochemical Reactions

In the COMSOL Multiphysics Reference Manual, see Table 2-4 for links to common
sections such as Discretization, Consistent Stabilization, Inconsistent Stabilization, and
Advanced Settings sections, some of them accessed by clicking the Show button ( )
and choosing the applicable option. You can also search for information: press F1 to
open the Help window or Ctrl+F1 to open the Documentation window.

The Creeping Flow Interface

The Creeping Flow (spf) interface ( ) is used for simulating fluid flows at very low
Reynolds numbers for which the inertial term in the Navier–Stokes equations can be
neglected. Creeping flow, also referred to as Stokes flow, occurs in systems with high
viscosity or small geometrical length scales (for example, in microfluidics and MEMS
devices). The fluid can be compressible or incompressible, as well as Newtonian or
non-Newtonian.

The equations solved by the Creeping Flow interface are the Stokes equations for
conservation of momentum and the continuity equation for conservation of mass.

The Creeping Flow interface can be used for stationary and time-dependent analyses.

The main feature is the Fluid Properties node, which adds the Stokes equations and
provides an interface for defining the fluid material and its properties. Except where
noted below, see The Laminar Flow Interface for all other settings.

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 429

 PHYSICAL MODEL
By default, the Neglect inertial term (Stokes flow) checkbox is selected. If unchecked,the
inertial terms are included in the computations.

DISCRETIZATION
By default, the Creeping Flow interface uses P2+P1 elements. Contrary to general
laminar and turbulent single-phase flow simulations employing purely linear P1+P1
elements, P2+P1 elements are well suited for most creeping flow simulations but are
not in general fully conservative. Full conservation properties and high accuracy is
obtained by selecting P2+P2 or P3+P3 elements.

Note that P1+P1, P2+P2, and P3+P3 discretizations (the so-called equal order
interpolations) require streamline diffusion to be active.

The Laminar Flow Interface

The Laminar Flow (spf) interface ( ) is used to compute the velocity and pressure
fields for the flow of a single-phase fluid in the laminar flow regime. A flow remains
laminar as long as the Reynolds number is below a certain critical value. At higher
Reynolds numbers, disturbances have a tendency to grow and cause transition to
turbulence. This critical Reynolds number depends on the model, but a classical
example is pipe flow, where the critical Reynolds number is known to be approximately
2000.

The physics interface supports incompressible flow, weakly compressible flow (the
density depends on temperature but not on pressure), and compressible flow at low
Mach numbers (typically less than 0.3). It also supports flow of non-Newtonian fluids.

The equations solved by the Laminar Flow interface are the Navier–Stokes equations
for conservation of momentum and the continuity equation for conservation of mass.

The Laminar Flow interface can be used for stationary and time-dependent analyses.
Time-dependent studies should be used in the high-Reynolds number regime as these
flows tend to become inherently unsteady.

When the Laminar Flow interface is added, the following default nodes are also added
in the Model Builder: Fluid Properties, Wall (the default boundary condition is No slip),
and Initial Values. Other nodes that implement, for example, boundary conditions and
volume forces, can be added from the Physics toolbar or from the context menu
displayed when right-clicking Laminar Flow.

430 | CHAPTER 6: FLUID FLOW INTERFACES

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Physics interface variables can be referred to using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is spf.

About the Physics Interface Label Names

The label for a single-phase flow interface (the node name in the Model Builder) is
dynamic and is reset according to the turbulence model and the neglect inertial term
(Stokes flow) property set at the physics interface level. The availability also depends
on the software license.

If the Neglect inertial term (Stokes flow) checkbox is selected, then the Label changes
to Creeping Flow, which is the same Label that displays when that interface is added
from the Model Wizard or Add Physics window.

PHYSICAL MODEL

Compressibility
Depending of the fluid properties and the flow regime, three options are available for
the Compressibility option. In general the computational complexity increases from
Incompressible flow to Weakly compressible flow to Compressible flow (Ma<0.3) but the
underlying hypotheses are increasingly more restrictive in the opposite direction.

When the Incompressible flow option (default) is selected, the incompressible form of
the Navier–Stokes and continuity equations is applied. In addition, the fluid density is
evaluated at the Reference pressure level defined in this section. The Reference
temperature is set to 293.15 K.

The Weakly compressible flow option models compressible flow when the pressure
dependency of the density can be neglected. When selected, the compressible form of
the Navier–Stokes and continuity equations is applied. In addition, the fluid density is
evaluated at the Reference pressure level defined in this section.

When the Compressible flow (Ma<0.3) option is selected, the compressible form of the
Navier–Stokes and continuity equations is applied. Ma  0.3 indicates that the inlet
and outlet conditions, as well as the stabilization, may not be suitable for transonic and
supersonic flow. For more information, see The Mach Number Limit.

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 431

 Porous Media Domains
With the addition of various modules, the Enable porous media domains checkbox is
available. Selecting this option, a Fluid and Matrix Properties node, a Mass Source node,
and a Forchheimer Drag subnode are added to the physics interface. These are described
for the Brinkman Equations interface in the respective module’s documentation. The
Fluid and Matrix Properties can be applied on all domains or on a subset of the domains.

Porous Treatment of No Slip Condition

Porous treatment of no slip condition is available when Enable porous media domains is
activated. It controls how the No Slip condition on Wall boundaries and Interior Wall
boundaries adjacent to Porous Medium should be treated. The options are Standard no
slip formulation (default) and Porous slip. The latter option provides a unified treatment
when the porous matrix is fully resolved as well as when it is under resolved ensuring
a smooth transition between regions with different resolutions; see No Slip under Wall.

Include Gravity
When the Include gravity checkbox is selected, a global Gravity feature is shown in the
interface model tree, and the buoyancy force is included in the Navier–Stokes
equations.

Also, when the Include gravity checkbox is selected, the Use reduced pressure option
changes the pressure formulation from using the total pressure (default) to using the
reduced pressure. This option is suitable for configurations where the density changes
are very small; otherwise, the default formulation can be used. For more information,
see Gravity

Reference Values
Reference values are global quantities used to evaluate the density of the fluid when
the Incompressible flow or the Weakly compressible flow option is selected.

Reference pressure level There are generally two ways to include the pressure in fluid
flow computations: either to use the absolute pressure pAppref, or the gauge
pressure p. When pref is nonzero, the physics interface solves for the gauge pressure
whereas material properties are evaluated using the absolute pressure. The reference
pressure level is also used to define the reference density.

Reference temperature The reference temperature is used to define the reference

density.

432 | CHAPTER 6: FLUID FLOW INTERFACES

Reference position When Include gravity is selected, the reference position can be
defined. It corresponds to the location where the total pressure (that includes the
hydrostatic pressure) is equal to the Reference pressure level.

DEPENDENT VARIABLES
The following dependent variables (fields) are defined for this physics interface — the
Velocity field u and its components, and the Pressure p.

If required, the names of the field, component, and dependent variable can be edited.
Editing the name of a scalar dependent variable changes both its field name and the
dependent variable name. If a new field name coincides with the name of another field
of the same type, the fields share degrees of freedom and dependent variable names. A
new field name must not coincide with the name of a field of another type or with a
component name belonging to some other field. Component names must be unique
within a model except when two fields share a common field name.

CONSISTENT STABILIZATION
To enable this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog.

There are two consistent stabilization methods: Streamline diffusion and Crosswind
diffusion. Usually, both checkboxes for these methods are selected by default and
should remain selected for optimal performance. Consistent stabilization methods do
not perturb the original transport equation. Streamline diffusion must be selected
when using equal-order interpolation for pressure and velocity.

Select the Use dynamic subgrid time scale checkbox to estimate the influence of
temporal variations in the velocity field, instead of using the inverse time step, in the
consistent stabilization terms. This checkbox is selected by default

Select the Limit small time steps effect on stabilization time scale checkbox to prevent
loss of stabilization at small time steps.

INCONSISTENT STABILIZATION
To enable this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog.

There is usually just one inconsistent stabilization method — Isotropic diffusion. This
method is equivalent to adding a term to the diffusion coefficient in order to dampen
the effect of oscillations by making the system somewhat less dominated by
convection. If possible, minimize the use of the inconsistent stabilization method

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 433

 because by using it you no longer solve the original problem. By default, the Isotropic
diffusion checkbox is not selected because this type of stabilization adds artificial
diffusion and affects the accuracy of the original problem. However, this option can be
used to get a good initial guess for underresolved problems.

If required, select the Isotropic diffusion checkbox and enter a Tuning parameter id as
a scalar positive value. The default value is 0.25 (a reasonable value to start with is
roughly 0.5 divided by the element order). A higher value adds more isotropic
diffusion.

• Numerical Stability — Stabilization Techniques for Fluid Flow

ADVANCED SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog. Normally these settings do not need
to be changed.

The Use pseudo time stepping for stationary equation form is per default set to Automatic
from physics. This option can add pseudo time derivatives to the equation when the
Stationary equation form is used in order to speed up convergence. Pseudo time
stepping is triggered when the Laminar Flow interface is selected in some multiphysics
coupling features. Set Automatic from physics to On to apply pseudo time stepping also
for laminar flows. Set it to Off to disable pseudo time stepping completely.

When Use pseudo time stepping for stationary equation form is set to Automatic from
physics or On, a CFL number expression should also be defined. For the default Automatic
option, the local CFL number (from the Courant–Friedrichs–Lewy condition) is
determined by a PID regulator.

The Use Block Navier-Stokes preconditioner checkbox under Linear solver is available
when the Compressibility option is set to Incompressible flow. When this checkbox is
selected, the default solver for time-dependent study steps will use the Block Navier–
Stokes preconditioner in iterative solvers for the velocity and pressure. Using this

434 | CHAPTER 6: FLUID FLOW INTERFACES

preconditioner may result in shorter solution times for large time dependent problems
with high Reynolds numbers.

• Pseudo Time Stepping for Laminar Flow Models

• Domain, Boundary, Pair, and Point Nodes for Single-Phase Flow
• Theory for the Coupling of Fluid Flow to Electrochemical Reactions
• Block Navier–Stokes in the COMSOL Multiphysics Reference
Manual.

DISCRETIZATION
The default discretization for Laminar Flow is P1+P1 elements — that is, piecewise
linear interpolation for velocity and pressure. This is suitable for most flow problems.

Some higher-order interpolations are also available, and they can be cost-effective
options to obtain high accuracy for flows with low Reynolds numbers.

The P2+P2 and P3+P3 options, the equal-order interpolation options, are the preferred
higher-order options because they have higher numerical accuracy than the
mixed-order options P2+P1 and P3+P2. The equal-order interpolation options do,
however, require streamline diffusion to be active.

Flow Past a Cylinder: Application Library path COMSOL_Multiphysics/

Fluid_Dynamics/cylinder_flow

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 435

 Domain, Boundary, Pair, and Point Nodes for Single-Phase Flow
The following nodes, listed in alphabetical order, are available from the Physics ribbon
toolbar (Windows users), Physics context menu (Mac or Linux users), or by
right-clicking to access the context menu (all users).

• Boundary Stress • Periodic Flow Condition

• Flow Continuity • Point Mass Source1
• Fluid Properties • Porous Interface
• Gravity • Pressure Point Constraint
• Initial Values • Symmetry
• Inlet • Volume Force
• Line Mass Source1 • Wall
• Open Boundary
• Outlet
1 A feature that may require an additional license

For 2D axisymmetric components, COMSOL Multiphysics takes the axial

symmetry boundaries (at r  0) into account and adds an Axial Symmetry
node that is valid on the axial symmetry boundaries only.

In the COMSOL Multiphysics Reference Manual, see Table 2-4 for links to common
sections and Table 2-5 to common feature nodes. You can also search for information:
press F1 to open the Help window or Ctrl+F1 to open the Documentation window.

For a detailed overview of the functionality available in each product, visit

[Link]

Fluid Properties
The Fluid Properties node adds the momentum and continuity equations solved by the
physics interface, except for volume forces, which are added by the Volume Force
feature. The node also provides an interface for defining the material properties of the
fluid.

436 | CHAPTER 6: FLUID FLOW INTERFACES

MODEL INPUTS
Fluid properties, such as density and viscosity, can be defined through user inputs,
variables, or by selecting a material. For the latter option, additional inputs, for
example temperature or pressure, may be required to define these properties.

Temperature
By default, the Temperature model input is set to Common model input, and the
temperature is controlled from Default Model Inputs under Global Definitions or by a
locally defined Model Input. If a Heat Transfer interface is included in the component,
it controls the temperature Common model input. Alternatively, the temperature field
can be selected from another physics interface. All physics interfaces have their own
tags (Name). For example, if a Heat Transfer in Fluids interface is included in the
component, the Temperature (ht) option is available for T.

You can also select User defined from the Temperature model input in order to manually
prescribe T.

Absolute Pressure
This input appears when a material requires the absolute pressure as a model input.
The absolute pressure is used to evaluate material properties, but it also relates to the
value of the calculated pressure field. There are generally two ways to calculate the
pressure when describing fluid flow: either to solve for the absolute pressure or for a
pressure (often denoted gauge pressure) that relates to the absolute pressure through
a reference pressure.

The choice of pressure variable depends on the system of equations being solved. For
example, in a unidirectional incompressible flow problem, the pressure drop over the
modeled domain is probably many orders of magnitude smaller than the atmospheric
pressure, which, when included, may reduce the stability and convergence properties
of the solver. In other cases, such as when the pressure is part of an expression for the
gas volume or the diffusion coefficients, it may be more convenient to solve for the
absolute pressure.

The default Absolute pressure pA is p  pref, where p is the dependent pressure variable
from the Navier–Stokes or RANS equations, and pref is from the user input defined at
the physics interface level. When pref is nonzero, the physics interface solves for a
gauge pressure. If the pressure field instead is an absolute pressure field, pref should be
set to 0.

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 437

 The Absolute pressure field can be edited by clicking Make All Model Inputs Editable
( ) and entering the desired value in the input field.

Model Inputs and Multiphysics Couplings in the COMSOL Multiphysics

Reference Manual

FLUID PROPERTIES

Density
The density can either be specified by a material, or by a User defined expression. The
density in a material can depend on temperature and pressure, and these dependencies
are automatically replaced by pref for weakly compressible flows and pref and Tref for
incompressible flows (as specified by the Compressibility setting at the physics interface
level). If density variations with respect to pressure are to be included in the
computations, Compressibility must be set to compressible. Any dependencies in the
density on quantities other than temperature and pressure must be consistent with the
Compressibility setting at the interface level.

Dynamic Viscosity
The Dynamic viscosity  describes the relationship between the shear rate and the shear
stresses in a fluid. Intuitively, water and air have low viscosities, and substances often
described as thick (such as oil) have higher viscosities.

Volume Force
The Volume Force node specifies the volume force F on the right-hand side of the
momentum equation.

u T 2
 +   u   u =   – pI +   u +  u   – ---     u I + F
t 3

If several volume-force nodes are added to the same domain, then the sum of all
contributions are added to the momentum equation.

Initial Values
The initial values serve as initial conditions for a transient simulation or as an initial
guess for a nonlinear solver in a stationary simulation. Note that for a transient
compressible-flow simulation employing a material for which the density depends on

438 | CHAPTER 6: FLUID FLOW INTERFACES

the pressure (such as air), discontinuities in the initial values trigger pressure waves
even when the Mach number is small. The pressure waves must be resolved and this
puts a restriction on the time step.

INITIAL VALUES
Initial values or expressions should be specified for the Velocity field u and the Pressure
p.

When Include gravity is selected and Use reduced pressure not selected in the interface
Physical model section, the Compensate for hydrostatic pressure approximation (named
Compensate for hydrostatic pressure for compressible flows) checkbox is available and
selected by default. When it is selected, the hydrostatic pressure is automatically added
to the pressure entered in p user input.

Wall
The Wall node includes a set of boundary conditions describing fluid-flow conditions
at stationary, moving, and leaking walls. For turbulent flow, the description may
involve wall functions and asymptotic expressions for certain turbulence variables.

BOUNDARY CONDITION
Select a Boundary condition for the wall.

• No Slip • Leaking Wall

• Slip • Navier Slip

No Slip
No slip is the default boundary condition to model solid walls. A no-slip wall is a wall
where the fluid velocity relative to the wall velocity is zero. For a stationary wall that
means that u = 0.

The option for Porous treatment of no slip condition is available when Enable porous
media domains is activated in the Physical Model section in the settings for the main
physics interface node. It specifies how Wall boundaries and Interior Wall boundaries
adjacent to porous domains are treated. When Standard no slip formulation is chosen,
a common no slip condition is applied on all solid walls. When the default Porous slip
is chosen, a blending analytic expression is instead applied on the corresponding wall
boundaries adjacent to the porous medium domain. It results in a no slip condition in
case the porous length scale is fully resolved by the mesh and a slip condition in the

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 439

 opposite limit when the mesh is much coarser than the porous scale. A smooth
transition between these limits is ensured. Note that the interpretation and usage of
nonzero slip at the wall is the same as in the Navier slip boundary condition. By default,
Velocity formulation is on and the treatment is based on an approximate reconstruction
of the far-field pressure gradient using the slip velocity at the wall. If the
Pressure-gradient formulation is chosen, the local pressure gradient at the wall is
employed.

Slip
The Slip option prescribes a no-penetration condition, u·n  . It is implicitly assumed
that there are no viscous effects at the slip wall and hence, no boundary layer develops.
From a modeling point of view, this can be a reasonable approximation if the main
effect of the wall is to prevent fluid from leaving the domain.

Leaking Wall
This boundary condition may be used to simulate a wall where fluid is leaking into or
leaving the domain with the velocity u = ul through a perforated wall. The
components of the Fluid velocity ul on the leaking wall should be specified.

Navier Slip
This boundary condition enforces no-penetration at the wall, u nwall0, and adds
a tangential stress


K nt = – --- u slip


where Knt  Kn  Kn  nwallnwallKn  Knwall, and K is the viscous stress tensor. is
a slip length, and uslip  u  u  nwallnwall is the velocity tangential to the wall.

The Slip length setting is per default set to Factor of minimum element length. The slip
length  is then defined as   fhhmin, where hmin is the smallest element side and fh
is a user input. Select User defined from the Slip length selection list in order to manually
prescribe (SI unit: m).

In cases where the wall movement is nonzero, select the Account for the translational
wall velocity of the wall in the friction force checkbox to use
u  ubnd  u ubnd  nwallnwall instead of uslip in the friction force.

The Navier slip option is not available when selecting a turbulence model.

440 | CHAPTER 6: FLUID FLOW INTERFACES

WALL MOVEMENT
This section contains controls to describe the wall movement relative to the lab (or
spatial) frame.

The Translational velocity setting controls the translational wall velocity, utr. The list is
per default set to Automatic from frame. The physics automatically detects if the spatial
frame moves. This can for example happen if an ALE interface is present in the model
component. If there is no movement utr  0. If the frame moves, utr becomes equal
to the frame movement. utr is accounted for in the actual boundary condition
prescribed in the Boundary condition section.

Select Zero (Fixed wall) from Translational velocity selection list to prescribe utr  0.

Select Manual from Translational velocity selection list in order to manually prescribe
Velocity of moving wall, utr. This can for example be used to model an oscillating wall
where the magnitude of the oscillations are very small compared to the rest of the
model. Specifying translational velocity manually does not automatically cause the
associated wall to move. An additional Moving Mesh node needs to be added from
Definitions to physically track the wall movement in the spatial reference frame.

The Sliding wall option is appropriate if the wall behaves like a conveyor belt with the
surface sliding in a tangential direction. A velocity is prescribed at the wall and the
boundary itself does not have to actually move relative to the reference frame.

• For 3D components, values or expressions for the Velocity of sliding wall uw should
be specified. If the velocity vector entered is not in the plane of the wall, COMSOL
Multiphysics projects it onto the tangential direction. Its magnitude is adjusted to
be the same as the magnitude of the vector entered.
• For 2D components, the tangential direction is unambiguously defined by the
direction of the boundary. For this reason, the sliding wall boundary condition has
different definitions in different space dimensions. A single entry for the Velocity of
the tangentially moving wall Uw should be specified in 2D.

CONSTRAINT SETTINGS
This section is displayed by clicking the Show More Options button ( ) and selecting
Advanced Physics Options in the Show More Options dialog. The Constraints settings can
be set to Automatic, Pointwise constraints, Nitsche constraints, or Weak constraints.
Mixed constraints can be selected when imposing a no slip condition exactly.

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 441

 Depending on the constraint method selected, the following settings are available:

• Apply reaction terms on can be set to Individual dependent variables (default) or All
physics (symmetric). This setting is not available when Nitsche constraints is selected.
• Select Elemental (default) or Nodal under Constraint method. This setting is not
available for Nitsche constraints or Weak constraints.

• Theory for the Wall Boundary Condition

• The Moving Mesh Interface in the COMSOL Multiphysics Reference
Manual

Inlet
This condition should be used on boundaries for which there is a net flow into the
domain. To obtain a numerically well-posed problem, it is advisable to also consider
the Outlet conditions when specifying an Inlet condition. For example, if the pressure
is specified at the outlet, the velocity may be specified at the inlet, and vice versa.
Specifying the velocity vector at both the inlet and the outlet may cause convergence
difficulties.

BOUNDARY CONDITION
The available Boundary condition options for an inlet are Velocity, Fully developed flow,
and Pressure. After selecting a Boundary Condition from the list, a section with the same
or a similar name displays underneath. For example, if Velocity is selected, a Velocity
section, where further settings are defined, is displayed.

VELOCITY
The Normal inflow velocity is specified as u = nU0, where n is the boundary normal
pointing out of the domain and U0 is the normal inflow speed.

The Velocity field option sets the velocity vector to u = u0. The components of the inlet
velocity vector u0 should be defined for this choice.

442 | CHAPTER 6: FLUID FLOW INTERFACES

PRESSURE CONDITIONS
This option specifies the normal stress, which in most cases is approximately equal to
the pressure.

• For incompressible flow, the Pressure list has two options, Static and Total. For
weakly compressible and compressible flow, the static pressure should be specified
in the text field.
- If Pressure is Static, and the reference pressure pref, defined at the physics
interface level, is equal to , the value of the pressure p0, at the boundary, is the
absolute pressure. Otherwise, p0 is the relative pressure at the boundary.
- If Pressure is Total, the Average checkbox is available and cleared by default to
prescribe the total pressure pointwise. If it is selected, the averaged total pressure
is imposed in the weak forms instead.
When Include gravity is selected and Use reduced pressure not selected in the interface
Physical model section, the Compensate for hydrostatic pressure approximation
(named Compensate for hydrostatic pressure for incompressible flows) checkbox is
available and selected by default. When it is selected, the hydrostatic pressure is
automatically added to the pressure entered in p0 user input.
• The Suppress backflow option adjusts the inlet pressure locally in order to reduce the
amount of fluid exiting the domain through the boundary. If you clear the suppress
backflow option, the inlet boundary can become an outlet depending on the
pressure field in the rest of the domain.
• Flow direction controls in which direction the fluid enters the domain.
- For Normal flow, it prescribes zero tangential velocity component.
- For User defined, an Inflow velocity direction du (dimensionless) should be
specified. The magnitude of du does not matter, only the direction. du must
point into the domain.

MASS FLOW
The mass flow at an inlet can be specified by the Mass flow rate, the Pointwise mass flux,
the Standard flow rate, or the Standard flow rate (SCCM).

The Apply condition on each disjoint selection separately checkbox is selected per
default. When this setting is selected, the mass flow condition is applied separately on
each disjoint selection. If this option is not selected, the condition is applied over the
whole feature selection. The Apply condition on each disjoint selection separately should
be disabled only if the flow conditions are known to be identical on each disjoint
boundary.

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 443

 Mass Flow Rate
The Mass flow rate option sets the integrated mass flow over the boundary selection,
the Normal mass flow rate to a specific value, m. The mass flow is assumed to be parallel
to the boundary normal, and the tangential flow velocity is set to zero.

For 2D components, the Channel thickness dbc is used to define the area across which
the mass flow occurs. This setting is not applied to the whole model. Line or surface
integrals of the mass flow over the boundary evaluated during postprocessing or used
in integration coupling operators do not include this scaling automatically. Such results
should be appropriately scaled when comparing them with the specified mass flow.

Pointwise Mass Flux

The Pointwise mass flux sets the mass flow parallel to the boundary normal. The
tangential flow velocity is set to zero. The mass flux is a model input, which means that
COMSOL Multiphysics can take its value from another physics interface when
available. When User defined is selected a value or function Mf should be specified for
the Mass flux.

Standard Flow Rate

The Standard flow rate Qsv sets a standard volumetric flow rate, according to the SEMI
standard E12-0303. The mass flow rate is specified as the volumetric flow rate of a gas
at standard density — the Mean molar mass Mn divided by a Standard molar volume Vm
(that is, the volume of one mole of a perfect gas at standard pressure and standard
temperature). The flow occurs across the whole boundary in the direction of the
boundary normal and is computed by a surface (3D) or line (2D) integral. The
tangential flow velocity is set to zero.

The standard density can be defined directly, or by specifying a standard pressure and
temperature, in which case the ideal gas law is assumed. The options in the
Standard flow rate defined by list are:

• Standard density, for which the Standard molar volume Vm should be specified.
• Standard pressure and temperature, for which the Standard pressure Pst and the
Standard temperature Tst should be defined.

For 2D components, the Channel thickness dbc is used to define the area across which
the mass flow occurs. This setting is not applied to the whole model. Line or surface
integrals of the mass flow over the boundary evaluated during postprocessing or used
in integration coupling operators do not include this scaling automatically. Such results
should be appropriately scaled when comparing them with the specified mass flow.

444 | CHAPTER 6: FLUID FLOW INTERFACES

Standard Flow Rate (SCCM)
The Standard flow rate (SCCM) boundary condition is equivalent to the Standard flow
rate boundary condition, except that the flow rate is entered directly in SCCMs
(standard cubic centimeters per minute) without the requirement to specify units.
Here, the dimensionless Number of SCCM units Qsccm should be specified.

FULLY DEVELOPED FLOW

The Fully developed flow option adds contributions to the inflow boundary, which force
the flow toward the solution for a fully developed channel flow. The channel can be
thought of as a virtual extrusion of the inlet cross section. The inlet boundary must
hence be flat in order for the fully developed flow condition to work properly. In 2D
axisymmetric models, the inlet normal must be parallel to the symmetry axis.

Select an option to control the flow rate at the inlet:

• Average velocity, Uav.

• Flow rate, V0. Two-dimensional models also require an Entrance thickness, Dz,
which is the out-of-plane thickness of the extruded entrance channel.
• Average pressure, Pav. Note that Pav is the average pressure on the inflow boundary.

The Apply condition on each disjoint selection separately checkbox is selected per
default. When this setting is selected, the fully developed flow condition is applied
separately on each disjoint selection. If this option is not selected, the condition is
applied over the whole feature selection. The Apply condition on each disjoint selection
separately should be disabled only if the flow conditions are known to be identical on
each disjoint boundary.

The fully developed flow condition requires any volume force to be approximately
aligned with the normal of the inlet boundary. The exception is gravity when the
Include gravity setting is selected in the physics interface settings. Unless Use reduced
pressure is also selected, an option to Compensate for hydrostatic pressure or
Compensate for hydrostatic pressure approximation becomes available. It is selected per
default and should only be deselected if the inlet normal is aligned with the gravity
force and you want to specify an average pressure that includes the hydrostatic
pressure.

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 445

 CONSTRAINT SETTINGS
This section is displayed by clicking the Show button ( ) and selecting Advanced
Physics Options.

• Prescribing Inlet and Outlet Conditions

• Normal Stress Boundary Condition

Fully Developed Flow (Inlet)

Outlet
This condition should be used on boundaries for which there is a net outflow from the
domain. To obtain a numerically well-posed problem, it is advisable to also consider
the Inlet conditions when specifying an Outlet condition. For example, if the velocity
is specified at the inlet, the pressure may be specified at the outlet, and vice versa.
Specifying the velocity vector at both the inlet and the outlet may cause convergence
difficulties. Selecting appropriate outlet conditions for the Navier–Stokes equations is
a nontrivial task. Generally, if there is something interesting happening at an outflow
boundary, the computational domain should be extended to include this
phenomenon.

BOUNDARY CONDITION
The available Boundary condition options for an outlet are Pressure, Velocity, Fully
developed flow, and Mass flow.

446 | CHAPTER 6: FLUID FLOW INTERFACES

PRESSURE CONDITIONS
This option specifies the normal stress, which in most cases is approximately equal to
the pressure. The tangential stress component is set to zero.

• For incompressible flow, the Pressure list has two options, Static and Total. For
weakly compressible and compressible flow, the static pressure should be specified
in the text field.
- If Pressure is Static, and the reference pressure pref, defined at the physics
interface level, is equal to , the value of the pressure p0, at the boundary, is the
absolute pressure. Otherwise, p0 is the relative pressure at the boundary.
- If Pressure is Total, the Average checkbox is available and selected by default to
prescribed the averaged total pressure in the weak forms. If it is cleared, the total
pressure is imposed pointwise.

When Include gravity is selected and Use reduced pressure not selected in the interface
Physical model section, the Compensate for hydrostatic pressure approximation (named
Compensate for hydrostatic pressure for incompressible flows) checkbox is available and
selected by default. When it is selected, the hydrostatic pressure is automatically added
to the pressure entered in p0 user input.

• The Normal flow option changes the no tangential stress condition to a no tangential
velocity condition. This forces the flow to exit (or enter) the domain perpendicularly
to the outlet boundary.
• The Suppress backflow checkbox is selected by default. This option adjusts the outlet
pressure in order to reduce the amount of fluid entering the domain through the
boundary.

VELOCITY
See the Inlet node Velocity section for the settings.

FULLY DEVELOPED FLOW

This boundary condition is applicable when the flow exits the domain into a long pipe
or channel, at the end of which a flow profile is fully developed.

The channel can be thought of as a virtual extrusion of the outlet cross section. The
outlet boundary must hence be flat in order for the fully developed flow condition to
work properly. In 2D axisymmetric models, the outlet normal must be parallel to the
symmetry axis.

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 447

 Select an option to control the flow rate at the outlet:

• Average velocity, Uav.

• Flow rate, V0. Two-dimensional models also require an Entrance thickness, Dz,
which is the out-of-plane thickness of the extruded entrance channel.
• Average pressure, Pav. Observe that Pav is the average pressure on the outflow.

The Apply condition on each disjoint selection separately checkbox is selected per
default. When this setting is selected, the fully developed flow condition is applied
separately on each disjoint selection. If this option is not selected, the condition is
applied over the whole feature selection. The Apply condition on each disjoint selection
separately should be disabled only if the flow conditions are known to be identical on
each disjoint boundary.

The fully developed flow condition requires any volume force to be approximately
aligned with the normal of the outlet boundary. The exception is gravity when the
Include gravity setting is selected in the physics interface settings. Unless Use reduced
pressure is also selected, an option to Compensate for hydrostatic pressure or
Compensate for hydrostatic pressure approximation becomes available. It is selected per
default and should only be deselected if the outlet normal is aligned with the gravity
force and you want to specify an average pressure that includes the hydrostatic
pressure.

MASS FLOW
See the Inlet node Mass Flow section for the settings.

CONSTRAINT SETTINGS
This section is displayed by clicking the Show button ( ) and selecting Advanced
Physics Options.

Prescribing Inlet and Outlet Conditions

Fully Developed Flow (Outlet)

448 | CHAPTER 6: FLUID FLOW INTERFACES

Symmetry
This node provides a boundary condition for symmetry boundaries. It should only be
used when the geometry and expected solution have mirror symmetry. By using
symmetries in a model its size can be reduced by one-half or more, making this an
efficient tool for solving large problems.

The Symmetry boundary condition prescribes no penetration and vanishing shear

stresses. The boundary condition is a combination of a Dirichlet condition and a
Neumann condition:

u  n = 0,  – pI +   u +  u  T  – 2
---     u I n = 0
 3 

u  n = 0,  – pI +   u +  u  T  n = 0

for the compressible and incompressible formulations. The Dirichlet condition takes
precedence over the Neumann condition, and the above equations are equivalent to
the following equation for both the compressible and incompressible formulations:

u  n = 0, K –  K  n n = 0
K =   u +  u  T n

BOUNDARY SELECTION
For 2D axial symmetry, a boundary condition does not need to be defined for the
symmetry axis at r  0. The software automatically provides a condition that prescribes
ur  0 and vanishing stresses in the z direction and adds an Axial Symmetry node that
implements these conditions on the axial symmetry boundaries only.

CONSTRAINT SETTINGS
This section is displayed by clicking the Show button ( ) and selecting Advanced
Physics Options.

Open Boundary
The Open Boundary condition describes boundaries in contact with a large volume of
fluid. Fluid can both enter and leave the domain on boundaries with this type of
condition.

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 449

 BOUNDARY CONDITIONS
The Boundary condition options for open boundaries are Normal stress and No viscous
stress.

Normal Stress
The Normal stress f0 condition implicitly imposes p  f0.

When Include gravity is selected and Use reduced pressure not selected in the interface
Physical model section, the Compensate for hydrostatic pressure approximation (named
Compensate for hydrostatic pressure for incompressible flows) checkbox is available and
selected by default. When it is selected, the hydrostatic pressure is automatically added
to the pressure entered in f0 user input.

No Viscous Stress
The No Viscous Stress condition specifies vanishing viscous stress on the boundary. This
condition does not provide sufficient information to fully specify the flow at the open
boundary and must at least be combined with pressure constraints at adjacent points.

The No viscous stress condition prescribes:

   u +  u  T  – 2
---     u I n = 0
 3 

  u +  u  T n = 0

for the compressible and the incompressible formulations. This condition can be useful
in some situations because it does not impose any constraint on the pressure. A typical
example is a model with volume forces that give rise to pressure gradients that are hard
to prescribe in advance. To make the model numerically stable, this boundary
condition should be combined with a point constraint on the pressure.

Boundary Stress
The Boundary Stress node adds a boundary condition that represents a general class of
conditions also known as traction boundary conditions.

BOUNDARY CONDITION
The Boundary condition options for the boundary stress are General stress, Normal
stress, and Normal stress, normal flow.

450 | CHAPTER 6: FLUID FLOW INTERFACES

General Stress
When General stress is selected, the components for the Stress F should be specified.
The total stress on the boundary is set equal to the given stress F:

 – pI +   u +  u  T  – 2
---     u I n = F
 3 

 – pI +   u +  u  T  n = F

for the compressible and the incompressible formulations.

This boundary condition implicitly sets a constraint on the pressure that for 2D flows is

u n
p = 2 ---------- – n  F (6-1)
n

If unn is small, Equation 6-1 states that p  n·F.

Normal Stress
Normal Stress is described for the Open Boundary node.

Normal Stress, Normal Flow

For Normal stress, normal flow, the magnitude of the Normal stress f0 should be
specified. The tangential velocity is set to zero on the boundary:

n  – pI +   u +  u  T  – ---     u I n = – f 0 ,
T 2
tu = 0
3

T
n  – pI +   u +  u  T  n = – f 0 , tu = 0

for the compressible and the incompressible formulations.

This boundary condition implicitly sets a constraint on the pressure that for 2D flows is

u n
p = 2 ---------- + f 0 (6-2)
n

If unn is small, Equation 6-2 states that p  f0.

When Include gravity is selected and Use reduced pressure not selected in the interface
Physical model section, the Compensate for hydrostatic pressure approximation (named
Compensate for hydrostatic pressure for incompressible flows) checkbox is available and
selected by default. When it is selected, the hydrostatic pressure is automatically added

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 451

 to the pressure entered in f0 or phydron is added to F depending on the selected
option.

CONSTRAINT SETTINGS
This section is displayed by clicking the Show button ( ) and selecting Advanced
Physics Options.

If Normal Stress, Normal Flow is selected as the Boundary condition, then to Apply
reaction terms on all dependent variables, the All physics (symmetric) option should be
selected. Alternatively, the Individual dependent variables could be selected to restrict
the reaction terms as needed.

Periodic Flow Condition

The Periodic Flow Condition splits its selection into a source group and a destination
group (the Destination Selection section is optional). Fluid that leaves the domain
through one of the destination boundaries enters the domain through the
corresponding source boundary. This corresponds to a situation where the geometry
is a periodic part of a larger geometry. If the boundaries are not parallel to each other,
the velocity vector is automatically transformed.

If the boundaries are curved, the orientation of the source must be specified manually
(see Orientation of Source).

No input is required when Compressible flow (Ma<0.3) is selected for Compressibility

under the Physical Model section for the physics interface. Typically when a periodic
boundary condition is used with a compressible flow, the pressure is the same at both
boundaries and the flow is driven by a volume force.

BOUNDARY SELECTION

The software usually automatically identifies the boundaries as either source boundaries or
destination boundaries, as indicated in the selection list. This works fine for cases like opposing
parallel boundaries. In other cases, right-click Periodic Flow Condition and select Manual Destination
Selection to control the destination. By default it contains the selection that COMSOL
Multiphysics identifies.

DESTINATION SELECTION
This section is available for specifying the destination boundaries, if needed, when the
Manual Destination Selection option is selected in the context menu for the Periodic Flow

452 | CHAPTER 6: FLUID FLOW INTERFACES

Condition node. You can only select destination boundaries from the union of all source
and destination boundaries.

FLOW CONDITION
This section is available when Incompressible flow is selected for Compressibility under
the Physical Model section for the physics interface.

The Flow Condition at the boundary is specified through a value or expression for either
ꞏ . The mass flow option ensures
the Pressure difference, psrc  pdst, or the Mass flow, m
a pressure difference is achieved across the source and destination boundaries. This
pressure difference can, for example, drive the fully developed flow in a channel.

To set up a periodic boundary condition, both boundaries must be selected in the

Periodic Flow Condition node. COMSOL Multiphysics automatically assigns one
boundary as the source and the other as the destination. To manually set the
destination selection, a Destination Selection subnode is available from the context
menu (by right-clicking the parent node) or from the Physics toolbar, Attributes menu.
All destination sides must be connected.

ORIENTATION OF SOURCE
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog. For information about the Orientation
of Source section, see Orientation of Source and Destination in the COMSOL
Multiphysics Reference Manual.

ORIENTATION OF DESTINATION
This section appears if the setting for Transform to intermediate map in the Orientation
of Source section is changed from the default value, Automatic, and Advanced Physics
Options is selected in the Show More Options dialog. For information about the
Orientation of Destination section, see Orientation of Source and Destination in the
COMSOL Multiphysics Reference Manual.

MAPPING BETWEEN SOURCE AND DESTINATION

To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog. For information about the Mapping
Between Source and Destination section, see Mapping Between Source and Destination
in the COMSOL Multiphysics Reference Manual.

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 453

 Porous Interface
The Porous Interface condition is designed to properly account for a sudden jump in
porosity, p, across a boundary between different domains. It describes possible
discontinuity of stress across the interface, and also reveals modified behavior of
tangential velocity due to boundary layer on the upstream side of the interface. Choose
among the following options under the Continuity conditions:

• Velocity and stress — stress continuity is invoked;

• Velocity and porosity-corrected stress (default)— the sum of stress and a contribution
due to the inertial term being constant across the interface;
• Velocity and porosity-corrected stress with loss — the sum of the stress and a
contribution due to the inertial term is lost in the downstream direction.

All three options give continuity of the velocity across the interface in constant-density
flows, while in case of a jump in density:

1 give continuity of the mass flux (product of density and normal velocity) and
tangential velocity;
2 account for the density-jump correction in the stress condition.

When Velocity and porosity-corrected stress with loss is chosen, a Pressure loss model
needs to be specified. The options are:

• User-defined loss coefficients (default). Enter values for Expansion loss coefficient and
Contraction loss coefficient;
• Borda–Carnot. Specify Vena contracta model.

Vena contracta model allows two options:

• User defined (default) — enter value of Vena contracta coefficient;

• Weisbach — enter values of Lower limit and Exponent.

Notice that the notions ‘expansion’ and ‘contraction’ used in this section correspond
not to geometrical changes, but to the variations of the microscopic structure across
the interface in the flow direction — increasing or decreasing porosity, p, respectively.

Flow Continuity
The Flow Continuity condition is suitable for pairs where the boundaries match; it
prescribes that the flow field is continuous across the pair.

454 | CHAPTER 6: FLUID FLOW INTERFACES

The parts of the boundary where a source boundary lacks a corresponding destination
boundary and vice versa will get conditions from other boundary conditions under the
physics interface acting as fallback boundary conditions on the nonoverlapping parts.
By default, the Wall node is used as fallback unless another feature has been selected
for the [Link] the condition is set for pairs adjacent to domains with different
porosity p, it imposes conditions identical to the Porous Interface feature, which is
designed to properly account for a sudden jump in porosity. See the section Porous
Interface above for further details (note that the Flow Continuity does not account for
the density-jump corrections),

About Pairs

Pressure Point Constraint

The Pressure Point Constraint condition can be used to specify the pressure level. If it
is not possible to specify the pressure level using a boundary condition, the pressure
level must be set in some other way, for example, by specifying a fixed pressure at a
point.

PRESSURE CONSTRAINT
The relative pressure value is set by specifying the Pressure p0. If the reference pressure
pref defined at the physics interface level is equal to zero, p0 represents the absolute
pressure.

When Include gravity is selected and Use reduced pressure not selected in the interface
Physical model section, the Compensate for hydrostatic pressure approximation (named
Compensate for hydrostatic pressure for incompressible flows) checkbox is available and
selected by default. When it is selected, the hydrostatic pressure is automatically added
to the pressure entered in p0 user input.

CONSTRAINT SETTINGS
This section is displayed by clicking the Show button ( ) and selecting Advanced
Physics Options.

The Apply reaction terms on setting is set per default to Individual dependent variables.
All physics (symmetric) cannot be used together with a segregated solver when fluid
flow is coupled with Moving Mesh, Level Set, or Phase Field in Fluids.

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 455

 Point Mass Source
This feature requires at least one of the following licenses: Battery Design Module,
CFD Module, Chemical Reaction Engineering Module, Corrosion Module,
Electrochemistry Module, Electrodeposition Module, Fuel Cell & Electrolyzer
Module, Microfluidics Module, Pipe Flow Module, or Subsurface Flow Module.

The Point Mass Source feature models mass flow originating from an infinitely small
domain centered around a point.

For the Reacting Flow in Porous Media, Diluted Species interface, which is available
with the CFD Module, Chemical Reaction Engineering Module, or Battery Design
Module, there are two nodes: one for the fluid flow (Fluid Point Source) and one for
the species (Species Point Source).

SOURCE STRENGTH
The source Mass flux, qꞏ p should be specified. A positive value results in mass being
ejected from the point into the computational domain. A negative value results in mass
being removed from the computational domain.

Point sources located on a boundary or on an edge affect the adjacent computational

domains. This has the effect, for example, that a point source located on a symmetry
plane has twice the given strength.

Mass Sources for Fluid Flow in the COMSOL Multiphysics Reference

Manual

Line Mass Source

This feature requires at least one of the following licenses: Battery Design Module,
CFD Module, Chemical Reaction Engineering Module, Corrosion Module,
Electrochemistry Module, Electrodeposition Module, Fuel Cell & Electrolyzer
Module, Microfluidics Module, Pipe Flow Module, or Subsurface Flow Module.

The Line Mass Source feature models mass flow originating from a tube region with
infinitely small radius.

For the Reacting Flow in Porous Media, Diluted Species interface, which is available
with the CFD Module, Chemical Reaction Engineering Module, or Battery Design
Module, there are two nodes, one for the fluid flow (Fluid Line Source) and one for the
species (Species Line Source).

456 | CHAPTER 6: FLUID FLOW INTERFACES

SELECTION
The Line Mass Source feature is available for all dimensions, but the applicable selection
differs between the dimensions.

MODEL DIMENSION APPLICABLE GEOMETRICAL ENTITY

2D Points
2D Axisymmetry Points not on the axis of symmetry
3D Edges

SOURCE STRENGTH
The source Mass flux, qꞏ l , should be specified. A positive value results in mass being
ejected from the line into the computational domain and a negative value means that
mass is removed from the computational domain.

Line sources located on a boundary affect the adjacent computational domains. This,
for example, has the effect that a line source located on a symmetry plane has twice the
given strength.

Mass Sources for Fluid Flow in the COMSOL Multiphysics Reference

Manual

Gravity
The Gravity global feature is automatically added when Include gravity is selected at the
interface level in the Physical Model settings. It defines the gravity forces from the
Acceleration of gravity value.

ACCELERATION OF GRAVITY
The Acceleration of gravity (SI unit m/s, default value gconstez in 2D axial symmetry
and 3D and gconstey in 2D) is used to define the gravity forces. It should be a global
quantity.

THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 457

 The Darcy’s Law Interface
The Darcy’s Law (dl) interface ( ), found under the Porous Media and Subsurface Flow
branch ( ) when adding a physics interface, is used to simulate fluid flow through
interstices in a porous medium. It can be used to model low-velocity flows or media
where the permeability and porosity are very small, and for which the pressure gradient
is the major driving force and the flow is mostly influenced by the frictional resistance
within the pores. Set up multiple Darcy’s Law interfaces to model multiphase flows
involving more than one mobile phase. The Darcy’s Law interface can be used for
stationary and time-dependent analyses.

The main feature is the Porous Medium Properties node, which provides interfaces for
defining the fluid material and the porous matrix properties.

When this physics interface is added, the following default nodes are also added in the
Model Builder — Porous Medium, No Flow (the default boundary condition), and Initial
Values. Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions and mass sources. You can also right-click Darcy’s Law to select
physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is dl.

PHYSICAL MODEL
Enter a Reference pressure level pref (SI unit: Pa). The default value is 1[atm].

GRAVITY EFFECTS
By default there are no gravity effects added to Darcy’s Law interface. Select the
checkbox Include gravity to activate the acceleration of gravity. When this checkbox is
selected, a global Gravity feature node is shown in the interface model tree.

458 | CHAPTER 6: FLUID FLOW INTERFACES

Enter a value for the acceleration of gravity. The default value g_const is the
predefined standard acceleration of gravity on earth.

For a detailed overview of the functionality available in each product, visit

[Link]

DISCRETIZATION
To display all settings available in this section, click the Show More Options button ( )
and select Advanced Physics Options in the Show More Options dialog.

You can choose the order of the shape functions used for the pressure variable solved
by the Darcy’s Law interface. The default shape functions are Quadratic Lagrange.

The Compute boundary fluxes checkbox is not activated by default. When this option is
selected, the solver computes variables storing accurate boundary fluxes from each
boundary into the adjacent domain.

If the checkbox is cleared, the COMSOL Multiphysics software instead computes the
flux variables from the dependent variables using extrapolation, which is less accurate
in postprocessing results but does not create extra dependent variables on the
boundaries for the fluxes.

Also, the Apply smoothing to boundary fluxes checkbox is available if the previous
checkbox is checked. The smoothing can provide a better behaved flux value close to
singularities.

For details about the boundary fluxes settings, see Computing Accurate Fluxes in the
COMSOL Multiphysics Reference Manual.

The Value type when using splitting of complex variables setting should in most pure
mass transport problems be set to Real which is the default. It makes sure that the
dependent variable does not get affected by small imaginary contributions, which can
occur, for example, when combining a Time Dependent or Stationary study with a
frequency-domain study. For more information, see Splitting Complex-Valued
Variables in the COMSOL Multiphysics Reference Manual.

THE DARCY’S LAW INTERFACE | 459

 DEPENDENT VARIABLES
The dependent variable (field variable) is the Pressure. The name can be changed but
the names of fields and dependent variables must be unique within a model.

• Domain, Boundary, Edge, Point, and Pair Nodes for the Darcy’s Law
Interface
• Theory for the Darcy’s Law Interface
• Physical Constants in the COMSOL Multiphysics Reference Manual

• Species Transport in the Gas Diffusion Layers of a PEM: Application

Library path Fuel_Cell_and_Electrolyzer_Module/Fuel_Cells/
pem_gdl_species_transport_2d
• Steam Reformer: Application Library path
Chemical_Reaction_Engineering_Module/Reactors_with_Porous_Catalysts/
steam_reformer
• Terzaghi Compaction: Application Library path
Subsurface_Flow_Module/Poromechanics/terzaghi_compaction

Domain, Boundary, Edge, Point, and Pair Nodes for the Darcy’s Law
Interface
The Darcy’s Law Interface has the following domain, boundary, edge, point, and pair
nodes, these nodes are available from the Physics ribbon toolbar (Windows users),
Physics context menu (Mac or Linux users), or right-click to access the context menu
(all users).

460 | CHAPTER 6: FLUID FLOW INTERFACES

DOMAIN

• Porous Medium • Mass Source

• Unsaturated Porous Medium • Cross Section
• Dual Porosity Medium • Initial Values
• Dual Permeability Medium • Unsaturated Porous Medium
• Unsaturated Dual Permeability • ThicknessContributing Velocity
Medium • Porous Electrode Coupling

BOUNDARY, EDGE, AND POINT

The following nodes (listed in alphabetical order) are available on exterior boundaries:

• Atmosphere/Gauge • Outlet
• Electrode Surface Coupling • Pervious Layer
• Flux Discontinuity • Point Mass Source
• Fracture • Precipitation
• Hydraulic Head • Pressure
• Inlet • Pressure Head
• Interior Wall • Reaction Coefficients1
• Line Mass Source • Symmetry
• Mass Flux • Thin Barrier
• No Flow
1
This node is described for the Transport of Diluted Species Interface

The relevant physics interface condition at interior boundaries is continuity:

n   1 u1 –2 u2  = 0

The continuity boundary condition ensures that the pressure and mass flux are
continuous. In addition, the Pressure boundary condition is available on interior
boundaries.

THE DARCY’S LAW INTERFACE | 461

 The continuity boundary condition provides continuity for the pressure and mass flux.
In addition, the following boundary conditions are available on interior boundaries:

• Pressure • Pervious Layer

• Pressure Head • Flux Discontinuity
• Hydraulic Head • Fracture
• Interior Wall • Thin Barrier

Note that some of the above-mentioned domain or boundary conditions

are only available with some products. For a detailed overview of the
functionality available in each product, visit [Link]
products/specifications/

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

For axisymmetric components, COMSOL Multiphysics takes the axial

symmetry boundaries (at r = 0) into account and automatically adds an
Axial Symmetry node that is valid on the axial symmetry boundaries only.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Porous Medium
The Porous Medium node adds the equations for modeling flow through porous
domains. By default these are Equation 6-3 and Equation 6-4 (excluding any mass
sources).

     +    u  = Q
(6-3)
t p m

462 | CHAPTER 6: FLUID FLOW INTERFACES

 
u = – --- p (6-4)


For a steady-state problem the first term in Equation 6-3 disappears. When Mass
Source is active Equation 6-4 is of the following form:


u = – ---  p – g  (6-5)


The Porous Medium feature has two default subfeatures — the Fluid subfeature where
the fluid density  (SI unit: kg/m3) and dynamic viscosity  (SI unit: Pa·s) are defined
and the Porous Matrix subfeature where the permeability  (SI unit: m2) and porosity
p (dimensionless) are specified.

COORDINATE SYSTEM SELECTION

Select a coordinate system from the Coordinate system list for the interpretation of
directions in anisotropic material properties. The default is the Global coordinate
system, and the list contains any additional orthonormal coordinate system (except
boundary coordinate systems) added under the Definitions node.

The subnodes inherit these coordinate system settings. In particular, the Permeability
or Hydraulic conductivity (in the Porous Matrix subnode) should be set according to
the coordinate system selected in this section.

FLOW MODEL
Define if the flow follows a linear or nonlinear pressure-velocity relationship.

• Darcian flow (default) defines a linear relationship using Equation 6-3 and
Equation 6-4.
• Non-Darcian flow defines a nonlinear relationship using Equation 6-3 and


p = – --- u –  u u


where the nonlinear parameter  (SI unit: 1/m) is further specified in the Porous
Matrix subfeature.

STORAGE MODEL
Define the time dependent storage term by selecting one of the following options from
the list:

• From density and porosity (default), uses the formulation of Equation 6-3

THE DARCY’S LAW INTERFACE | 463

 • Linearized storage, where the following linear equation is used to define the storage:
S p =  p  f +  1 –  p  p

• Poroelastic storage, where Sppf

• Quasistatic, where the storage term is set to zero
• User defined to directly specify the storage coefficient Sp

• Porous Material in the COMSOL Multiphysics Reference Manual

• About Darcian and Non-Darcian Flow in the Porous Media Flow
Module User’s Guide
• Storage Models in the Subsurface Flow Module User’s Guide

Note that some options are only available in some modules. For a detailed
overview of the functionality available in each product, visit https://
[Link]/products/specifications/

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Porous Medium
Richards’ Equation > Porous Medium

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equation selected:

Domains > interface > Porous Medium

Fluid
This node defines the material properties of the fluid of the Porous Medium, the
Unsaturated Porous Medium, the Dual Porosity Medium, the Dual Permeability
Medium, or the Unsaturated Porous Medium parent node.

MODEL INPUT
This section contains fields and values that are inputs for expressions defining material
properties. If such user-defined property groups are added, the model inputs appear
here.

464 | CHAPTER 6: FLUID FLOW INTERFACES

FLUID PROPERTIES
The available options for the Fluid type are Incompressible, Compressible, Compressible,
linearized, or Ideal gas. The selection determines how the density and compressibility
of the fluid are set.

Incompressible
The default Reference density ref uses values From material. For User defined enter
another value or expression.

Compressible
The default Density  uses values From material. For User defined enter another value or
expression as a function of the absolute pressure. The compressibility of the fluid is
computed from it.

Compressible, Linearized
The default Reference density ref and the Fluid compressibility fuse values From
material. For User defined enter another value or expression. The fluid density increases
linearly as a function of the fluid pressure.

Ideal Gas
For Ideal gas it uses the ideal gas law to describe the gas density. In this case, specify
the thermodynamics properties. Select a Gas constant type — Specific gas constant Rs
(the default) or Mean molar mass Mn (SI unit: J/(mol·K)). For Mean molar mass the
universal gas constant R  8.314 J/(mol·K) is used as the built-in physical constant.
For both properties, the defaults use values From material. For User defined enter
another value or expression.

Dynamic Viscosity
For all fluid types specify the Dynamic viscosity  (SI unit: Pa·s) uses values From
material. For User defined enter another value or expression.

Constitutive Relation
The Polymer Flow Module allows the use of inelastic non-Newtonian models. For
Darcian Flow, you can choose between two options: Specify dynamic viscosity (default)
or Inelastic non-Newtonian. Use the default option to either get the dynamic viscosity
from a material, or to add a user defined expression. The default option lets you either
use the dynamic viscosity from a material or define it with a custom expression. The
Inelastic non-Newtonian option provides several models and defines the Apparent shear
rate.

THE DARCY’S LAW INTERFACE | 465

 Inelastic Non-Newtonian
The following inelastic non-Newtonian Models are available for the Fluid feature under
the Porous Medium feature:

• Power Law
• Carreau
• Carreau-Yasuda
• Cross
• Cross-Williamson
• Sisko

The setup of these models in porous media is similar to their definition under the Fluid
Properties feature.

Apparent Shear Rate

ꞏ
To define the apparent shear rate  app , either enter a user-defined expression for the
correction factor () or choose the Capillary bundle approach from the list and provide
a value for the Tortuosity factor (C).

• Read more about the theory of Non-Newtonian Fluids in Porous

Media in the Polymer Flow Module User’s Guide

LOCATION IN USER INTERFACE

A default Fluid node is automatically added to the Porous Medium, the Unsaturated
Porous Medium, the Dual Porosity Medium, the Dual Permeability Medium, and the
Unsaturated Dual Permeability Medium node.

Context Menus
Darcy’s Law > porous medium feature > Fluid
Richards’ Equation > porous medium feature > Fluid

Porous Matrix
This node defines the porosity and material properties of the solid matrix of the Porous
Medium or Unsaturated Porous Medium parent node.

466 | CHAPTER 6: FLUID FLOW INTERFACES

MODEL INPUT
This section contains fields and values that are inputs for expressions defining material
properties. If such user-defined property groups are added, the model inputs appear
here.

MATRIX PROPERTIES
The default Porosity p of the solid matrix is taken From material. In a Porous Material
node, the porosity is defined as follows:

p = 1 –  si – imfi
i i

where si and imfi are the porosities of the Solid and Immobile Fluids subnodes under
the Porous Material node. See Porous Material in the COMSOL Multiphysics
Reference Manual for more information.

Select User defined to enter a value or expression for p.

If required by a Storage model in the parent node, specify the Effective compressibility
of the porous matrix. The default value is p  104 1/Pa.

Select a Permeability model to specify the capacity of the porous material to transmit
flow. Which options are available depends on the Flow model setting in the parent node.
Note that for the Unsaturated Porous Medium parent node only the Darcian Flow
models are available.

Darcian Flow
If Darcian flow is selected in the parent Porous medium node, select between the
following options for the Permeability model:

• Select Permeability (default) to define the permeability of the porous matrix. The
default Permeability  (SI unit: m2) uses the value From material, as defined by the
Porous material. For User defined select Isotropic to define a scalar value or Diagonal
or Symmetric to define a tensor value.
• Select Hydraulic conductivity or define a combination of fluid permeability and
dynamic viscosity. For Hydraulic conductivity K (SI unit: m/s) select Isotropic to
define a scalar value or Diagonal or Symmetric to define a tensor value. The default is
2.94·104 m/s.
• Select Kozeny-Carman to define the permeability from the porosity and mean particle
diameter. Enter the mean Particle diameter dp (SI unit: m), the default value is
0.5 mm.

THE DARCY’S LAW INTERFACE | 467

 Non-Darcian Flow
If Non-Darcian flow is selected in the parent node, select between the following options
for the Permeability model:

• For Forchheimer, the default Permeability  (SI unit: m2) uses the value From
material. Furthermore, the dimensionless Forchheimer parameter cF can be defined.
The default value is 0.55.
• If Ergun is selected, enter the mean Particle diameter dp (default value: 0.5 mm).
• For Burke–Plummer, specify the mean Particle diameter dp (default value: 0.5 mm).
• For Klinkenberg, the default Permeability  (SI unit: m2) uses the value From
material. Enter the Klinkenberg parameter bK. Its default value is 103 Pa.

• About Darcian and Non-Darcian Flow in the Porous Media Flow

Module User’s Guide
• Permeability Models in the Porous Media Flow Module User’s Guide

Note that some options are only available in some modules. For a detailed
overview of the functionality available in each product, visit https://
[Link]/products/specifications/

RETENTION MODEL
For an Unsaturated Porous Medium select between the following options for the
retention model:

• van Genuchten, to specify the retention model using the van Genuchten relationship.
Enter the constitutive relation constants  (the default value is 1, and the SI unit is
1/m), n (default value is 2), and l (default value is 0.5). The constitutive parameter
m is equal to 1  1n.
• Brooks and Corey to use the retention model according to the Brooks and Corey
relationships.
• User defined to use a different retention model. Enter user defined expressions for
the Effective saturation Se, the Liquid volume fraction l. The default is p (the
porosity variable). Specify the Specific moisture capacity Cm (SI unit: 1/m). The
default value is 0 (1/m). Enter an expression for the Relative permeability r. The
default value is 1.

468 | CHAPTER 6: FLUID FLOW INTERFACES

If there is a residual volume of liquid that cannot move through the pore network, also
enter a value between 0 and 1 for the Residual liquid volume fraction r. The default
value is 0.

Retention and Permeability Relationships in the Subsurface Flow Module

User’s Guide

LOCATION IN USER INTERFACE

A default Porous Matrix node is automatically added to Porous Medium or the
Unsaturated Porous Medium node.

Context Menus
Darcy’s Law > porous medium feature > Fluid
Richards’ Equation > porous medium feature > Fluid

Unsaturated Porous Medium

The Unsaturated Porous Medium node adds Richards’ equation (Equation 6-6) for flow
in variably saturated porous media.

Cm p r 
- + S e S p ------ +     – --------  p +  g D  = Q m
  ------- (6-6)
 g  t   

The Unsaturated Porous Medium feature has two default subfeatures: the Fluid
subfeature, where the fluid density  (SI unit: kg/m3) and viscosity  (SI unit: Pa·s)
are defined, and the Porous Matrix subfeature, where the permeability  (SI unit: m2)
and porosity p (dimensionless) are specified as well as the retention model.

COORDINATE SYSTEM SELECTION

Select a coordinate system from the Coordinate system list for the interpretation of
directions in anisotropic material properties. The default is the Global coordinate
system, and the list contains any additional orthonormal coordinate system (except
boundary coordinate systems) added under the Definitions node.

The subnodes inherit these coordinate system settings. In particular, the Permeability
or Hydraulic conductivity (in the Porous Matrix subnode) should be set according to
the coordinate system selected in this section.

THE DARCY’S LAW INTERFACE | 469

 STORAGE MODEL
With this option, define the first term in Equation 6-6. Therefore, choose from a list
one of the following formulations:

• From retention model

• Linearized storage (default)
• From liquid content
• User defined

If From retention model is chosen, the first term of Equation 6-6 accounts only for the
specific moisture capacity Cm which is defined by the Retention Model in the Porous
Matrix subnode:

C m p
-------- .
g t

For the option Linearized storage the storage coefficient Sp in Equation 6-6 is defined
as

S p =  p  f +  1 –  p  p .

From liquid content enforces the first term of Equation 6-6 to be defined in terms of
liquid volume fraction as

  l 
----------------
t

and the User defined option allows to define the storage coefficient Sp in Equation 6-6
directly.

Note that some options are only available in some modules. For a detailed
overview of the functionality available in each product, visit https://
[Link]/products/specifications/

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Unsaturated Porous Medium
Richards’ Equation > Unsaturated Porous Medium

470 | CHAPTER 6: FLUID FLOW INTERFACES

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equation selected:

Domains > interface > Unsaturated Porous Medium

Dual Porosity Medium

Use the Dual Porosity Medium if the flow through macropores is governed by Darcy’s
law and the micropores do not significantly contribute to the overall flow in the porous
medium.

The feature has three default subfeatures —the Fluid subfeature where the fluid density
 (SI unit: kg/m3) and dynamic viscosity  (SI unit: Pa·s) are defined, the Macropores
subfeature where the hydraulic properties of the macroporous matrix are defined, and
the Micropores where the hydraulic properties of the microporous matrix are defined.

COORDINATE SYSTEM SELECTION

Select a coordinate system from the Coordinate system list for the interpretation of
directions in anisotropic material properties. The default is the Global coordinate
system, and the list contains any additional orthonormal coordinate system (except
boundary coordinate systems) added under the Definitions node.

The subnodes inherit these coordinate system settings. In particular, the Permeability
or Hydraulic conductivity (in the Macropores and Micropores subnodes) should be set
according to the coordinate system selected in this section.

STORAGE MODEL
Choose from a list one of the following formulations:

• From density and porosity (default), uses the formulation equivalent to Equation 6-3
• Linearized storage, where the following linear equation is used to define the storage
in the macropores as Sp  pf  1  pp and in the micropores as
Sp,m  p,mf  1  p,mp.
• Poroelastic storage, where Sp = pf in the macropores and Sp,m = p,mf in the
micropores.

THE DARCY’S LAW INTERFACE | 471

 • Quasistatic, where the storage term is set to zero
• User defined to directly specify the storage coefficient Sp and Sp,m for the macro-
and micropores respectively.

• Storage Models in the Subsurface Flow Module User’s Guide

INTERPOROSITY FLOW
Specify the Fluid Transfer Function w which is used to calculate the interporosity flow,
the mass exchange between macro- and microscale system:

Q ip =  w  p M – p m  (6-7)

with pM being the pressure within the macropores and pm the pressure within the
micropores.

Find more information in the Subsurface Flow Module User’s Guide

• Dual Porosity & Dual Permeability Modeling

• Dual Porosity and Dual Permeability Theory

Note that some options are only available in some modules. For a detailed
overview of the functionality available in each product, visit https://
[Link]/products/specifications/

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Dual Porosity/Permeability > Dual Porosity Medium
Richards’ Equation > Dual Porosity/Permeability > Dual Porosity Medium

Ribbon
Physics tab with Darcy’s Law or Richards’ Equation selected:

Domains > Dual Porosity/Permeability > Dual Porosity Medium

472 | CHAPTER 6: FLUID FLOW INTERFACES

Dual Permeability Medium
Use the Dual Permeability Medium feature to model transport in a saturated dual
permeability medium where Darcy’s Law describes the transport in each system.

The feature has three default subfeatures —the Fluid subfeature where the fluid density
 (SI unit: kg/m3) and dynamic viscosity  (SI unit: Pa·s) are defined, the Macropores
subfeature where the hydraulic properties of the macroporous matrix are defined, and
the Micropores where the hydraulic properties of the microporous matrix are defined.

COORDINATE SYSTEM SELECTION

Select a coordinate system from the Coordinate system list for the interpretation of
directions in anisotropic material properties. The default is the Global coordinate
system, and the list contains any additional orthonormal coordinate system (except
boundary coordinate systems) added under the Definitions node.

The subnodes inherit these coordinate system settings. In particular, the Permeability
or Hydraulic conductivity (in the Macropores and Micropores subnodes) should be set
according to the coordinate system selected in this section.

STORAGE MODEL
With this option under the Porous Medium section, Equation 6-3 is formulated with
respect to a storage term (analogous to Equation 4-4). Therefore, choose from a list
one of the following formulations:

• From density and porosity (default), uses the formulation of Equation 4-4
• Linearized storage, where the following linear equation is used to define the storage:
S p,i =  p,i  f +  1 –  p,i  p,i

with i=M for the macropores and i=m for the micropores.

• Poroelastic storage, where Sp,i = p,if

• Quasistatic, where the storage term is set to zero
• User defined to directly specify the storage coefficient Sp

• Storage Models in the Subsurface Flow Module User’s Guide

THE DARCY’S LAW INTERFACE | 473

 INTERPOROSITY FLOW
Specify the Fluid Transfer Function w which is used to calculate the interporosity flow,
which is the mass exchange between macro- and microscale system:

Q ip =  w  p M – p m  (6-8)

with pM being the pressure in the macropores and pm the pressure in the micropores.

Find more information in the Subsurface Flow Module User’s Guide

• Dual Porosity & Dual Permeability Modeling

• Dual Porosity and Dual Permeability Theory

Note that some options are only available in some modules. For a detailed
overview of the functionality available in each product, visit https://
[Link]/products/specifications/

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Dual Porosity/Permeability > Dual Permeability Medium
Richards’ Equation > Dual Porosity/Permeability > Dual Permeability Medium

Ribbon
Physics tab with Darcy’s Law or Richards’ Equation selected:

Domains > Dual Porosity/Permeability > Dual Permeability Medium

Unsaturated Dual Permeability Medium

Use the Unsaturated Dual Permeability Medium feature to model the transport in an
unsaturated dual permeability medium where Richards’ Equation (Equation 6-6)
describes the transport in each system.

C m,M p M
 M    S e,M S p,M + -------------
- ----------- +  u M  = – Q ip +  M Q m
  g  t 
C m,m p m (6-9)
 1 –  M     S e,m S p,m + ------------
- ---------- +  u m  = Q ip +  1 –  M Q m
  g  t 
p =  M p M +  1 –  M p m

474 | CHAPTER 6: FLUID FLOW INTERFACES

M is the volume fraction of the macropores and Qip is the interporosity flow, which is
the mass transfer from the macroscale to the microscale system. The indices M and m
stand for macropores and micropores, respectively.

The feature has three default subfeatures —the Fluid subfeature where the fluid density
 (SI unit: kg/m3) and dynamic viscosity  (SI unit: Pa·s) are defined, the Macropores
subfeature where the hydraulic properties of the macroporous matrix are defined, and
the Micropores where the hydraulic properties of the microporous matrix are defined.

COORDINATE SYSTEM SELECTION

Select a coordinate system from the Coordinate system list for the interpretation of
directions in anisotropic material properties. The default is the Global coordinate
system, and the list contains any additional orthonormal coordinate system (except
boundary coordinate systems) added under the Definitions node.

The subnodes inherit these coordinate system settings. In particular, the Permeability
or Hydraulic conductivity (in the Macropores and Micropores subnodes) should be set
according to the coordinate system selected in this section.

STORAGE MODEL
With this option, define the first term in Equation 6-9. Therefore, choose from a list
one of the following formulations:

• From retention model

• Linearized storage (default)
• From liquid content
• User defined

If From retention model is chosen, the first term of Equation 6-9 accounts only for the
specific moisture capacity Cm,i which is defined by the Retention Model in the
Macropores (i=M) or Micropores (i=m) subnode:

C m i p
----------- .
g t

For the option Linearized storage the storage coefficient Sp,i in Equation 6-9 is defined
as

S p,i =  p,i  f +  1 –  p,i  p,i .

THE DARCY’S LAW INTERFACE | 475

 From liquid content enforces the first term of Equation 6-9 to be defined in terms of
liquid volume fraction as

  l,i 
------------------
t

• and the User defined option allows to define the storage coefficient Sp,i in
Equation 6-9 directly.

• Storage Models in the Subsurface Flow Module User’s Guide

INTERPOROSITY FLOW
Specify the Fluid Transfer Function w which is used to calculate the interporosity flow,
which is the mass exchange between macro- and microscale system:

Q ip =  w  p M – p m  (6-10)

with pM being the pressure within the macroscale system and pm the pressure within
the microscale system.

• Dual Porosity & Dual Permeability Modeling in the Subsurface Flow

Module User’s Guide

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Dual Porosity/Permeability > Unsaturated Dual Permeability Medium
Richards’ Equation > Dual Porosity/Permeability > Unsaturated Dual Permeability Medium

Ribbon
Physics tab with Darcy’s Law or Richards’ Equation selected:

Domains > Dual Porosity/Permeability > Unsaturated Dual Permeability Medium

Macropores
This node defines the porosity and material properties of the macroporous matrix of
the Dual Porosity Medium, the Dual Permeability Medium, or the Unsaturated Dual
Permeability Medium parent node.

476 | CHAPTER 6: FLUID FLOW INTERFACES

MODEL INPUT
This section contains fields and values that are inputs for expressions defining material
properties. If such user-defined property groups are added, the model inputs appear
here.

VOLUME FRACTION
Here enter the Volume Fraction of Macropores, M.

MATRIX PROPERTIES
The default Porosity p,M of the solid matrix of the macropore system is taken From
material. Select User defined to enter a value or expression for p,M.

If required by a Storage model in the parent node, specify the Effective compressibility
of the porous matrix of the macropore system. The default value is p,M  104 1/Pa.

Select a Permeability model to specify the capacity of the porous material to transmit
flow. For the macropore settings only the Darcian Flow models are available. Select
between the following options for the Permeability model:

• Select Permeability (default) to define the permeability of the porous matrix. The
default Permeability M (SI unit: m2) uses the value From material, as defined by the
Porous material. For User defined select Isotropic to define a scalar value or Diagonal
or Symmetric to define a tensor value.
• Select Hydraulic conductivity or define a combination of fluid permeability and
dynamic viscosity. For Hydraulic conductivity KM (SI unit: m/s) select Isotropic to
define a scalar value or Diagonal or Symmetric to define a tensor value. The default is
2.94·104 m/s.
• Select Kozeny-Carman to define the permeability from the porosity and mean particle
diameter. Enter the mean Particle diameter dp,M (SI unit: m), the default value is
0.5 mm.

RETENTION MODEL
For an Unsaturated Dual Permeability Medium select between the following options
for the retention model:

• van Genuchten, to specify the retention model using the van Genuchten relationship.
Enter the constitutive relation constants  (the default value is 1, and the SI unit is
1/m), n (default value is 2), and l (default value is 0.5). The constitutive parameter
m is equal to 1  1n.

THE DARCY’S LAW INTERFACE | 477

 • Brooks and Corey to use the retention model according to the Brooks and Corey
relationships.
• User defined to use a different retention model. Enter user defined expressions for
the Effective saturation Se,M, the Liquid volume fraction l,M. The default is p,M (the
porosity variable). Specify the Specific moisture capacity Cm,M (SI unit: 1/m). The
default value is 0 (1/m). Enter an expression for the Relative permeability r,M. The
default value is 1.

If there is a residual volume of liquid that cannot move through the pore network, also
enter a value between 0 and 1 for the Residual liquid volume fraction r. The default
value is 0.

Retention and Permeability Relationships in the Subsurface Flow Module

User’s Guide

LOCATION IN USER INTERFACE

A default Macropores node is automatically added to the Dual Porosity Medium, the
Dual Permeability Medium, and the Unsaturated Dual Permeability Medium node.

Context Menus
Darcy’s Law > porous medium feature > Macropores
Richards’ Equation > porous medium feature > Macropores

Micropores
This node defines the porosity and material properties of the microporous matrix of
the Dual Porosity Medium, Dual Permeability Medium, or the Unsaturated Dual
Permeability Medium parent node.

MODEL INPUT
This section is only available for the Unsaturated Dual Permeability Medium parent
node and contains fields and values that are inputs for expressions defining material
properties. If such user-defined property groups are added, the model inputs appear
here.

MATRIX PROPERTIES
The default Porosity p,m of the solid matrix of the micropore system is taken From
material. Select User defined to enter a value or expression for p,m.

478 | CHAPTER 6: FLUID FLOW INTERFACES

If required by a Storage model in the parent node, specify the Effective compressibility
of the microscale porous matrix. The default value is p,m  104 1/Pa.

Select a Permeability model to specify the capacity to transmit flow for the porous
material of the microscale system. Here, only the Darcian Flow models are available.
Select between the following options for the Permeability model:

• Select Permeability (default) to define the permeability of the porous matrix. The
default Permeability m (SI unit: m2) uses the value From material, as defined by the
Porous material. For User defined select Isotropic to define a scalar value or Diagonal
or Symmetric to define a tensor value.
• Select Hydraulic conductivity or define a combination of fluid permeability and
dynamic viscosity. For Hydraulic conductivity Km (SI unit: m/s) select Isotropic to
define a scalar value or Diagonal or Symmetric to define a tensor value. The default is
2.94·104 m/s.
• Select Kozeny-Carman to define the permeability from the porosity and mean particle
diameter. Enter the mean Particle diameter dp,m (SI unit: m), the default value is
0.5 mm.

RETENTION MODEL
For an Unsaturated Dual Permeability Medium select between the following options
for the retention model:

• From macropores to take over the retention model settings from the Macropores
node.
• van Genuchten, to specify the retention model using the van Genuchten relationship.
Enter the constitutive relation constants  (the default value is 1, and the SI unit is
1/m), n (default value is 2), and l (default value is 0.5). The constitutive parameter
m is equal to 1  1n.
• Brooks and Corey to use the retention model according to the Brooks and Corey
relationships.
• User defined to use a different retention model. Enter user defined expressions for
the Effective saturation Se,m, the Liquid volume fraction l,m. The default is p,m (the
porosity variable). Specify the Specific moisture capacity Cm,m (SI unit: 1/m). The
default value is 0 (1/m). Enter an expression for the Relative permeability r,m. The
default value is 1.

THE DARCY’S LAW INTERFACE | 479

 If there is a residual volume of liquid that cannot move through the pore network, also
enter a value between 0 and 1 for the Residual liquid volume fraction r. The default
value is 0.

Retention and Permeability Relationships in the Subsurface Flow Module

User’s Guide

INITIAL VALUES
Define the initial pressure of the microscale system. The default is to use the initial
values defined in the Initial Values feature node which means that the same initial
values apply to the macroscale and the microscale system.

Select User Defined to specify an initial pressure for the microscale system pm0,init.

Note that some options are only available in some modules. For a detailed
overview of the functionality available in each product, visit https://
[Link]/products/specifications/

LOCATION IN USER INTERFACE

A default Micropores node is automatically added to the Dual Porosity Medium, the
Dual Permeability Medium, and the Unsaturated Dual Permeability Medium node.

Context Menus
Darcy’s Law > porous medium feature > Micropores
Richards’ Equation > porous medium feature > Micropores

Mass Source
The Mass Source node adds a mass source Qm, which appears on the right-hand side of
the Darcy’s Law equation (Equation 6-3).

    +    u  = Q
(6-11)
t m

For the Fracture Flow interface, the mass source is applicable to the right-hand side of
Equation 4-40 in the Subsurface Flow Module User’s Guide (where it is multiplied
with the fracture thickness).

480 | CHAPTER 6: FLUID FLOW INTERFACES

MASS SOURCE
Enter a value or expression for the Mass source Qm (SI unit: kg/(m3·s)). The default
is 0 kg/(m3·s).

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Mass Source
Richards’ Equation > Mass Source

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equation selected:

Domains > interface > Mass Source

Contributing Velocity
The Contributing Velocity node adds a velocity contribution uctrb to the velocity field
ud that is computed from the flow model selected in the parent Porous Medium,
Unsaturated Porous Medium, or Dual Porosity Medium feature. The total velocity
field used in the continuity equation is therefore

u = u d + u ctrb (6-12)

Such velocity contributions can arise in multiphysics models that involve, for example,
transport of solutes. If several Contributing Velocity nodes are added under the same
parent feature, the contributions are treated additively. The sum of all contributing
velocity fields is available in the variables [Link] (material frame) and
[Link] (spatial frame), with name being the Name of the physics interface (for
example, dl).

COORDINATE SYSTEM SELECTION

Select a coordinate system from the Coordinate system list to specify the coordinate
system in which the contributing velocity field is given. The default is the Global
coordinate system. Note that the coordinate system used to specify the components of
the contributing velocity field can be selected independently of the coordinate system
of the parent feature.

THE DARCY’S LAW INTERFACE | 481

 CONTRIBUTING VELOCITY
The default contributing velocity field uctrb is User defined, which lets you enter the
components of the contributing velocity field explicitly. If a matching velocity field is
announced by another feature, it will be available in the selection list.

ADVANCED
Select the Frame in which the contributing velocity field is specified. The default is the
Reference configuration, which means that the velocity field follows the material under,
for example, a rotation of the domain. If instead your contributing velocity field
follows a fixed direction in space, select the Deformed configuration. This section is only
displayed if Advanced Physics Options is activated and the spatial and material frame of
the component differ, such as in the presence of a Moving Mesh or Solid Mechanics
interface.

LOCATION IN USER INTERFACE

Context Menus
If the selected porous medium feature is Porous Medium, Unsaturated Porous
Medium, or Dual Porosity Medium:

Darcy’s Law > porous medium feature > Contributing Velocity

Richards’ Equation > porous medium feature > Contributing Velocity

Ribbon
Physics tab with Porous Medium, Unsaturated Porous Medium, or Dual Porosity Medium
selected:

Attributes > Contributing Velocity

Initial Values
The Initial Values node adds an initial value for the pressure that can serve as an initial
condition for a transient simulation or as an initial guess for a nonlinear solver.

If you have the Subsurface Flow Module licensed, you can specify the pressure either
directly or as an expression for the pressure head, Hp, or the hydraulic head, H; the
hydraulic head and the pressure head relate to the pressure p as

p;
H p = ------ H = Hp + D
g

482 | CHAPTER 6: FLUID FLOW INTERFACES

where  is the fluid density (SI unit: kg/m3); g denotes the acceleration of gravity
(SI unit: m/s2); and D is the elevation (SI unit: m).

INITIAL VALUES
Enter a value or expression for the initial value of the Pressure p (SI unit: Pa). The
default value is 0 Pa.

In case the Subsurface Flow Module is available, click the Pressure head button to enter
a value or expression for Hp (SI unit: m). The default is 0 m. Click the Hydraulic head
button to enter a value or expression for H (SI unit: m). The default is 0 m.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Initial Values
Richards’ Equation > Initial Values

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equation selected:

Domains > interface > Initial Values

Gravity
The Gravity node is automatically added when Include gravity is selected at interface
level in the Darcy’s Law settings and it is active in all domains in which the Darcy’s Law
interface is applied.

COORDINATE SYSTEM SELECTION

The Global coordinate system is selected by default. The Coordinate system list contains
any additional coordinate systems that the model includes. It can be used when
prescribing the direction of the gravitational forces.

GRAVITY
Specify either the acceleration vector or — in case you have a Subsurface Flow Module
license — the elevation.

THE DARCY’S LAW INTERFACE | 483

 When Acceleration is selected from the list, enter the components of the Gravity vector
g. The default value is g_const which is the physical constant having the value
9.8066 m/s2.

• For 3D and 2D axisymmetric models, the default value is -g_const in the

z direction.
• For 2D models, the default value is -g_const in the y direction.

When Elevation is selected from the list, specify the Elevation D (SI unit: m). The
Acceleration of gravity is taken From physics interface by default but could be overridden
when User defined is selected.

When the checkbox Include gravity is not selected in the Darcy’s Law interface Settings,
the elevation D is set equal to zero.

Select the checkbox Specify reference position to define a reference position for the
gravity calculation.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Gravity
Richards’ Equation > Gravity

Cross Section
Use this node with 1D components to model domains with another cross-sectional
area than the global one that is used in the interface Physical Model section. In 1D
geometries, the pressure is assumed to be constant in the radial direction, and Darcy’s
Law accounts for that.

CROSS SECTION
Enter values for the Cross-sectional area Ac to set the cross section of the domain in the
plane perpendicular to the 1D geometry.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Cross Section
Richards’ Equation > Cross Section

484 | CHAPTER 6: FLUID FLOW INTERFACES

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equation selected:

Domains > interface > Cross Section

Thickness
Use this node with 2D and 1D axisymmetric components to model domains with
another thickness than the overall thickness that is specified in the interface Physical
Model section. In 2D geometries, the pressure is assumed to be constant in the
out-of-plane direction (the z direction with default spatial coordinate names). In 1D
axisymmetric geometries the thickness represents the z direction.

THICKNESS
Specify a value for the Thickness dz of the domain in the out-of-plane direction. This
value replaces the overall thickness in the domains that are selected in the Domain
Selection section, and is used to multiply some terms into the heat equation.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Thickness
Richards’ Equation > Thickness

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equation selected:

Domains > interface > Thickness

Porous Electrode Coupling

Use the Porous Electrode Coupling node to define a mass source based on the volumetric
current densities of one or multiple Porous Electrode Reaction nodes in an
Electrochemistry interface.

The source (or sink) is proportional to the Molar mass (kg/mol) of the reacting species,
the current densities and the stoichiometric coefficients according to Faraday’s law as
defined by summation over the Reaction Coefficient subnodes.

THE DARCY’S LAW INTERFACE | 485

 SPECIES
Based on the number of species required for the model, use the Add ( ) and
Delete ) buttons under the table to add or remove Species. Then enter a value for
the Molar mass (kg/mol) in the applicable rows for each species.

The setting in the table will determine the number of available fields for entering the
stoichiometric coefficients of the Reaction Coefficient subnodes.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Porous Electrode Coupling
Richards’ Equation > Porous Electrode Coupling

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equation selected:

Domains > interface > Porous Electrode Coupling

Electrode Surface Coupling

Use the Electrode Surface Coupling node to define a combined wall and inflow/outflow
boundary condition based on current densities of one or multiple Electrode Reaction
nodes in an Electrochemistry interface.

The flow is proportional to the Molar mass (kg/mol) of the reacting species, the current
densities and the stoichiometric coefficients according to Faraday’s law as defined by
summation over the Reaction Coefficient subnodes.

SPECIES
Based on the number of species required for the model, use the Add ( ) and
Delete ) buttons under the table to add or remove Species. Then enter a value for
the Molar mass (kg/mol) in the applicable rows for each species.

The setting in the table will determine the number of available field for entering the
stoichiometric coefficients of the Reaction Coefficient subnodes.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Electrode Surface Coupling
Richards’ Equation > Electrode Surface Coupling

486 | CHAPTER 6: FLUID FLOW INTERFACES

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equation selected:

Boundaries > interface > Electrode Surface Coupling

Pressure
Use the Pressure node to specify the pressure on a boundary. In many cases the
distribution of pressure is known, giving a Dirichlet condition p = p0 where p0 is a
known pressure given as a number, a distribution, or an expression involving time, t,
for example.

PAIR SELECTION
If this node is selected from the Pairs menu, choose the pair to apply this condition to.
A pair must be created first. See Identity, Contact, and Sector Symmetry Pairs in the
COMSOL Multiphysics Reference Manual for more details.

PRESSURE
Enter a value or expression for the Pressure p0 (SI unit: Pa). Enter a relative pressure
value in p0 (SI unit: Pa).

For the Subsurface Flow Module, the Pressure node provides the pressure P0 as a
condition on edges in 3D models. Then select the edges under Edge Selection.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

As an attribute to the Fracture boundary condition or as boundary condition to The

Fracture Flow Interface (which are available in the Subsurface Flow Module and the
Porous Media Flow Module), the pressure can also be specified on edges (in 3D) or
points (in 2D) surrounding the fracture domain. In other cases, pressure constraints
on edges or points are not implemented.

If the boundary condition is adjacent to an Unsaturated Dual

Permeability Medium feature it constrains the pressures in the
Macropores and Micropores to p0.

THE DARCY’S LAW INTERFACE | 487

 LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Pressure
Richards’ Equation > Pressure

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equation selected:

Boundaries > interface > Pressure

Pairs > interface > Pressure

Mass Flux
Use the Mass Flux node to specify the mass flux into or out of the model domain
through some of its boundaries. It is often possible to determine the mass flux from
the pumping rate or from measurements. With this boundary condition, positive
values correspond to flow into the model domain:


n   ---  p +  g D  = N 0


where N0 is a value or expression for the specified inward (or outward) Darcy flux. D
is the elevation head which is set to zero for other than Subsurface Flow Module
applications.

MASS FLUX
Enter a value or expression for the Inward mass flux N0. A positive value of N0
represents an inward mass flux whereas a negative value represents an outward mass
flux. The units are based on the geometric entity: Boundaries: (SI unit: kg/(m2·s)).

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Mass Flux
Richards’ Equation > Mass Flux

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equation selected:

Boundaries > interface > Mass Flux

488 | CHAPTER 6: FLUID FLOW INTERFACES

Line Mass Source
The Line Mass Source node adds mass flow originating from a tube of infinitely small
radius.

SELECTION
The Line Mass Source feature is available for all dimensions, but the applicable selection
differs between the dimensions.

MODEL DIMENSION APPLICABLE GEOMETRICAL ENTITY

2D Points
2D Axisymmetry Points not on the symmetry axis
3D Edges

LINE MASS SOURCE

Enter a value or expression for the source strength, N0 (SI unit: kg/(m·s)). A positive
value results in mass injection from the line into the computational domain, and a
negative value means that the mass is removed from the computational domain.

Line sources located on a boundary affect the adjacent computational domains. This
effect makes the physical strength of a line source located in a symmetry plane twice
the given strength.

Fracture Flow
The equivalent feature available for fracture flow is the Mass Source attribute of the
Fracture boundary condition and the Mass Source point feature of the Fracture Flow
Interface, both available in 3D only.

LOCATION IN USER INTERFACE

According to the model dimension as mentioned in Selection, this feature is available
for the geometrical entities Edges or Points.

Context Menus
Darcy’s Law > geometrical entity > Line Mass Source
Richards’ Equation > geometrical entity > Line Mass Source

Ribbon
Physics tab with Darcy’s Law or Richards’ Equation selected:

geometrical entity > Line Mass Source

THE DARCY’S LAW INTERFACE | 489

 Point Mass Source
The Point Mass Source node models mass flow originating from an infinitely small
sphere centered around a point. It is available for points in 3D geometries.

POINT MASS SOURCE

Enter a value or expression for the source strength, N0 (SI unit: kg/(s)). A positive
value results in mass injection from the point into the computational domain, and a
negative value means that the mass is removed from the computational domain.

Point sources located on a boundary or on an edge affect the adjacent computational

domains. This has the effect, for example, that the physical strength of a point source
located in a symmetry plane is twice the given strength.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Points > Point Mass Source
Richards’ Equation > Points > Point Mass Source

Ribbon
Physics tab with Darcy’s Law or Richards’ Equation selected:

Points > Point Mass Source

Inlet
The Inlet node adds a boundary condition for the inflow (or outflow) perpendicular
(normal) to the boundary. It has three options that can be used to specify inlet
condition on a boundary, as follows:

VELOCITY
Enter a value or expression for the Normal inflow velocity U0 (SI unit: m/s). A positive
value of U0 represents an inflow velocity. A negative value represents an outflow
velocity. The inlet velocity boundary condition is implemented as;


n   ---  p +  g D  = U 0


where U0 is a value or expression for the specified inward (or outward) Darcy velocity.
A positive value of the velocity U0 corresponds to flow into the model domain whereas

490 | CHAPTER 6: FLUID FLOW INTERFACES

a negative value represents an outflow. D is the elevation head which is set to zero for
other than Subsurface Flow Module applications.

PRESSURE
Use the Pressure option to specify the inlet pressure on a boundary. In many cases the
distribution of pressure is known, giving a Dirichlet condition p = p0 where p0 is a
known pressure given as a number, a distribution, or an expression involving time, t,
for example. Enter a value or expression for the Pressure p0 (SI unit: Pa).

MASS FLOW
If you select Mass flow as the inlet condition, specify the total Mass flow rate M0 (SI
unit: kg/s), or the Pointwise mass flux N0 (SI unit: kg/(m2·s)).

With Mass flow rate boundary condition, positive values correspond to flow into the
model domain:

–    u  n  dS = M0


where M0 is a value or expression for the specified inward (or outward) Darcy flux.

Pointwise mass flux boundary condition, positive values correspond to flow into the
model domain:


n   ---  p +  g D  = N 0


where N0 is a value or expression for the specified inward (or outward) Darcy flux.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Inlet
Richards’ Equation > Inlet

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equation selected:

Boundaries > interface > Inlet

THE DARCY’S LAW INTERFACE | 491

 Symmetry
The Symmetry node describes a symmetry boundary. The following condition
implements the symmetry condition on an axis or a flow divide:


n  --- p = 0



n  ---  p +  g D  = 0


For axisymmetric components, COMSOL Multiphysics takes the axial symmetry

boundaries (at r = 0) into account and automatically adds an Axial Symmetry node that
is valid on the axial symmetry boundaries only.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Symmetry
Richards’ Equation > Symmetry

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equation selected:

Boundaries > interface > Symmetry

No Flow
The No Flow node is the default boundary condition stating that there is no flow across
impermeable boundaries. The mathematical formulation is:


n   ---  p +  g D  = 0


where n is the vector normal to the boundary. D is the elevation head which is set to
zero for any other than Subsurface Flow Module applications.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > No Flow
Richards’ Equation > No Flow

492 | CHAPTER 6: FLUID FLOW INTERFACES

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equation selected:

Boundaries > interface > No Flow

Flux Discontinuity
Use the Flux Discontinuity node to specify a mass flux discontinuity through an interior
boundary. The condition is represented by the following equation:

– n   u 1 –  u 2  = N 0

In this equation, n is the vector normal (perpendicular) to the interior boundary,  is

the fluid density, u1 and u2 are the Darcy velocities in the adjacent domains (as defined
in Equation 6-13) and N0 is a specified value or expression for the flux discontinuity.


u = – --- p (6-13)


For this boundary condition, a positive value of N0 corresponds to a flow discontinuity

in the opposite direction to the normal vector of the interior boundary.

MASS FLUX
Enter a value or expression for the Inward mass flux N0 (SI unit: kg/(m2·s)). A positive
value of N0 represents a mass flux discontinuity in the opposite direction to the normal
vector of the interior boundary.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Interior Surfaces > Flux Discontinuity
Richards’ Equation > Interior Surfaces > Flux Discontinuity

Ribbon
Physics tab with Darcy’s Law or Richards’ Equation selected:

Boundaries > Interior Surfaces > Flux Discontinuity

Pairs > Flux Discontinuity

THE DARCY’S LAW INTERFACE | 493

 Outlet
The Outlet node adds a boundary condition for the outflow (or inflow) perpendicular
(normal) to the boundary. It has different options that can be used to specify inlet
condition on a boundary, as follows:

VELOCITY
Enter a value or expression for the Normal outflow velocity U0 (SI unit: m/s). A positive
value of U0 represents an outflow velocity whereas a negative value represents an
inflow velocity.


– n   --- p = U 0


where U0 is a specified value or expression for the outward (or inward) Darcy velocity.

PRESSURE
Similar to the inlet Pressure option, the outlet pressure on a boundary can be specified.
Enter a value or expression for the Pressure p0 (SI unit: Pa).

DISCHARGE
Enter a value or expression for the discharge QD to specify the volumetric flow rate
through the boundary. The mass flux u is related to the discharge as follows:

 u  ds = QD . (6-14)


The mean normal outflow velocity at the boundary can then be calculated as Uout =
QD/A, where A is the cross sectional area of the boundary.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Outlet
Richards’ Equation > Outlet

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equation selected:

Boundaries > interface > Outlet

494 | CHAPTER 6: FLUID FLOW INTERFACES

Precipitation
The Precipitation node adds a boundary condition for the inflow (of rain)
perpendicular (normal) to the boundary.

Enter a value or expression for the Precipitation rate which then contributes to the mass
flux at the selected boundaries.

Select the Slope correction checkbox to apply a correction that depends on the normal
to the boundary and the vertical axis. This is needed when the boundary is inclined.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Hydraulics > Precipitation
Richards’ Equation > Hydraulics > Precipitation

Ribbon
Physics tab with Darcy’s Law or Richards’ Equation selected:

Boundaries > Hydraulics > Precipitation

Interior Wall
The Interior Wall boundary condition can only be applied on interior boundaries. It is
similar to the No Flux boundary available on exterior boundaries except that it applies
on both sides of an interior boundary. It allows discontinuities of velocity and pressure
across the boundary. The Interior Wall boundary condition can be used to avoid
meshing thin structures by applying no-flux condition on interior curves and surfaces
instead.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Interior Surfaces > Interior Wall
Richards’ Equation > Interior Surfaces > Interior Wall

Ribbon
Physics tab with Darcy’s Law or Richards’ Equation selected:

Boundaries > Interior Surfaces > Interior Wall

THE DARCY’S LAW INTERFACE | 495

 Thin Barrier
The Thin Barrier boundary condition models interior permeable walls, membranes,
geotextiles, or perforated plates as thin permeable barriers. The Thin Barrier boundary
condition can only be applied on interior boundaries.

WALL
Enter a value or expression for the Thickness db (SI unit: m, the default is 0.1 m) and
for the Permeability b (SI unit: m2). The default Permeability b uses the value From
material. For User defined select Isotropic to define a scalar value or Diagonal, Symmetric,
or Full to define a tensor value and enter another value or expression in the field or
matrix.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Interior Surfaces > Thin Barrier
Richards’ Equation > Interior Surfaces > Thin Barrier

Ribbon
Physics tab with Darcy’s Law or Richards’ Equation selected:

Boundaries > Interior Surfaces > Thin Barrier

Pressure Head
Use the Pressure Head node to specify the pressure head (instead of the pressure) on a
boundary. It adds this boundary condition for the pressure head Hp  Hp0, where Hp0
is a known pressure head given as a number, a distribution, or an expression involving
time, t, for example. The dimension of the pressure head is length (SI unit: m).

PRESSURE HEAD
Enter a value or expression for the Pressure head Hp0 (SI unit: m). The default is 0 m.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

496 | CHAPTER 6: FLUID FLOW INTERFACES

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Hydraulics > Pressure Head
Richards’ Equation > Hydraulics > Pressure Head

Ribbon
Physics tab with Darcy’s Law or Richards’ Equation selected:

Boundaries > Hydraulics > Pressure Head

Pairs > Pressure Head

Hydraulic Head
Use the Hydraulic Head node to specify the hydraulic head (instead of the pressure) on
a boundary. This adds the Dirichlet condition for the hydraulic head H  H0 where H0
is a known hydraulic head given as a number, a distribution, or an expression involving
time, t, for example.

HYDRAULIC HEAD
Enter a value or expression for the Hydraulic head H0 (SI unit: m). The default is 0 m.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Hydraulics > Hydraulic Head
Richards’ Equation > Hydraulics > Hydraulic Head

Ribbon
Physics tab with Darcy’s Law or Richards’ Equation selected:

Boundaries > Hydraulics > Hydraulic Head

Pairs > Hydraulic Head

Atmosphere/Gauge
The Atmosphere/Gauge node specifies an atmospheric pressure or gauges the pressure
to the atmospheric value. This means that the total hydraulic potential reduces to the

THE DARCY’S LAW INTERFACE | 497

 gravitational pressure at the free surface. At a free surface, such as a spring or a seepage
face, the pressure is atmospheric. If the pressures in the model is gauged to the
atmospheric value (p  0), the total hydraulic potential reduces to the gravitational
potential at the free surface — for example, the height of the free surface multiplied by
the fluid weight, or f gD. This boundary condition sets the pressure at the boundary
to zero and p  0.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog.

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Hydraulics > Atmosphere/Gauge
Richards’ Equation > Hydraulics > Atmosphere/Gauge

Ribbon
Physics tab with Darcy’s Law or Richards’ Equation selected:

Boundaries > Hydraulics > Atmosphere/Gauge

Pervious Layer
The Pervious Layer node provides a boundary condition that describes a mass flux
through a semi-pervious layer connected to an external fluid source at different
pressure, pressure head, or hydraulic head. The model domain might connect to a
larger body of water through the semi-pervious layer. This condition is represented
with the following boundary condition:

  pb – p 
n   ---  p +  g D  = R b -------------------
- +  Db – D  (6-15)
 g

In this equation, pb (SI unit: Pa) and Db (SI unit: m) are the pressure and the elevation
of the distant fluid source, respectively, and Rb (SI unit: 1/s) is the conductance of
materials between the source and the model domain (conductance to flow in the
semi-pervious layer adjacent to the boundary). Typically Rb  KB, where K is the
hydraulic conductivity (SI unit: m/s) of the layer and B (SI unit: m) is its thickness.
Using logical relationships, it is possible to activate these expressions at different times
or under various flow conditions.

498 | CHAPTER 6: FLUID FLOW INTERFACES

When the pressure head Hp is specified instead of the pressure, the boundary condition
is the following:

n  K  H p + D  = R b   H pb – H p  +  D b – D   (6-16)

Hpb is the pressure head (SI unit: m) at the edge of the layer.

When the hydraulic head H is specified instead of the pressure head, the boundary
condition becomes:

n  KH = R b  H b – H  (6-17)

Hb is the hydraulic head (SI unit: m) at the edge of the layer.

K in Equation 6-16 and Equation 6-17 is the hydraulic conductivity.

At a free surface, such as a water table or seepage face, the pressure is atmospheric (here
taken to be zero), so the total hydraulic potential equals gravitational potential, which
is defined on D.

Gravity effects are not active by default. Select the checkbox Include
gravity to activate the acceleration of gravity. Setting the elevation D to
zero also turns off gravity effects.

PERVIOUS LAYER
Specify the material properties whether to specify an external pressure, pressure head,
or hydraulic head.

Select an External variable to specify — Pressure and elevation to specify the external
pressure, Pressure head and elevation to specify the external pressure head, or Hydraulic
head. For all selections, enter a value for the Conductance Rb (SI unit: 1/s).

• For Pressure and elevation enter an External pressure pb (SI unit: Pa) and External
elevation Db (SI unit: m).
• For Pressure head and elevation enter an External pressure head Hpb (SI unit: m), Hpb
(SI unit: m) and External elevation Db (SI unit: m).
• For Hydraulic head enter the Hydraulic head Hb (SI unit: m).

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Hydraulics > Pervious Layer

THE DARCY’S LAW INTERFACE | 499

 Richards’ Equation > Hydraulics > Pervious Layer

Ribbon
Physics tab with Darcy’s Law or Richards’ Equation selected:

Boundaries > Hydraulics > Pervious Layer

Well
The Well feature is intended to model injection or production wells and is available
with the Subsurface Flow Module.

WELL
Enter a value or expression for the Well diameter dw (SI unit: m, the default is 0.1 m).
Select the Well type from the list: Production or Injection.

Specify either the Pressure (SI unit: Pa), Pressure head (SI unit: m), Hydraulic head (SI
unit: m), or the Mass flow. If you select Mass flow, specify the total Mass flow rate (SI
unit: kg/s), the Mass flow rate per unit length (SI unit: kg/(m·s)) or the Mass flux (SI
unit: kg/(m2·s)).

LOCATION IN USER INTERFACE

According to the model dimension this feature is available for the geometrical entities
Edges (3D) or Points (2D).

Context Menus
Darcy’s Law > geometrical entity > Well
Richards’ Equation > geometrical entity > Well

Ribbon
Physics tab with Darcy’s Law or Richards’ Equation selected:

geometrical entity > Well

Fracture
Use the Fracture node to model flow along a boundaries representing fractures within
a porous medium. By default, it adds the tangential form of the continuity equation
and of Darcy’s Law.


d f -----   p   +  T   d f u  = d f Q m (6-18)
t

500 | CHAPTER 6: FLUID FLOW INTERFACES

 
u = – --- Tp (6-19)


The Fracture feature has two default subfeatures — the Fluid (Fracture) subfeature
where the fluid density  (SI unit: kg/m3) and dynamic viscosity  (SI unit: Pa·s) are
defined and the Fracture Material subfeature where the permeability  (SI unit: m2)
and porosity p (dimensionless) are specified.

COORDINATE SYSTEM SELECTION

Select a coordinate system from the Coordinate system list for the interpretation of
directions in anisotropic material properties. The default is the Global coordinate
system, and the list contains any additional orthonormal coordinate system (except
boundary coordinate systems) added under the Definitions node.

The subnodes inherit these coordinate system settings. In particular, the Permeability
or Hydraulic conductivity (in the Fracture Material subnode) should be set according to
the coordinate system selected in this section.

APERTURE
Enter a value for the Fracture thickness df.

FLOW MODEL
Define if the flow follows a linear or nonlinear pressure-velocity relationship.

• Darcian flow (default) defines a linear relationship using Equation 6-18 and
Equation 6-19.
• Non-Darcian flow defines a nonlinear relationship using Equation 6-18 and the
pressure-velocity relationship


Tp = – --- u –  u u


where the nonlinear parameter  (SI unit: 1/m) is further specified in the Fracture
Material subfeature.

STORAGE MODEL
Equation 6-18 can also be formulated with respect to a storage term.

p 
S ------ +    – ---  p  = Q m
t 

THE DARCY’S LAW INTERFACE | 501

 The storage coefficient S can be interpreted as the weighted compressibility of the
porous material and the fluid. Define the time dependent storage term by selecting one
of the following options from the list:

• From density and porosity (default), uses the formulation of Equation 6-18
• Linearized storage, where the following linear equation is used to define the storage:
S =  p  f +  1 –  p  p

• User defined to directly specify the storage coefficient S

• About Darcian and Non-Darcian Flow in the Porous Media Flow

Module User’s Guide
• Storage Models in the Subsurface Flow Module User’s Guide

FRACTURE MODEL
Choose between two options to model fractures that have different characteristics.

• Highly conductive fracture represents a fracture filled with a highly permeable

material.
• Thin conducting barrier represents a fracture that is filled with a low permeable
material and therefore acts as a barrier for the flow which adds a jump in the pressure
across the boundary:

 p d,u – p u,d   b
– n u,d   u  u,d =  ------------------------------- ------
db 

The barrier thickness db and permeability b are defined in the Fracture Material
subnode.
In addition, either on the upside, downside, or on both sides a highly conductive
fracture material is present and Equation 6-18 and Equation 6-19 are solved for the
highly conductive side. Further specifications are done in the Fracture Material
subnode.

Note that the option to specify the Fracture Model is only available for the
Fracture feature within the Darcy’s Law interface.

For the Fracture boundary node several subnodes are available from the context menu
(right-click the parent node) or from the Physics toolbar, Attributes menu. As they are

502 | CHAPTER 6: FLUID FLOW INTERFACES

almost identical to the boundary conditions of the Fracture Flow interface, see
Domain, Boundary, Edge, Point, and Pair Nodes for the Fracture Flow Interface in
the Subsurface Flow User’s Guide for further description. Note that Precipitation is
only available for the Fracture Flow interface.

This feature requires a specific license. For a detailed overview of the

functionality available in each product, visit [Link]
products/specifications/.

Flow in a Fractured Reservoir: Application Library path

Subsurface_Flow_Module/Fluid_Flow/fractured_reservoir_flow

LOCATION IN USER INTERFACE

Context Menus
Darcy’s Law > Fracture
Richards’ Equation > Fracture
Fracture Flow > Fracture

Ribbon
Physics tab with interface as Darcy’s Law or Richards’ Equations selected:

Boundaries > interface > Fracture

With Fracture Flow selected:

Boundaries > Fracture

Fluid (Fracture)
This node defines the material properties of the fluid of the Fracture parent node. The
fluid can be specified as general gas or liquid or as an ideal gas.

MODEL INPUT
This section contains fields and values that are inputs for expressions defining material
properties. If such user-defined property groups are added, the model inputs appear
here.

THE DARCY’S LAW INTERFACE | 503

 FLUID PROPERTIES
The available options for the Fluid type are Gas/Liquid (default) or Ideal gas. The
selection determines how the density and compressibility of the fluid are set. The latter
is only required if a linearized storage model has been selected in the parent Fracture
node. For all fluid types specify the dynamic viscosity  (SI unit: Pa·s).

Gas/Liquid
This option specifies the Density and the Compressibility for a general gas or liquid.

Ideal Gas
For Ideal gas it uses the ideal gas law to describe the fluid. In this case, specify the
thermodynamics properties. Select a Gas constant type — Specific gas constant Rs (the
default) or Mean molar mass Mn (SI unit: J/(mol·K)). For Mean molar mass the
universal gas constant R  8.314 J/(mol·K) is used as the built-in physical constant.
For both properties, the defaults use values From material. For User defined enter
another value or expression.

If required by a Storage Model, the Compressibility of the ideal gas is calculated as f
1pA with the absolute pressure pA (Pa).

If Hydraulic conductivity is selected as the Permeability model (see the

Fracture Material section), dynamic viscosity is not available. The
hydraulic conductivity is defined using a combination of fluid and fracture
material properties and replaces the need of defining the dynamic
viscosity.

LOCATION IN USER INTERFACE

A default Fluid node is automatically added to the Fracture node.

Context Menus
Darcy’s Law > Fracture > Fluid
Richards’ Equation > Fracture > Fluid
Fracture Flow > Fracture > Fluid

Fracture Material
This node defines the porosity and material properties of the solid matrix of the
Fracture parent node.

504 | CHAPTER 6: FLUID FLOW INTERFACES

MODEL INPUT
This section contains fields and values that are inputs for expressions defining material
properties. If such user-defined property groups are added, the model inputs appear
here.

BARRIER PROPERTIES
This section appears if the Thin conducting barrier option for the Fracture model is
chosen in the parent Fracture feature.

Enter the values for the barrier thickness db and permeability b.

Note that the Thin Conductive barrier option is only available for the Fracture feature
within the Darcy’s Law interface.

FRACTURE MATERIAL PROPERTIES

If the Thin conducting barrier option for the Fracture model is chosen in the parent
Fracture feature you can specify the Conductive layer position here. Choose between
Both sides, Upside, and Downside. A sketch above shows the position of conductive
layers and barrier within the fracture.

The Porosity p of the fracture material is taken From material by default. Select User
defined to enter a value or expression for p.

If required by a Storage model in the parent node, specify the Effective compressibility
of the porous matrix. The default value is p  104 1/Pa.

Select a Permeability model to specify the capacity of the fracture material to transmit
flow. Which options are available depends on the Flow model setting in the parent node.

• Select Permeability (default) to define the permeability of the fracture material.

• Select Hydraulic conductivity or define a combination of fluid permeability and
dynamic viscosity. For Hydraulic conductivity K (SI unit: m/s) select Isotropic to
define a scalar value or Diagonal or Symmetric to define a tensor value. The default is
2.94·104 m/s.
• Select Cubic law to define the permeability from the fracture thickness specified in
the parent node and the roughness factor ff.

THE DARCY’S LAW INTERFACE | 505

 If the Non-Darcian Flow Model is selected in the parent Fracture feature enter a value
for the Inertial resistance coefficient .

• About Darcian and Non-Darcian Flow in the Porous Media Flow

Module User’s Guide
• Permeability Models in the Porous Media Flow Module User’s Guide

LOCATION IN USER INTERFACE

A default Fracture Material node is automatically added to the Fracture node.

Context Menus
Darcy’s Law > Fracture > Fracture Material
Richards’ Equation > Fracture > Fracture Material
Fracture Flow > Fracture > Fracture Material

506 | CHAPTER 6: FLUID FLOW INTERFACES

The Free and Porous Media Flow,
Br i nk m a n Interfac e
The Free and Porous Media Flow, Brinkman (fp) interface ( ) is found under the Porous
Media and Subsurface Flow branch ( ) when adding a physics interface. It is used to
compute fluid velocity and pressure fields of single-phase flow where free flow is
connected to porous media. The Free and Porous Media Flow, Brinkman interface is
used over at least two different domains: a free channel and a porous medium. The
physics interface is well suited for transitions between slow flow in porous media,
governed by the Brinkman equations, and fast flow in channels described by the
Navier–Stokes equations. Fluids with varying density can be included at Mach
numbers below 0.3. Also the viscosity of a fluid can vary, for example, to describe
non-Newtonian fluids. The physics interface can be used for stationary and
time-dependent analyses.

When this physics interface is added, the following default nodes are also added in the
Model Builder — Fluid Properties, Wall, and Initial Values. Then, from the Physics
toolbar, add a Porous Medium node to be used on the domain selection corresponding
to the porous media, or add other nodes that implement, for example, boundary
conditions and volume forces. You can also right-click Free and Porous Media Flow,
Brinkman to select physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is fp.

PHYSICAL MODEL

Compressibility
By default the physics interface uses the Incompressible flow formulation of the Navier–
Stokes and Brinkman equations to model constant density flow. If required, select

THE FREE AND POROUS MEDIA FLOW, BRINKMAN INTERFACE | 507

 Weakly compressible flow from the Compressibility list, to account for small variations in
the density, which are only dependent on the temperature (nonisothermal flow) or
Compressible flow (Ma<0.3) for fully compressible flow. However, for the flow modeled
with this physics interface, the Mach number must be below 0.3.

Swirl Flow
For 2D axisymmetric components, select the Swirl flow checkbox to include the swirl
velocity component, that is the velocity component u in the azimuthal direction.
While u can be nonzero, there can be no gradients in the  direction. Note that this
feature is only available for specific modules. Visit [Link]
products/specifications/ for a detailed overview.

Neglect Inertial Term

Select the Neglect inertial term (Stokes flow) checkbox if the inertial forces are small
compared to the viscous forces.

Enable porous media domains

The Enable porous media domains checkbox is selected by default to solve the Brinkman
equations in porous domains. If it is unchecked, the Porous Medium node is no longer
available in this interface.

Include Gravity
Gravity is not included by default. Select the Include gravity checkbox to activate the
acceleration of gravity. This automatically adds a global Gravity feature node to the
interface model tree, and the buoyancy force is included in the Equations.

Also, when the Include gravity checkbox is selected, the Use reduced pressure option
changes the pressure formulation from using the total pressure (default) to using the
reduced pressure. This option is suitable for configurations where the density changes
are very small; otherwise, the default formulation can be used. For more information,
see Gravity.

Porous Treatment of No Slip Condition

Choose how the No Slip condition on internal boundaries should be treated. The
options are Standard no slip formulation and Porous slip (default). The latter option
provides a unified treatment when the porous matrix is fully resolved as well as when
it is under resolved ensuring a smooth transition between regions with different
resolutions; see No Slip under Wall in The Laminar Flow and Creeping Flow Interfaces
section.

Reference Pressure Level

Enter a Reference pressure level pref (SI unit: Pa). The default value is 1[atm].

508 | CHAPTER 6: FLUID FLOW INTERFACES

Reference Temperature
Enter a Reference temperature Tref (SI unit: K). The default value is 293.15[K].

Reference Position
If Include gravity is selected, a Reference position rref (SI unit: m) can be specified which
is then used for the calculation of the hydrostatic pressure.

TURBULENCE
Turbulent flow can be simulated by changing the Turbulence model type to RANS-EVM
(Reynolds-Averaged Navier–Stokes, eddy-viscosity model), RANS-RSM
(Reynolds-Averaged Navier–Stokes, Reynolds stress model), or Large Eddy Simulation
(which is only available in 3D). If turbulent flow is activated, you can choose from
different Turbulence models and options for Wall treatment. For a description of the
different turbulence models, wall treatment options, and turbulence model
parameters, see Theory for the Turbulent Flow Interfaces in the CFD Module User’s
Guide.

DEPENDENT VARIABLES
The following dependent variables (fields) are defined for this physics interface — the
Velocity field u (SI unit: m/s) and its components, and the Pressure p (SI unit: Pa).

• Domain, Boundary, Point, and Pair Nodes for the Free and Porous
Media Flow, Brinkman Interface
• Theory for the Free and Porous Media Flow, Brinkman Interface

Current Density Distribution in a Solid Oxide Fuel Cell: Application

Library path Fuel_Cell_and_Electrolyzer_Module/Fuel_Cells/sofc_unit_cell

Domain, Boundary, Point, and Pair Nodes for the Free and Porous
Media Flow, Brinkman Interface
The Free and Porous Media Flow, Brinkman Interface has the following domain,
boundary, point, and pair nodes, listed in alphabetical order, available from the Physics

THE FREE AND POROUS MEDIA FLOW, BRINKMAN INTERFACE | 509

 ribbon toolbar (Windows users), Physics context menu (Mac or Linux users), or
right-click to access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• Electrode Surface Coupling • Porous Medium

• Fluid Properties • Mass Source
• Gravity • Volume Force
• Initial Values • Wall
• Porous Electrode Coupling

Note that some features are only available with certain COMSOL products (see
[Link]

The following nodes (listed in alphabetical order) are described for the Laminar Flow
interface in the COMSOL Multiphysics Reference Manual:

• No Viscous Stress • Open Boundary

• Flow Continuity • Periodic Flow Condition
• Inlet • Point Mass Source
• Line Mass Source • Pressure Point Constraint
• Outlet • Symmetry

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Fluid Properties
Use the Fluid Properties node to define the fluid material, density, and dynamic
viscosity.

510 | CHAPTER 6: FLUID FLOW INTERFACES

FLUID PROPERTIES
The default Fluid material uses the Domain material (the material defined for the
domain). Select another material as needed.

The default Density  (SI unit: kg/m3) uses values From material based on the Fluid
material selection. For User defined enter another value or expression. The default is
0 kg/m3.

The Dynamic viscosity  (SI unit: Pa·s) uses values From material based on the Fluid
material selection. For User defined enter another value or expression. The default is
0 Pa·s.

Porous Medium
The Porous Medium node adds the Brinkman equations (excluding any mass sources)
to the interface. It has two default subfeatures — the Fluid subfeature where the fluid
density (SI unit: kg/m3) and dynamic viscosity  (SI unit: Pa·s) are defined and the
Porous Matrix subfeature where the permeability  (SI unit: m2) and porosity p
(dimensionless) are specified.

FLOW MODEL
Define if the flow follows a linear or nonlinear pressure-velocity relationship.

• Darcian flow (default) defines a linear relationship characteristic for low Reynolds
number laminar flows.
• Non-Darcian flow defines a nonlinear relationship due to inertial or turbulence
effects.

POROUS MEDIUM TURBULENCE MODEL

Defines a model which complements the basic turbulence model and accounts for the
influence of the porous matrix on turbulent flow in Porous Media.

• Default and Nakayama-Kuwahara specify Coefficient of subgrid turbulence generation by

porous matrix Cstpm which by default is set to 0.212.
• Pedras-de Lemos asks to specify Choice of the Pedras-de Lemos model coefficient. When
User defined is chosen, specify Coefficient of subgrid turbulence generation by porous
matrix Cstpm with default value 0.212. Choosing Original results in
C stpm = 0.28   p , while choosing Recalibrated results in C stpm = 0.18 .

THE FREE AND POROUS MEDIA FLOW, BRINKMAN INTERFACE | 511

 Fluid
This node defines the material properties of the fluid of the Porous Medium parent
node.

MODEL INPUT
This section contains fields and values that are inputs for expressions defining material
properties. If such user-defined property groups are added, the model inputs appear
here.

FLUID PROPERTIES
Specify the Density  and the Dynamic viscosity  (SI unit: Pa·s) of the fluid. The
defaults use values From material, for User defined enter a value or expression.

The dynamic viscosity describes the relationship between the shear stresses and the
shear rate in a fluid. Intuitively, water and air have a low viscosity, and substances often
described as thick, such as oil, have a higher viscosity.

If the Polymer Flow Module license is available and for Darcian flow, also
non-Newtonian models are available and the Constitutive Relation can be switched
between Specify dynamic viscosity (default) or Inelastic non-Newtonian.

Constitutive Relation
Use the default option to either get the dynamic viscosity from a material, or to add a
user defined expression. The default option lets you either use the dynamic viscosity
from a material or define it with a custom expression. The Inelastic non-Newtonian
option provides several models and defines the Apparent shear rate.

Inelastic Non-Newtonian
The following inelastic non-Newtonian Models are available for the Fluid feature under
the Porous Medium feature:

• Power Law
• Carreau
• Carreau-Yasuda
• Cross
• Cross-Williamson
• Sisko

The setup of these models in porous media is similar to their definition under the Fluid
Properties feature.

512 | CHAPTER 6: FLUID FLOW INTERFACES

Apparent Shear Rate
ꞏ
To define the apparent shear rate  app , either enter a user-defined expression for the
correction factor () or choose the Capillary bundle approach from the list and provide
a value for the Tortuosity factor (C).

• Read more about the theory of Non-Newtonian Fluids in Porous

Media in the Polymer Flow Module User’s Guide

Porous Matrix
This node defines the porosity and material properties of the solid matrix of the Porous
Medium parent node.

MODEL INPUT
This section contains fields and values that are inputs for expressions defining material
properties. If such user-defined property groups are added, the model inputs appear
here.

MATRIX PROPERTIES
The Porosity p (a dimensionless number between 0 and 1) uses by default the value
From material. For User defined the default value is 1 which is equivalent to free flow.

Select a Permeability model to specify the capacity of the porous material to transmit
flow. Depending on the Flow model selection in the parent Porous Medium feature node
you can choose between the following options:

• Permeability (the default) to directly enter the permeability of the porous matrix or
Kozeny-Carman to define the permeability from the porosity and mean Particle
diameter for Darcian flow or
• Forchheimer or Ergun for Non-Darcian flow.

Note that some of these options are only available for certain modules. For a detailed
overview of the functionality available in each product, visit https://
[Link]/products/specifications/.

The default Permeability (SI unit: m2) uses the value From material, as defined by the
Porous material list. For User defined select Isotropic to define a scalar value or Diagonal
or Symmetric to define a tensor value.

THE FREE AND POROUS MEDIA FLOW, BRINKMAN INTERFACE | 513

 For Kozeny-Carman, enter the mean Particle diameter dp (SI unit: m), the default value
is 0.5 mm.

In case a Non-Darcian flow model is chosen in the parent node, Forchheimer is the
default selection for the Permeability model. For Forchheimer, the default Permeability
(SI unit: m2) uses the value From material. Furthermore, the dimensionless
Forchheimer parameter cF can be defined. The default value is 0.55.

If Ergun is selected, enter the mean Particle diameter dp (default value: 0.5 mm). The
permeability  is then calculated using Equation 2-4 in the Porous Media Flow
Module User’s Guide.

• About Darcian and Non-Darcian Flow in the Porous Media Flow

Module User’s Guide
• Permeability Models in the Porous Media Flow Module User’s Guide

Mass Source
Enter a value or expression for an optional mass source (or sink) Source term Qm (SI
unit: kg/(m3·s)). This term accounts for mass deposit or mass creation within porous
domains. The physics interface assumes that the mass exchange occurs at zero velocity.

Effective Mass Transport Parameters

A correction factor (Bruggeman, No Correction, or User defined) to the mass transport
parameters (defined in the Transport Properties node) can be applied for the porous
domain. Species diffusivities and mobilities are automatically adjusted by the porous
media corrections. For User defined enter a value or expression for the Conversion factor
feff. The default is 1. Species diffusivities and mobilities are automatically adjusted by
the porous media corrections.

Volume Force
The Volume Force node specifies the force F on the right-hand side of the Navier–
Stokes or Brinkman equations, depending on whether a Fluid Properties or Porous
Medium node is active on the domain. It then acts on each fluid element in the
specified domains.

VOLUME FORCE
Enter the components of the Volume force F (SI unit: N/m3).

514 | CHAPTER 6: FLUID FLOW INTERFACES

Porous Electrode Coupling
Use the Porous Electrode Coupling node to define a mass source based on the volumetric
current densities of one or multiple Porous Electrode Reaction nodes in an
Electrochemistry interface.

The source (or sink) is proportional to the Molar mass (kg/mol) of the reacting species,
the current densities and the stoichiometric coefficients according to Faraday’s law as
defined by summation over the Reaction Coefficient subnodes.

SPECIES
Based on the number of species required for the model, use the Add ( ) and
Delete ) buttons under the table to add or remove Species. Then enter a value for
the Molar mass (kg/mol) in the applicable rows for each species.

The setting in the table will determine the number of available fields for entering the
stoichiometric coefficients of the Reaction Coefficient subnodes.

Initial Values
The Initial Values node adds initial values for the velocity field and the pressure that can
serve as an initial condition for a transient simulation or as an initial guess for a
nonlinear solver.

INITIAL VALUES
Enter initial values or expressions for the Velocity field u (SI unit: m/s) and for the
Pressure p (SI unit: Pa). The default values are 0 m/s and 0 Pa, respectively.

Gravity
The Gravity global feature is automatically added when Include gravity is selected at the
interface level in the Physical Model settings. It defines the gravity forces from the
Acceleration of gravity value. When a turbulence model that solves for the turbulent
kinetic energy, k, is used, the option Include buoyancy-induced turbulence is available.
When selected, the Buoyancy contribution is by default set to Automatic from
multiphysics. Contributions are only obtained from multiphysics couplings that
support buoyancy-induced turbulence, such as Nonisothermal Flow. If the Buoyancy
contribution is switched to User defined, a text field for the Turbulent Schmidt number
appears. See also Theory for Buoyancy-Induced Turbulence in the CFD Module
User’s Guide.

THE FREE AND POROUS MEDIA FLOW, BRINKMAN INTERFACE | 515

 This feature requires a specific license. For a detailed overview of the functionality
available in each product visit [Link]

ACCELERATION OF GRAVITY
The Acceleration of gravity (SI unit m/s, default value gconstez in 2D axial symmetry
and 3D and gconstey in 2D) is used to define the gravity forces. It should be a global
quantity.

Electrode-Electrolyte Interface Coupling

Use the Electrode-Electrolyte Interface Coupling node to define a combined wall and
inflow/outflow boundary condition based on current densities of one or multiple
Electrode Reaction nodes in an Electrochemistry interface.

The flow is proportional to the Molar mass (kg/mol) of the reacting species, the current
densities and the stoichiometric coefficients according to Faraday’s law as defined by
summation over the Reaction Coefficient subnodes.

TANGENTIAL VELOCITY CONDITION

For information about this section, see the Wall node. No slip is the default, but Slip
may in some cases be a more applicable, for instance if a gas diffusion electrode is
modeled as a boundary condition.

SPECIES
Based on the number of species required for the model, use the Add ( ) and
Delete ) buttons under the table to add or remove Species. Then enter a value for
the Molar mass (kg/mol) in the applicable rows for each species.

The setting in the table will determine the number of available field for entering the
stoichiometric coefficients of the Reaction Coefficient subnodes.

Wall
The Wall node includes a set of boundary conditions describing fluid-flow conditions
at stationary, moving, and leaking walls.

516 | CHAPTER 6: FLUID FLOW INTERFACES

BOUNDARY CONDITION
Select a Boundary condition for the wall.

• No Slip1 • Leaking Wall

• Slip

No Slip
No slip is the default boundary condition for a stationary solid wall for laminar flow
(and SST, Low Re k-, Algebraic yPlus, L-VEL, and Spalart–Allmaras turbulence
models). The condition prescribes u = 0; that is, the fluid at the wall is not moving.

Slip
The Slip option prescribes a no-penetration condition, u·n  . It is implicitly assumed
that there are no viscous effects at the slip wall and hence, no boundary layer develops.
From a modeling point of view, this can be a reasonable approximation if the main
effect of the wall is to prevent fluid from leaving the domain.

Leaking Wall
This boundary condition can be used to simulate a wall where fluid is leaking into or
leaving the domain with the velocity u = ul through a perforated wall. The
components of the Fluid velocity ul on the leaking wall should be specified.

CONSTRAINT SETTINGS
This section is displayed by clicking the Show button ( ) and selecting Advanced
Physics Options.

• The Moving Mesh Interface in the COMSOL Multiphysics Reference

Manual

THE FREE AND POROUS MEDIA FLOW, BRINKMAN INTERFACE | 517

 The Free and Porous Media Flow,
Darcy Interface
The Free and Porous Media Flow, Darcy interface ( ) combines the Darcy’s Law and
Laminar Flow interfaces to compute fluid velocity and pressure fields where free flow is
connected to porous media. The physics interface is suited for situations when both
the free flow and the flow in the porous media are laminar.

When a predefined Free and Porous Media Flow, Darcy interface is added from the Porous
Media and Subsurface Flow branch ( ) of the Model Wizard or Add Physics window, the
Darcy’s Law and Laminar Flow interfaces are added to the Model Builder. In addition,
the Multiphysics node is added, which automatically includes the Free and Porous Media
Flow Coupling multiphysics coupling.

On the Constituent Physics Interfaces

The Darcy’s Law interface is used to simulate fluid flow through interstices in a porous
medium. It can be used to model low-velocity flows in media where the permeability
and porosity are very small, and for which the pressure gradient is the major driving
force and the flow is mostly influenced by the frictional resistance within the pores.

The Laminar Flow interface is used to compute the velocity and pressure for a fluid flow
in the laminar flow regime. A flow will remain laminar as long as the Reynolds number
is below a certain critical value. At higher Reynolds numbers, disturbances have a
tendency to grow and cause transition to turbulence.

SETTINGS FOR PHYSICS INTERFACES AND COUPLING FEATURES

When physics interfaces are added using the predefined couplings, for example Free
and Porous Media Flow, Darcy, specific settings are included with the physics interfaces
and the coupling features.

However, if physics interfaces are added one at a time, followed by the coupling
features, these modified settings are not automatically included.

Free and Porous Media Flow Coupling (Multiphysics Coupling)

The Free and Porous Media Flow Coupling multiphysics coupling node ( )
bidirectionally links the Laminar Flow and Darcy’s Law interfaces.

518 | CHAPTER 6: FLUID FLOW INTERFACES

SETTINGS
The Label is the default multiphysics coupling feature name.

The Name is used primarily as a scope prefix for variables defined by the coupling node.
Refer to such variables in expressions using the pattern <name>.<variable_name>. In
order to distinguish between variables belonging to different coupling nodes or physics
interfaces, the name string must be unique. Only letters, numbers, and underscores (_)
are permitted in the Name field. The first character must be a letter.

The default Name (for the first multiphysics coupling feature in the model) is nsd1.

DOMAIN SELECTION
When nodes are added from the context menu, you can select Manual (the default)
from the Selection list to choose specific domains to define the multiphase flow
coupling, or select All domains as needed.

When the Free and Porous Media Flow Coupling is added as an effect of adding a Free
and Porous Media Flow, Darcy multiphysics interface, the selection is the intersection
of the boundaries of the participating physics interfaces.

Only boundaries that are active in the participating physics interfaces (which are
chosen from the Coupled Interfaces selection list) can be selected.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Free flow
and Porous media flow lists include all applicable physics interfaces.

The default values depend on how the coupling node is created.

• If it is added from the Physics ribbon (Windows users), Physics contextual toolbar
(Mac and Linux users), or context menu (all users), then the first physics interface
of each type in the component is selected as the default.
• If it is added automatically when a multiphysics interface is selected in the Model
Wizard or Add Physics window, then the two participating physics interfaces are
selected.

You can also select None from either list to uncouple the Free and Porous Media Flow
Coupling node from a physics interface. If the physics interface is removed from the
Model Builder, for example Darcy’s Law is deleted, then the Porous media flow list
defaults to None as there is nothing to couple to.

THE FREE AND POROUS MEDIA FLOW, DARCY INTERFACE | 519

 COUPLING SETTINGS
This section defines the settings for the Free flow boundary condition.

• When No slip is selected, the tangential component of the fluid velocity at the
free-porous interface on the free-flow side (computed in the Free flow interface) is
set to zero.
• When Continuous tangential velocity is selected, the tangential velocity component
of the flow on both sides of the free-porous interface is continuous. In addition,
continuity of normal stresses and conservation of mass across the interface are
ensured.
• When Beavers-Joseph is selected, the difference of the tangential velocity
components of the flow on both sides of the free-porous interface is proportional to
the shear stress at the free-porous interface on the free-flow side. For this option,
enter a value or expression for the Beavers and Joseph parameter aBJ (dimensionless)
that determines the proportionality. In addition, select the checkbox Use projected
permeability to use only the component of the permeability tensor in the direction
of the difference of the tangential velocities on both sides of the interface, instead
of the trace of the permeability tensor, in determining the proportionality.
• When Beavers-Joseph-Saffman is selected, the tangential velocity component of the
flow at the free-porous interface on the free-flow side is proportional to the shear
stress at the interface on the free-flow side. For this option also, enter a value or
expression for the Beavers and Joseph parameter aBJ (dimensionless) that determines
the proportionality. In addition, select the checkbox Use projected permeability to
use only the component of the permeability tensor in the direction of the difference
of the tangential velocities on both sides of the interface, instead of the trace of the
permeability tensor, in determining the proportionality.
• When Viscous slip is selected, the difference of the tangential velocity components
of the flow on the free-flow side and a user-supplied interfacial velocity is
proportional to the shear stress at the free-porous interface on the free-flow side.
Enter values for the Slip length Ls (SI unit: m, the default is 1e-7 m), and an
Interfacial velocity ub (SI unit: m/s, the default is 0 m/s).

520 | CHAPTER 6: FLUID FLOW INTERFACES

In all the above cases, additionally continuity of normal stresses and conservation of
mass across the interface are ensured.

Please find related information in the CFD Module User’s Guide:

• The Darcy’s Law Interface

• Theory for the Darcy’s Law Interface
• The Laminar Flow Interface
• Theory for the Single-Phase Flow Interfaces
• Theory for the Free and Porous Media Flow, Darcy Interface

THE FREE AND POROUS MEDIA FLOW, DARCY INTERFACE | 521

 The Brinkman Equations Interface
The Brinkman Equations (br) interface ( ), found under the Porous Media and
Subsurface Flow branch ( ) when adding a physics interface, is used to compute fluid
velocity and pressure fields of single-phase flow in porous media in the laminar flow
regime. The physics interface extends Darcy’s law to describe the dissipation of the
kinetic energy by viscous shear, similar to the Navier–Stokes equations. Fluids with
varying density can be included at Mach numbers below 0.3. Also the viscosity of a
fluid can vary, for example, to describe non-Newtonian fluids. To simplify the
equations, select the Stokes–Brinkman flow feature to reduce the dependence on
inertial effects when the Reynolds number is significantly less than 1. The physics
interface can be used for stationary and time-dependent analyses.

The main node is the Porous Medium feature, which adds the Brinkman equations and
provides interfaces for defining the fluid material and the porous matrix.

When this physics interface is added, the following default nodes are also added in the
Model Builder — Porous Medium, Wall (the default boundary type, using No slip as the
default boundary condition), and Initial Values. Then, from the Physics toolbar, add
other nodes that implement, for example, boundary conditions and volume forces.
You can also right-click Brinkman Equations to select physics features from the context
menu.

The boundary conditions are essentially the same as for the Laminar Flow interface.
Differences exist for the following boundary types: Outlet, Symmetry, Open
Boundary, and Boundary Stress where the viscous part of the stress is divided by the
porosity to appear as

1 T 2 
-----    u +  u   – ---     u I 
p  3 

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links to common
sections such as Discretization, Consistent Stabilization, Inconsistent Stabilization, and
Advanced Settings, all accessed by clicking the Show button ( ) and choosing the
applicable option. You can also search for information: press F1 to open the Help
window or Ctrl+F1 to open the Documentation window.

SETTINGS
The Label is the default physics interface name.

522 | CHAPTER 6: FLUID FLOW INTERFACES

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is br.

PHYSICAL MODEL
This node specifies the properties of the Brinkman Equations interface, which describe
the overall type of fluid flow model.

Compressibility
By default the physics interface uses the Incompressible flow formulation of the
Brinkman equations to model constant density flow. Alternatively, from the
Compressibility list select Weakly compressible flow, which should be used when the
pressure dependency of the density can be neglected, or Compressible flow (Ma<0.3) to
solve for the full compressible flow of which, however, the Mach number must be
below 0.3.

Swirl Flow
For 2D axisymmetric models, select the Swirl flow checkbox to include the swirl
velocity component, that is the velocity component u in the azimuthal direction.
While u can be nonzero, there can be no gradients in the  direction.

Note that this feature is only available for specific modules. See https://
[Link]/products/specifications/ for a detailed overview.

Neglect Inertial Term (Stokes–Brinkman Flow)

The Neglect inertial term (Stokes–Brinkman) checkbox is selected by default to model
flow at low Reynolds numbers for which the inertial term can be neglected. This results
in the linear Stokes–Brinkman equations.

Include Gravity
Gravity is not included by default. Select the Include gravity checkbox to activate the
acceleration of gravity. This automatically adds a global Gravity feature node to the
interface model tree, and the buoyancy force is included in the equations.

Also, when the Include gravity checkbox is selected, the Use reduced pressure option
changes the pressure formulation from using the total pressure (default) to using the
reduced pressure. This option is suitable for configurations where the density changes

THE BRINKMAN EQUATIONS INTERFACE | 523

 are very small; otherwise, the default formulation can be used. For more information,
see Gravity.

Porous Treatment of No Slip Condition

Choose how the No Slip condition on Wall boundaries and Interior Wall boundaries
adjacent to Porous Medium should be treated. The options are Standard no slip
formulation and Porous slip (the default). The latter option provides a unified treatment
when the porous matrix is fully resolved as well as when it is under resolved ensuring
a smooth transition between regions with different resolutions; see No Slip under Wall
in The Laminar Flow and Creeping Flow Interfaces section.

Reference Pressure Level

Enter a Reference pressure level pref (SI unit: Pa). The default value is 1[atm].

Reference Temperature Level

Enter a Reference temperature level Tref (SI unit: K). The default value is 293.15[K].

Reference Position
If Include gravity is selected, a Reference position rref (SI unit: m) can be specified which
is then used for the calculation of the hydrostatic pressure.

TURBULENCE
Turbulent flow can be simulated by changing the Turbulence model type to RANS-EVM
(Reynolds-Averaged Navier–Stokes, eddy-viscosity model). If RANS-EVM is activated,
you can set the Turbulence model either to one of the algebraic turbulence models,
Algebraic yPlus (the default) or L-VEL, or to one of the following two-equation
turbulence models: k-Low Reynolds Number k-. The Wall treatment is set to
Automatic (default) or Low Re while Wall functions is the only option for k-.

For more information about turbulence modeling, see Theory for the Turbulent Flow
Interfaces in the CFD Module User’s Guide.

DEPENDENT VARIABLES
The following dependent variables (fields) are defined for this physics interface — the
Velocity field u (SI unit: m/s) and its components, and the Pressure p (SI unit: Pa).

ADVANCED SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog. Normally these settings do not need
to be changed.

524 | CHAPTER 6: FLUID FLOW INTERFACES

The Use pseudo time stepping for stationary equation form option adds pseudo time
derivatives to the equation when the Stationary equation form is used in order to speed
up convergence. When selected, a CFL number expression should also be defined. For
the default Automatic option, the local CFL number (from the Courant–Friedrichs–
Lewy condition) is determined by a PID regulator.

• Domain, Boundary, Point, and Pair Nodes for the Brinkman Equations
Interface
• Theory for the Brinkman Equations Interface

In the COMSOL Multiphysics Reference Manual:

• Pseudo Time Stepping for Laminar Flow Models and Pseudo Time
Stepping
• Numerical Stability — Stabilization Techniques for Fluid Flow
• Nitsche Constraints

Domain, Boundary, Point, and Pair Nodes for the Brinkman

Equations Interface
The Brinkman Equations Interface has the following domain, boundary, point, and
pair nodes, listed in alphabetical order, available from the Physics ribbon toolbar
(Windows users), Physics context menu (Mac or Linux users), or right-click to access
the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

These nodes are described in this section:

• Porous Medium • Volume Force

• Initial Values • Fluid Properties
• Mass Source

THE BRINKMAN EQUATIONS INTERFACE | 525

 The following nodes (listed in alphabetical order) are described for the Laminar Flow
interface in the COMSOL Multiphysics Reference Manual:

• Flow Continuity • Periodic Flow Condition

• Inlet • Point Mass Source
• Line Mass Source • Pressure Point Constraint
• Boundary Stress • Symmetry
• Outlet • Wall
• Open Boundary

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Porous Medium
The Porous Medium node adds the Brinkman equations: Equation 4-34 and
Equation 4-35 (excluding any mass sources). It has two default subfeatures — the
Fluid subfeature where the fluid density (SI unit: kg/m3) and dynamic viscosity  (SI
unit: Pa·s) are defined and the Porous Matrix subfeature where the permeability  (SI
unit: m2) and porosity p (dimensionless) are specified.

COORDINATE SYSTEM SELECTION

Select a coordinate system from the Coordinate system list for the interpretation of
directions in anisotropic material properties. The default is the Global coordinate
system, and the list contains any additional orthonormal coordinate system (except
boundary coordinate systems) added under the Definitions node.

The subnodes inherit these coordinate system settings. In particular, the Permeability
(in the Porous Matrix subnode) should be set according to the coordinate system
selected in this section.

526 | CHAPTER 6: FLUID FLOW INTERFACES

FLOW MODEL
Define if the flow follows a linear or nonlinear pressure-velocity relationship.

• Darcian flow (default) defines a linear relationship characteristic for low Reynolds
number laminar flows.
• Non-Darcian flow defines a nonlinear relationship due to inertial or turbulence
effects.

POROUS MEDIUM TURBULENCE MODEL

If in the parent feature node a turbulence model other than Algebraic yPlus or L-VEL is
selected, define a model which complements the basic turbulence model and accounts
for the influence of the porous matrix on turbulent flow in porous media.

• For Default and Nakayama-Kuwahara specify the Coefficient of subgrid turbulence

generation by porous matrix Cstpm which by default is set to 0.212.
• For Pedras-de Lemos, specify the Choice of the Pedras-de Lemos model coefficient.
When User defined is chosen, specify the Coefficient of subgrid turbulence generation
by porous matrix Cstpm with default value 0.212. Choosing Original results in
C stpm = 0.28   p , while choosing Recalibrated results in Cstpm  0.18.
• If Turbulence model type is RANS-RSM, specify the Production of Reynolds stresses by
porous matrix. The default is Isotropic and the other option is Streamwise.

DISTANCE EQUATION
This section is available for the turbulence models Algebraic yPlus or L-VEL. Select how
the Reference length scale lref (SI unit: m) is defined — Automatic (default) or Manual:

• For Automatic the wall distance is automatically evaluated to one tenth of the
shortest side of the geometry bounding box. This is usually quite accurate but it can
sometimes give a too high value if the geometry consists of several slim entities. In
such cases, define the reference length scale manually.
• For Manual it defines a different value or expression for the length scale. The default
is 1 m.

lref controls the result of the distance equation. Objects that are much smaller than lref
are effectively be diminished while the distance to objects much larger than lref are
accurately represented.

THE BRINKMAN EQUATIONS INTERFACE | 527

 LOCATION IN USER INTERFACE

Context Menus
Brinkman Equations > Porous Medium

Ribbon
Physics tab with Brinkman Equations selected:

Domains > Porous Medium

Fluid
This node defines the material properties of the fluid of the Porous Medium parent
node.

MODEL INPUT
This section contains fields and values that are inputs for expressions defining material
properties. If such user-defined property groups are added, the model inputs appear
here.

FLUID PROPERTIES
Specify the Density  and the Dynamic viscosity  (SI unit: Pa·s) of the fluid. The
defaults use values From material, for User defined enter a value or expression.

The dynamic viscosity describes the relationship between the shear stresses and the
shear rate in a fluid. Intuitively, water and air have a low viscosity, and substances often
described as thick, such as oil, have a higher viscosity.

If the Polymer Flow Module license is available and for Darcian flow, also
non-Newtonian models are available and the Constitutive Relation can be switched
between Specify dynamic viscosity (default) or Inelastic non-Newtonian.

Constitutive Relation
Use the default option to either get the dynamic viscosity from a material, or to add a
user defined expression. The default option lets you either use the dynamic viscosity
from a material or define it with a custom expression. The Inelastic non-Newtonian
option provides several models and defines the Apparent shear rate.

Inelastic Non-Newtonian
The following inelastic non-Newtonian Models are available for the Fluid feature under
the Porous Medium feature:

• Power Law

528 | CHAPTER 6: FLUID FLOW INTERFACES

• Carreau
• Carreau-Yasuda
• Cross
• Cross-Williamson
• Sisko

The setup of these models in porous media is similar to their definition under the Fluid
Properties feature.

Apparent Shear Rate

ꞏ
To define the apparent shear rate  app , either enter a user-defined expression for the
correction factor () or choose the Capillary bundle approach from the list and provide
a value for the Tortuosity factor (C).

• Read more about the theory of Non-Newtonian Fluids in Porous

Media in the Polymer Flow Module User’s Guide

LOCATION IN USER INTERFACE

A default Fluid node is automatically added to the Porous Medium node.

Context Menus
Brinkman Equations > Porous Medium > Fluid

Porous Matrix
This node defines the porosity and material properties of the solid matrix of the Porous
Medium parent node.

MODEL INPUT
This section contains fields and values that are inputs for expressions defining material
properties. If such user-defined property groups are added, the model inputs appear
here.

MATRIX PROPERTIES
The Porosity p (a dimensionless number between 0 and 1) uses by default the value
From material. For User defined the default value is 1.

THE BRINKMAN EQUATIONS INTERFACE | 529

 Select a Permeability model to specify the capacity of the porous material to transmit
flow. Depending on the Flow model selection in the parent Porous Medium feature node
you can choose between the following options:

• Permeability (the default) to directly enter the permeability of the porous matrix or
Kozeny-Carman to define the permeability from the porosity and mean Particle
diameter for Darcian flow or
• Forchheimer or Ergun for Non-Darcian flow.

Note that some of these options are only available for certain modules. For a detailed
overview of the functionality available in each product, visit https://
[Link]/products/specifications/.

The default Permeability  (SI unit: m2) uses the value From material, as defined by the
Porous material list. For User defined select Isotropic to define a scalar value or Diagonal
or Symmetric to define a tensor value.

For Kozeny-Carman, enter the mean Particle diameter dp (SI unit: m), the default value
is 0.5 mm.

In case a Non-Darcian flow model is chosen in the parent node, Forchheimer is the
default selection for the Permeability model. For Forchheimer, the default Permeability
(SI unit: m2) uses the value From material. Furthermore, the dimensionless
Forchheimer parameter cF can be defined. The default value is 0.55.

If Ergun is selected, enter the mean Particle diameter dp (default value: 0.5 mm). The
permeability  is then calculated using Equation 2-4 in the Porous Media Flow
Module User’s Guide.

• About Darcian and Non-Darcian Flow in the Porous Media Flow

Module User’s Guide
• Permeability Models in the Porous Media Flow Module User’s Guide

LOCATION IN USER INTERFACE

A default Porous Matrix node is automatically added to the Porous Medium node.

Context Menus
Brinkman Equations > Porous Medium > Porous Matrix

530 | CHAPTER 6: FLUID FLOW INTERFACES

Mass Source
The Mass Source node adds a mass source (or mass sink) Qm to the right-hand side of
the continuity equation: Equation 6-72. This term accounts for mass deposit and/or
mass creation in porous domains. The physics interface assumes that the mass
exchange occurs at zero velocity.

 
   +    u  = Q m (6-20)
t p

DOMAIN SELECTION
Only Porous Matrix domains are available.

MASS SOURCE
Enter a value or expression for the Source term Qbr (SI unit: kg/(m3·s)). The default
is 0 kg/(m3·s).

LOCATION IN USER INTERFACE

Context Menus
Brinkman Equations > Mass Source

Ribbon
Physics tab with Brinkman Equations selected:

Domains > Mass Source

Volume Force
Use the Volume Force node to specify the force F on the right-hand side of
Equation 6-73. It then acts on each fluid element in the specified domains.

-  u u
---- +  u    ----- =
p   t  p
(6-21)
1 T 2   Q m
–  p +   -----    u +  u   – ---     u I  –   – 1  + -------
- u + F
p  3    p2 

VOLUME FORCE
Enter the components of Volume force F (SI unit: N/m3).

THE BRINKMAN EQUATIONS INTERFACE | 531

 LOCATION IN USER INTERFACE

Context Menus
Brinkman Equations > Volume Source

Ribbon
Physics tab with Brinkman Equations selected:

Domains > Volume Source

Initial Values
The Initial Values node adds initial values for the velocity field and the pressure that can
serve as an initial condition for a transient simulation or as an initial guess for a
nonlinear solver.

INITIAL VALUES
Enter initial values or expressions for the Velocity field u (SI unit: m/s) and the Pressure
p (SI unit: Pa). The default values are 0 m/s and 0 Pa, respectively.

LOCATION IN USER INTERFACE

Context Menus
Brinkman Equations > Initial Values

Ribbon
Physics tab with Brinkman Equations selected:

Domains > Initial Values

Fluid Properties
The Fluid Properties node adds the momentum and continuity equations to solve for
free flow in nonporous domains. The node also provides an interface for defining the
material properties of the fluid.

MODEL INPUTS
Fluid properties, such as density and viscosity, can be defined through user inputs,
variables, or by selecting a material. For the latter option, additional inputs — for
example, temperature or pressure — may be required to define these properties.

532 | CHAPTER 6: FLUID FLOW INTERFACES

Temperature
By default, the single-phase flow interfaces are set to model isothermal flow. Hence,
the Temperature is User defined and defaults to 293.15 K. If a Heat Transfer interface
is included in the component, the temperature may alternatively be selected from this
physics interface. All physics interfaces have their own tags (Name). For example, if a
Heat Transfer in Fluids interface is included in the component, the Temperature (ht)
option is available.

Absolute Pressure
This input appears when a material requires the absolute pressure as a model input.
The absolute pressure is used to evaluate material properties, but it also relates to the
value of the calculated pressure field. There are generally two ways to calculate the
pressure when describing fluid flow: either to solve for the absolute pressure or for a
pressure (often denoted gauge pressure) that relates to the absolute pressure through
a reference pressure.

The choice of pressure variable depends on the system of equations being solved. For
example, in a unidirectional incompressible flow problem, the pressure drop over the
modeled domain is probably many orders of magnitude smaller than the atmospheric
pressure, which, when included, may reduce the stability and convergence properties
of the solver. In other cases, such as when the pressure is part of an expression for the
gas volume or the diffusion coefficients, it may be more convenient to solve for the
absolute pressure.

The default Absolute pressure pA is p  pref where p is the dependent pressure variable
from the Navier–Stokes equations, and pref is from the user input defined at the
physics interface level. When pref is nonzero, the physics interface solves for a gauge
pressure. If the pressure field instead is an absolute pressure field, pref should be set to
0.

The Absolute pressure field can be edited by clicking Make All Model Inputs Editable
( ) and entering the desired value in the input field.

FLUID PROPERTIES
If density variations with respect to pressure are to be included in the computations,
the flow must be set to compressible.

The Dynamic viscosity  describes the relationship between the shear rate and the shear
stresses in a fluid. Intuitively, water and air have low viscosities, and substances often
described as thick (such as oil) have higher viscosities.

THE BRINKMAN EQUATIONS INTERFACE | 533

 LOCATION IN USER INTERFACE

Context Menus
Brinkman Equations > Fluid Properties

Ribbon
Physics tab with Brinkman Equations selected:

Domains > Fluid Properties

Gravity
The Gravity global feature is automatically added when Include gravity is selected at the
interface level in the Physical Model settings. It defines the gravity forces from the
Acceleration of gravity value. When a turbulence model that solves for the turbulent
kinetic energy, k, is used, the option Include buoyancy-induced turbulence is available.
When selected, the Buoyancy contribution is by default set to Automatic from
multiphysics. Contributions are only obtained from multiphysics couplings that
support buoyancy-induced turbulence, such as Nonisothermal Flow. If the Buoyancy
contribution is switched to User defined, a text field for the Turbulent Schmidt number
appears. See also Theory for Buoyancy-Induced Turbulence in the CFD Module
User’s Guide.

This feature requires a specific license. For a detailed overview of the functionality
available in each product visit [Link]

ACCELERATION OF GRAVITY
The Acceleration of gravity (SI unit m/s, default value gconstez in 2D axial symmetry
and 3D and gconstey in 2D) is used to define the gravity forces. It should be a global
quantity.

LOCATION IN USER INTERFACE

Context Menus
Brinkman Equations > Gravity

534 | CHAPTER 6: FLUID FLOW INTERFACES

The Multiphase Flow in Porous Media
Interface
The Multiphase Flow in Porous Media interface ( ) combines the Darcy’s Law interface
with the Phase Transport in Porous Media interface to model the flow and transport of
multiple immiscible phases in a porous medium.

When a predefined Multiphase Flow in Porous Media interface is added from the Porous
Media and Subsurface Flow branch ( ) of the Model Wizard or Add Physics window,
Phase Transport and Darcy’s Law interfaces are added to the Model Builder. In addition,
the Multiphysics node is added, which automatically includes the Multiphase Flow in
Porous Media multiphysics coupling.

On the Constituent Physics Interfaces

The Phase Transport in Porous Media interface is used to simulate the transport of
immiscible multiple species in a porous medium and solves for their averaged volume
fractions (also called saturations). The interfaces between the different phases are not
tracked explicitly, although microscopic interfacial effects are taken into account in the
macroscopic equations through the capillary pressure functions.

The Darcy’s Law interface is used to simulate fluid flow through interstices in a porous
medium. It can be used to model low-velocity flows in media where the permeability
and porosity are very small, and for which the pressure gradient is the major driving
force and the flow is mostly influenced by the frictional resistance within the pores.

SETTINGS FOR PHYSICS INTERFACES AND COUPLING FEATURES

When physics interfaces are added using the predefined couplings, for example
Multiphase Flow in Porous Media, specific settings are included with the physics
interfaces and the coupling features.

However, if physics interfaces are added one at a time, followed by the coupling
features, these modified settings are not automatically included.

For example, if single Phase Transport in Porous Media and Darcy’s Law interfaces are
added, an empty Multiphysics node appears in the model tree. You can choose from the

THE MULTIPHASE FLOW IN POROUS MEDIA INTERFACE | 535

 available coupling features but the settings in the constituent interfaces are not
modified.
TABLE 6-1: MODIFIED SETTINGS FOR A MULTIPHASE FLOW IN POROUS MEDIA INTERFACE.

PHYSICS INTERFACE OR MODIFIED SETTINGS

COUPLING FEATURE

Darcy’s Law In the Darcy’s Law interface, linear shape functions are
used for the discretization of the pressure.
Multiphase Flow in Porous The Domain Selection is the same as that of the
Media participating physics interfaces.
The corresponding Phase Transport in Porous Media and
Darcy’s Law interfaces are preselected in the Coupled
Interfaces section.

Multiphase Flow in Porous Media (Multiphysics Coupling)

The Multiphase Flow in Porous Media multiphysics coupling node ( ) links
bidirectionally the Phase Transport in Porous Media and Darcy’s Law interfaces.

SETTINGS
The Label is the default multiphysics coupling feature name.

The Name is used primarily as a scope prefix for variables defined by the coupling node.
Refer to such variables in expressions using the pattern <name>.<variable_name>. In
order to distinguish between variables belonging to different coupling nodes or physics
interfaces, the name string must be unique. Only letters, numbers, and underscores (_)
are permitted in the Name field. The first character must be a letter.

The default Name (for the first multiphysics coupling feature in the model) is mfpm1.

DOMAIN SELECTION
When nodes are added from the context menu, you can select Manual (the default)
from the Selection list to choose specific domains to define the multiphase flow
coupling, or select All domains as needed.

When Multiphase Flow in Porous Media is added as an effect of adding a Multiphase Flow
in Porous Media interface, the selection is the same as for the participating physics
interfaces.

Only domains that are active in the participating physics interfaces (which are chosen
from the Coupled Interfaces selection list) can be selected.

536 | CHAPTER 6: FLUID FLOW INTERFACES

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Phase
transport and Flow in porous media lists include all applicable physics interfaces.

The default values depend on how the coupling node is created.

• If it is added from the Physics ribbon (Windows users), Physics contextual toolbar
(Mac and Linux users), or context menu (all users), then the first physics interface
of each type in the component is selected as the default.
• If it is added automatically when a multiphysics interface is selected in the Model
Wizard or Add Physics window, then the two participating physics interfaces are
selected.

You can also select None from either list to uncouple the Multiphase Flow in Porous
Media node from a physics interface. If the physics interface is removed from the Model
Builder, for example Darcy’s Law is deleted, then the Flow in porous media list defaults
to None as there is nothing to couple to.

Please find related information under

• The Phase Transport Interfaces and

• Theory for the Phase Transport Interfaces

in the CFD Module User’s Guide.

Well (Multiphysics Coupling)

The Multiphase Flow in Porous Media ( ) interface contains an optional Well
multiphysics coupling feature on points (in 2D) or edges (in 3D). The Well coupling
node is intended to model injection or production wells and links bidirectionally the
Phase Transport in Porous Media and Darcy’s Law interfaces.

The Subsurface Flow Module license is required to use this coupling feature (see
[Link]

SETTINGS
The Label is the default multiphysics coupling feature name.

The default Name (for the first Well multiphysics coupling feature in the model) is
wellmpp1.

THE MULTIPHASE FLOW IN POROUS MEDIA INTERFACE | 537

 POINT OR EDGE SELECTION
When nodes are added from the context menu, you can select Manual (the default)
from the Selection list to choose specific points or edges, or select All points or All edges
as needed.

COUPLED INTERFACES
This section controls which individual interfaces are coupled by the current coupling
feature. If a physics interface is deleted and then added to the model again, then in
order to reestablish the coupling, you need to choose the correct physics interface
again from the Phase transport or Flow in porous media lists

WELL
Select whether the well is a production or injection well: for an injection well positive
mass fluxes signify inflowing fluxes and for a production well positive mass fluxes are
outflowing fluxes.

Either specify the total Total mass flow rate M0 (SI unit: kg/s), the Total mass flow rate
per unit length Ml (SI unit: kg/(m·s)), the Pressure p0 (SI unit: Pa), the Pressure head
Hp0 (SI unit: m), or the Hydraulic head H0 (SI unit: m). The elevation D used to
convert the hydraulic head to a pressure, is defined by the gravity effects in the coupled
Darcy’s Law interface.

Select the Specify well diameter checkbox to define the Well diameter dw (SI unit: m).
This averages the mass source on the boundary of a cylinder around the line. This
option avoids obtaining an increasing pressure value at the line when meshing finer
than this diameter. It makes use of the circavg operator.

If the well diameter is defined, the option to specify the Total mass flux N0 (SI unit:
kg/(m2·s)) is also available.

PHASES
The number of Phase sections depends on the number of phases defined in the coupled
Phase Transport in Porous Media interface: the number of sections is equal to the
number of phases not computed from the volume constraint, which is in turn equal to
the number of phases defined in the coupled Phase Transport in Porous Media
interface minus one.

Specify the boundary condition for each phase not computed from the volume
constraint. You can choose to enter a value for the Mass flow rate M0,i (SI unit: kg/s),
the Mass flow rate per unit length Ml,i (SI unit: kg/(m·s)), or the Volume fraction s0,i
(dimensionless). If the well diameter is specified, the option to specify the Mass flux N0,i

538 | CHAPTER 6: FLUID FLOW INTERFACES

(SI unit: kg/(m2·s)) is also available. For a production well, you can additionally
choose the Outflow option. For an injection well you can additionally choose to specify
the Fraction of total volumetric flow rate s0,i (dimensionless).

Fraction of total volumetric flow rate

When the Fraction of total volumetric flow rate s0,i is specified for a phase, a mass flow
rate per unit length Ml,i is supplied for that phase, which is given by

M l i =  i s 0 i V l (6-22)

where Vl is the total volumetric flow rate per unit length, which is computed as follows:
assume that there are N phases, and that phase 1 is the phase computed from the
volume constraint, that for phases i, for i    p, the fraction of total volumetric flow
rate s0,i is supplied, and that for the phases i, for i  p+1  N, the mass flow rate per
unit length Ml,i is given. Then Vl can be computed from the relation

p N
M l 1 M l i
V l = -----------
1
-+  V l s i 0 +  -----------
i
(6-23)
i=2 i = p+1

where the mass flow rate Ml,1 for the phase computed from the volume constraint is
given by

p N

M l 1 = M l –  V l  i s i 0 –  M l i (6-24)
i=2 i = p+1

Note that the total mass flow rate per unit length Ml is readily available when any of
the Total mass flow rate, Total mass flow rate per unit length, or the Total mass flux
options is chosen. Furthermore, the total mass flow rate per unit length Ml is also
available as a Lagrange multiplier in case any of the Pressure, the Pressure head, or the
Hydraulic head is given, as these conditions are implemented as weak constraints. For
the mass flow rates per unit length Ml,i of the individual phases, it holds similarly that
they are readily available when any of the Mass flow rate, Mass flow rate per unit length,
or the Mass flux options are chosen, and that for the Volume fraction option,
implemented as a weak constraint, the mass flow rate per unit length Ml,i is computed
as a Lagrange multiplier.

Outflow
When the Outflow option is specified for a phase si, an outflowing mass flow rate per
unit length Ml,i is supplied for that phase, which is given by

THE MULTIPHASE FLOW IN POROUS MEDIA INTERFACE | 539

 N
 i  ri  i  ri
M l i = ------------ M l
r i
with  r =  -----------
i
- (6-25)
i=1

where ri denotes the relative permeability of phase i.

Thin Barrier (Multiphysics Coupling)

The Multiphase Flow in Porous Media ( ) interface contains an optional Thin Barrier
multiphysics coupling feature on edges (in 2D) or surfaces (in 3D). The Thin Barrier
coupling node can be used to avoid meshing interior permeable structures that are less
permeable than the adjacent porous domains. It links bidirectionally the Phase
Transport in Porous Media and Darcy’s Law interfaces.

SETTINGS
The Label is the default multiphysics coupling feature name.

The default Name (for the first Thin Barrier multiphysics coupling feature in the model)
is tbls1.

BOUNDARY SELECTION
When nodes are added from the context menu, you can select Manual (the default)
from the Selection list to choose specific boundaries, or select All boundaries as needed.

COUPLED INTERFACES
This section controls which individual interfaces are coupled by the current coupling
feature. If a physics interface is deleted and then added to the model again, then in
order to reestablish the coupling, you need to choose the correct physics interface
again from the Phase transport or Flow in porous media lists

SHELL PROPERTIES
The Thin Barrier multiphysics coupling feature is designed to be only applicable for a
Single Layer Material. Select the Single Layer Material to be used for the thin barrier from
the Layer list. The Thickness of the thin barrier is set in the Layer Definition section of
the Shell Property Group window of the Single Layer Material. Also use this section to
set the number of Mesh elements that are used in the Extra Dimension to solve the
equations through the thickness of the thin barrier layer.

540 | CHAPTER 6: FLUID FLOW INTERFACES

MODEL INPUT
The material properties of the different phases, such as density and viscosity, can be
defined through user inputs, variables, or by selecting a boundary material. For the
latter option, additional inputs, for example temperature or pressure, may be required
to define these properties. These model inputs appear in this section.

MATRIX PROPERTIES
Select the material to use as porous matrix inside the thin barrier. Select Boundary
material (the default) from the Porous material list to use the material defined for the
thin barrier boundary. Select another material to use that material’s properties.

The default Porosity p (a dimensionless number between 0 and 1) uses the value From
material, defined by the Porous material selected. For User defined the default is 0.5.

The default Permeability (SI unit: m2) uses the value From material, as defined by the
Porous material selected. For User defined the default is 1e-9 m2.

CAPILLARY PRESSURE
Select a Capillary pressure model — User defined (the default), van Genuchten, Brooks
and Corey, or Capillary diffusion. The van Genuchten and Brooks and Corey options are
only available when there are not more than two phases present in the model.

• For User defined, enter an expression for the Capillary pressure pc (SI unit: Pa) for
the phases that are not computed from the volume constraint. The pressure of the
corresponding phase is equal to the pressure of the phase computed from the
volume constraint plus the capillary pressure supplied for this phase.

• For van Genuchten, enter an Entry capillary pressure pec (SI unit: Pa), the Constitutive
relations constant mvG (dimensionless), and Constitutive relations constant lvG
(dimensionless).
• For Brooks and Corey, enter an Entry capillary pressure pec (SI unit: Pa) and the Pore
size distribution index p (dimensionless).
• For Capillary diffusion, enter a capillary diffusion coefficient Dc (SI unit: m2/s) for
the phases that are not computed from the volume constraint.

PHASE PROPERTIES
Select the material to use for each phase. Select Boundary material (the default) from
the Fluid list to use the material defined for the domain. Select another material to use
that material’s properties.

THE MULTIPHASE FLOW IN POROUS MEDIA INTERFACE | 541

 Enter the Density  (SI unit: kg/m3), Dynamic viscosity (SI unit: Pa·s) and Relative
permeability r (a dimensionless number between 0 and 1) for each phase. The defaults
for the Density and Dynamic viscosity are From material, and User defined for the Relative
permeability. For User defined, the default values are 1000 kg/m3, 0.001 Pa·s. and s i2 ,
respectively.

The Relative permeability input fields are only available when the Capillary pressure
model is set to User defined or Capillary diffusion. In the other cases (van Genuchten or
Brooks and Corey) the relative permeabilities are defined by the van Genuchten or Brooks
and Corey model. For the van Genuchten and Brooks and Corey model, specify the
Residual saturation sri (dimensionless, default value 0) for each phase.

INITIAL VALUES
Enter a value or expression in the field for the volume fraction inside the thin barrier
of each phase, except for the one computed from the volume constraint, and in the
field for the pressure of the phase computed from the volume constraint.

Please find related information under

• Single Layer Materials in the COMSOL Multiphysics Reference

Manual and
• Theory for the Multiphase Flow in Porous Media Interface in the
Subsurface Flow Module User’s Guide.

542 | CHAPTER 6: FLUID FLOW INTERFACES

T he M ulti pha s e Free an d Porou s
Media Flow Interface
The Multiphase Free and Porous Media Flow interface ( ) combines the Phase
Transport in Free and Porous Media interface with the Darcy’s Law and Laminar Flow
interfaces to model the transport of multiple immiscible phases in coupled free and
porous media flow.

When a predefined Multiphase Free and Porous Media Flow interface is added from the
Porous Media and Subsurface Flow branch ( ) of the Model Wizard or Add Physics
window, the Phase Transport in Free and Porous Media, Darcy’s Law and Laminar Flow
interfaces are added to the Model Builder. In addition, the Multiphysics node is added,
which automatically includes the Multiphase Flow in Porous Media, the Mixture Model,
and the Free and Porous Media Flow Coupling multiphysics couplings.

On the Constituent Physics Interfaces

The Phase Transport in Free and Porous Media Flow interface is used to simulate the
transport of multiple immiscible species in free and porous media flow and solves for
their averaged volume fractions (also called saturations). The interfaces between the
different phases are not tracked explicitly, although microscopic interfacial effects are
taken into account in the macroscopic equations through the capillary pressure
functions in porous domains.

The Darcy’s Law interface is used to simulate fluid flow through interstices in a porous
medium. It can be used to model low-velocity flows in media where the permeability
and porosity are very small, and for which the pressure gradient is the major driving
force and the flow is mostly influenced by the frictional resistance within the pores.

The Laminar Flow interface is used to compute the velocity and pressure for a fluid flow
in the laminar flow regime. A flow will remain laminar as long as the Reynolds number
is below a certain critical value. At higher Reynolds numbers, disturbances have a
tendency to grow and cause transition to turbulence.

SETTINGS FOR PHYSICS INTERFACES AND COUPLING FEATURES

When physics interfaces are added using the predefined couplings, for example
Multiphase Free and Porous Media Flow, specific settings are included with the physics
interfaces and the coupling features.

THE MULTIPHASE FREE AND POROUS MEDIA FLOW INTERFACE | 543

 However, if physics interfaces are added one at a time, followed by the coupling
features, these modified settings are not automatically included.
TABLE 6-2: MODIFIED SETTINGS FOR A MULTIPHASE FREE AND POROUS MEDIA FLOW INTERFACE.

PHYSICS INTERFACE OR MODIFIED SETTINGS

COUPLING FEATURE

Darcy’s Law In the Darcy’s Law interface, linear shape functions are
used for the discretization of the pressure.
Multiphase Flow in Porous The Domain Selection is the same as that of the
Media participating physics interfaces.
The corresponding Phase Transport in Free and Porous
Media Flow and Darcy’s Law interfaces are preselected in
the Coupled Interfaces section.
Mixture Model The Domain Selection is the same as that of the
participating physics interfaces.
The corresponding Phase Transport in Free and Porous
Media Flow and Laminar Flow interfaces are preselected
in the Coupled Interfaces section.
Free and Porous Media The Boundary Selection is the same as that of the
Flow Coupling participating physics interfaces.
The corresponding Laminar Flow and Darcy’s Law
interfaces are preselected in the Coupled Interfaces
section.

Coupling Features
Three coupling features are added: The Multiphase Flow in Porous Media
(Multiphysics Coupling), the Free and Porous Media Flow Coupling (Multiphysics
Coupling), and the Mixture Model multiphysics coupling.

Please find related information under

• Theory for the Phase Transport Mixture Model Interfaces

• Theory for the Free and Porous Media Flow, Darcy Interface
• Theory for the Multiphase Flow in Porous Media Interface

in the CFD Module User’s Guide.

544 | CHAPTER 6: FLUID FLOW INTERFACES

The Bubbly Flow Interface
In this section:

• The Laminar Bubbly Flow Interface

• Domain and Boundary Nodes for Bubbly Flow

The Laminar Bubbly Flow Interface

The Laminar Bubbly Flow (bf) interface ( ), found under the Multiphase Flow > Bubbly
Flow branch ( ) when adding a physics interface, is used to model the flow of liquids
with dispersed bubbles at low and moderate Reynolds numbers.

It is assumed that the bubbles only occupy a small volume fraction and that they always
travel with their terminal velocity. It is thereby possible to solve only one set of Navier–
Stokes equations for the liquid phase and to let the velocity of the bubbles be guided
by a slip model. The pressure distribution is computed from a mixture-averaged
continuity equation. The volume fraction of bubbles is tracked by solving a transport
equation for the effective gas density.

The physics interface can also model the distribution of the number density, that is, the
number of bubbles per unit volume which in turn can be used to calculate the
interfacial area, useful when simulating chemical reactions in the mixture.

The main physics node is the Fluid Properties feature, which adds the equations for
laminar bubbly flow and provides an interface for defining the fluid materials for the
liquid and the gas and the slip velocity model to use.

When this physics interface is added, the following default physics nodes are also added
in the Model Builder — Laminar Bubbly Flow, Fluid Properties, Wall (the default boundary
types are No slip for the liquid and No gas flux for the gas), and Initial Values. Then, from
the Physics toolbar, add other nodes that implement, for example, boundary
conditions and volume forces. You can also right-click Laminar Bubbly Flow to select
physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern

THE BUBBLY FLOW INTERFACE | 545

 <name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is bf.

PHYSICAL MODEL
Specify if the gas concentration is low and whether or not to solve for the interfacial
area.

Low Gas Concentration

The Low gas concentration checkbox is selected by default. This approximation is valid
if the gas volume fraction is low (g less than a few percent) and its density does not
have any significant effects on the continuity equation. It is then generally valid to
replace the continuity equation in The Bubbly Flow Equations.

Solve for Interfacial Area

To add a transport equation for the bubble density in order to determine the interfacial
area, select the Solve for interfacial area checkbox.

Reference Values
Reference values are global quantities used to evaluate the density of the liquid and the
absolute pressure pA.

Reference pressure level There are generally two ways to include the pressure in fluid
flow computations: either to use the absolute pressure pA  p  pref, or the gauge
pressure p. When pref is nonzero, the physics interface solves for the gauge pressure
whereas material properties are evaluated using the absolute pressure. The reference
pressure level is also used to define the density of the liquid. The default Reference
pressure level pref (SI unit: Pa) is 1 atm.

Reference temperature The reference temperature is used to define the density of the
liquid. The default Reference temperature Tref (SI unit: K) is 293.15 K.

Swirl Flow
For 2D axisymmetric models, select the Swirl flow checkbox to include the swirl
velocity component — that is, the velocity component u in the azimuthal direction.
While u can be nonzero, there can be no gradients in the  direction. Also see in the
CFD Module User’s Guide.

546 | CHAPTER 6: FLUID FLOW INTERFACES

DEPENDENT VARIABLES
The dependent variables (field variables) are the Velocity field, liquid phase u (SI unit:
m/s), the Pressure p (SI unit: Pa), the Effective gas density rhogeff (SI unit: kg/m3),
and the Number density, gas phase nd (SI unit: 1/m3). The names can be changed but
the names of fields and dependent variables must be unique within a component.

CONSISTENT STABILIZATION AND INCONSISTENT STABILIZATION

To display this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog. Consistent streamline and crosswind diffusion is
applied by default to both gas and momentum transport.

Select the Use dynamic subgrid time scale checkbox to estimate the influence of
temporal variations in the velocity field, instead of using the inverse time step, in the
consistent stabilization terms. This checkbox is selected by default

Select the Limit small time steps effect on stabilization time scale checkbox to prevent
loss of stabilization at small time steps.

Additional inconsistent stabilization terms may be added when required as isotropic

diffusion. The Inconsistent Stabilization section contains the settings for stabilization of
the momentum transport (the fluid flow) in the Momentum transport area and
stabilization of the equation for the dispersed phase in the Gas phase transport area.

ADVANCED SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog. Normally these settings do not need
to be changed.

Select the Use pseudo time stepping for stationary equation form checkbox to add
pseudo time derivatives to the equation when the Stationary equation form is used.
When selected, also choose a CFL number expression — Automatic (the default) or
Manual. Automatic sets the local CFL number (from the Courant–Friedrichs–Lewy

THE BUBBLY FLOW INTERFACE | 547

 condition) to the built-in variable CFLCMP which in turn triggers a PID regulator for
the CFL number. For Manual enter a Local CFL number CFLloc (dimensionless).

• Pseudo Time Stepping for Laminar Flow Models in this guide and
Pseudo Time Stepping in the COMSOL Multiphysics Reference
Manual
• Domain and Boundary Nodes for Bubbly Flow
• Theory for the Bubbly Flow Interface

Domain and Boundary Nodes for Bubbly Flow

The following domain and boundary nodes are available from the Physics ribbon
toolbar (Windows users), Physics context menu (Mac or Linux users), or right-click to
access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

DOMAIN
• Fluid Properties
• Volume Force (described for The Laminar Flow and Creeping Flow Interfaces)
• Mass Transfer
• Gravity
• Initial Values

BOUNDARY

The sections describe the available boundary conditions for the liquid and
the gas. In all equations, n denotes the outward pointing unit vector
normal to the boundary.

The boundary types for the liquid flow variables, described in this section, are:

• Wall (the default boundary condition)

548 | CHAPTER 6: FLUID FLOW INTERFACES

• Inlet
• Outlet
• Symmetry
• Periodic Flow Condition
• Interior Wall

PAIR
• Flow Continuity
In addition to the boundary conditions for the liquid, the following boundary
conditions for the gas are available for all boundary condition types except symmetry
and interior wall:

• Gas Concentration (the default condition for inlets)

• Gas Flux
• Gas Outlet (the default condition for outlets)
• No Gas Flux (the default condition for walls)
• Symmetry

Gas Boundary Condition Equations

For 2D axisymmetric components, COMSOL Multiphysics takes the axial

symmetry boundaries (at r = 0) into account and automatically adds an
Axial Symmetry node that is valid on the axial symmetry boundaries only.

Fluid Properties
The Fluid Properties node contains the material properties for the liquid and the gas. It
also contains settings for the slip model.

MODEL INPUTS
Fluid properties, such as the gas density and liquid viscosity, can be defined through
user inputs, variables, or by selecting a material. For the latter option, additional
inputs, for example temperature or pressure, may be required to define these
properties.

THE BUBBLY FLOW INTERFACE | 549

 Temperature
By default, the Temperature model input is set to Common model input, and the
temperature is controlled from Default Model Inputs under Global Definitions or by a
locally defined Model Input. If a Heat Transfer interface is included in the component,
it controls the temperature Common model input. Alternatively, the temperature field
can be selected from another physics interface. All physics interfaces have their own
tags (Name). For example, if a Heat Transfer in Fluids interface is included in the
component, the Temperature (ht) option is available for T.

You can also select User defined from the Temperature model input in order to manually
prescribe T.

Absolute Pressure
This input appears when a material requires the absolute pressure as a model input.
The absolute pressure is used to evaluate material properties, but it also relates to the
value of the calculated pressure field. There are generally two ways to calculate the
pressure when describing fluid flow: either to solve for the absolute pressure or for a
pressure (often denoted gauge pressure) that relates to the absolute pressure through
a reference pressure.

The default Absolute pressure pA is ppref, where p is the dependent pressure variable
from the Navier–Stokes or RANS equations, and pref is from the user input defined at
the physics interface level. When pref is nonzero, the physics interface solves for a
gauge pressure. If the pressure field instead is an absolute pressure field, pref should be
set to 0.

The Absolute pressure field can be edited by clicking Make All Model Inputs Editable
( ) and entering the desired value in the input field.

Model Inputs and Multiphysics Couplings in the COMSOL Multiphysics

Reference Manual

MATERIALS
Select the materials to use for the material properties of the liquid and the gas (when
they are set to take their value from the material). The default is to use the Domain
material for the Liquid and None for the Gas. Select another material to use that
material’s properties for the liquid or gas as needed.

550 | CHAPTER 6: FLUID FLOW INTERFACES

LIQUID PROPERTIES
The default Density, liquid phase l (SI unit: kg/m3) uses values From material. For User
defined enter another value or expression.

The density in a material can depend on temperature and/or pressure and these
dependencies are automatically replaced by pref and Tref, which are specified at the
physics interface level.

The default Dynamic viscosity, liquid phase l (SI unit: Pa·s) uses values From material;
the value is then defined for the material selected in the Materials section for the
continuous phase. For User defined enter another value or expression.

The dynamic viscosity describes the relationship between the shear stresses and the
shear rate in a fluid. Intuitively, water and air have a low viscosity, and substances often
described as thick, such as oil, have a higher viscosity.

GAS PROPERTIES
The default Density, gas phase g (SI unit: kg/m3) uses values From material. For User
defined enter another value or expression. Alternatively, select Calculate from ideal gas
law and enter the Molecular weight M (SI unit: kg/mol) of the gas.

Enter the Bubble diameter db (SI unit: m). The default value is 103 m (1 mm).

SLIP MODEL
Select a Slip model — Homogeneous flow (the default), Pressure-drag balance, or User
defined.

Homogeneous flow assumes that the velocity of the two phases are equal; that is,
uslip  0. For User defined enter different values or expressions for the components of
the Slip velocity uslip (SI unit: m/s).

For Pressure-drag balance it uses a model based on the assumption that the pressure
forces on a bubble are balanced by the drag force:

3 C
--- ------d-  l u slip u slip = – p
4 db

Here db (SI unit: m) is the bubble diameter, and Cd (dimensionless) is the drag
coefficient.

THE BUBBLY FLOW INTERFACE | 551

 Select a Drag coefficient model:

• Small spherical bubbles (Hadamard-Rybczynski) for bubbles with a diameter smaller

than 2 mm.
• Large bubbles for gas bubbles with a diameter larger than 2 mm. Then enter the
Surface tension coefficient  (SI unit N/m). The default is 0.07 N/m.
• Air bubbles in tap water (Schwarz-Turner) for air bubbles of 1–10 mm mean diameter
in water.
User defined to enter a different value or expression for the Drag coefficient Cd
(dimensionless). The default value is 1.

Mass Transfer
Add a Mass Transfer node to include mass transfer from the gas to the liquid.

MASS TRANSFER
Select a Mass transfer model — Two-film theory, Nonequilibrium model, or User defined.
For User defined enter a value or expression for the Mass transfer from gas to liquid mgl
(SI unit: kg/(m3·s)). The default is 0 kg/(m3·s).

Nonequilibrium Model
For Nonequilibrium model enter values or expressions for each of the following:

• Equilibrium volume fraction in dispersed phase eq. The default is 0.

• Inverse time scales fl and fg (SI unit: 1/s). The default is 0 s-1.
The mass transfer is modeled as

m gl =  g  f l max   g –  eq  – f g max   eq –  g  

Two-Film Theory
For Two-film theory enter values for the Mass transfer coefficient k (SI unit: m/s),
Henry’s constant H (SI unit: Pa·m3/mol), the Dissolved gas concentration c (SI unit:
mol/m3), and the Molecular weight of species M (SI unit: kg/mol). Refer to the theory
below for more information.

Henry’s law gives the equilibrium concentration c* of gas dissolved in liquid:

p + p ref
c = ------------------
-
H

552 | CHAPTER 6: FLUID FLOW INTERFACES

where H is Henry’s constant. The molar flux per interfacial area, N (SI unit: mol/
(s·m2)), is determined by

N = k  c – c 

where k is the Mass transfer coefficient and c is the Dissolved gas concentration in
liquid.

The mass transfer from gas to liquid, mgl, is given by

m gl = NMa

where M is the Molecular weight of species and a is the interfacial area per volume
(SI unit: m2/m3).

Also see Theory for the Bubbly Flow Interface for details about how a is
computed.

For two-film theory, you also need to solve for the concentration of the dissolved gas;

c  cu =  m gl
+   l   Dc  + ---------
t M

which can be done by adding a Transport of Diluted Species interface.

Gravity
The Gravity node adds the force g to the right-hand side of the momentum transport
equation. Gravity nodes are mutually exclusive, that is, there can only be one active
Gravity node per domain.

GRAVITY
Enter the components of the Gravity vector, g (SI unit: m/s2). For 2D components
the default Gravity vector is defined as 0g_const. Here g_const is a physical
constant equal to 9.8066 m/s2. For 3D and 2D axisymmetric components, the gravity
acts in the negative z direction by default.

THE BUBBLY FLOW INTERFACE | 553

 Initial Values
The Initial Values node adds initial values for the flow variables and the effective gas
density that can serve as initial conditions for a transient simulation or as an initial guess
for a nonlinear solver.

INITIAL VALUES
Enter initial values or expressions for the Velocity field, liquid phase u (SI unit: m/s)
and for the Pressure p (SI unit: Pa). The default values are 0. Also enter a value or
expression for the Effective gas density rhogeff (SI unit: kg/m3). The default is 0 kg/
m3.

If the Solve for interfacial area checkbox is selected, enter an initial value for the Number
density, gas phase nd (SI unit: 1/m3). The default is 0 1/m3.

Wall
The Wall node adds a selection of boundary conditions that describe the existence of a
solid wall. The Wall node by default specifies no gas flux for the gas phase.

Gas Boundary Condition Equations

LIQUID BOUNDARY CONDITION

Select a Liquid boundary condition — No slip (the default) or Slip.

No Slip
No slip is the default boundary condition for the liquid. A no slip wall is a wall where
the liquid velocity relative to the wall velocity is zero. For stationary walls, it sets the
liquid velocity to zero at the wall:

ul = 0

Slip
Sets the velocity component normal to the wall to zero:

ul  n = 0

554 | CHAPTER 6: FLUID FLOW INTERFACES

GAS BOUNDARY CONDITION
From the list, select a Gas boundary condition for the gas phase on the wall — No gas
flux (the default), Gas concentration, Gas outlet, Gas flux, or Symmetry.

Gas Concentration
For Gas concentration enter the Effective gas density gg0 (SI unit: kg/m3). The
default is 0 kg/[Link] Solve for interfacial area is selected, also select either the Bubble
number density (the default) or the Bubble diameter and gas density button.

• For Bubble number density enter the Bubble number density n0 (SI unit: 1/m3). The
default is 1000 1/m3.
• For Bubble diameter and gas density enter the Bubble diameter db (SI unit: m) (the
default is 1 mm) and Density, gas phase g (SI unit: kg/m3) (the default is 1 kg/
m3).

Gas Flux
For Gas flux enter the Gas mass flux N  g  g (SI unit: kg/(m2·s)) (the default is 0 kg/
(m2·s)) and if the Solve for interfacial area checkbox is selected on the physics interface,
the Number density flux Nn (SI unit: 1/(m2·s)). The default is 0 (1/(m2·s)).

WALL MOVEMENT
This section contains controls to describe the wall movement relative to the spatial
frame.

The Translational velocity setting controls the translational velocity, utr. The list is per
default set to Automatic from frame. When Automatic from frame is selected, the physics
automatically detects if the spatial frame moves. If there is no movement, utr is set to
0. If the frame moves, utr becomes equal to the frame movement. utr is accounted for
in the actual boundary condition prescribed in the Liquid boundary condition section.

When Manual is selected from Translational velocity selection list, the Velocity of moving
wall, utr can be prescribed manually. Specifying the translational velocity manually
does not automatically cause the associated wall to move. An additional Moving Mesh
interface needs to be added to physically track the wall movement in the spatial
reference frame.

The Sliding wall option is appropriate if the surface is sliding in its tangential direction.
A velocity is prescribed at the wall and the boundary itself does not have to actually
move relative to the reference frame.

THE BUBBLY FLOW INTERFACE | 555

 CONSTRAINT SETTINGS
This section is displayed by clicking the Show More Options button ( ) and selecting
Advanced Physics Options.

Wall in the Single-Phase Flow interfaces

Inlet
The Inlet node adds a selection of boundary conditions that describe inlets in fluid-flow
simulations.

Gas Boundary Condition Equations

LIQUID BOUNDARY CONDITION

Select a Liquid boundary condition — Velocity (the default), Pressure, or Fully developed
flow.

VELOCITY
For Velocity click the Normal inflow velocity (the default) or Velocity field buttons.

• For Normal inflow velocity enter a value or expression for U0 (SI unit: m/s).
• For Velocity field specify that the velocity at the boundary is equal to a given u0 (SI
unit: m/s) and enter the components in the matrix:
ul = u0

PRESSURE
This condition specifies the normal stress which in most cases is approximately equal
to the pressure. Enter the Pressure p0 (SI unit: Pa) at the boundary. The default is 0 Pa.
Enter a relative pressure value in p0 (SI unit: Pa). Or if the reference pressure pref
defined at the physics interface level is equal to 0 Pa, enter an absolute pressure in p0.

• The Suppress backflow checkbox is selected by default. This option adjusts the inlet
pressure, p̂ 0  p 0 , locally in order to reduce the amount of fluid exiting the domain
through the boundary. If suppress backflow is deselected, the inlet boundary can
become an outlet depending on the pressure field in the rest of the domain.

556 | CHAPTER 6: FLUID FLOW INTERFACES

• Flow direction controls in which direction the fluid enters the domain.
- For Normal flow (the default) it prescribes zero tangential velocity component.
- For User defined define an Inflow velocity direction, du (dimensionless). The
magnitude of du does not matter, only the direction. du must point into the
domain.

FULLY DEVELOPED FLOW

The Fully developed flow option adds contributions to the inflow boundary, which force
the flow toward the solution for a fully developed channel flow. The channel can be
thought of as a virtual extrusion of the inlet cross section. The velocity is defined by
the boundary condition of the adjacent boundary in the model.

The boundary conditions for the virtual inlet channel are inherited from the
boundaries adjacent to the inlet channel. Virtual boundaries adjacent to walls (except
slip walls) are treated as no-slip walls. Virtual boundaries adjacent to any other type of
boundary are treated as slip-walls (or equivalently as symmetry boundaries).

Select a Fully developed flow option — Average velocity (the default), Flow rate, or
Average pressure.

The Fully developed flow option adds contributions to the inflow boundary, which force
the flow toward the solution for a fully developed channel flow.

Select an option to control the flow rate at the inlet:

• For Average velocity enter Uav.

• Flow rate, V0. Two-dimensional models also require an Entrance thickness, Dz,
which is the out-of-plane thickness of the extruded entrance channel.
• Average pressure, Pav. Observe that Pav is the average pressure on the inflow
boundary.

Fully developed inflow can be used in conjunction with the turbulence models. No
additional inputs are required for the turbulence variables. They are instead solved for
on the inlet boundary to be consistent with the fully developed flow profile.

A fully developed flow boundary cannot be adjacent to any Interior Wall feature

GAS BOUNDARY CONDITION

These settings are the same as for Wall. See Gas Boundary Condition. The only
difference is that Gas concentration is the default.

THE BUBBLY FLOW INTERFACE | 557

 Outlet
The Outlet node adds a set of boundary conditions that describe outlets in fluid-flow
simulations; that is, the conditions at boundaries where the fluid exits the domain.

LIQUID BOUNDARY CONDITION

Select a Liquid boundary condition — Pressure (the default), Velocity, or Fully developed
flow.

PRESSURE CONDITIONS
The Pressure condition specifies the normal stress which in most cases is approximately
equal to the pressure. The tangential stress component is set to 0 N/m2.

• Enter the Pressure p0 (SI unit: Pa) at the boundary. The default is 0 Pa. Enter a
relative pressure value in p0 (SI unit: Pa). Or if the reference pressure pref defined
at the physics interface level is equal to 0 Pa, enter an absolute pressure in p0.
• Select the Normal flow checkbox to change the no tangential stress condition to a no
tangential velocity condition. This forces the flow to exit (or enter) the domain
perpendicularly to the outlet boundary.
• The Suppress backflow checkbox is selected by default. This option adjusts the outlet
pressure in order to reduce the amount of fluid entering the domain through the
boundary.

VELOCITY
For Velocity click the Velocity field (the default) or Normal outflow velocity buttons.

• For Velocity field specify that the velocity at the boundary is equal to a given u0 (SI
unit: m/s) and enter the components in the matrix:
ul = u0

• For Normal outflow velocity enter a value or expression for U0 (SI unit: m/s).

FULLY DEVELOPED FLOW

This boundary condition is applicable when the flow exits the domain into a long pipe
or channel, at the end of which a flow profile is fully developed.

The boundary conditions for the virtual inlet channel are inherited from the
boundaries adjacent to the outlet channel. Virtual boundaries adjacent to walls (except
slip walls) are treated as no-slip walls. Virtual boundaries adjacent to any other type of
boundary are treated as slip-walls (or equivalently as symmetry boundaries).

558 | CHAPTER 6: FLUID FLOW INTERFACES

Select a Fully developed flow option — Average velocity (the default), Flow rate, or
Average pressure.

GAS BOUNDARY CONDITION

The settings for the Gas boundary condition are the same as for Wall. See Gas Boundary
Condition.

When the Gas boundary condition is set to Gas outlet, the Exterior gas condition(s)
section contains an input field for the Effective gas density on the downstream side of
the outlet. If Solve for interfacial area is selected at the physics interface level, additional
input fields appear for the Bubble number density or the Bubble diameter and gas density
on the downstream side of the outlet.

See in the Single-Phase Flow interfaces

Symmetry
The Symmetry node adds boundary conditions that describe symmetry boundaries in
fluid-flow simulations. The boundary condition for symmetry boundaries prescribes
no penetration and vanishing shear stresses:

t T   l   l +  T   u l + u l – ---    u l I  n = 0
T 2
u l  n = 0,
  3 

BOUNDARY SELECTION
A boundary condition for 2D axial symmetry is not required.

For the symmetry axis at r  0, the software automatically provides a condition that
prescribes ur  0 and vanishing stresses in the z direction and adds an Axial Symmetry
node that is valid on the axial symmetry boundaries only.

Periodic Flow Condition

The Periodic Flow Condition splits its selection into a source group and a destination
group. Fluid that leaves the domain through one of the destination boundaries enters
the domain through the corresponding source boundary. This corresponds to a
situation where the geometry is a periodic part of a larger geometry. If the boundaries
are not parallel to each other, the velocity vector is automatically transformed.

THE BUBBLY FLOW INTERFACE | 559

 If the boundaries are curved, the orientation of the source must be specified manually
(see ).

PRESSURE DIFFERENCE
A value or expression should be specified for the Pressure difference, psrc  pdst. This
pressure difference can, for example, drive the fully developed flow in a channel.

To set up a periodic boundary condition, both boundaries must be selected in the

Periodic Flow Condition node. COMSOL Multiphysics automatically assigns one
boundary as the source and the other as the destination. To manually set the
destination selection, a Destination Selection section is available when the Manual
Destination Selection option is selected in the context menu for the Periodic Flow
Condition node.

CONSTRAINT SETTINGS
This section is displayed by clicking the Show More Options button ( ) and selecting
Advanced Physics Options in the Show More Options dialog.

ORIENTATION OF SOURCE
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog. For information about the Orientation
of Source section, see Orientation of Source and Destination in the COMSOL
Multiphysics Reference Manual.

ORIENTATION OF DESTINATION
This section appears if the setting for Transform to intermediate map in the Orientation
of Source section is changed from the default value, Automatic, and Advanced Physics
Options is selected in the Show More Options dialog. For information about the
Orientation of Destination section, see Orientation of Source and Destination in the
COMSOL Multiphysics Reference Manual.

Interior Wall
The Interior Wall node adds a wall boundary condition on interior boundaries. It is
similar to the Wall boundary condition available on exterior boundaries except that it
applies on both sides of an interior boundary. It allows discontinuities of the dependent

560 | CHAPTER 6: FLUID FLOW INTERFACES

variables across the boundary. The Interior Wall node specifies no gas flux for the gas
phase.

Gas Boundary Condition Equations

LIQUID BOUNDARY CONDITION

Select a Liquid boundary condition — No slip (the default) or Slip.

No Slip
No slip is the default boundary condition for the liquid. A no slip wall is a wall where
the liquid velocity relative to the wall velocity is zero. For interior stationary walls, it
sets the liquid velocity to zero on both sides of the wall:

ul = 0

For turbulent bubbly flows, the no slip condition may either be prescribed exactly or
modeled using automatic wall treatment or wall functions depending on the Wall
Treatment setting in the Turbulence section of the interface settings.

Slip
The Slip condition prescribes a no-penetration condition. For an interior stationary
wall it sets the velocity component normal to the interior wall to zero on both sides of
the wall:

ul  n = 0

WALL MOVEMENT
This section contains controls to describe the wall movement relative to the spatial
frame.

The Translational velocity setting controls the translational velocity, utr. The list is per
default set to Automatic from frame. When Automatic from frame is selected, the physics
automatically detects if the spatial frame moves. If there is no movement, utr is set to
0. If the frame moves, utr becomes equal to the frame movement. utr is accounted for
in the actual boundary condition prescribed in the Liquid boundary condition section.

When Manual is selected from Translational velocity selection list, the Velocity of moving
wall, utr can be prescribed manually. Specifying the translational velocity manually
does not automatically cause the associated wall to move. An additional Moving Mesh

THE BUBBLY FLOW INTERFACE | 561

 interface needs to be added to physically track the wall movement in the spatial
reference frame.

CONSTRAINT SETTINGS
This section is displayed by clicking the Show More Options button ( ) and selecting
Advanced Physics Options.

Wall

Flow Continuity
The Flow Continuity condition is suitable for pairs where the boundaries match; it
prescribes that the flow field is continuous across the pair.

The parts of the boundary where a source boundary lacks a corresponding destination
boundary and vice versa will get conditions from other boundary conditions under the
physics interface acting as fallback boundary conditions on the nonoverlapping parts.
By default, the Wall node is used as fallback unless another feature has been selected
for the boundary.

About Pairs

Gas Boundary Condition Equations

In addition to the boundary conditions for the liquid, specify boundary conditions for
the gas on Wall, Inlet, Outlet, and Interior Wall nodes. Select a Gas Boundary Condition:

Gas Concentration
Using this boundary condition, specify the effective gas density.

0
̃ g = ̃ g

Gas Outlet
This boundary condition is appropriate for boundaries where the gas phase flows
outward with the gas velocity, ug, at the boundary.

562 | CHAPTER 6: FLUID FLOW INTERFACES

Gas Flux
Using this boundary condition, specify the gas mass flux through the boundary:

– n   ̃ g u g  = N g  g

Symmetry
This boundary condition, which is useful on boundaries that represent a symmetry line
for the gas flow, sets the gas flux through the boundary to zero:

n   ̃ g u g  = 0

No Gas Flux
This boundary condition represents boundaries where the gas flux through the
boundary is zero:

n   g ug  = 0

In walls moving with a boundary velocity ubnd, the gas flux through the boundary is:

n    g u g  = n    g u bnd 

THE BUBBLY FLOW INTERFACE | 563

 The Mixture Model Interface
In this section:

• The Mixture Model, Laminar Flow Interface

• Domain and Boundary Nodes for the Mixture Model Interface

The Mixture Model, Laminar Flow Interface

The Mixture Model, Laminar Flow (mm) interface ( ), found under the Multiphase Flow
> Mixture Model branch ( ) when adding a physics interface, is used to model the
flow at low and moderate Reynolds numbers of liquids containing a dispersed phase.
The dispersed phase can be bubbles, liquid droplets, or solid particles, which are
assumed to always travel with their terminal velocity.

The Mixture Model, Laminar Flow interface solves one set of Navier–Stokes equations
for the momentum of the mixture. The pressure distribution is calculated from a
mixture-averaged continuity equation and the velocity of the dispersed phase is
described by a slip model. The volume fraction of the dispersed phase is tracked by
solving a transport equation for the volume fraction.

The physics interface can also model the distribution of the number density, which in
turn can be used to calculate the interfacial area, which is useful when simulating
chemical reactions in the mixture.

The main physics node is the Mixture Properties feature. It adds the equations for the
mixture and provides an interface for defining the fluid materials for the continuous
and dispersed phases as well as which slip model and mixture viscosity model to use.

When this physics interface is added, the following default physics nodes are also added
in the Model Builder — Mixture Properties, Wall, and Initial Values. Then, from the
Physics toolbar, add other nodes that implement, for example, boundary conditions
and volume forces. You can also right-click Mixture Model, Laminar Flow to select
physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern

564 | CHAPTER 6: FLUID FLOW INTERFACES

<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is mm.

PHYSICAL MODEL
Specify the characteristics of the dispersed phase, the model for the slip velocity, and
whether or not to solve for the interfacial area.

Dispersed Phase
To characterize the Dispersed phase, select Solid particles (the default) or Liquid
droplets/bubbles.

The selection from this list is further defined for the Mixture Properties node under
the Mixture Model section.

Slip Model
To compute the slip velocity jslip (SI unit: m/s), select a Slip model — Homogeneous
flow (the default), Hadamard-Rybczynski, Schiller-Naumann, Haider-Levenspiel, or User
defined.

• The Homogeneous flow model assumes that the velocities of the two phases are equal,
that is, uslip = 0.
• In most cases there is a significant difference in the velocity fields due to the
buoyancy of the dispersed phase. Use one of the predefined slip models for such
cases.
• For User defined specify an arbitrary expression for the relative velocity. For example,
give a constant velocity based on experimental data. Enter the Slip velocity field uslip
(SI unit: m/s) or Slip flux jslip (SI unit: m/s) in the Mixture Properties node under
the Mixture Model section.

Solve For Interfacial Area

To add a transport equation for the number density of the dispersed particles, in order
to determine the interfacial area, select the Solve for interfacial area checkbox (by
default not selected).

For the Mass Transfer rate, use a two-film theory model, which includes the interfacial
area per unit volume between the two phases. It is possible to compute the interfacial
area per unit volume if the number density n (that is, the number of dispersed particles

THE MIXTURE MODEL INTERFACE | 565

 per volume) is known. Select the Solve for interfacial area checkbox to add the
following equation for the number density n:

n j  n  nu
+ +   slip  c  = 0
t

This equation states that a dispersed phase particle cannot disappear, appear, or merge
with other particles, although it can expand or shrink.

The Mixture Model, Laminar Flow Interface calculates the interfacial area a
(SI unit: m2/m3) from

13 23
a =  4n   3 d 

Reference values
Reference values are global quantities used to evaluate the density of both phases and
the absolute pressure pA.

Reference pressure level There are generally two ways to include the pressure in fluid
flow computations: either to use the absolute pressure pA  p  pref, or the gauge
pressure p. When pref is nonzero, the physics interface solves for the gauge pressure
whereas material properties are evaluated using the absolute pressure. The reference
pressure level is also used to define the density of both phases. The default Reference
pressure level pref (SI unit: Pa) is 1[atm].

Reference temperature The reference temperature is used to define the density of

both phases. The default Reference temperature Tref (SI unit: K) is 293.15[K].

Swirl Flow
For 2D axisymmetric components, select the Swirl flow checkbox to include the swirl
velocity component — that is, the velocity component j in the azimuthal direction.
While j can be nonzero, there can be no gradients in the  direction.

General Single-Phase Flow Theory

DEPENDENT VARIABLES
Enter values for the dependent variables (field variables):

• Velocity field, mixture j (SI unit: m/s)

566 | CHAPTER 6: FLUID FLOW INTERFACES

• Pressure p (SI unit: Pa)
• Volume fraction, dispersed phase phid (dimensionless)
• Squared slip velocity slipvel
• Number density, dispersed phase nd (SI unit: 1/m3).

The names can be changed but the names of fields and dependent variables must be
unique within a component.

CONSISTENT STABILIZATION AND INCONSISTENT STABILIZATION

To display this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog.

The consistent stabilizations Streamline diffusion and Crosswind diffusion are by default
applied to the Navier-Stokes and dispersed phase transport equations. In addition, when
the flow is turbulent, the consistent stabilizations are also applied to the Turbulence.
Additional inconsistent stabilization terms may be added when required as isotropic
diffusion.

ADVANCED SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog. Normally these settings do not need
to be changed.

Penalty Diffusion can be used to suppress negative values of the dispersed volume
fraction. Including this term has been observed to slow down convergence and it is
therefore disabled by default.

Select the Use pseudo time stepping for stationary equation form checkbox to add
pseudo time derivatives to the equation when the Stationary equation form is used.
When selected, also choose a CFL number expression — Automatic (the default) or
Manual. Automatic sets the local CFL number (from the Courant–Friedrichs–Lewy
condition) to the built-in variable CFLCMP which in turn triggers a PID regulator for
the CFL number. For Manual enter a Local CFL number CFLloc (dimensionless).

• Pseudo Time Stepping for Laminar Flow Models in this guide and
Pseudo Time Stepping in the COMSOL Multiphysics Reference
Manual
• Domain and Boundary Nodes for the Mixture Model Interface
• Slip Velocity Models and Theory for the Mixture Model Interface

THE MIXTURE MODEL INTERFACE | 567

 Polymer Electrolyte Membrane Electrolyzer: Application Library path
Fuel_Cell_and_Electrolyzer_Module/Electrolyzers/pem_electrolyzer

Domain and Boundary Nodes for the Mixture Model Interface

The following domain and boundary nodes are available from the Physics ribbon
toolbar (Windows users), Physics context menu (Mac or Linux users), or right-click to
access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

DOMAIN
• Mixture Properties
• Mass Transfer
• Volume Force (described for The Laminar Flow Interface)
• Gravity
• Initial Values

BOUNDARY
The boundary types for the mixture flow variables, described in this section, are:

• Wall (the default boundary condition feature)

• Inlet
• Outlet
• Symmetry
• Periodic Flow Condition
• Interior Wall

PAIR
• Mixture Continuity
For the flow variables, the velocity and the pressure, the boundary conditions
correspond to those defined for single phase flow (see ). There is also a point constraint

568 | CHAPTER 6: FLUID FLOW INTERFACES

for the pressure, which is useful if no other boundary condition in the model includes
a pressure level constraint.

The following boundary conditions for the dispersed phase are available for all
boundary condition types except symmetry and interior wall. These are described in
more detail in the .

• Dispersed phase concentration (the default condition for inlets)

• Dispersed phase flux
• Dispersed phase outlet (the default condition for outlets)
• No dispersed phase flux (the default condition for walls)
• Symmetry

For 2D axisymmetric components, COMSOL Multiphysics takes the axial

symmetry boundaries (at r = 0) into account and automatically adds an
Axial Symmetry node to the component that is valid on the axial symmetry
boundaries only.

Mixture Properties
The Mixture Properties node contains the material properties for the continuous phase
and the dispersed phase. It also contains settings for the viscosity model. For the
Mixture Model, Turbulent Flow interfaces, the Mixture Properties node also adds the
equations for the turbulence transport equations.

MODEL INPUTS
The viscosity of each phase can be defined through user inputs, variables, or by
selecting a material. For the latter option, additional inputs, for example temperature
or pressure, may be required to define these properties.

Temperature
By default, the Temperature model input is set to Common model input, and the
temperature is controlled from Default Model Inputs under Global Definitions or by a
locally defined Model Input. If a Heat Transfer interface is included in the component,
it controls the temperature Common model input. Alternatively, the temperature field
can be selected from another physics interface. All physics interfaces have their own
tags (Name). For example, if a Heat Transfer in Fluids interface is included in the
component, the Temperature (ht) option is available for T.

THE MIXTURE MODEL INTERFACE | 569

 You can also select User defined from the Temperature model input in order to manually
prescribe T.

Absolute Pressure
This input appears when a material requires the absolute pressure as a model input.
The absolute pressure is used to evaluate material properties, but it also relates to the
value of the calculated pressure field. There are generally two ways to calculate the
pressure when describing fluid flow: either to solve for the absolute pressure or for a
pressure (often denoted gauge pressure) that relates to the absolute pressure through
a reference pressure.

The default Absolute pressure pA is p  pref, where p is the dependent pressure variable
from the Navier–Stokes or RANS equations, and pref is from the user input defined at
the physics interface level. When pref is nonzero, the physics interface solves for a
gauge pressure. If the pressure field instead is an absolute pressure field, pref should be
set to 0.

The Absolute pressure field can be edited by clicking Make All Model Inputs Editable
( ) and entering the desired value in the input field.

Model Inputs and Multiphysics Couplings in the COMSOL Multiphysics

Reference Manual

MATERIALS
Select the fluid materials to use for the material properties. The default material used
for the Continuous phase is the Domain material. This corresponds to the material
currently applied to the domain in question. The Dispersed phase uses None per default.
A valid material must be selected instead.

CONTINUOUS PHASE PROPERTIES

The default Density, continuous phase c (SI unit: kg/m3) uses values From material (as
selected in the Materials section). For User defined enter another value or expression. In
this case the default is 0 kg/m3.

The density in a material can depend on temperature and/or pressure and these
dependencies are automatically replaced by pref and Tref, which are specified at the
physics interface level.

The default Dynamic viscosity, continuous phase c (SI unit: Pa·s), uses values From
material. It describes the relationship between the shear stresses and the shear rate in a

570 | CHAPTER 6: FLUID FLOW INTERFACES

fluid. Intuitively, water and air have a low viscosity, and substances often described as
thick, such as oil, have a higher viscosity. For User defined enter another value or
expression. In this case, the default is 0 Pa·s.

DISPERSED PHASE PROPERTIES

The default Density, dispersed phase d (SI unit: kg/m3) uses values From material (as
selected in the Materials section). For User defined enter another value or expression. In
this case, the default is 0 kg/m3.

The density in a material can depend on temperature and/or pressure and these
dependencies are automatically replaced by pref and Tref, which are specified at the
physics interface level.

Enter the Diameter of particles/droplets dd (SI unit: m). The default is 103 m (1 mm).
If Haider-Levenspiel is selected for the Slip model under Physical Model, enter a value
between 0 and 1 for the Sphericity (dimensionless). The default is 1.

If Liquid droplets/bubbles is selected from the Dispersed phase list in the interface, then
Dynamic viscosity, dispersed phase d (SI unit: Pa·s) is also available. The default uses
values From material (as selected in the Materials section) or select User defined to enter
another value or expression. In this case, the default is 0 Pa·s.

MIXTURE MODEL
The options in this section are based on the selection made from the Dispersed phase
list for the Mixture Model interfaces.

Slip Velocity Field

When a User defined Slip model is selected for the physics interface, specify an arbitrary
expression for the relative velocity. For example, give a constant velocity based on
experimental data.

• When Slip velocity is set to Specify slip velocity field, enter the slip velocity between
the two phases, uslip (SI unit: m/s).
• When Slip velocity is set to Specify slip flux, enter the slip flux jslip (SI unit: m/s).

• The Mixture Model Equations in Theory for the Mixture Model

Interface

THE MIXTURE MODEL INTERFACE | 571

 Mixture Viscosity Model
Select the Mixture viscosity model.

• When Solid particles is the Dispersed phase, select either Krieger type (the default) or
User defined.
• When Liquid droplets/bubbles is the Dispersed phase, select Krieger type (the default),
User defined, or Volume averaged.

For User defined enter a value or expression for the Dynamic viscosity  (SI unit: Pas).
The default is 0 Pas. When using this option, make sure to limit the viscosity to
bounded positive values.

When Krieger type is selected, enter a value or expression for the Maximum packing
concentration max (dimensionless). The default is 0.62.

Select Krieger type to model the most generally valid expression for the mixture
viscosity:

 d  – 2,5 max 
 =  c  1 – -----------
-
  
max

where max is the maximum packing concentration, which for solid particles is
approximately 0.62. The dimensionless parameter *  1 for solid particles and

 d + 0,4 c
 = --------------------------
-
d + c

for droplets and bubbles. When applying the Krieger type viscosity model, d is
replaced by min(d, 0.999max) for better robustness.

Select Volume averaged to model the mixture viscosity of liquid-liquid mixtures, which
uses the following equation for the viscosity:

 = d d + c c

The Mixture Model interfaces always employ the mixture viscosity in the particle
Reynolds number expression used to calculate the slip velocity, thereby accounting for
the increase in viscous drag due to particle-particle interactions.

Mass Transfer
Use the Mass Transfer node to include mass transfer from the dispersed phase to the
continuous phase.

572 | CHAPTER 6: FLUID FLOW INTERFACES

MASS TRANSFER
Select a Mass transfer model — User defined (the default), Nonequilibrium model, or
Two-film theory. For User defined enter a value or expression for the Mass transfer from
dispersed to continuous phase mdc (SI unit: kg/(m3·s)). The default is 0 kg/(m3·s).

Nonequilibrium Model
For Nonequilibrium model enter values or expressions for each of the following:

• Equilibrium volume fraction in dispersed phase eq. The default is 0.

• Inverse time scales fc and fd (SI unit: 1/s). The default is 0 s1.
The mass transfer is modeled as

m dc =  d  max  f c   d –  eq  0  – max  f d   eq –  d  0  

Two-Film Theory
For Two-film theory enter values or expressions for each of the following:

• Mass transfer coefficient k (SI unit: m/s). The default is 0 m/s.

• Species concentration in dispersed phase cd (SI unit: mol/m3). The default is 0 mol/
m3.
• Species concentration in continuous phase cc (SI unit: mol/m3). The default is
0 mol/m3.
• Molecular weight of species M (SI unit: kg/mol). The default is 0 kg/mol.

The mass transfer is modeled as

m dc = k  c d – c c Ma

where k denotes the mass transfer rate, and cd and cc are the species concentrations in
the dispersed and the continuous phase, respectively. M is the species’ molecular
weight, and a is the interfacial area per unit volume between the two phases. The Mass
transfer model — Two-film theory is available when the Solve for interfacial area
checkbox is selected (under the Physical Model section).

Gravity
The Gravity node adds the force g to the right-hand side of the momentum transport
equation.

THE MIXTURE MODEL INTERFACE | 573

 GRAVITY
Enter the components of the Gravity vector, g (SI unit: m/s2). The default value is
(0,-g_const) where g_const is a physical constant equal to 9.8066 m/s2.

For 2D components, gravity acts in the negative y direction by default.

For example, in a 2D component, the y-component is -g_const and the
other component is 0. In this setting, g_const is a predefined physical
constant for the standard gravity (acceleration due to gravity at sea level).

For 3D and 2D axisymmetric components, gravity acts in the negative

z direction by default. For example, in a 3D component, the z-component
is -g_const and the other components are 0. In this setting, g_const is
a predefined physical constant for the standard gravity (acceleration due
to gravity at sea level).

Initial Values
The Initial Values node adds initial values for the mixture velocity, pressure, and volume
fraction of the dispersed phase, that can serve as an initial condition for a transient
simulation or as an initial guess for a nonlinear solver.

INITIAL VALUES
Enter values or expressions for the following dependent variables:

• Velocity field, mixture, the components of j (SI unit: m/s). The defaults are 0 m/s.
• Pressure p (SI unit: Pa). The default is 0 Pa.
• Volume fraction, dispersed phase (dimensionless). The default is 0.
• If the Solve for interfacial area checkbox is selected for the physics interface, enter an
initial value for the Number density, dispersed phase nd (SI unit: 1/m3). The default
is 0 1/m3.
• If a Schiller-Naumann or a Haider-Levenspiel slip model is used in the physics interface,
enter an initial value for the Squared slip velocity slipvel (SI unit: m2/s2). The
default is 0.081 m2/s2.

For the Mixture Model, Turbulent Flow interfaces, initial values for the turbulence
variables should be also specified. Additional initial values may be required for:

• Turbulent kinetic energy k

574 | CHAPTER 6: FLUID FLOW INTERFACES

• Turbulent dissipation rate ep
• Reciprocal wall distance G
• Specific dissipation rate om
• Undamped turbulent kinematic viscosity 0
• Wall distance in viscous units yPlus
• Tangential velocity in viscous units uPlus
• Turbulent relative fluctuations zeta
• Elliptic blending function alpha

The following sections describe the available boundary conditions for the
mixture and the dispersed phase volume fraction. In all equations, n
denotes the outward pointing unit vector normal to the boundary.

Wall
The Wall node has boundary conditions available that describe the existence of a solid
wall.

MIXTURE BOUNDARY CONDITION

Select a Mixture boundary condition for the wall.

• No Slip
• Slip
• Navier Slip

No Slip
No slip is the default boundary condition for the liquid. A no slip wall is a wall where
the mixture velocity relative to the wall velocity is zero. For stationary walls, it sets the
mixture velocity to zero at the wall:

j = 0

Slip
Select Slip to set the velocity component normal to the wall to zero j  nwall  0.

THE MIXTURE MODEL INTERFACE | 575

 Navier Slip
This boundary condition enforces no-penetration at the wall, j  nwall  0, and adds a
friction force of the form


F fr = – --- j tang


where  is a slip length, and jtang  j  j  nwallnwall is the velocity tangential to the
wall. The boundary condition does not set the tangential velocity component to zero;
however, the extrapolated tangential velocity component is 0 at a distance  outside of
the wall. In cases where the wall movement is nonzero, select the Account for the
translational wall velocity in the friction force checkbox to use
j  ubnd  j  ubnd  nwallnwall instead of jtang in the friction force. Then, the
extrapolated tangential velocity component is ubnd at a distance  outside of the wall.
Note that the Velocity of sliding wall uw is always accounted for in the friction force.

The Navier Slip boundary condition is suitable for walls adjacent to a free surface when
solving for laminar flow. Applying this boundary condition, the contact line
(fluid-fluid-solid interface) is free to move along the wall. Note that in problems with
contact lines, the tangential velocity of the wall typically represents the movement of
the contact line but the physical wall is not moving. In such cases, Account for the
translational wall velocity in the friction force should not be selected.

The Slip length setting is per default set to Factor of minimum element length. The slip
length  is then defined as   fhhmin, where hmin is the smallest element side
(corresponds to the element size in the wall normal direction for boundary layer
elements) and fh is a user input. Select User defined from the Slip length list to manually
prescribe 

DISPERSED PHASE BOUNDARY CONDITION

Select a Dispersed phase boundary condition — No dispersed phase flux (the default),
Dispersed phase concentration, Dispersed phase outlet, Dispersed phase flux, or Symmetry.

• For Dispersed phase concentration enter a Dispersed phase volume fraction d0
(dimensionless). The default is 0. When Solve for interfacial area is selected under
the Physical Model section, click either the Dispersed phase number density n0
(SI unit: 1/m3) or Diameter of particles/droplets dd (SI unit: m) button and enter a
value or expression for n0 (the default is 5·106 1/m3) or dd (the default is 1 mm).
• For Dispersed phase flux enter values or expression for the Dispersed phase flux Nd
(SI unit: m/s) and, if Solve for interfacial area is selected under the Physical Model

576 | CHAPTER 6: FLUID FLOW INTERFACES

 section, the Number density flux n (SI unit: 1/(m2s)). The defaults are 0 m/s and
0 1/(m2s), respectively.

Dispersed Phase Boundary Condition Equations

WALL MOVEMENT
This section contains controls to describe the wall movement relative to the lab (or
spatial) frame.

The Translational velocity setting controls the translational wall velocity, utr. The list is
per default set to Automatic from frame. The physics automatically detects if the spatial
frame moves. This can for example happen if an ALE interface is present in the model
component. If there is no movement utr  0. If the frame moves, utr becomes equal
to the frame movement. utr is accounted for in the actual boundary condition
prescribed in the Boundary condition section.

Select Manual from Translational velocity selection list in order to manually prescribe
Velocity of moving wall, utr. This can for example be used to model an oscillating wall
where the magnitude of the oscillations are very small compared to the rest of the
model. Specifying translational velocity manually does not automatically cause the
associated wall to move. An additional Moving Mesh interface needs to be added to
physically track the wall movement in the spatial reference frame.

The Sliding wall option is appropriate if the wall behaves like a conveyor belt; that is,
the surface is sliding in its tangential direction. A velocity is prescribed at the wall and
the boundary itself does not have to actually move relative to the reference frame.

• For 3D components, values or expressions for the Velocity of sliding wall uw should
be specified. If the velocity vector entered is not in the plane of the wall, COMSOL
Multiphysics projects it onto the tangential direction. Its magnitude is adjusted to
be the same as the magnitude of the vector entered.
• For 2D components, the tangential direction is unambiguously defined by the
direction of the boundary. For this reason, the sliding wall boundary condition has
different definitions in different space dimensions. A single entry for the Velocity of
the tangentially moving wall Uw should be specified in 2D.

THE MIXTURE MODEL INTERFACE | 577

 CONSTRAINT SETTINGS
This section is displayed by clicking the Show More Options button ( ) and selecting
Advanced Physics Options.

Wall in the Single-Phase Flow interfaces

Inlet
The Inlet node adds a set of boundary conditions that describe inlets in fluid-flow
simulations.

MIXTURE BOUNDARY CONDITION

Select a Mixture boundary condition for the inlet — Velocity (the default) or Pressure.

VELOCITY
For Velocity click the Normal inflow velocity (the default) or Velocity field buttons.

• For Normal inflow velocity enter a value or expression for J0 (SI unit: m/s).
• For Velocity field specify that the velocity at the boundary is equal to a given j0 (SI
unit: m/s) and enter the components in the matrix:
j = j0

PRESSURE
This condition specifies the normal stress which in most cases is approximately equal
to the pressure. Enter the Pressure p0 (SI unit: Pa) at the boundary. The default is 0 Pa.
Enter a relative pressure value in p0 (SI unit: Pa). Or if the reference pressure pref
defined at the physics interface level is equal to 0 Pa, enter an absolute pressure in p0.

• The Suppress backflow checkbox is selected by default. This option adjusts the inlet
pressure, p̂ 0  p 0 , locally in order to reduce the amount of fluid exiting the domain
through the boundary. If suppress backflow is deselected, the inlet boundary can
become an outlet depending on the pressure field in the rest of the domain.

578 | CHAPTER 6: FLUID FLOW INTERFACES

• Flow direction controls in which direction the fluid enters the domain.
- For Normal flow (the default) it prescribes zero tangential velocity component.
- For User defined define an Inflow velocity direction, dj (dimensionless). The
magnitude of dj does not matter, only the direction. dj must point into the
domain.

DISPERSED PHASE BOUNDARY CONDITION

The default is Dispersed phase concentration. See Wall>Dispersed Phase Boundary
Condition for the settings.

See Dispersed Phase Boundary Condition Equations for more information.

See Inlet in the Single-Phase Flow interfaces

Outlet
The Outlet node adds a selection of boundary conditions that describe outlets in
fluid-flow simulations; that is, the conditions at boundaries where the fluid exits the
domain.

MIXTURE BOUNDARY CONDITION

Select a Mixture boundary condition for the outlet — Pressure (the default) or Velocity.

PRESSURE CONDITIONS
The Pressure condition specifies the normal stress which in most cases is approximately
equal to the pressure. The tangential stress component is set to 0 N/m2.

• Enter the Pressure p0 (SI unit: Pa) at the boundary. The default is 0 Pa. Enter a
relative pressure value in p0 (SI unit: Pa). Or if the reference pressure pref defined
at the physics interface level is equal to 0 Pa, enter an absolute pressure in p0.
• Select the Normal flow checkbox to change the no tangential stress condition to a no
tangential velocity condition. This forces the flow to exit (or enter) the domain
perpendicularly to the outlet boundary.
• The Suppress backflow checkbox is selected by default. This option adjusts the outlet
pressure in order to reduce the amount of fluid entering the domain through the
boundary.

THE MIXTURE MODEL INTERFACE | 579

 VELOCITY
For Velocity click the Normal outflow velocity (the default) or Velocity field buttons.

• For Normal outflow velocity enter a value or expression for J0 (SI unit: m/s).
• For Velocity field specify that the velocity at the boundary is equal to a given j0 (SI
unit: m/s) and enter the components in the matrix:
j = j0

DISPERSED PHASE BOUNDARY CONDITION

The default is Dispersed phase outlet. See Wall>Dispersed Phase Boundary Condition
for the settings.

When the Dispersed phase boundary condition is set to Dispersed phase outlet, the
Exterior dispersed phase condition(s) section contains an input field for the Dispersed
phase volume fraction on the downstream side of the outlet. If Solve for interfacial area
is selected at the physics interface level, additional input fields appear for the Dispersed
phase number density or the Diameter of particles/droplets on the downstream side of
the outlet.

See Dispersed Phase Boundary Condition Equations for more information.

See Outlet in the Single-Phase Flow interfaces

Symmetry
The Symmetry node adds boundary conditions that describe symmetry boundaries in
fluid-flow simulations. The boundary condition for symmetry boundaries prescribes
no penetration and vanishing tangential stress:

T
j  n = 0, t T    c +  d –  u slip j slip +  Gm n = 0

BOUNDARY SELECTION
A boundary condition for 2D axial symmetry is not required.

For the symmetry axis at r  0, the program automatically provides a condition that
prescribes ur  0 and vanishing stresses in the z direction and adds an Axial Symmetry
node that is valid on the axial symmetry boundaries only.

580 | CHAPTER 6: FLUID FLOW INTERFACES

Periodic Flow Condition
The Periodic Flow Condition splits its selection into a source group and a destination
group. Fluid that leaves the domain through one of the destination boundaries enters
the domain through the corresponding source boundary. This corresponds to a
situation where the geometry is a periodic part of a larger geometry. If the boundaries
are not parallel to each other, the velocity vector is automatically transformed.

If the boundaries are curved, the orientation of the source must be specified manually
(see Orientation of Source).

PRESSURE DIFFERENCE
A value or expression should be specified for the Pressure difference, psrc  pdst. This
pressure difference can, for example, drive the fully developed flow in a channel.

To set up a periodic boundary condition, both boundaries must be selected in the

Periodic Flow Condition node. COMSOL Multiphysics automatically assigns one
boundary as the source and the other as the destination. To manually set the
destination selection, a Destination Selection subnode is available from the context
menu (by right-clicking the parent node) or from the Physics toolbar, Attributes menu.
All destination sides must be connected.

CONSTRAINT SETTINGS
This section is displayed by clicking the Show More Options button ( ) and selecting
Advanced Physics Options.

ORIENTATION OF SOURCE
For information about the Orientation of Source section, see Orientation of Source and
Destination in the COMSOL Multiphysics Reference Manual.

Interior Wall
The Interior Wall node adds a wall boundary condition on interior boundaries. It is
similar to the Wall boundary condition available on exterior boundaries except that it
applies on both sides of an interior boundary. It allows discontinuities of the dependent

THE MIXTURE MODEL INTERFACE | 581

 variables across the boundary. The Interior Wall node specifies no flux for the dispersed
phase.

Dispersed Phase Boundary Condition Equations

MIXTURE BOUNDARY CONDITION

Select a Mixture boundary condition — No slip (the default) or Slip.

No Slip
No slip is the default boundary condition for the mixture. A no slip wall is a wall where
the mixture velocity relative to the wall velocity is zero. For an interior stationary wall,
it sets the mixture velocity to zero at the wall:

j = 0

Slip
Select Slip to set the velocity component normal to the wall to zero j  n  0.

Navier Slip
The Navier slip condition enforces no-penetration at the wall, j  n  0, and adds a
friction force of the form


F fr = – --- j


where is a slip length. For more information, see the Navier Slip option in the Wall
feature.

WALL MOVEMENT
This section contains controls to describe the wall movement relative to the lab (or
spatial) frame.

The Translational velocity setting controls the translational wall velocity, utr. The list is
per default set to Automatic from frame. The physics automatically detects if the spatial
frame moves. This can for example happen if an ALE interface is present in the model
component. If there is no movement utr = 0. If the frame moves, utr becomes equal
to the frame movement. utr is accounted for in the actual boundary condition
prescribed in the Boundary condition section.

582 | CHAPTER 6: FLUID FLOW INTERFACES

Select Manual from Translational velocity selection list in order to manually prescribe
Velocity of moving wall, utr. This can for example be used to model an oscillating wall
where the magnitude of the oscillations are very small compared to the rest of the
model. Specifying translational velocity manually does not automatically cause the
associated wall to move. An additional Moving Mesh interface needs to be added to
physically track the wall movement in the spatial reference frame.

CONSTRAINT SETTINGS
This section is displayed by clicking the Show More Options button ( ) and selecting
Advanced Physics Options.

Wall

Mixture Continuity
The Mixture Continuity condition is suitable for pairs where the boundaries match; it
prescribes that the flow field is continuous across the pair.

The parts of the boundary where a source boundary lacks a corresponding destination
boundary and vice versa will get conditions from other boundary conditions under the
physics interface acting as fallback boundary conditions on the nonoverlapping parts.
By default, the Wall node is used as fallback unless another feature has been selected
for the boundary.

About Pairs

THE MIXTURE MODEL INTERFACE | 583

 T he E ule r - E ul er Mod el In t erface
In this section:

• The Euler–Euler Model, Laminar Flow Interface

• Domain, Boundary, Point, and Pair Nodes for the Euler–Euler Model Interfaces

The Euler–Euler Model, Laminar Flow Interface

The Euler-Euler Model, Laminar Flow (ee) interface ( ), found under the Multiphase
Flow > Euler-Euler Model branch ( ) when adding a physics interface, can be used to
simulate the flow of two continuous and fully interpenetrating incompressible phases
(see Ref. 1 under the Theory for the Euler–Euler Model Interfaces). The physics
interface can model flow at low and moderate Reynolds numbers. Typical applications
are fluidized beds (solid particles in gas), sedimentation (solid particles in liquid), or
transport of liquid droplets or bubbles in a liquid.

The physics interface solves two sets of Navier–Stokes equations, one for each phase,
in order to calculate the velocity field for each phase. The phases interchange
momentum as described by a drag model. The pressure is calculated from a
mixture-averaged continuity equation and the volume fraction of the dispersed phase
is tracked with a transport equation.

When this physics interface is added, the following default physics nodes are also added
in the Model Builder — Phase Properties, Wall, and Initial Values. Then, from the Physics
toolbar, add other nodes that implement, for example, boundary conditions and
volume forces. You can also right-click Euler-Euler Model, Laminar Flow to select physics
features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is ee.

584 | CHAPTER 6: FLUID FLOW INTERFACES

DISPERSED PHASE
Select a Dispersed phase — Solid particles or Liquid droplets/bubbles.

When Solid particles is selected, the section is available in the Phase Properties node.
Also see Dispersed Phase in the theory section.

PHYSICAL MODEL

Dispersed Phase
Select a Dispersed phase — Solid particles or Liquid droplets/bubbles.

When Solid particles is selected, the section is available in the Phase Properties node.
Also see Dispersed Phase in the theory section.

Reference Values
Reference values are global quantities used to evaluate the density of the fluid and
define the absolute pressure.

Reference pressure level There are generally two ways to include the pressure in fluid
flow computations: either to use the absolute pressure pA  p  pref, or the gauge
pressure p. When pref is nonzero, the physics interface solves for the gauge pressure
whereas material properties are evaluated using the absolute pressure. The reference
pressure level is also used to define the density of incompressible fluids.

Reference temperature The reference temperature is used to define the reference

density.

DEPENDENT VARIABLES
The dependent variables (field variables) are:

• Velocity field, continuous phase uc

• Velocity field, dispersed phase ud
• Pressure p
• Volume fraction, dispersed phase phid

The names can be changed but the names of fields and dependent variables must be
unique within a component.

CONSISTENT STABILIZATION
To display this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog.

THE EULER-EULER MODEL INTERFACE | 585

 INCONSISTENT STABILIZATION
To display this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog.

Inconsistent stabilization can be activated independently for the momentum equation

for the continuous phase, the momentum equation for the dispersed phase, and for the
dispersed phase transport equation respectively by selecting the corresponding
checkbox. Each inconsistent stabilization contribution has a tuning parameter.

ADVANCED SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog. Normally these settings do not need
to be changed.

The Residue volume fraction, continuous phase, c,res, and Residue volume fraction,
dispersed phase, d,res, set the smallest values used to avoid division by zero when
evaluating terms that involve 1c and 1d. Observe that this value does not prevent
c or d from becoming smaller than c,res and d,res, respectively.

Select the Use pseudo time stepping for stationary equation form checkbox to add
pseudo time derivatives to the equation when the Stationary equation form is used.
When selected, also choose a CFL number expression — Automatic (the default) or
Manual. Automatic sets the local CFL number (from the Courant–Friedrichs–Lewy
condition) to the built-in variable CFLCMP which in turn triggers a PID regulator for
the CFL number. For Manual enter a Local CFL number CFLloc (dimensionless).

DISCRETIZATION
The Euler-Euler Model, Laminar Flow interface supports three options for the basis
functions: P1+P1 (the default option), P2+P1, and P3+P2. They all represent
Lagrangian basis functions of different orders:

• P1+P1 – Linear basis functions for all degrees of freedom. Linear basis functions are
computationally less expensive than the higher-order options and are also more
robust. This option requires that Streamline diffusion is activated for both of the
momentum equations.
• P2+P1 – Quadratic basis functions for all degrees of freedom except the pressure
which is described by linear basis functions. Higher order elements, as compared to
linear elements, are a computationally effective way to obtain high accuracy but only

586 | CHAPTER 6: FLUID FLOW INTERFACES

 if the flow is well resolved. This requirement is most likely fulfilled for flows with
very low velocities and/or small length scales.
• P3+P2 – Cubic basis functions for all degrees of freedom except the pressure which
is described by quadratic basis functions. This option is computationally very
expensive and the least robust one but it is also the option with the highest formal
accuracy.

To see all settings available in this section, click the Show More Options button ( )
and select Advanced Physics Options.

• Pseudo Time Stepping for Laminar Flow Models in this guide and
Pseudo Time Stepping in the COMSOL Multiphysics Reference
Manual
• Isotropic Diffusion in the COMSOL Multiphysics Reference Manual
• Domain, Boundary, Point, and Pair Nodes for the Euler–Euler Model
Interfaces
• Theory for the Euler–Euler Model Interfaces

Domain, Boundary, Point, and Pair Nodes for the Euler–Euler Model
Interfaces
have the following domain, boundary, point, and pair nodes available from the Physics
ribbon toolbar (Windows users), Physics context menu (Mac or Linux users), or
right-click to access the context menu (all users).

• Flow Continuity • Mass Transfer

• Gravity • Outlet
• Initial Values • Phase Properties
• Inlet • Volume Force
• Interior Wall • Wall

The following nodes are described for the Laminar Flow interface:

• Pressure Point Constraint

THE EULER-EULER MODEL INTERFACE | 587

 • Symmetry

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Phase Properties
The Phase Properties node has the material properties settings for the pure continuous
and dispersed phases. It also has settings for the viscosity models for each of the two
interpenetrating phases and the drag model. When the dispersed phase consists of solid
particles, it also has settings for the solid pressure model.

MODEL INPUTS
Fluid properties, such as density and viscosity, can be defined through user inputs,
variables, or by selecting a material. For the latter option, additional inputs, for
example temperature or pressure, may be required to define these properties.

Temperature
By default, the Temperature model input is set to Common model input, and the
temperature is controlled from Default Model Inputs under Global Definitions or by a
locally defined Model Input. If a Heat Transfer interface is included in the component,
it controls the temperature Common model input. Alternatively, the temperature field
can be selected from another physics interface. All physics interfaces have their own
tags (Name). For example, if a Heat Transfer in Fluids interface is included in the
component, the Temperature (ht) option is available for T.

You can also select User defined from the Temperature model input in order to manually
prescribe T.

Absolute Pressure
This input appears when a material requires the absolute pressure as a model input.
The absolute pressure is used to evaluate material properties, but it also relates to the
value of the calculated pressure field. There are generally two ways to calculate the
pressure when describing fluid flow: either to solve for the absolute pressure or for a
pressure (often denoted gauge pressure) that relates to the absolute pressure through
a reference pressure.

588 | CHAPTER 6: FLUID FLOW INTERFACES

The default Absolute pressure pA is p  pref, where p is the dependent pressure variable
from the Navier–Stokes or RANS equations, and pref is from the user input defined at
the physics interface level. When pref is nonzero, the physics interface solves for a
gauge pressure. If the pressure field instead is an absolute pressure field, pref should be
set to 0.

The Absolute pressure field can be edited by clicking Make All Model Inputs Editable
( ) and entering the desired value in the input field.

Model Inputs and Multiphysics Couplings in the COMSOL Multiphysics

Reference Manual

MATERIALS
Select the materials to use for the material properties of the continuous phase and the
dispersed phase respectively (when they are defined to take their values From material).
By default, the default material used for the Continuous phase is the Domain material.
This corresponds to the material currently applied to the domain in question. The
Dispersed phase uses None per default. A valid material must be selected instead.

Other materials present can be selected from the Continuous phase or Dispersed phase
list in order to use that material’s properties for the liquid or gas, respectively.

CONTINUOUS PHASE PROPERTIES

Density, Continuous Phase

The default Density, continuous phase c (SI unit: kg/m3) uses values From material.
For User defined enter another value or expression.

Dynamic Viscosity, Pure Continuous Phase

The dynamic viscosity describes the relationship between the shear stresses and the
shear rate in a fluid. Intuitively, water and air have a low viscosity, and substances often
described as thick, such as oil, have a higher viscosity. The dynamic viscosity of the pure
continuous phase refers to the viscosity the fluid would have in the absence of the
dispersed phase.

The default Dynamic viscosity, pure continuous phase c (SI unit: Pa·s) uses values From
material. The value of the viscosity is then the value defined for the material selected in
the Materials section for the continuous phase. For User defined define a different value
or expression.

THE EULER-EULER MODEL INTERFACE | 589

 DISPERSED PHASE PROPERTIES
The default Density, dispersed phase d (SI unit: kg/m3) uses values From material. For
User defined enter another value or expression.

Enter a value or expression for the Diameter of particles/droplets dd (SI unit: m). The
default is 103 m (1 mm).

The Dynamic viscosity, pure dispersed phase d (SI unit: Pa·s) field is
available when Liquid droplets/bubbles is selected as the Dispersed phase in .

The default uses values From material. The value of the viscosity is then
the value defined for the material selected in the Materials section for the
dispersed phase. For User defined define a different value or expression.

VISCOSITY MODEL
Select a model for the Mixture viscosity — Krieger type (the default), Volume average
(available for liquid droplets and bubbles) or User defined. Krieger type is appropriate
for dispersed phases that do not form any pure phase regions. Volume average is more
appropriate when the dispersed phase effectively can become the only phase in some
regions.

• For Krieger type, select an option for Maximum packing concentration. The default is
to use d,max = 1.0 for droplets and bubbles and d,max = 0.62 for particles. Select
User defined to enter another value or expression for d,max.
• For User defined enter a value or expression for the Mixture viscosity. The default
expression corresponds to the Krieger type viscosity with maximum packing equal
to 0.62.

Select a Dynamic viscosity model, continuous phase — Mixture viscosity (default), Pure
phase value or User defined. The Mixture viscosity option is appropriate when the
maximum packing is controlled by the mixture viscosity. The Pure phase value option
is more appropriate if the maximum packing is controlled in some other way, for
example by solid pressure. Select User defined to enter an expression for the Dynamic
viscosity model, continuous phase.

Select a Dynamic viscosity model, dispersed phase — Mixture viscosity (default), Pure
phase value (available for bubbles and droplets), Gidaspow (available for solid particles)
m
or User defined. The Gidaspow model prescribes  d = 0.5 d , which basically
m
prescribes a small value for  d and should be combined with a solid pressure model.

590 | CHAPTER 6: FLUID FLOW INTERFACES

DRAG MODEL
The models are described in . When the dispersed phase corresponds to Solid particles,
the Schiller–Naumann, Hadamard–Rybczynzki, Ishii–Zuber, Gidaspow, and Haider–
Levenspiel models are available. For Liquid droplets/bubbles, the Schiller–Naumann,
Hadamard–Rybczynzki, Ishii–Zuber, and Tomiyama et al. models are available. Select the
User-defined drag coefficient option to specify an expression for the drag coefficient CD.

• The Schiller–Naumann is valid for dispersed, rigid spheres and can be used for particle
Reynolds numbers less than approximately 2.5·105. The Schiller–Naumann model
is a general purpose model for diluted flows.
• The Hadamard–Rybczynzki is appropriate for diluted flows at low Reynolds numbers.
• The Ishii–Zuber model can for solid particles be regarded as a generalization of the
Schiller–Naumann model in that it can be used also for dense flows. For droplets
and bubbles, the Ishii–Zuber model also takes deformation of the fluid particles into
account. The deformation depends on the Surface tension coefficient, . Use a
predefined expression, select Library coefficient, liquid/gas interface or Library
coefficient, liquid/liquid interface. Then select an option from the list that displays
below (for example, Water/Air, Glycerol/Air and so forth). For User defined enter a
value or expression for the surface tension coefficient  (SI unit: N/m).
• The Gidaspow model describes the drag force coefficient for dispersed phase flows
for all concentrations of the dispersed solid, including high concentrations.
• The Haider-Levenspiel model describes nonspherical particles for low and moderate
particle Reynolds numbers. Enter a value between 0 and 1 for the Sphericity Sp
(dimensionless).
• The Tomiyama et al. model describes the drag for diluted droplets and bubbles for
any particle Reynolds number. The correlation is available in three versions, Pure
fluid, Slightly contaminated fluid and Contaminated fluid. The Tomiyama et al. model
depends on the Surface tension coefficient,  that is selected in the same way as for
the Ishii–Zuber model.

SOLID PRESSURE MODEL

The solid pressure models the particle dispersion due to collisions and friction between
the solid particles. Details of the implemented models are described in .

THE EULER-EULER MODEL INTERFACE | 591

 This section is available when Solid particles is selected as the Dispersed phase in . Select
a Solid pressure model ps.

• The No solid pressure (the default) disables solid pressure and leaves the collision to
be described by a mixture viscosity. This option is appropriate for solid suspension,
typically solids suspended in liquids.
• The other options, Gidaspow-Ettehadieh, Gidaspow, Ettehadieh, and User-defined
modulus of elasticity describes particle-particle interaction for dispersed solids,
typically solid particles in gas flows. These options should be combined with the
m
Gidaspow model or a small user-defined value for  d .
• For User-defined modulus of elasticity enter an expression for the Modulus of elasticity,
G (SI unit: N/m2).

MIXING LENGTH LIMIT

When the Mixing length limit lmix,lim is set to Automatic, the mixing length limit is
evaluated as the shortest side of the geometry bounding box. If the geometry is, for
example, a complicated system of slim entities, this measure can be too high. In such
cases, it is recommended that the mixing length limit is defined manually. Select Manual
to enter a different value or expression.

Initial Values
The Initial Values node adds initial values for the velocity fields for the continuous and
dispersed phases, the pressure, and the volume fraction of the dispersed phase that can
serve as initial conditions for a transient simulation or as an initial guess for a nonlinear
solver.

INITIAL VALUES
Enter initial values or expressions in the tables for each component of:

• Velocity field, continuous phase uc

• Velocity field, dispersed phase ud

Enter initial values or expressions for:

• Pressure p
• Volume fraction, dispersed phase phid

592 | CHAPTER 6: FLUID FLOW INTERFACES

Mass Transfer
Use the Mass Transfer node to include mass transfer from the dispersed phase to the
continuous phase.

MASS TRANSFER
Select a Mass transfer model — User defined (the default) or Nonequilibrium model. For
User defined enter a value or expression for the Mass transfer from dispersed to
continuous phase mdc (SI unit: kg/(m3·s)). The default is 0 kg/(m3·s).

Nonequilibrium Model
For Nonequilibrium model enter values or expressions for each of the following:

• Equilibrium volume fraction in dispersed phase eq. The default is 0.

• Inverse time scales fc and fd (SI unit: 1/s). The default is 0 s-1.
The mass transfer is modeled as

m dc =  d  max  f c   d –  eq  0  – max  f d   eq –  d  0  

Interphase Velocity
Select Dispersed phase velocity field (the default) or User defined. For User defined enter
a value or expression for each component of the interphase velocity uint.

Gravity
The Gravity node adds the gravity force vector to the right-hand side of the momentum
equations; cg for the continuous phase and dg for the dispersed phase.

Gravity nodes are mutually exclusive, that is, there can only be one active Gravity node
per domain.

GRAVITY
Enter the components of the Gravity vector, g (SI unit: m/s2). For 2D components
the default Gravity vector is defined as 0-g_const. Here g_const is a physical
constant equal to 9.8066 m/s2. For 3D and 2D axisymmetric components, the gravity
acts in the negative z direction by default.

Volume Force
The Volume Force node specifies the volume forces on the right-hand side of the
momentum equations; Fc for the continuous phase and Fd for the dispersed phase. It

THE EULER-EULER MODEL INTERFACE | 593

 may for example be used to add user defined interaction forces (in addition to the
interphase drag force).

If several volume force nodes are added to the same domain, the sum of all
contributions are added to the respective momentum equations.

Wall
The Wall node includes a set of boundary conditions to describe the flow conditions
at a solid wall. The conditions for each phase are selected separately.

CONTINUOUS PHASE BOUNDARY CONDITION

Select a Continuous Phase Condition — No slip (the default), or Slip.

No Slip
A No slip condition models a wall where the velocity of the continuous phase relative
to the wall velocity is zero. For a stationary wall, it sets the velocity of the continuous
phase to zero:

uc = 0

For turbulent flows, wall functions are used to model the thin region near solid (no
slip) walls with high gradients in the flow variables

Slip
The Slip option prescribes a no-penetration condition. For a stationary wall, uc  n  0.
It is implicitly assumed that there are no viscous effects at the slip wall and hence, no
boundary layer develops.

Leakage
ꞏ
Select the Leakage checkbox to specify an Inward continuous phase mass flux, m c (SI
2 ꞏ
unit: kg(m ·s)). This option is intended for small values of m c . For large mass fluxes,
select an Inlet feature instead.

DISPERSED PHASE BOUNDARY CONDITION

Select a Dispersed velocity boundary condition — No slip (the default), or Slip. The
conditions prescribed for the velocity of the dispersed phase correspond to those
described for the continuous phase above.

594 | CHAPTER 6: FLUID FLOW INTERFACES

Dispersed Phase Boundary Condition
The Dispersed Phase Boundary Condition defines the condition at the wall for the
volume fraction of the dispersed phase.

• No dispersed phase flux prescribes a zero flux condition at the wall.

ꞏ
• Select Leakage to prescribe an Inward dispersed phase mass flux, m d (SI unit: kg
2 ꞏ
(m ·s)). This option is intended for small values of m d . For large mass fluxes, select
an Inlet feature instead.

WALL MOVEMENT
This section contains controls to describe the wall movement relative to the lab (or
spatial) frame.

The Translational velocity setting controls the translational wall velocity, utr. The list is
per default set to Automatic from frame. The physics automatically detects if the spatial
frame moves. This can for example happen if an ALE interface is present in the model
component. If there is no movement utr  0. If the frame moves, utr becomes equal
to the frame movement. utr is accounted for in the actual boundary condition
prescribed in the Boundary condition section.

Select Manual from Translational velocity selection list in order to manually prescribe
Velocity of moving wall, utr. This can for example be used to model an oscillating wall
where the magnitude of the oscillations are very small compared to the rest of the
model. Specifying translational velocity manually does not automatically cause the
associated wall to move. An additional Moving Mesh interface needs to be added to
physically track the wall movement in the spatial reference frame.

The Sliding wall option is appropriate if the wall behaves like a conveyor belt; that is,
the surface is sliding in its tangential direction. A velocity is prescribed at the wall and
the boundary itself does not have to actually move relative to the reference frame.

• For 3D components, values or expressions for the Velocity of sliding wall uw should
be specified. If the velocity vector entered is not in the plane of the wall, COMSOL
Multiphysics projects it onto the tangential direction. Its magnitude is adjusted to
be the same as the magnitude of the vector entered.
• For 2D components, the tangential direction is unambiguously defined by the
direction of the boundary. For this reason, the sliding wall boundary condition has
different definitions in different space dimensions. A single entry for the Velocity of
the tangentially moving wall Uw should be specified in 2D.

THE EULER-EULER MODEL INTERFACE | 595

 CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options.

Inlet
The Inlet node prescribes the flow and turbulence conditions (when applicable) at a
boundary where one or both phases enter the adjacent domain.

TWO-PHASE INLET TYPE

Selecting Mixture as the Two-Phase Inlet Type, inlet conditions are prescribed for both
phases. Selecting Continuous phase or Dispersed phase, inlet conditions are prescribed
for the corresponding phase, while wall conditions are applied to the remaining phase.

CONTINUOUS PHASE
The options available in this section are based on the Two-phase inlet type selected.

When Mixture or Continuous phase is selected as the Two-phase inlet type the Velocity
field, continuous phase uc0 is prescribed.

When Dispersed phase is selected as the Two-phase inlet type, the conditions applied to
the continuous phase are No slip (uc0) or Slip (uc·n0, and no viscous stresses in the
tangential directions).

DISPERSED PHASE
The options available in this section are based on the Two-phase inlet type selected.

When Mixture or Dispersed phase is selected as the Two-phase inlet type, the Velocity
field, dispersed phase ud0.

The Dispersed phase boundary condition can be Volume fraction, Mass flux, or No flux.

• For Volume fraction, specify the Dispersed phase volume fraction, d0, at the inlet.
This option locks the value of d, but the exact flux depends on the dispersed phase
velocity, ud0.
ꞏ
• For Mass flux, prescribe describe the Inward dispersed phase mass flux, m d (SI unit:
2
kg(m ·s)). This option prescribes the flux for the dispersed phase, but the resulting
volume fraction at the inlet depends on the dispersed phase velocity, ud0.
When Continuous phase is selected as the Two-phase inlet type, the conditions applied
to the dispersed phase are No slip (ud0) or Slip (ud·n0, and no viscous stresses in the
tangential directions).

596 | CHAPTER 6: FLUID FLOW INTERFACES

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options.

Outlet
This node is used to prescribe conditions at boundaries where the two-phase mixture
exits the computational domain. The settings available in the node specifies conditions
for the entire mixture, as opposed to the Inlet, Outlet and Wall nodes where the
conditions for can be specified for each phase individually.

MIXTURE BOUNDARY CONDITION

Select Mixture boundary conditions — Pressure, no viscous stress, Pressure, normal flow,
Velocity, Degassing, or Filtration.

Pressure Conditions
When Pressure, no viscous stress or Pressure, normal flow is selected as the Mixture
boundary condition, specify the Pressure p0 (SI unit: Pa) at the outlet. Note that the
prescribed pressure should be consistent with the reference pressure level specified at
the interface level as well as consistent with present volume forces (including gravity).

Velocity
Select Velocity to specify the Velocity field, continuous phase uc0 (SI unit: m/s) together
with the Velocity field, dispersed phase ud (SI unit: m/s) at the outlet.

Degassing and Filtration

Selecting Degassing allows the dispersed phase to exit the domain but prevents the
continuous phase from exiting. This is achieved through a pressure (p0), no viscous
stress condition for the dispersed phase, in combination with a no penetration
condition for the continuous phase.

Selecting Filtration on the other hand allows the continuous phase to exit the domain
but prevents the dispersed phase from exiting. This is achieved through a pressure (p0),

THE EULER-EULER MODEL INTERFACE | 597

 no viscous stress condition for the continuous phase, in combination with a no
penetration condition for the dispersed phase.

Pressure, no viscous stress is the least constraining option and allows the
flow to exit the domain in nonnormal directions.

Pressure, normal flow restricts the flow to be completely normal to the

outlet boundary. This option can influence the flow quite a bit upstream
of the inlet.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options.

Interior Wall
The Interior Wall boundary condition can only be applied on interior boundaries.

It is similar to the Wall boundary condition available on exterior boundaries except

that it applies on both sides of an interior boundary. It allows discontinuities of the
dependent variables across the boundary. The Interior Wall boundary condition can be
used to avoid meshing thin structures by applying no-slip conditions on interior curves
and surfaces instead. Slip conditions and conditions for a moving wall can also be
prescribed. It is compatible with laminar and turbulent flow. The conditions for each
phase are selected separately.

CONTINUOUS PHASE BOUNDARY CONDITION

Select a Continuous Phase Condition — No slip (the default), or Slip.

No Slip
A No slip condition models a wall where the velocity of the continuous phase relative
to the wall velocity is zero. For an interior stationary wall, it sets the velocity of the
continuous phase to zero on both sides of the wall:

uc = 0

For turbulent flows, wall functions are used to model the thin region near solid walls
with high gradients in the flow variables.

598 | CHAPTER 6: FLUID FLOW INTERFACES

Slip
Slip prescribes a no-penetration condition. For a stationary interior wall, uc  n  0 on
both sides.

DISPERSED PHASE BOUNDARY CONDITION

Select a Dispersed velocity boundary condition — No slip (the default), or Slip. The
conditions prescribed for the velocity of the dispersed phase correspond to those
described for the continuous phase above.

Dispersed Phase Boundary Condition

The Dispersed phase boundary condition is set to No dispersed phase flux.

WALL MOVEMENT
This section contains controls to describe the wall movement relative to the lab (or
spatial) frame.

The Translational velocity setting controls the translational wall velocity, utr. The list is
per default set to Automatic from frame. The physics automatically detects if the spatial
frame moves. This can for example happen if an ALE interface is present in the model
component. If there is no movement utr  0. If the frame moves, utr becomes equal
to the frame movement. utr is accounted for in the actual boundary condition
prescribed in the Boundary condition section.

Select Manual from Translational velocity selection list in order to manually prescribe
Velocity of moving wall, utr. This can for example be used to model an oscillating wall
where the magnitude of the oscillations are very small compared to the rest of the
model. Specifying translational velocity manually does not automatically cause the
associated wall to move. An additional Moving Mesh interface needs to be added to
physically track the wall movement in the spatial reference frame.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options.

Flow Continuity
The Flow Continuity condition is suitable for pairs where the boundaries match; it
prescribes that the flow field is continuous across the pair.

The parts of the boundary where a source boundary lacks a corresponding destination
boundary and vice versa will get conditions from other boundary conditions under the

THE EULER-EULER MODEL INTERFACE | 599

 physics interface acting as fallback boundary conditions on the nonoverlapping parts.
By default, the Wall node is used as fallback unless another feature has been selected
for the boundary.

About Pairs

600 | CHAPTER 6: FLUID FLOW INTERFACES

Theory for the Laminar Flow and
Creeping Flow Interfaces
The theory for the Single-Phase Flow, Laminar Flow interface is described in this
section:

• General Single-Phase Flow Theory

• Compressible Flow
• Weakly Compressible Flow
• The Mach Number Limit
• Incompressible Flow
• The Reynolds Number
• Gravity
• Theory for the Wall Boundary Condition
• Prescribing Inlet and Outlet Conditions
• Mass Flow
• Fully Developed Flow (Inlet)
• Fully Developed Flow (Outlet)
• No Viscous Stress
• Normal Stress Boundary Condition
• Porous Interface
• Mass Sources for Fluid Flow
• Numerical Stability — Stabilization Techniques for Fluid Flow
• Solvers for Laminar Flow
• Pseudo Time Stepping for Laminar Flow Models
• Nitsche Constraints
• Particle Tracing in Fluid Flow
• References for the Single-Phase Flow, Laminar Flow Interfaces

The theory about most boundary conditions is found in Ref. 2.

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 601
 General Single-Phase Flow Theory
The Single-Phase Fluid Flow interfaces are based on the Navier–Stokes equations,
which in their most general form read


------ +    u  = 0 (6-26)
t

u
 ------ +   u   u =    – pI + K  + F (6-27)
t

T  p
C p  T
------- +  u   T = –    q  + K:S – ---- -------  ------ +  u   p + Q (6-28)
t  T p t

where

•  is the density (SI unit: kg/m3)

• u is the velocity vector (SI unit: m/s)
• p is pressure (SI unit: Pa)
• I is the identity matrix (unitless)
• K is the viscous stress tensor (SI unit: Pa)
• F is the volume force vector (SI unit: N/m3)
• Cp is the specific heat capacity at constant pressure (SI unit: J/(kg·K))
• T is the absolute temperature (SI unit: K)
• q is the heat flux vector (SI unit: W/m2)
• Q contains the heat sources (SI unit: W/m3)
• S is the strain-rate tensor:

1
S = ---  u +  u  T 
2

The operation “:” denotes a contraction between tensors defined by

a:b =   anm bnm (6-29)

n m

This is sometimes referred to as the double dot product.

Equation 6-26 is the continuity equation and represents conservation of mass.

Equation 6-27 is a vector equation which represents conservation of momentum.

602 | CHAPTER 6: FLUID FLOW INTERFACES

Equation 6-28 describes the conservation of energy, formulated in terms of
temperature. This is an intuitive formulation that facilitates boundary condition
specifications.

To close the equation system, Equation 6-26 through Equation 6-28, constitutive
relations are needed.

For a Newtonian fluid, which has a linear relationship between stress and strain, Stokes
(Ref. 1) deduced the following expression:

2
K = 2S – ---     u I (6-30)
3

The dynamic viscosity,  (SI unit: Pa·s), for a Newtonian fluid is allowed to depend on
the thermodynamic state but not on the velocity field. All gases and many liquids can
be considered Newtonian.

For an inelastic non-Newtonian fluid, the relationship between stress and strain rate is
nonlinear, and an apparent viscosity is introduced instead of the dynamic viscosity.
Examples of non-Newtonian fluids are honey, mud, blood, liquid metals, and most
polymer solutions.

In theory, the same equations describe both laminar and turbulent flows. In practice,
however, the mesh resolution required to simulate turbulence with the Laminar Flow
interface makes such an approach impractical.

There are several books where derivations of the Navier–Stokes equations

and detailed explanations of concepts such as Newtonian fluids can be
found. See, for example, the classical text by Batchelor (Ref. 3) and the
more recent work by Panton (Ref. 4).

Many applications describe isothermal flows for which Equation 6-28 is decoupled
from Equation 6-26 and Equation 6-27.

2D AXISYMMETRIC FORMULATIONS
A 2D axisymmetric formulation of Equation 6-26 and Equation 6-27 requires  to
be zero. That is, there must be no gradients in the azimuthal direction. A common
additional assumption is, however, that u  0. In such cases, the -equation can be
removed from Equation 6-27. The resulting system of equations is both easier to
converge and computationally less expensive compared to retaining the -equation.

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 603
 The default 2D axisymmetric formulation of Equation 6-26 and Equation 6-27
therefore assumes that

   = 0
u = 0

Compressible Flow
The equations of motion for a single-phase fluid are the continuity equation:


------ +    u  = 0 (6-31)
t

and the momentum equation:

u
 ------ + u  u = – p +      u +  u  T  – ---     u I + F
2
(6-32)
t 3

These equations are applicable for incompressible as well as for compressible flow with
density and viscosity variations.

Weakly Compressible Flow

The same equations as for Compressible Flow are applied for weakly compressible flow.
The only difference is that the density is evaluated at the reference pressure. The
density may be a function of other quantities, in particular it may be temperature
dependent.

The weakly compressible flow equations are valid for incompressible as well as
compressible flow with density variations independent of the pressure.

Provided that the densities dependency pressure is specified through model inputs, the
density is automatically evaluated at the reference pressure level.

The Mach Number Limit

An important dimensionless number in fluid dynamics is the Mach number, Ma,
defined by

u-
Ma = -----
a

604 | CHAPTER 6: FLUID FLOW INTERFACES

where a is the speed of sound. A flow is formally incompressible when Ma  0. This is
theoretically achieved by letting the speed of sound tend to infinity. The Navier–Stokes
equations then have the mathematical property that pressure disturbances are
instantaneously propagated throughout the entire domain. This results in a parabolic
equation system.

The momentum equation, Equation 6-32, is parabolic for unsteady flow and elliptic
for steady flow, whereas the continuity equation, Equation 6-31, is hyperbolic for both
steady and unsteady flow. The combined system of equations is thus hybrid
parabolic-hyperbolic for unsteady flow and hybrid elliptic-hyperbolic for steady flow.
An exception occurs when the viscous term in Equation 6-32 becomes vanishingly
small, such as at an outflow boundary, in which case the momentum equation becomes
locally hyperbolic. The number of boundary conditions to apply on the boundary then
depends on the number of characteristics propagating into the computational domain.
For the purely hyperbolic system, the number of characteristics propagating from the
boundary into the domain changes as the Mach number passes through unity. Hence,
the number of boundary conditions required to obtain a numerically well-posed
system must also change. The compressible formulation of the laminar and turbulent
interfaces uses the same boundary conditions as the incompressible formulation, which
implies that the compressible interfaces are not suitable for flows with a Mach number
larger than or equal to one. Yet, the practical Mach number limit is lower than one.
The main reason is that the numerical scheme (stabilization and boundary conditions)
of the Laminar Flow interface does not recognize the direction and speed of pressure
waves. The fully compressible Navier–Stokes equations do, for example, start to display
very sharp gradients already at moderate Mach numbers. But the stabilization for the
single-phase flow interface does not necessarily capture these gradients. It is impossible
to give an exact limit where the low Mach number regime ends and the moderate
Mach number regime begins, but a rule of thumb is that the Mach number effects start
to appear at Ma  0.3. For this reason, the compressible formulation is referred to as
Compressible flow (Ma<0.3) in COMSOL Multiphysics.

Incompressible Flow
When the temperature variations in the flow are small, a single-phase fluid can often
be assumed incompressible; that is,  is constant or nearly constant. This is the case for
all liquids under normal conditions and also for gases at low velocities. For constant ,
Equation 6-31 reduces to

  u = 0 (6-33)

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 605
 and Equation 6-32 becomes

u T
 +   u   u =    – pI +   u +  u    + F (6-34)
t

Provided that the densities dependency on temperature and pressure is specified

through model inputs, the density is evaluated at the reference pressure level and at the
reference temperature. However, if the density is a function of other quantities such as
a concentration field, or if the density is specified by a user defined expression, the user
has to make sure that the density is defined as constant when the incompressible flow
formulation is used.

The Reynolds Number

A fundamental characteristic in analyses of fluid flow is the Reynolds number:

Re = UL
------------


where U denotes a velocity scale, and L denotes a representative length. The Reynolds
number represents the ratio between inertial and viscous forces. At low Reynolds
numbers, viscous forces dominate and tend to damp out all disturbances, which leads
to laminar flow. At high Reynolds numbers, the damping in the system is very low,
giving small disturbances the possibility to grow by nonlinear interactions. If the
Reynolds number is high enough, the flow field eventually ends up in a chaotic state
called turbulence.

Note that the Reynolds number can have different meanings depending on the length
scale and velocity scale. To be able to compare two Reynolds numbers, they must be
based on equivalent length and velocity scales.

The Fluid Flow interfaces automatically calculate the local cell Reynolds number
Rec  uh2 using the element length h for L and the magnitude of the velocity
vector u for the velocity scale U. This Reynolds number is not related to the character
of the flow field, but to the stability of the numerical discretization. The risk for
numerical oscillations in the solution increases as Rec grows. The cell Reynolds
number is a predefined quantity available for visualization and evaluation (typically it
is available as: [Link]).

606 | CHAPTER 6: FLUID FLOW INTERFACES

Gravity

DEFINITION
The gravity force is defined from the acceleration of gravity vector, g, and the density,
. Under usual conditions and in Cartesian coordinates with the z-axis in the vertical
direction,

 0 
 
g =  0 .
 
 – g const 

When gravity is considered, a volume force equal to g is included in the momentum

equation. For example, for laminar weakly compressible flow, it reads:

u
 ------ + u  u =    – pI +   u +  u  T  – ---     u I + F + g
2
(6-35)
t  3 

Introducing a constant reference density ref, and assuming that g is homogeneous,

this equation is equivalently written:

u
 ------ + u  u=
t
(6-36)
   – pI +   u +  u  T  – ---     u I +  ref g   r – r ref  + F +   –  ref g
2
 3 

where r is the position vector and rref is an arbitrary reference position vector.

From this equation, it is convenient to define the reduced pressure which accounts for
the hydrostatic pressure, p̃ = p –  ref g   r – r ref 

u
 ------ + u  u =    – p̃I +   u +  u  T  – ---     u I +
2
t 3 (6-37)
F +   –  ref g

In Equation 6-37, the gravity force is written   refg.

PRESSURE FORMULATION
When the relative pressure is used (default option) the interface dependent variable
represents the relative pressure and the absolute pressure is defined as pA  pref  p.
When the pressure is used to define a boundary condition (for example, when p0

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 607
 defines the pressure condition at an outlet), it represents the relative pressure. Hence
defining the outlet pressure as phydro,approx  refg  r  rref compensates for the
gravity force for an ambient reference pressure of 0 Pa when the density is constant,
there is no external force, and provided pref, g, and r0 are defined consistently.

When the reduced pressure is used, the interface dependent variable (named p by
default) represents the reduced pressure. The absolute pressure is then defined as
pA  pref  refg  r  rref  p. In this case when the pressure is used to define a
boundary condition (for example, to define a pressure condition at an outlet), its value
corresponds to the reduced pressure. Hence, the prescribed pressure compensates for
an approximate hydrostatic pressure, phydro,approx  refg  r  rref, which is exact
only when the density is constant and there is no external force.

PRESSURE BOUNDARY CONDITION

For an immobile fluid the momentum equation simplifies to   pI  F g or

   – p̃I  = F +   –  ref g depending on the pressure formulation.

For incompressible flow, assuming there are no external forces, this leads respectively
to p   refg  r  rref  p0 or p̃ = p 0 .

Theory for the Wall Boundary Condition

See Wall for the node settings.

SLIP
The Slip condition assumes that there are no viscous effects at the slip wall and hence,
no boundary layer develops. From a modeling point of view, this is a reasonable
approximation if the important effect of the wall is to prevent fluid from leaving the
domain. Mathematically, the constraint can be formulated as:

u  n = 0,  – pI +   u +  u  T  n = 0

The no-penetration term takes precedence over the Neumann part of the condition
and the above expression is therefore equivalent to

u  n = 0, K n –  K n  n n = 0
K n =   u +  u  T n

expressing that there is no flow across the boundary and no viscous stress in the
tangential direction.

608 | CHAPTER 6: FLUID FLOW INTERFACES

For a moving wall with translational velocity utr, u in the above equations is replaced
by the relative velocity urel  u  utr.

POROUS SLIP
The Porous slip option, which can be chosen in the Porous treatment of no slip condition
list when Enable porous media domains is activated, results in a special treatment of Wall
boundaries and Interior Wall boundaries adjacent to porous domains (Porous Medium
feature). Namely, similar to the Navier slip boundary condition, no penetration and
tangential stress conditions are applied at the wall:

 u slip
u  n = 0 K n –  K n  n n = – ----- ---------- ,
 p y ps
 T
u slip = u –  u  n n , K n = -----  u +  u  n
p

Here, Kn is the viscous wall traction, n is the wall normal, uslip is the tangential velocity
at the wall while real no slip is assumed to be applied at a distance dw (half-height of
the first cell adjacent to the wall) outside the wall, and is the porous slip length. An
analytical derivation of the velocity profile in the boundary layer where the pressure
gradient is balanced by the sum of the Darcy term, the Forchheimer drag, and the
viscous term (neglecting convective terms) leads to the following expression for yps:

 1 + X e – p  1 + X e – p 2 d l pore

= l̃  --------------------------  -------------------------- e p – 1 ,
p p w
y ps

 p = ------ , l̃ = ------------------------
–p  1 + X p  l̃ 1 + 2c 
 1 – Xp e 

1 + 2c  – 1 + 4--- c  1 + 4c D – 1  ND  2
3
X p = --------------------------------------------------- , c  = -------------------------------- , c D = ----------- ------ p - g
4 2  
1 + 2c  + 1 + --- c 
3

where l pore =    p is the porous length scale,  ND = c F   is the non-Darcian

coefficient (cF is the Forchheimer coefficient), g is the gravity vector, and p, l, Xp, c,
cD, are intermediate variables. This formula is used when the Pressure-gradient
formulation is chosen and it uses the pressure gradient at the wall. By default, the
Velocity formulation is activated and the corresponding formula is:

 wd l pore
y ps = l̃  e p – 1  ,  p = ------ , l̃ = ------------------------
l̃ 1 + 2c 
 ND  u slip l pore 2 13
c  =  ---------------------------- ----------   2
  dw 

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 609
 Although this formulation is an approximation, since it uses the slip velocity at the wall
to reconstruct the pressure gradient, it is rather accurate. Moreover, the influence of
the convective terms is partially accounted for in this formulation.

SLIDING WALL
The sliding wall option is appropriate if the wall behaves like a conveyor belt; that is,
the surface is sliding in its tangential direction. The wall does not have to actually move
in the coordinate system.

• In 2D, the tangential direction is unambiguously defined by the direction of the

boundary, but the situation becomes more complicated in 3D. For this reason, this
boundary condition has slightly different definitions in the different space
dimensions.
• For 2D and 2D axisymmetric components, the velocity is given as a scalar Uw and
the condition prescribes
u  n = 0, u  t = Uw

where t  (ny , nx) for 2D and t  (nznr) for axial symmetry.

• For 3D components, the velocity is set equal to a given vector uw projected onto
the boundary plane:
u w –  n  u w n
u = ---------------------------------------- u w
u w –  n  u w n

The normalization makes u have the same magnitude as uw even if uw is not exactly
parallel to the wall.

Navier Slip
This boundary condition enforces no-penetration at the wall, u  nwall  0, and adds a
tangential stress


K nt = – --- u slip


where Knt  Kn  Kn  nwallnwallKn  Knwall, and K is the viscous stress tensor. is
a slip length, and uslip  u  u  nwallnwall is the velocity tangential to the wall. The
boundary condition does not set the tangential velocity component to zero; however,
the extrapolated tangential velocity component is 0 at a distance  outside the wall.

The Slip Length setting is per default set to Factor of minimum element length. The slip
length  is then defined as   fhhmin, where hmin is the smallest element side

610 | CHAPTER 6: FLUID FLOW INTERFACES

(corresponds to the element size in the wall normal direction for boundary layer
elements) and fh is a user input.

In cases where the wall movement is nonzero, Account for the translational wall velocity
in the friction force may be selected to use u  ubnd  u ubnd  nwallnwall instead
of uslip in the friction force. Then, the extrapolated tangential velocity component is
ubnd at a distance  outside of the wall. Note that the Velocity of sliding wall uw is
always accounted for in the friction force.

The Navier Slip option is not available when selecting a turbulence model.

CONSTRAINT SETTINGS
The wall feature uses three different techniques to constrain the velocity field:

• Pointwise constraints is the standard technique to enforce strong constraints in the

finite element method. The desired value of the velocity is prescribed at each node
point in the mesh. Since the constraint is enforced locally at each node, only local
values are affected by the constraint and the constraints are independent of each
other. The solvers can therefore eliminate both the constrained degrees of freedom
and the constraint force terms, effectively reducing the number of degrees of
freedom being solved for.
The main advantage of pointwise constraints is that they enforce the constraint
exactly pointwise. This means that they do not introduce any leakage of the velocity
across the wall, unless specified. The main disadvantage of pointwise constraints is
that they introduce locking effects when trying to impose a no-penetration
condition for the velocity, u · n  0, on curved walls or walls with sharp corners.
• Weak constraints is an alternative method to prescribe the velocity. It consists on
enforcing the boundary condition for the velocity via Lagrange multipliers. Their
main advantage is that the Lagrange multiplier can provide an accurate
representation of the reaction flux at the wall. Their main disadvantage is that they
introduce extra unknowns, and are usually difficult to combine with other constraint
methods on adjacent boundaries. Moreover, they may require extra constraints for
the Lagrange multipliers. For more information, see Weak Constraints in the
COMSOL Multiphysics Reference Manual.
• Nitsche constraints use a numerical flux to prescribe the velocity at the wall. They
impose the constraint in an integral sense rather than pointwise, and do not suffer
from the locking effects introduced by pointwise constraints when trying to
prescribe a no penetration condition for the velocity. They are also better behaved
when prescribing nonlinear constraints. Their main disadvantage is that the

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 611
 constraint is only imposed approximately, and may produce small leaks. For more
information, see Nitsche Constraints.

The following combination of Constraint techniques can be selected in the Constraint

Setting sections of Wall boundary conditions:

• Automatic (default) use different constraint methods depending on whether only the
normal component of the velocity is prescribed, such as in the no penetration
condition, u · n  0, imposed, for example, in Slip walls or No Slip walls using Wall
Functions or Automatic Wall Treatment, or both tangential and normal
components are prescribed, as is the case of No Slip walls in laminar flow.
Nitsche constraints are used to impose the no penetration condition for Slip walls.
When a No Slip condition is prescribed, pointwise constraints are used except for
moving walls where Nitsche constraints are used.
• Pointwise constraints
• Nitsche constraints
• Weak constraints — these are not available on Interior Walls.
• Mixed constraints — this option is only available when both the tangential and
normal components of the velocity need to be prescribed. The velocity on the wall
normal direction is imposed via pointwise constraints. The constraint for the
tangential directions is relaxed, and Nitsche constraints are used instead. This
provides improved accuracy and performance when working with coarse boundary
layer meshes. For more information, see Ref. 18.

Prescribing Inlet and Outlet Conditions

The Navier–Stokes equations can show large variations in mathematical behavior,
ranging from almost completely elliptic to almost completely hyperbolic. This has
implications when it comes to prescribing admissible boundary conditions. There is
also a discrepancy between mathematically valid boundary conditions and practically
useful boundary conditions. See Inlet and Outlet for the node settings.

INLET CONDITIONS
An inlet requires specification of the velocity components. The most robust way to do
this is to prescribe a velocity field using a Velocity condition.

A common alternative to prescribing the complete velocity field is to prescribe a

pressure and all but one velocity component. The pressure cannot be specified

612 | CHAPTER 6: FLUID FLOW INTERFACES

pointwise because this is mathematically over-constraining. Instead the pressure can be
specified via a stress condition:

u n
– p + 2 --------- = F n (6-38)
n

where unn is the normal derivative of the normal velocity component.

Equation 6-38 is prescribed by the Pressure condition in the Inlet and Outlet features
and the Normal stress condition in the Open Boundary and Boundary Stress features.
Equation 6-38 is mathematically more stringent compared to specifying the pressure
pointwise and at the same time cannot guarantee that p obtains the desired value. In
practice, p is close to Fn, except for low Reynolds number flows where viscous effects
are the only effects that balance the pressure. In addition to Equation 6-38, all but one
velocity component must be specified. For low Reynolds numbers, this can be specified
by a vanishing tangential stress condition:

u t
 -------- = 0
n

which is what the Normal stress condition does. Vanishing tangential stress becomes a
less well-posed inlet condition as the Reynolds number increases. The Pressure
condition in the Inlet feature therefore requires a flow direction to be prescribed,
which provides a well-posed condition independent of Reynolds number.

OUTLET CONDITIONS
The most common approach is to prescribe a pressure via a normal stress condition on
the outlet. This is often accompanied by a vanishing tangential stress condition:

u t
 -------- = 0
n

where utn is the normal derivative of the tangential velocity field. It is also possible
to prescribe ut to be zero. The latter option should be used with care since it can have
a significant effect on the upstream solution.

The elliptic character of the Navier–Stokes equations mathematically permits

specifying a complete velocity field at an outlet. This can, however, be difficult to apply
in practice. The reason being that it is hard to prescribe the outlet velocity so that it is
consistent with the interior solution at each point. The adjustment to the specified
velocity then occurs across an outlet boundary layer. The thickness of this boundary

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 613
 layer depends on the Reynolds number; the higher the Reynolds number, the thinner
the boundary layer.

ALTERNATIVE FORMULATIONS
COMSOL Multiphysics provides several specialized boundary conditions that either
provide detailed control over the flow at the boundary or that simulate specific devices.
In practice they often prescribe a velocity or a pressure but calculate the prescribed
values using, for example, ODEs.

Normal Stress Boundary Condition

Mass Flow
The Mass flow boundary condition constrains the mass flowing into the domain across
an inlet or outlet boundary. The mass flow can be specified in a number of ways.

POINTWISE MASS FLUX

The pointwise mass flux sets the velocity at the boundary to

mf
u = – ------- n


where mf is the normal mass flux and is the density.

MASS FLOW RATE

The mass flow rate boundary condition sets the total mass flow through the inlet
boundary according to

–  dbc   u  n  dS = m


where dbc (only present in the 2D Cartesian axis system) is the boundary thickness
normal to the fluid-flow domain and m is the total mass flow rate. For the outlet
boundary, the mass flow rate boundary condition sets the total mass flow as

 dbc   u  n  dS = m


614 | CHAPTER 6: FLUID FLOW INTERFACES

In addition to the constraint on the total flow across the boundary, the tangential
velocity components are set to zero on the boundary

un = 0 (6-39)

STANDARD FLOW RATE

The standard flow rate boundary condition specifies the mass flow as a standard
volumetric flow rate. The mass flow through the inlet boundary is set by the equation


–  dbc ------st-  u  n  dS = Qsv


where dbc (only present in the 2D component Cartesian axis system) is the boundary
thickness normal to the fluid-flow domain, st is the standard density, and Qsv is the
standard flow rate. The standard density is defined by one of the following equations:

M
 st = -------n-
Vn

p st M n
 st = ---------------
-
RT st

where Mn is the mean molar mass of the fluid, Vn is the standard molar volume, pst is
the standard pressure, R is the universal molar gas constant, and Tst is the standard
temperature. For the outlet boundary, the mass flow rate is set by the equation


 dbc ------st-  u  n  dS = Q sv


Equation 6-39 or Equation 6-40 is also enforced for compressible and incompressible
flow, respectively, ensuring that the normal component of the viscous stress and the
tangential component of the velocity are zero at the boundary.

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 615
 Fully Developed Flow (Inlet)
The Fully developed flow boundary condition can be understood from the following
figure:

pinl 

The flow to the domain  is assumed to enter through a straight channel of length L.
The channel is a virtual extrusion of the inlet cross section and a pressure constant
pressure Pinl is applied on the inlet of the virtual channel.

The Fully Developed Flow boundary condition prescribes that the tangential flow
component on the boundary is zero:

u –  u  n n = 0

The momentum equation for a fully developed flow in the virtual extrusion of the inlet
cross section can then be projected onto the inlet boundary with the following weak
equation as the result:

 p + P inl 
   L  –  t u +  t u   + ----------------------- I :   t û  – P inl  n  û 
T
2


The exact value of L is somewhat arbitrary as long as it is not too high or too low. L is
therefore set to ten times the inlet edge length in 2D and to ten times the square root
of the inlet area in 2D axisymmetry and in 3D.

The fact that the velocity profile is not prescribed, but rather the solution of a projected
weak contribution, means that the actual velocity profile that is obtained on the inlet
can deviate from the analytical fully developed flow profile (in cases such an analytical
solution exists) if required by the solution inside the computational domain, . This is
most notably if the flow is strongly curved just downstream of the inlet. This can, for
example, be the case if an object is positioned just downstream of the inlet or if the
adjacent boundaries are not orthogonal to the inlet. The analytical solution can in
these cases be recovered by including a little bit of the inlet channel in the
computational domain.

616 | CHAPTER 6: FLUID FLOW INTERFACES

The inlet pressure, Pinl, must be solved for and the its equation is a discrete algebraic
equation (DAE) for Pinl. When, for example, the average velocity is specified, the DAE
reads

  u  n + U av P̂ inl

where  denotes the average over the inlet. Since the equation for Pinl is a DAE (the
equation for Pinl does not contain Pinl), it must be solved coupled to Navier–Stokes
and it must be treated by a Vanka presmoother and postsmoother if iterative solvers are
used.

The boundary conditions for the virtual inlet channel are inherited from the
boundaries adjacent to the inlet channel. Virtual boundaries adjacent to walls (except
slip walls) are treated as no-slip walls. Virtual boundaries adjacent to any other type of
boundary are treated as slip-walls (or equivalently as symmetry boundaries). So in the
figure above, if the lower boundary of  is a no-slip wall and the top boundary is a
symmetry boundary, the lower boundary of the virtually extruded channel becomes a
no-slip wall and the upper boundary becomes a symmetry boundary.

Algebraic turbulence models need additional equations or constraints. For all other
turbulence models, their weak equations in the virtual inlet channel are projected onto
the inlet of the computational domain. The projections are regularized to prevent the
production to fall to zero, so the solutions of the projects can show slight deviation
from the correct solutions close to no-slip walls.

Fully Developed Flow (Outlet)

The Fully developed flow boundary condition can be understood from the following
figure:

 pexit

The flow to the domain  is assumed to enter through a straight channel of length L.
The channel is a virtual extrusion of the inlet cross section and a pressure constant
pressure Pexit is applied on the inlet of the virtual channel.

The Fully Developed Flow boundary condition prescribes that the tangential flow
component on the boundary is zero:

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 617
 u –  u  n n = 0

The momentum equation for a fully developed flow in the virtual extrusion of the inlet
cross section can then be projected onto the inlet boundary with the following weak
equation as the result:

 p + P exit 
   L  –  t u +  t u   + -------------------------- I :   t û  – P exit  n  û 
T
2   


The exact value of L is somewhat arbitrary as long as it is not too high or too low. L is
therefore set to ten times the inlet edge length in 2D and to ten times the square root
of the inlet area in 2D axisymmetry and in 3D.

The fact that the velocity profile is not prescribed, but rather the solution of a projected
weak contribution, means that the actual velocity profile that is obtained on the outlet
can deviate from the analytical fully developed flow profile (in cases such an analytical
solution exists) if required by the solution inside the computational domain, . The
inlet pressure, Pexit, must be solved for and the its equation is a discrete algebraic
equation (DAE) for Pexit. When, for example, the average velocity is specified, the
DAE reads

  u  n + U av P̂ exit

where  denotes the average over the inlet. Since the equation for Pext is a DAE, it
must be solved coupled to Navier–Stokes and it must be treated by a Vanka
presmoother and postsmoother if iterative solvers are used.

The boundary conditions for the virtual outlet channel are inherited from the
boundaries adjacent to the outlet channel. Virtual boundaries adjacent to walls (except
slip walls) are treated as no-slip walls. Virtual boundaries adjacent to any other type of
boundary are treated as slip-walls (or equivalently as symmetry boundaries).

No Viscous Stress
For this module, and in addition to the Pressure, No Viscous Stress boundary
condition, the viscous stress condition sets the viscous stress to zero:

   u +  u  T  – 2
---     u I n = 0
 3 

   u +  u  T  n = 0

618 | CHAPTER 6: FLUID FLOW INTERFACES

using the compressible/weakly compressible and the incompressible formulation,
respectively.

The condition is not a sufficient outlet condition since it lacks information about the
outlet pressure. It must hence be combined with pressure point constraints on one or
several points or lines surrounding the outlet.

This boundary condition is numerically the least stable outlet condition, but can still
be beneficial if the outlet pressure is nonconstant due to, for example, a nonlinear
volume force.

Normal Stress Boundary Condition

The total stress on the boundary is set equal to a stress vector of magnitude f0, oriented
in the negative normal direction:

 – pI +    u +  u  T  – 2
---     u I  n = – f 0 n
  3 

 – pI +   u +  u  T  n = – f 0 n

using the compressible/weakly compressible and the incompressible formulation,

respectively.

This implies that the total stress in the tangential direction is zero. This boundary
condition implicitly sets a constraint on the pressure which for 2D flows is

u n
p = 2 ---------- + f 0 (6-40)
n

If unn is small, Equation 6-40 states that p  f0.

The Normal Stress condition is the mathematically correct version of the Pressure
Conditions condition (Ref. 4), but it is numerically less stable.

Pressure Boundary Condition

For single-phase flow, a mathematically correct natural boundary condition for outlets
is

 – pI +   u +  u  T  – 2
---     u I n = – p 0 n (6-41)
 3 

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 619
  – pI +   u +  u  T  n = – p 0 n (6-42)

using the compressible/weakly compressible and the incompressible formulation,

respectively.

This is a normal stress condition together with a no-tangential-stress condition. When

  ,Equation 6-41 or Equation 6-42 can be supplemented with a tangential velocity
condition

u  t  (6-43)

If so, the no-tangential-stress condition is overridden. An issue with Equation 6-41 or

Equation 6-42 is that it does not strongly enforce unidirectional flow on the boundary.
If the prescribed pressure on an outlet is too high, parts of the outlet can actually have
inflow. This is not as much of an issue for the Navier–Stokes equations as it is an issue
for scalar transport equations solved along with the Navier–Stokes equations. Hence,
when applying the Pressure boundary condition at an outlet or inlet you can further
constrain the flow. With the Suppress backflow option

 – pI +   u +  u  T  – 2
---     u I n = – p̂ 0 n
 3 
, (6-44)
 – pI +   u +  u  T  n = – p̂ 0 n
p̂ 0  p 0

the normal stress is adjusted to keep

un0 (6-45)

Equation 6-44 effectively means that the prescribed pressure is p0 if u  n  , but

smaller at locations where u  n  . This means that Equation 6-44 does not
completely prevent backflow, but the backflow is substantially reduced. Backflow is
suppressed also when external forces are acting on the fluid, provided the magnitude
of these forces are of the same order as the dynamic pressure at the outlet.

A pressure condition can also be applied at an inlet. In this case, either the normal stress
is prescribed

620 | CHAPTER 6: FLUID FLOW INTERFACES

 n  – pI +   u +  u  T  – ---     u I n = – p̂ 0
T 2
3
T
(6-46)
n  – pI +   u +  u  T  n = – p̂ 0
p̂ 0  p 0

together with the tangential condition in Equation 6-43, or, a general flow direction
is prescribed.

T  – pI +   u +  u  T  – 2
ru ---     u I n = – p̂ 0  r  n 
 3  u
T  – pI +   u +  u  T  n = – p̂  r  n 
ru 0 u
(6-47)
p̂ 0  p 0
du
u –  u  r u r u = 0 r u = ----------
-
du

The “>” option is used with suppress backflow to have u  n   or u  ru  .

For incompressible single-phase flow, it is also allowed to specify the total pressure,
ptot, instead of the static pressure, pstat, on inlet and outlet boundaries. It is more
useful, for example, in pump applications. The pressure is then prescribed at the
boundaries using Bernoulli’s principle,

1 2
p = p stat = p tot – ---  u (6-48)
2

The equation is imposed with two options: Average and Pointwise.

In the first option, pstat is prescribed by:

T T
n  – pI +   u +  u  T  n = n  – p stat I +   u +  u  T  n (6-49)

where ptot and u2 are averaged over the boundaries using the aveop operator:

1 2
p stat = aveop  p tot  – ---   aveop  u 
2

For the second option, Equation 6-48 is prescribed pointwise.

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 621
 See Inlet, Outlet, Open Boundary, and No Viscous Stress for the individual node
settings. Note that some modules have additional theory sections describing options
available with that module.

Porous Interface
This condition properly accounts for a sudden jump in porosity p across boundary
between different domains, giving the corresponding jump in stress across the
interface. The emergence of a boundary layer on the upstream side of the interface is
also revealed, which provides modifications to the behavior of the tangential velocity
near the interface. The feature helps to avoid the need to add a very thin transition
region, in which p,upstream changes to p,downstream continuously, while solving the
Brinkman equation in such situations. Below, ‘’ and ‘’ (used as subscripts and
superscripts) refer to the upstream and downstream side of the interface, respectively.

Whatever the option for Continuity conditions is chosen, the velocity conditions are


 
u  n = 0 ut   u –  u  n n  =0
 


That is, the mass flux and tangential velocity across the interface are constant, and n is
assumed to be a normal pointing from ‘’ to ‘’. Velocity and stress option just imposes
usual stress continuity adding a correction for a possible jump in density:


pn – Kn +   u  n   -----
1  2u = 0
p 
av


The option Velocity and porosity-corrected stress (default) adds a contribution due to
the inertial term, which is nontrivial (gives nonzero effect) when p jumps across the
interface:


pn – Kn +   u  n   -----
1  2 u +   u  n   --------
1 -
 u av = 0
p   2 2p 
av


=  ----- + -----   2 , u av =  u  + u    2 and so on.

1 1 1
Notice that -----
 p av p  p 
Then, the Velocity and porosity-corrected stress with loss option has a ‘pressure loss’
term:

622 | CHAPTER 6: FLUID FLOW INTERFACES

 
pn – Kn +   u  n   -----
1  2 u +   u  n   --------
1 
- u av +p
(loss)
n = 0
 p   2 2p 
av


This loss occurs in the downstream direction. Its mechanism is similar to the
mechanism of common engineering ‘total pressure loss’. Indeed, ‘expansion’ and
‘contraction’ across the interface experienced by the flow on the microscopic level (due
to increasing or decreasing porosity, p, respectively), are roughly equivalent to
expansion and contraction of flow in a pipe with geometrical variations. Accordingly,
it is assumed that only the normal velocity component is responsible for the loss. In
the case of normal flow and negligible viscous term the pipe-analogy is quite exact.

Subsequently, Pressure loss model should be specified. Choosing User-defined loss

coefficients requires values of Expansion loss coefficient and Contraction loss coefficient
(Kexp  1 and Kcontr  0.5 are default values). The resulting losses on the ‘expansion’
and ‘contraction’ boundaries are:

(loss,expansion)     u  n  2 
p -  
= K exp   -----------------------
  2 2p  

(loss,contraction)    u  n  2
p = K contr  -----------------------
-
 2 2p  

Borda–Carnot option results in the following losses:

 av  u  n  u  n 2
= -------  ----------------- – -----------------  
(loss,expansion)
p
2 p  p 
1    u  n  -  1
2 2
  ------- – 1
(loss,contraction)
p = ---  -----------------------
2 p
2
  c vc

and requires to choose Vena contracta model. The User defined option asks for value of
Vena contracta coefficient (cvc  0.5 by default). Weisbach option computes the
coefficient based on the ratio of porosities across the interface:

  p,  n vc
c vc = c vc,ll +  1 – c vc,ll   ----------
  p, 

and requires inputs for Lower limit and Exponent (cvc,ll  0.63 and nvc  3 are default
values).

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 623
 Although the velocity conditions formally look the same for all Continuity conditions
options, the latter two (containing the inertial correction due to the porosity jump)
lead to an extra nontrivial effect. Namely, the tangential velocity exponentially evolves
over a boundary layer formed on the upstream side of the interface boundary. This
boundary layer has a characteristic length equal to

 p, 
l  = --------------------
-
u  n

The ratio of the tangential velocities right after the interface (‘downstream’) and
outside the boundary layer (‘upstream infinity’) is

u t,  2 p    p 
- = --------------------------------------------------------
------------------------------------------ -
u t,upstream infinity  p    p  +  p    p 

In terms of the interstitial velocity, u(i) = u/p, this results in:

(i)
u t,  2
- = ---------------------------------------------------------  1
-----------------------------------------------
(i)
u t,upstream infinity  p  p 
  +  p    p 

Obviously, tangential component of the interstitial kinetic energy is lost across the
interface. For low viscosity and high normal-flow component, when lu is much lower
than lgeom (the characteristic geometry length) there are apparent variations in ut and
u(i)t in a thin boundary layer on the upstream side of the interface. However, for the
cases with high viscosity or nearly interface-parallel flow, lu is much larger than lgeom,
and any variations in ut and u(i)t are virtually invisible.

Notice that the feature does not take into account possible underresolved porous scale,
as is described in Theory for the Free and Porous Media Flow, Darcy Interface.

624 | CHAPTER 6: FLUID FLOW INTERFACES

Mass Sources for Fluid Flow
There are two types of mass sources in a Single-Phase Flow interface: point sources and
line sources.

These features require at least one of the following licenses: Battery

Design Module, CFD Module, Chemical Reaction Engineering Module,
Corrosion Module, Electrochemistry Module, Electrodeposition
Module, Fuel Cell & Electrolyzer Module, Microfluidics Module, Pipe
Flow Module, Polymer Flow Module, or Subsurface Flow Module.

POINT SOURCE
ꞏ
A point source is theoretically formed by taking a mass injection/ejection, Q (SI unit:
3
kg/(m ·s)), in a small volume V and then letting the size of the volume tend to zero
while keeping the total mass flux constant. Given a point source strength, qꞏ p (SI unit:
kg/s), this can be expressed as

ꞏ
lim
V  0 Q = qꞏ p (6-50)
V

An alternative way to form a point source/sink is to assume that mass is injected/

extracted through the surface of a small object. Letting the object surface area tend to
zero while keeping the mass flux constant, results in the same point source. For this
alternative approach, effects resulting from the physical object volume, such as drag
and fluid displacement, need to be neglected.

The weak contribution

qꞏ p test  p 
ꞏ
is added to a point in the geometry. As can be seen from Equation 6-50, Q must tend
to plus or minus infinity as V tends to zero. This means that in theory the pressure
also tends to plus or minus infinity.

Observe that “point” refers to the physical representation of the source. A point source
can therefore only be added to points in 3D components and to points on the
symmetry axis in 2D axisymmetry components. Other geometrical points in 2D
components represent physical lines.

The finite element representation of Equation 6-50 corresponds to a finite pressure in

a point with the effect of the point source spread out over a region around the point.

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 625
 The size of the region depends on the mesh and on the strength of the source. A finer
mesh gives a smaller affected region, but also a more extreme pressure value. It is
important not to mesh too finely around a point source since the resulting pressure can
result in unphysical values for the density, for example. It can also have a negative effect
on the condition number for the equation system.

LINE SOURCE
ꞏ
A line source can theoretically be formed by assuming a source of strength Q (SI unit:
3
kg/(m ·s)), located within a tube with cross-sectional area S and then letting S tend
to zero, while keeping the total mass flux per unit length constant. Given a line source
strength, qꞏ l (SI unit: kg/(m·s)), this can be expressed as

ꞏ
lim
S  0  Q = qꞏ l (6-51)
S

As in the point source case, an alternative approach is to assume that mass is injected/
extracted through the surface of a small object. This results in the same mass source,
but requires that effects on the fluid resulting from the physical object volume are
neglected.

The weak contribution

qꞏ l test  p 

is added to lines in 3D or to points in 2D (which represent cut-through views of lines).

Line sources can also be added to the axisymmetry line in 2D axisymmetry
components. It cannot, however, be added to geometrical lines in 2D because they
represent physical planes.

As with a point source, it is important not to mesh too finely around the line source.

For feature node information, see Line Mass Source and Point Mass
Source.

626 | CHAPTER 6: FLUID FLOW INTERFACES

 For the Reacting Flow in Porous Media, Diluted Species interface, which
is available with the CFD Module, Chemical Reaction Engineering
Module, or Battery Design Module, these shared physics nodes are
renamed as follows:

• The Line Mass Source node is available as two nodes, one for the fluid
flow (Fluid Line Source) and one for the species (Species Line Source).
• The Point Mass Source node is available as two nodes, one for the fluid
flow (Fluid Point Source) and one for the species (Species Point Source).

Numerical Stability — Stabilization Techniques for Fluid Flow

The momentum equation (Equation 6-32 or Equation 6-34) is a (nonlinear)
convection-diffusion equation. Such equations can easily become unstable if
discretized using the Galerkin finite element method. Stabilized finite element
methods are usually necessary in order to obtain physical solutions. The stabilization
settings are found in the main Fluid Flow interface features. To display this section,
click the Show More Options button ( ) and select Stabilization in the Show More
Options dialog.

There are three types of stabilization methods available for Navier–Stokes —

streamline diffusion, crosswind diffusion, and isotropic diffusion. Streamline
diffusion and crosswind diffusion are consistent stabilization methods, whereas
isotropic diffusion is an inconsistent stabilization method.

For optimal functionality, the exact weak formulations of and constants in the
streamline diffusion and crosswind diffusion methods depend on the order of the
shape functions (basis functions) for the elements. The values of constants in the
streamline diffusion and crosswind diffusion methods follow Ref. 5 and Ref. 6.

STREAMLINE DIFFUSION
For strongly coupled systems of equations, the streamline diffusion method must be
applied to the system as a whole rather than to each equation separately. These ideas
were first explored by Hughes and Mallet (Ref. 8) and were later extended to Galerkin
least-squares (GLS) applied to the Navier–Stokes equations (Ref. 9). This is the
streamline diffusion formulation that COMSOL Multiphysics supports.

Streamline diffusion is active by default because it is necessary when convection is

dominating the flow.

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 627
 The governing equations for incompressible flow are subject to the Babuška–Brezzi
condition, which states that the shape functions (basis functions) for pressure must be
of lower order than the shape functions for velocity. If the incompressible Navier–
Stokes equations are stabilized by streamline diffusion, it is possible to use equal-order
interpolation. Hence, streamline diffusion is necessary when using first-order elements
for both velocity and pressure. This applies also if the model is solved using geometric
multigrid (either as a solver or as a preconditioner) and at least one multigrid hierarchy
level uses linear Lagrange elements.

CROSSWIND DIFFUSION
Crosswind diffusion can also be formulated for systems of equations, and when applied
to the Navier–Stokes equations it becomes a shock-capturing operator. COMSOL
Multiphysics supports the formulation in Ref. 9 with a shock-capturing viscosity of the
Hughes–Mallet type Ref. 8.

Incompressible flows do not contain shock waves, but crosswind diffusion is still useful
for introducing extra diffusion in sharp boundary layers and shear layers that otherwise
would require a very fine mesh to resolve.

Crosswind diffusion is active by default as it makes it easier to obtain a solution even if

the problem is fully resolved by the mesh. Crosswind diffusion also enables the iterative
solvers to use inexpensive presmoothers. If crosswind diffusion is deactivated, more
expensive preconditioners must be used instead.

STABILIZATION TIME SCALES

The time-scale tensor for time-dependent problems should in theory depend on a
time-scale that many references just set to the time-step taken by the time-solver, t
(see, for example, Ref. 9 and Ref. 10). This simple approach does not, however,
necessarily reflect on the actual time-scales in the physics. A typical example is reacting
flows where the time step is often guided by fast reactions, while the flow develops
relatively slowly. The COMSOL Multiphysics software can therefore replace t2 in the
time-scale tensor with measures of type  1   t̃  2  –1 , which are calculated from
projections of weak expressions in a fashion similar to those in Ref. 11. These measures
of the time scale are used when Use dynamic subgrid time scale is activated.

The Limit small time steps effect on stabilization time scale option can be used to ramp
down the time-step contribution in the stabilization when the time step becomes much
smaller than the convective and diffusive time scales. This option prevents loss of
pressure stabilization at such small time steps.

628 | CHAPTER 6: FLUID FLOW INTERFACES

ISOTROPIC DIFFUSION
Isotropic diffusion adds diffusion to the Navier–Stokes equations. Isotropic diffusion
significantly reduces the accuracy of the solution but does a very good job at reducing
oscillations. The stability of the continuity equation is not improved.

Numerical Stabilization and Iterative

Solvers for Laminar Flow

The Navier–Stokes equations constitute a nonlinear equation system. A nonlinear
solver must hence be applied to solve the problem. The nonlinear solver iterates to
reach the final solution. In each iteration, a linearized version of the nonlinear system
is solved using a linear solver. In the time-dependent case, a time marching method
must also be applied. The default suggestions for each of these solver elements are
discussed below.

NONLINEAR SOLVER
The nonlinear solver method depends on if the model solves a stationary or a
time-dependent problem.

Stationary Solver
In the stationary case, a fully coupled, damped Newton method is applied. The initial
damping factor is low since a full Newton step can be harmful unless the initial values
are close to the final solution. The nonlinear solver algorithm automatically regulates
the damping factor in order to reach a converged solution.

For advanced models, the automatically damped Newton method might not be robust
enough. A pseudo time-stepping algorithm can then be invoked. See Pseudo Time
Stepping for Laminar Flow Models.

Time-Dependent Solver
In the time-dependent case, the initial guess for each time step is (loosely speaking) the
previous time step, which is a very good initial value for the nonlinear solver. The
automatic damping algorithm is then not necessary. The damping factor in the
Newton method is instead set to a constant value slightly smaller than one. Also, for
the same reason, it suffices to update the Jacobian once per time step.

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 629
 It is seldom worth the extra computational cost to update the Jacobian more than once
per time step. For most models it is more efficient to restrict the maximum time step
or possibly lower the damping factor in the Newton method.

LINEAR SOLVER
The linearized Navier–Stokes equation system has saddle point character, unless the
density depends on the pressure. This means that the Jacobian matrix has zeros on the
diagonal. Even when the density depends on the pressure, the equation system
effectively shares many numerical properties with a saddle point system.

For small 2D and 3D models, the default solver suggestion is a direct solver. Direct
solvers can handle most nonsingular systems and are very robust and also very fast for
small models. Unfortunately, they become slow for large models and their memory
requirement scales as somewhere between N1.5 and N2, where N is the number of
degrees of freedom in the model. The default suggestion for large 2D and 3D models
is therefore the iterative GMRES solver. The memory requirement for an iterative
solver optimally scales as N.

GMRES is accelerated by a multigrid method, per default the smoothed aggregation

algebraic multigrid (SAAMG) method. The cost of SAAMG is typically very low
compared to the number of GMRES iterations necessary if no multigrid method is
used. As the name implies, SAAMG builds its coarser meshes algebraically, so the
application requires no additional meshes in order to employ SAAMG. In contrast, the
geometric multigrid (GMG) method requires actual meshes. If a sufficient number of
multigrid levels can be constructed, GMG is often faster than SAAMG. GMG is also
superior for cluster computations and for shared-memory computations with many
cores. When the default linear solver is GMRES, an optional, but deactivated, linear
solver node is available where GMRES is accelerated by GMG.

Multigrid methods need smoothers, but the saddle point character of the linear system
restricts the number of applicable smoothers. The choices are further restricted by the
anisotropic meshes frequently encountered in fluid-flow problems. The efficiency of
the smoothers is highly dependent on the numerical stabilization. Iterative solvers
perform at their best when both Streamline Diffusion and Crosswind Diffusion are
active.

The default smoother for P1+P1 elements is SCGS. This is an efficient and robust
smoother specially designed to solve saddle point systems on meshes that contain
anisotropic elements. The SCGS smoother works well even without crosswind
diffusion. SCGS can sometimes work for higher-order elements, especially if Method in
the SCGS settings is set to Mesh element lines. But there is no guarantee for this, so the

630 | CHAPTER 6: FLUID FLOW INTERFACES

default smoother for higher order elements is an SOR Line smoother. SOR Line
handles mesh anisotropy but does not formally address the saddle point character. It
does, however, function in practice provided that streamline diffusion and crosswind
diffusion are both active.

A different kind of saddle point character can arise if the equation system contains
ODE variables. Some advanced boundary conditions can add equations with such
variables. These variables must be treated with the Vanka algorithm. SCGS includes an
option to invoke Vanka. Models with higher-order elements must apply SCGS or use
the Vanka smoother. The latter is the default suggestion for higher-order elements, but
it does not work optimally for anisotropic meshes.

TIME-DEPENDENT SOLVERS
The default time-dependent solver for Navier–Stokes is the BDF method with
maximum order set to two. Higher BDF orders are not stable for transport problems
in general nor for Navier–Stokes in particular.

BDF methods have been used for a long time and are known for their stability.
However, they can have severe damping effects, especially the lower-order methods.
Hence, if robustness is not an issue, a model can benefit from using the generalized-
method instead. Generalized- is a solver that has properties similar to those of the
second-order BDF solver but it is much less diffusive.

Both BDF and generalized- are per default set to automatically adjust the time step.
While this works well for many models, extra efficiency and accuracy can often be
gained by specifying a maximum time step. It is also often beneficial to specify an initial
time step to make the solver progress smoothly in the beginning of the time series.

• Time-Dependent Solver
• Multigrid, Direct, Iterative, SCGS, SOR Line, and Vanka
• Stationary Solver

Pseudo Time Stepping for Laminar Flow Models

A stationary formulation has per definition no time derivatives and Equation 6-34
reduces to:

T
  u   u =    – pI +   u +  u    + F (6-52)

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 631
 Solving Equation 6-52 requires a starting guess that is close enough to the final
solution. If no such guess is at hand, the fully transient problem can be solved instead.
This is, however, a rather costly approach in terms of computational time. An
intermediate approach is to add a fictitious time derivative to Equation 6-52:

u – nojac  u  T
 ------------------------------- +   u   u =    – pI +   u +  u    + F
t̃

where t̃ is a pseudo time step. Since unojac(u) is always zero, this term does not
affect the final solution. It does, however, affect the discrete equation system and
effectively transforms a nonlinear iteration into a step of size t̃ of a time-dependent
solver.

Pseudo time stepping is not active per default. The pseudo time step t̃ can be chosen
individually for each element based on the local CFL number:

h
t̃ = CFL loc ------
u

where h is the mesh cell size. A small CFL number means a small time step. It is
practical to start with a small CFL number and gradually increase it as the solution
approaches steady state.

If the automatic expression for CFLloc is set to the built-in variable CFLCMP, then the
automatic setting suggests a PID regulator for the pseudo time step in the default
solver. The PID regulator starts with a small CFL number and increases CFLloc as the
solution comes closer to convergence.

The default manual expression is

1.3 min  niterCMP 9  +

if  niterCMP  20 9  1.3 min  niterCMP – 20 9  0  + (6-53)
if  niterCMP  40 90  1.3 min  niterCMP – 40 9  0 

The variable niterCMP is the nonlinear iteration number. It is equal to one for the first
nonlinear iteration. CFLloc starts at 1.3 and increases by 30% each iteration until it
reaches 1.39 10.6. It remains there until iteration number 20 at which it starts to
increase until it reaches approximately 106. A final increase after iteration number 40

632 | CHAPTER 6: FLUID FLOW INTERFACES

then takes it to 1060. Equation 6-53 can, for some advanced flows, increase CFLloc
too slowly or too quickly. CFLloc can then be tuned for the specific application.

For details about the CFL regulator, see Pseudo Time Stepping.

Nitsche Constraints
Some boundary conditions are implemented using Nitsche constraints. These
boundary conditions include

• Wall — Slip, sliding walls, and moving walls (that is, walls with nonzero
translational velocity).
• Periodic Flow Condition
• Flow Continuity

The formulation used in the Fluid Flow interfaces in single-phase flow interfaces
corresponds to the Symmetric Interior Penalty Galerkin method (SIPG). The SIPG
method can be regarded to satisfy the boundary conditions in an integral sense rather
than pointwise. More information on SIPG can be found in Ref. 15.

In particular, the SIPG formulation includes a penalty parameter that must be large
enough for the formulation to be coercive. The higher the value, the better the
boundary condition is fulfilled, but a too high value results in an ill-conditioned
equation system. The penalty parameter is implemented according to Ref. 16.

Particle Tracing in Fluid Flow

The Particle Tracing Module is available to assist with these types of modeling
problems.

It is possible to model particle tracing with COMSOL Multiphysics provided that the
impact of the particles on the flow field is negligible. First compute the flow field, and
then, as an analysis step, calculate the motion of the particles. The motion of a particle
is defined by Newton’s second law

2
d x
= F  t x 
dx
m
dt2 dt

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 633
 where x is the position of the particle, m the particle mass, and F is the sum of all forces
acting on the particle. Examples of forces acting on a particle in a fluid are the drag
force, the buoyancy force, and the gravity force. The drag force represents the force
that a fluid exerts on a particle due to a difference in velocity between the fluid and the
particle. It includes the viscous drag, the added mass, and the Basset history term.
Several empirical expressions have been suggested for the drag force. One of those is
the one proposed by Khan and Richardson (Ref. 13). That expression is valid for
spherical particles for a wide range of particle Reynolds numbers. The particle
Reynolds number is defined as

u – u p 2r
Re p = ----------------------------


where u is the velocity of the fluid, up the particle velocity, r the particle radius,  the
fluid density, and  the dynamic viscosity of the fluid. The empirical expression for the
drag force according to Khan and Richardson is

2 -0.31 3.45
F = r  u – u p  u – u p   1.84Re p + 0.293Re p0.06 

The model Flow Past a Cylinder (Application Library path

COMSOL_Multiphysics/Fluid_Dynamics/cylinder_flow) demonstrates how to
add and set up particle tracing in a plot group using the Particle Tracing
with Mass node. It uses the predefined Khan–Richardson model for the
drag force and neglects gravity and buoyancy forces.

References for the Single-Phase Flow, Laminar Flow Interfaces

1. G.G. Stokes, Trans. Camb. Phil. Soc., vol. 8, pp. 287–305, 1845.

2. P.M. Gresho and R.L. Sani, Incompressible Flow and the Finite Element Method,
Volume 2: Isothermal Laminar Flow, John Wiley & Sons, 2000.

3. G.K. Batchelor, An Introduction To Fluid Dynamics, Cambridge University Press,

1967.

4. R.L. Panton, Incompressible Flow, 2nd ed., John Wiley & Sons, 1996.

5. I. Harari and T.J.R. Hughes, “What are C and h? Inequalities for the Analysis and
Design of Finite Element Methods,” Comput. Methods Appl. Mech. Eng., vol. 97,
pp. 157–192, 1992.

634 | CHAPTER 6: FLUID FLOW INTERFACES

6. Y. Bazilevs, V.M. Calo, T.E. Tezduyar, and T.J.R. Hughes, “YZ Discontinuity
Capturing for Advection-dominated Processes with Application to Arterial Drug
Delivery,” Int. J. Num. Meth. Fluids, vol. 54, pp. 593–608, 2007.

7. R.B. Bird, W.E. Stewart, and E.N. Lightfoot, Transport Phenomena, 2nd ed., John
Wiley & Sons, 2007.

8. T.J.R. Hughes and M. Mallet, “A New Finite Element Formulation for

Computational Fluid Dynamics: III. The Generalized Streamline Operator for
Multidimensional Advective-Diffusive System,” Comput. Methods Appl. Mech. Eng.,
vol. 58, pp. 305–328, 1986.

9. G. Hauke and T.J.R. Hughes, “A Unified Approach to Compressible and

Incompressible Flows,” Comput. Methods Appl. Mech. Eng., vol. 113, pp. 389–395,
1994.

10. G. Hauke, “Simple Stabilizing Matrices for the Computation of Compressible

Flows in Primitive Variables”, Comput. Methods Appl. Mech. Eng., vol. 190,
pp. 6881–6893, 2001.

11. M.-C. Hsu, Y. Bazilevs, V.M. Cali, T.E. Tezduyar, and T.J.R. Hughes, “Improving
stability of stabilized and multiscale formulations in flow simulations at small time
steps,” Comput. Methods Appl. Mech. Eng., vol. 199, pp. 828–840, 2010.

12. D.J. Tritton, Physical Fluid Dynamics, 2nd ed., Oxford University Press, 1988.

13. J.M. Coulson and J.F. Richardson, “Particle Technology and Separation
Processes,” Chemical Engineering, Volume 2, Butterworth-Heinemann, 2002.

14. J.L. Guermond, P. Minev, and J. Shen, “An overview of projection methods for
incompressible flows,” Comput. Methods Appl. Mech. Eng., vol. 195, pp. 6011–
6045, 2006.

15. B. Rivière, Discontinuous Galerkin Methods for Solving Elliptic and Parabolic
Equations, SIAM, 2008.

16. Y. Epshteyn and B. Rivière, “Estimation of penalty parameters for symmetric

interior penalty Galerkin methods,” J. Comput. Appl. Math., vol. 206, pp. 843–872,
2007.

17. R.P. Chhabra and J.F. Richardson, Non-Newtonian Flow and Applied Rheology,
2nd ed., Elsevier, 2008.

18. Y. Bazilevs and T.J.R. Hughes, “Weak imposition of Dirichlet boundary conditions
in fluid mechanics,” Comput. Fluids, vol. 36, pp. 12–26, 2007.

THEORY FOR THE LAMINAR FLOW AND CREEPING FLOW INTERFACES | 635
636 | CHAPTER 6: FLUID FLOW INTERFACES
Theory for the Darcy’s Law Interface
The Darcy’s Law Interface theory is described in this section:

• About Darcy’s Law

• Darcy’s Law — Equation Formulation
• Storage Models
• Average Linear Velocity
• Dual Porosity and Dual Permeability Theory
• References for the Darcy’s Law Interface

About Darcy’s Law

In a porous medium, the global transport of momentum by shear stresses in the fluid
is often negligible, because the pore walls impede momentum transport to the fluid
outside the individual pores. A detailed description, down to the resolution of every
pore, is not practical in most applications. A homogenization of the porous and fluid
media into a single medium is a common alternative approach. Darcy’s law together
with the continuity equation and equation of state for the pore fluid (or gas) provide
a complete mathematical model suitable for a wide variety of applications involving
porous media flows, for which the pressure gradient is the major driving force.

Darcy’s law describes fluid movement through interstices in a porous medium.

Because the fluid loses considerable energy to frictional resistance within pores, flow
velocities in porous media are very low. The Darcy’s Law interface can be applied to
small-scale tasks in chemical engineering as well as to large-scale geophysical or
hydrological tasks such as water moving in an aquifer or stream bank, oil migrating to
a well, and even magma rising through the earth to a chamber in a volcano (see Ref. 1,
Ref. 2, Ref. 3, and Ref. 4). Also set up multiple Darcy’s Law interfaces to model
multiphase flows involving more than one mobile phase.

Darcy’s law describes flow in porous media driven by gradients in the hydraulic
potential field, which has units of pressure. For many applications it is convenient to
represent the total hydraulic potential or the pressure and the gravitational
components with equivalent heights of fluid or head. Division of potential by the fluid
weight can simplify modeling because units of length make it straightforward to
compare to many physical data. Consider, for example, fluid levels in wells, stream
heights, topography, and velocities. The physics interface also supports specifying

THEORY FOR THE DARCY’S LAW INTERFACE | 637

 boundary conditions and result evaluation using hydraulic head and pressure head. In
the physics interface, pressure is always the dependent variable.

Darcy’s Law — Equation Formulation

Darcy’s law states that the velocity field is determined by the pressure gradient, the
fluid viscosity, and the structure of the porous medium:


u = – --- p (6-54)


In this equation, u is the Darcy’s velocity or specific discharge vector (SI unit: m/s);
 is the permeability of the porous medium (SI unit: m2);  is the fluid’s dynamic
viscosity (SI unit: Pa·s); p is the pore pressure (SI unit: Pa) and  is the density of the
fluid (SI unit: kg/m3).

• The above equation formulation is called linear Darcy’s law. The

applicability of this formulation is discussed in the chapter About
Darcian and Non-Darcian Flow in the Porous Media Flow Module
User’s Guide.
• Permeability can be defined in different ways, either directly or based
on predefined permeability models. These are described in detail in the
chapter Permeability Models.
• For a detailed overview of the permeability models available in each
product visit [Link]

The Darcy’s Law interface combines Darcy’s law with the continuity equation

 
 p  +    u  = Q m (6-55)
t

In the above equation,  is the fluid density (SI unit: kg/m3), p is the porosity, and
Qm is a mass source term (SI unit: kg/(m3·s)). Porosity is defined as the fraction of
the control volume that is occupied by pores. Thus, the porosity can vary from zero
for pure solid regions to unity for domains of free flow.

For large-scale applications it might be necessary to take gravity effects into account.
Darcy’s law then applies when the gradient in hydraulic potential drives fluid
movement in the porous medium. Darcy’s law then has the form:

638 | CHAPTER 6: FLUID FLOW INTERFACES

 
u = – ---  p –  g  (6-56)


• See the section Gravity Effects in the Subsurface Flow Module User’s
Guide.
• Note that gravity effects are only included in some modules. For a
detailed overview of the functionality available in each product visit
[Link]
• Gravity effects are not active by default. Select the checkbox Include
gravity to activate the acceleration of gravity. Setting the elevation D to
zero also turns off gravity effects.

Storage Models
The continuity equation (Equation 6-55) allows flexible specification of relationships
between parameters like density or permeability with respect to variables such as
pressure, temperature, concentration, and more.

Expanding the time-derivative term and defining the porosity and density as functions
of the pressure and using the chain rule results in

    =  -----
-  p - -----
p-  p p
p t +  t =  p p t +  p t
-------- ----- -------- ------
t p

By inserting the definition of fluid compressibility fp to the right-hand

side and rearrange the following expression is obtained

  p   p p
---------------- =    p  f + -------p- ------ = S p ------
t  p  t t

Using this relation, the generalized governing equation (Equation 6-55) takes the
following form:

p
S p ------ +    u  = Q m (6-57)
t

In this equation, Sp is the storage coefficient (SI unit: 1/Pa), which can be interpreted
as the weighted compressibility of the porous material and the fluid. The storage can
be an expression involving results from a solid-deformation equation or an expression
involving temperatures and concentrations from other analyses. The Darcy’s Law

THEORY FOR THE DARCY’S LAW INTERFACE | 639

 interface implements Equation 6-57 in the Porous Medium node, which explicitly
includes different storage models to choose from:

• From density and porosity, where the storage directly depends on changes in porosity
or fluid density
• Linearized storage, which defines Sp using the compressibility of the fluid f and of
the porous matrix p according to
S p =  p  f +  1 –  p  p

• Poroelastic storage where

Sp = p f .

Storage models are only available for the Subsurface Flow Module and
the Porous Media Flow Module. For a detailed overview of the
functionality available in each product visit [Link]
products/specifications/.

Average Linear Velocity

Because fluids typically occupy only 10% to 50% of a porous medium, it follows that
velocities within the pore channels exceed the Darcy velocity, u, on the order of two
to ten times. For clarity, the physics interface includes the average linear velocity within
the pores, u, (also termed the seepage velocity) defined as u  up, where p is the
porosity.

Dual Porosity and Dual Permeability Theory

The first well known publication on this topic is “The Behavior of Naturally Fractured
Reservoirs” by J.E. Warren and P.J. Root from 1963. Since then the researchers further
developed the idea of a dual porosity/permeability type of a porous medium. One of
the well established models is based on the findings of H. Gerke and M. van
Genuchten (Ref. 9). Basically the idea behind is to solve for each part of the porous
medium (macro- and micropores) the relevant equations and connect them via an
exchange term.

DUAL POROSITY MEDIUM

Darcy’s Law is solved for the macropores and taking into account their volume fraction
M the equation reads:

640 | CHAPTER 6: FLUID FLOW INTERFACES

 p M
 M  S p,M ----------- +  u M  =  M Q M – Q ip (6-58)
t

The term Qip describes the mass source from the Interporosity Flow which is the mass
transfer from the macropores to the micropores system. An additional ODE for the
(passive) micropores is solved which only act as additional storage volume:

p m
 1 –  M   S p,m ---------- =  1 –  M Q M + Q ip .
 t 

The indices M and m stand for macropores and micropores, respectively.

DUAL PERMEABILITY MEDIUM

Darcy’s law has to be solved for both the macroscale system and the microscale system:


 M    p,M   +    u M  = – Q ip +  M Q m
t
(6-59)
 1 –  M      p,m   +    u m  = Q ip +  1 –  M Q m
t
p =  M p M +  1 –  M p m

M is the volume fraction of the macropores, Qip denotes the interporosity flow,
signifying the mass transfer from the macropores (identified by index M) to the
micropores (identified by index m).

Solving Darcy’s law for both systems entails solving two equations, each associated
with dependent pressure variables pm and pM. These equations are coupled via Qip.
From this, average quantities for the dual permeability medium as a whole are
computed. The average pressure of the dual permeability medium is defined as:

p =  M p M +  1 –  M p m

Likewise, average values for permeability, storage, velocity and other relevant
parameters are computed.

UNSATURATED DUAL PERMEABILITY MEDIUM

Use the Unsaturated Dual Permeability Medium feature to model the transport in an
unsaturated dual permeability medium where Richards’ Equation (Equation 4-26)
describes the transport in each system.

THEORY FOR THE DARCY’S LAW INTERFACE | 641

 C m,M p M
 M    S e,M S p,M + -------------
- ----------- +  u M  = – Q ip +  M Q m
g  t
C m,m p m (6-60)
 1 –  M     S e,m S p,m + ------------
- ---------- +  u m  = Q ip +  1 –  M Q m
g  t
p =  M p M +  1 –  M p m

M is the volume fraction of the macropores, Qip is the interporosity flow which is the
mass transfer from the macroscale to the microscale system. The indices M and m stand
for macropores and micropores, respectively.

INTERPOROSITY FLOW
Specify the Fluid Transfer Function w which is used to calculate the interporosity flow,
which is the mass exchange between macro- and microscale system:

Q ip =  w  p M – p m  (6-61)

with pM being the pressure within the macroscale system and pm the pressure within
the microscale system

References for the Darcy’s Law Interface

1. J. Bear, Hydraulics of Groundwater, McGraw Hill, 1979.

2. S.E. Ingebritsen and W.E. Sanford, Groundwater in Geologic Processes, Cambridge

University Press, 1998.

3. N.H. Sleep and K. Fujita, Principles of Geophysics, Blackwell Science, 1997.

4. D.L. Turcotte and G. Schubert, Geodynamics, Cambridge University Press, 2002.

5. J. Bear, Dynamics of Fluids in Porous Media, Elsevier Scientific Publishing, 1972.

6. Ö. Akgiray and A.M. Saatçı, “A New Look at Filter Backwash Hydraulics,” Water
Sci. Technol. Water Supply, vol. 1, no. 2, pp. 65–72, 2001.

7. L. Klinkenberg, “The permeability of porous media to liquids and gases,” Drilling

and Production Practice, Am. Petroleum Inst., pp. 200–213, 1941.

8. Yu-Shu Wu and others, “Gas Flow in Porous Media with Klinkenberg Effects,”
Transp. Porous Media, vol. 32, pp. 117–137, 1998.

642 | CHAPTER 6: FLUID FLOW INTERFACES

9. H. Gerke and M. van Genuchten, “A dual-porosity model for simulating the
preferential movement of water and solutes in structured porous media,” Water
Resour. Res., vol. 29. pp. 305–319. 1993.

THEORY FOR THE DARCY’S LAW INTERFACE | 643

 Theory for the Free and Porous
Media Flow, Brinkman Interface
The Free and Porous Media Flow, Brinkman Interface uses the Navier–Stokes
equations to describe the flow in open regions, and the Brinkman equations to
describe the flow in porous regions.

The same fields, u and p, are solved for in both the free flow domains and in the porous
domains. This means that the pressure in the free fluid and the pressure in the pores is
continuous over the interface between a free flow domain and a porous domain. It also
means that continuity is enforced between the fluid velocity in the free flow and the
Darcy velocity in the porous domain. This treatment is one of several possible models
for the physics at the interface. Examples of other models can be found in Ref. 1.

The continuity in u and p implies a stress discontinuity at the interface between a

free-flow domain and a porous domain. The difference corresponds to the stress
absorbed by the rigid porous matrix, which is a consequence implicit in the
formulations of the Navier–Stokes and Brinkman equations.

Reference for the Free and Porous Media Flow, Brinkman Interface
1. M.L. Bars and M.G. Worster, “Interfacial Conditions Between a Pure Fluid and a
Porous Medium: Implications for Binary Alloy Solidification,” J. Fluid Mech.,
vol. 550, pp. 149–173, 2006.

644 | CHAPTER 6: FLUID FLOW INTERFACES

Theory for the Free and Porous
M e di a Flo w, Darc y In t erface
The Free and Porous Media Flow, Darcy interface uses the Navier–Stokes equations to
describe the flow in the free-flow region, implemented by the coupled Laminar Flow
interface, and Darcy’s law, implemented by the coupled Darcy’s Law interface, to
describe the flow in the porous region.

In the free-flow region, the velocity u and pressure p are solved for, whereas in the
porous region only the pressure p is solved for, and the velocity is calculated from the
pressure by Darcy’s law.

The coupling between the free and porous medium flow, implemented by the Free and
Porous Media Flow Coupling, ensures conservation of mass across the free-porous
interface, which is enforced by the condition

n    ns u ns  = n    d u d  (6-62)

Here ns and uns are the density and velocity on the free-flow side of the interface, d
and ud the density and velocity on the porous side, and n is a vector normal to the
interface. In addition the coupling enforces the balance of the normal forces by:

p ns – n  Kn = p d (6-63)

where pns and pd are the pressure on the free-flow side and porous side of the
interface, respectively, and K is the viscous stress tensor (defined only on the free-flow
side).

For the coupling condition for the tangential velocity component, there are several
options. One of the most widely used conditions is the Beavers-Joseph interface
condition (Ref. 1):

tr     n
 I – n  n   u ns – u d  = ------------------------  I – n  n Kn (6-64)
 ns  BJ

where aBJ is the Beavers and Joseph parameter,  is the permeability tensor of the
porous medium (defined on the porous side of the interface), ns is the fluid’s dynamic
viscosity on the free-flow side of the interface, and n is the number of space
dimensions.

THEORY FOR THE FREE AND POROUS MEDIA FLOW, DARCY INTERFACE | 645
 When the velocity in the porous domain is much smaller than the velocity in the free
flow, one can use a modification of the Beavers–Joseph condition which is given by the
Beavers–Joseph–Saffman interface condition (Ref. 2):

tr     n
 I – n  n u ns = ------------------------  I – n  n Kn (6-65)
 ns  BJ

For cases when the permeability tensor is not isotropic, the Use projected permeability
option allows to take into account only the permeability in the direction of the
tangential velocity component. This is accomplished by replacing the tr     n term
on the rights hand side of Equation 6-64 and Equation 6-65 by:

t  t (6-66)

with t a unit vector tangent to the interface for 2D and 2D-axisymmetric cases, and for
3D cases by:

v  v (6-67)

with v given by:

 I – n  n Kn -
v = ------------------------------------------ (6-68)
 I – n  n Kn

If one considers the limit of the Beavers–Joseph condition for large values of the
Beavers and Joseph parameter aBJ, one obtains the Continuous tangential velocity
condition:

 I – n  n   u ns – u d  = 0 (6-69)

The same limit for the Beavers–Joseph–Saffman condition results in the No slip
condition:

 I – n  n u ns = 0 (6-70)

The Viscous slip condition allows for a more general coupling condition. This interface
condition is given by:

Ls
 I – n  n   u ns – u b  = ---------  I – n  n Kn (6-71)
 ns

646 | CHAPTER 6: FLUID FLOW INTERFACES

Here Ls is a slip length, and ub a user-defined interfacial velocity.

Please find related information in the CFD Module User’s Guide:

• The Darcy’s Law Interface

• The Thin-Film and Porous Media Flow Interface
• The Laminar Flow Interface
• Theory for the Single-Phase Flow Interfaces

References
1. G.S. Beavers and D.D. Joseph, “Boundary Conditions at a Naturally Permeable
Wall,” J. Fluid Mech., vol. 30, 1967.

2. P.G. Saffman, “On the Boundary Condition at the Surface of a Porous Medium,”
Stud. Appl. Math., vol. 50, no. 2, pp. 93–101, 1971.

THEORY FOR THE FREE AND POROUS MEDIA FLOW, DARCY INTERFACE | 647
 Theory for the Brinkman Equations
Interface
theory is described in this section:

• About the Brinkman Equations

• Brinkman Equations Theory
• References for the Brinkman Equations Interface

About the Brinkman Equations

In porous domains, the flow variables and fluid properties are defined at any point
inside the medium by means of averaging of the actual variables and properties over a
certain volume surrounding the point. This control volume must be small compared
to the typical macroscopic dimensions of the problem, but it must be large enough to
contain many pores and solid matrix elements.

Porosity is defined as the fraction of the control volume that is occupied by pores.
Thus, the porosity can vary from zero for pure solid regions to unity for domains of
free flow.

The physical properties of the fluid, such as density and viscosity, are defined as
intrinsic volume averages that correspond to a unit volume of the pores. Defined this
way, they present the relevant physical parameters that can be measured experimentally,
and they are assumed to be continuous with the corresponding parameters in the
adjacent free flow.

The flow velocity is defined as a superficial volume average, and it corresponds to a

unit volume of the medium including both the pores and the matrix. It is sometimes
called the Darcy velocity, defined as the volume flow rate per unit cross section of the
medium. Such a definition makes the velocity field continuous across the boundaries
between porous regions and regions of free flow.

Brinkman Equations Theory

The dependent variables in the Brinkman equations are the Darcy velocity and the
pressure. The flow in porous media is governed by a combination of the continuity
equation and the momentum equation, which together form the Brinkman equations:

648 | CHAPTER 6: FLUID FLOW INTERFACES

      +    u  = Q (6-72)
t p m

-  u
+  u    ----- =
u
----
p   t  p
(6-73)
1 T 2   Q m
–  p +   -----    u +  u   – ---     u I  –   – 1  + -------- u+F
p  3    p2 

In these equations:

•  (SI unit: kg/(m·s)) is the dynamic viscosity of the fluid

• u (SI unit: m/s) is the velocity vector
•  (SI unit: kg/m3) is the density of the fluid
• p (SI unit: Pa) is the pressure
• p is the porosity
•  (SI unit: m2) is the permeability of the porous medium, and
• Qm (SI unit: kg/(m3·s)) is a mass source or sink

Influence of gravity and other volume forces can be accounted for via the force term
F (SI unit: kg/(m2·s2)).

When the Neglect inertial term (Stokes flow) checkbox is selected, the term (u · u
p) on the left-hand side of Equation 6-73 is disabled.

The mass source, Qm, accounts for mass deposit and mass creation within the domains.
The mass exchange is assumed to occur at zero velocity.

The Forchheimer and Ergun drag options add a viscous force proportional to the
square of the fluid velocity, F  uu, on the right-hand side of Equation 6-73 .

In case of a flow with variable density, Equation 6-72 and Equation 6-73 must be
solved together with the equation of state that relates the density to the temperature
and pressure (for instance the ideal gas law).

For incompressible flow, the density stays constant in any fluid particle, which can be
expressed as


    + u   = 0
t p

and the continuity equation (Equation 6-72) reduces to

THEORY FOR THE BRINKMAN EQUATIONS INTERFACE | 649

   u = Q m

References for the Brinkman Equations Interface

1. D. Nield and A. Bejan, Convection in Porous Media, 3rd ed., Springer, 2006.

2. M. Le Bars and M.G. Worster, “Interfacial Conditions Between a Pure Fluid and a
Porous Medium: Implications for Binary Alloy Solidification,” J. Fluid Mech.,
vol. 550, pp. 149–173, 2006.

650 | CHAPTER 6: FLUID FLOW INTERFACES

Theory for the Multiphase Flow in
Porous Media Interface
The model equations that are solved in the Multiphase Flow in Porous Media interface
are based on the mass conservation of each phase and on an extended Darcy’s law. The
mass conservation equation for each phase is given by


   s  +    i ui  = Qi (6-74)
t p i i

where p (dimensionless) is the porosity, and the vector ui should be interpreted as the
volumetric flux of phase i (SI unit m3/(m2·s) or m/s). The volumetric fluxes are
determined using the extended Darcy’s law (Ref. 2)

 ri
u i = – -------   p i –  i g  (6-75)
i

where  denotes the permeability (SI unit: m2) of the porous medium, g the
gravitational acceleration vector (SI unit: m/s2), and i the dynamic viscosity (SI unit:
kg/(m·s)), pi the pressure field (SI unit: Pa), and ri the relative permeability
(dimensionless) of phase i, respectively. The phase pressures pi are related through the
capillary pressure functions p cij :

p i – p j = p cij  s 1  s N  ij (6-76)

One phase pressure can be chosen independently so that N  1 capillary pressure

relations are needed to define the other phase pressures. In addition it is assumed that
all phases together fill the pore space completely, so that we have

 si = 1 (6-77)
i=1

Pressure-Saturation Formulation
The algebraic relations in Equation 6-76 and Equation 6-77 allow for a reduction of
the number of dependent variables. An often used way is to eliminate N  1 phase
pressures and one of the saturations. This results in a so-called pressure-saturation

THEORY FOR THE MULTIPHASE FLOW IN POROUS MEDIA INTERFACE | 651

 formulation (Ref. 3). The procedure in the Multiphase Flow in Porous Media interface is
to pick one of the phases, let us say phase i c , then express the volume fraction, or
saturation, s ic , of this phase in terms of the saturations of the other phases and in
addition to use the pressure, p ic , of this phase to define the other phase pressures:

N
 
 s i
sic = 1–
  
(6-78)
 i = 1 i  i c 

p i = p ic + p c  s 1  s N  for i  i c (6-79)
i

The equations for the saturations si, (i  ic) are given by Equation 6-74 and
Equation 6-75. To arrive at an equation for p ic , the conservation equations of all
phases are summed. This results in

N
   N 

p  i si +   
  u 
i i = Q tot (6-80)
t  
 i=1  i = 1 

where the total mass source Qtot is given by

Q tot =  Qi (6-81)
i=1

The equations for the saturations si are solved in the Phase Transport in Porous Media
interface (Equation 6-74 and Equation 6-75). The Equation 6-80 for the pressure
field p i c , needed as an input to the Phase Transport in Porous Media interface, is solved
for in the Darcy’s Law interface: the Multiphase Flow in Porous Media multiphysics
coupling interface replaces the equation

 1
    +     tot u d  = Q m where u d = – ---   p –  tot g  (6-82)
 t p tot 

which is originally implemented in the Darcy’s Law interface, with Equation 6-80 by
adding the following terms to the left-hand side of Equation 6-82:

 N 
     i u i –  tot u d (6-83)
i = 1 

652 | CHAPTER 6: FLUID FLOW INTERFACES

and by setting the right-hand side Qm to be equal to Qtot. The volume averaged
density tot is given by

 tot =  i si (6-84)
i=1

Furthermore, the Multiphase Flow in Porous Media multiphysics coupling interface

couples the user inputs for the porosity and permeability of the porous matrix in the
Darcy’s Law interface to the corresponding user input fields in the Phase and Porous
Media Transport Properties feature, and couples the pressure field computed for in the
Darcy’s Law interface to the user input field for the pressure of the phase computed
from the volume constraint.

Note that when the hydraulic conductivity of a domain is specified in the Darcy’s Law
interface (instead of the permeability), the coupled Phase and Porous Media Transport
Properties feature is supplied with a permeability that is computed using a reference
kinematic viscosity of 1.004*106 m2·s1 (kinematic viscosity of water at 293.15 K).

In the other direction, the multiphysics coupling node provides the averaged density
(denoted by tot in Equation 6-83 above) and effective viscosity to the Darcy’s Law
interface.

Mass Source and Boundary Mass Source

When adding a Mass Source node in the coupled Phase Transport interface, it is by
default assumed that the net mass source of all phases, Qtot, equals zero. This implies

Qic = –  Qi (6-85)
i = 1 i  i c

However, when mass is transferred from or to phases not accounted for in the phase
transport interface — for instance, when mass is released or absorbed by the solid part
of the porous matrix — select the Mass transfer to other phases checkbox. In this case
the mass source for the phase calculated from the volume constraint should also be
specified, and the net total mass source Qtot is supplied as a right-hand side to the
Darcy Equation 6-80 for p ic .

THEORY FOR THE MULTIPHASE FLOW IN POROUS MEDIA INTERFACE | 653

 Similarly, when a Boundary Mass Source node is added in the coupled Phase Transport
interface, the sum of the specified mass fluxes for all phases is supplied as a mass flux
boundary condition for the Darcy Equation 6-80.

Boundary Conditions
When supplying boundary conditions for the coupled Phase Transport in Porous Media
and Darcy’s Law interfaces, bear in mind that the Pressure boundary condition in the
Darcy’s Law interface affects the pressure of the phase from the volume constraint, p i c .
If the pressure of another phase needs to supplied, then take into account the capillary
pressure.

Since the coupled Darcy’s Law interface uses the conservation equation for the total
mass (Equation 6-80) to solve for p ic , the Mass Flux boundary condition in the Darcy’s
Law interface should be used to prescribe the mass flux of all phases together.

If the mass flux of the phase computed from the volume constraint needs to be
prescribed at a boundary, and the total mass flux is not known a priori, it might be
necessary to compute for the total mass flux in the form of a Lagrange multiplier by
using the weak constraint formulation for the pressure boundary condition in the
Darcy’s Law interface.

Thin Barrier (Multiphysics Coupling)

The Thin Barrier multiphysics coupling uses the Extra Dimension tool to solve the
equations for the pressure and saturations through the thin barrier structure’s
thickness. The thin structure has its domain represented by the product space between
the lumped boundary and the additional dimension for the thickness. Inside this
product space, it is convenient to split the gradient operator into a tangential part and
a normal part: tn.

The following assumptions, used in deriving the Thin Barrier equations from the
equations solved in the Multiphase Flow in Porous Media interface, justify the
simplification n:

• A thin barrier can be considered as a boundary of thickness significantly smaller than

the typical lengths of the overall model.
• The permeability of the porous material inside the thin barrier is significantly smaller
than the permeability of the porous material in the adjacent domains.

• There is no mass source or sink inside the thin barrier.

654 | CHAPTER 6: FLUID FLOW INTERFACES

In addition, these assumptions imply that the effect of gravity on the phase transport
can be neglected. This results in the following set of equations for the saturations si,
(i  ic) and the pressure p i c :


   s  +  n    i u i  = 0 for i  i c (6-86)
t p i i

 ri
u i = – -------  n p i (6-87)
i

N
   N 

p i si + n  
   i u i = 0
 (6-88)
t   
 i=1  i = 1 

Together with the algebraic relations

N
 
 s i
sic = 1–
  
(6-89)
 i = 1 i  i c 

p i = p ic + p c  s 1  s N  for i  i c (6-90)
i

Here p (dimensionless) denotes the porosity and  the permeability (SI unit: m2) of
the porous material in the thin barrier. Further more i denotes the density (SI unit:
kg/m3), i the dynamic viscosity (SI unit: kg/(m·s)), pi the pressure field (SI unit:
Pa), ui the (normal) volumetric flux (SI unit: m/s) and ri the relative permeability
(dimensionless) of phase i, respectively.

Since the interfaces between the thin barrier and the adjacent domains are boundaries
between a more permeable and a less permeable porous domain, conditions analogous
to the Porous Medium Discontinuity Boundary Condition are needed to couple the
equation Equation 6-86 for the saturations inside the thin barrier to the equations in
the adjacent domains. These condition at the extremities of the extra dimension can
be written as:

 t d d t
t t  p ec i if p ci  s i   p ec i
p ci  si  L = 0 =  for i  i c (6-91)
 p dc  s di  if
d d
p ci  s i   p ec i
t
 i

where the superscript t is used to indicate the saturation and capillary pressure inside
the thin barrier (the low permeable side of the interface), and the superscript d

THEORY FOR THE MULTIPHASE FLOW IN POROUS MEDIA INTERFACE | 655

 indicates the domain adjacent to the downside of the thin barrier (the high permeable
side), and

 t u u t
t t  p ec i if p c i  s i   p ec i
p ci  s i  L = ds =  for i  i c (6-92)
 p uc  s ui  if
u u t
p ci  s i   p ec i
 i

where the superscript u indicates the domain adjacent to the upside of the thin barrier
(again the high permeable side) and where ds is the length of the extra dimension, or
equivalently, the thickness of the thin structure.

The additional conditions are continuity of the fluxes for all phases and continuity of
the pressure of the phase ic computed from the volume constraint. Note that this last
condition assumes that this phase is present both inside the thin barrier and in the
adjacent domains. Furthermore, this boundary condition assumes that the phase ic
computed from the volume constraint is the wetting phase and it is necessary that the
settings for the van Genuchten or Brooks and Corey capillary pressure model in the
adjacent domains match this assumption.

Please find related information under

• Theory for the Phase Transport Interfaces in the CFD Module User’s
Guide and
• Using Extra Dimensions in the COMSOL Multiphysics Reference
Manual.

References
1. J. Bear, Dynamics of Fluids in Porous Media, Elsevier Scientific Publishing, 1972.

2. Z. Chen, G. Huan, and Y. Ma, Computational Methods for Multiphase Flows in

Porous Media, Philadelphia: Society for Industrial and Applied Mathematics, 2006.

3. R. Helmig, Multiphase Flow and Transport Processes in the Subsurface – A

Contribution to the Modeling of Hydrosystems, Springer–Verlag, 1997.

656 | CHAPTER 6: FLUID FLOW INTERFACES

T he o r y f o r th e Bu b b l y Fl ow In t erface
The Bubbly Flow interfaces are suitable for macroscopic modeling of mixtures of
liquids and gas bubbles. These physics interfaces solve for the averaged volume fraction
occupied by each of the two phases, rather than tracking each bubble in detail. Each
phase has its own velocity field.

In this section:

• The Bubbly Flow Equations

• Turbulence Modeling in Bubbly Flow Applications
• References for the Bubbly Flow Interfaces

The Bubbly Flow Equations

The two-fluid Euler–Euler Model is a general, macroscopic model for two-phase fluid
flow. It treats the two phases as interpenetrating media, tracking the averaged
concentration of the phases. One velocity field is associated with each phase. A
momentum balance equation and a continuity equation describe the dynamics of each
of the phases. The bubbly flow model is a simplification of the two-fluid model, relying
on the following assumptions:

• The gas density is negligible compared to the liquid density.

• The motion of the gas bubbles relative to the liquid is determined by a balance
between viscous drag and pressure forces.
• The two phases share the same pressure field.

Based on these assumptions, the momentum and continuity equations for the two
phases can be combined and a gas phase transport equation is kept in order to track
the volume fraction of the bubbles. The momentum equation is

u l
l l +  l  l u l  u l =
t
(6-93)
= – p +    l   l +  T   u l + u l – ---    u l I +  l  l g + F
T 2
3

In Equation 6-93, the variables are as follows:

• ul is the velocity vector (SI unit: m/s)

THEORY FOR THE BUBBLY FLOW INTERFACE | 657

 • p is the pressure (SI unit: Pa)
• is thephase volume fraction (SI unit: m3/m3)
•  is the density (SI unit: kg/m3)
• g is the gravity vector (SI unit: m/s2)
• F is any additional volume force (SI unit: N/m3)
• l is the dynamic viscosity of the liquid (SI unit: Pa·s), and
• T is the turbulent viscosity (SI unit: Pa·s)

The subscripts “l” and “g” denote quantities related to the liquid phase and the gas
phase, respectively.

The continuity equation is

 
  + g g  +    l l ul + g g ug  = 0 (6-94)
t l l

and the gas phase transport equation is

 g  g
--------------- +     g  g u g  = – m gl (6-95)
t

where mgl is the mass transfer rate from the gas to the liquid (SI unit: kg/(m3·s)).

For low gas volume fractions (  g  0.01 ), you can replace the momentum equations,
Equation 6-93, and the continuity equation, Equation 6-94, by

u l
l +  l u l  u l =
t (6-96)
T
= – p +      l +  T   u l + u l   +  l  l g + F

l   ul = 0 (6-97)

By default, the Laminar Bubbly Flow interface uses Equation 6-96 and 6-97. To
switch to Equation 6-93 and 6-94, click to clear the Low gas concentration checkbox
under the Physical Model section.

The physics interface solves for ul, p, and

̃ g =  g  g

the effective gas density. The gas velocity ug is the sum of the following velocities:

658 | CHAPTER 6: FLUID FLOW INTERFACES

 u g = u l + u slip + u drift (6-98)

where uslip is the relative velocity between the phases and udrift is a drift velocity (see
Turbulence Modeling in Bubbly Flow Applications). The physics interface calculates
the gas density from the ideal gas law:

 p + p ref M
 g = -----------------------------
-
RT

where M is the molecular weight of the gas (SI unit: kg/mol), R is the ideal gas
constant (8.314472 J/(mol·K)), pref a reference pressure (SI unit: Pa), and T is
temperature (SI unit: K). pref is a scalar variable, which by default is 1 atm (1
atmosphere or 101,325 Pa). The liquid volume fraction is calculated from

l = 1 – g

When there is a drift velocity, it has the form

 g
u drift = – D gc ---------- (6-99)
g

where

T
D gc = -----------
-
l T

Here, T is the turbulent viscosity, and T is the turbulent Schmidt number.

Inserting Equation 6-99 and Equation 6-98 into Equation 6-95 gives

 g  g
--------------- +     g  g  u l + u slip   =    D gc  g  g  – m gl
t

That is, the drift velocity introduces a diffusive term in the gas transport equation. This
is how the equation for the transport of the volume fraction of gas is actually
implemented in the physics interface.

MASS TRANSFER AND INTERFACIAL AREA

It is possible to account for mass transfer between the two phases by specifying an
expression for the mass transfer rate from the gas to the liquid mgl (SI unit: kg/
(m3·s)).

THEORY FOR THE BUBBLY FLOW INTERFACE | 659

 The mass transfer rate typically depends on the interfacial area between the two phases.
An example is when gas dissolves into a liquid. In order to determine the interfacial
area, it is necessary to solve for the bubble number density (that is, the number of
bubbles per volume) in addition to the phase volume fraction. The Bubbly Flow
interface assumes that the gas bubbles can expand or shrink but not completely vanish,
merge, or split. The conservation of the number density n (SI unit: 1/m3) then gives

n  n u
+    l + u slip   =    D gc n 
t

The number density and the volume fraction of gas give the interfacial area per unit
volume (SI unit: m2/m3):

13 23
a =  4n   3 g 

SLIP MODEL
The simplest possible approximation for the slip velocity uslip is to assume that the
bubbles always follow the liquid phase; that is, uslip  0. This is known as
homogeneous flow.

A better model can be obtained from the momentum equation for the gas phase.
Comparing size of different terms, it can be argued that the equation can be reduced
to a balance between the viscous drag force, fD and the pressure gradient (Ref. 3), a so
called pressure-drag balance:

 l p = f D (6-100)

Here fD can be written as

3 l
f D = – C d --- ------ u slip u slip (6-101)
4 db

where in turn db (SI unit: m) is the bubble diameter, and Cd (dimensionless) is the
viscous drag coefficient. Given Cd and db, Equation 6-100 can be used to calculate the
slip velocity. In practice, Equation 6-100 is multiplied by l to reduce the slip velocity
for large values of g.

Schwarz and Turner (Ref. 4) proposed a linearized version of Equation 6-101

appropriate for air bubbles of 1–10 mm mean diameter in water:

660 | CHAPTER 6: FLUID FLOW INTERFACES

 kg -
f D = – C W u slip C W = 5  10 4 ----------- (6-102)
m 3 ·s

The Hadamard–Rybczynski model is appropriate for small spherical bubbles with

diameter less than 2 mm, and bubble Reynolds number less than one. The model uses
the following expression for the drag coefficient (Ref. 5):

16 d b  l u slip
C d = ---------- , Re b = --------------------------
-
Re b l

For bubbles with diameter larger than 2 mm, the model suggested by Sokolichin,
Eigenberger, and Lapin (Ref. 1) is a more appropriate choice:

0,622 -
C d = ----------------------------
1
------- + 0,235
Eö

where Eö is the Eötvös number

2
g l d b
Eö = --------------
-


Here, g is the magnitude of the gravity vector and  the surface tension coefficient.

Turbulence Modeling in Bubbly Flow Applications

For most bubbly flow applications the flow field is turbulent. In that case, use a
turbulence model and solve for the averaged velocity field. The Bubbly Flow,
Turbulent Flow interfaces include the turbulence models described in Theory for the
Turbulent Flow Interfaces, but the material properties of the liquid phase are used. In
addition to the options of the single-phase flow model, in , , , , and it is also possible
to account for bubble-induced turbulence — that is, extra production of turbulence
due to relative motion between the gas bubbles and the liquid. The transport equation
for the turbulent kinetic energy, k, includes a source term Sk which accounts for
bubble-induced turbulence and is given by

S k = – C k  g p  u slip

The transport equation for the turbulent energy’s dissipation rate,  includes the
following source term:

THEORY FOR THE BUBBLY FLOW INTERFACE | 661

 
C  S k ---
k

The following source term is added to the transport equation for the specific
dissipation rate, :


  S k ----
k

Suitable values for the model parameters Ck, C, and  are not as well established as
the parameters for single-phase flow. In the literature, values within the ranges
0.01 < Ck < 1, 1  C  1.92 have been suggested (Ref. 1), and  can be defined as
  C  1.

In the v2-f turbulence model, the -equation also gets a contribution,

u slip  l w
 2 ------------------------------ S
- –  -----k-
 u slip  l w  k

where lw is the wall distance variable and its gradient gives the wall-normal direction.
Such modeling assumes that the bubble-induced turbulence kinetic energy is pumped
into the component of k aligned with uslip.

The turbulent viscosity appears in the momentum equation and when adding a drift
term to the gas velocity:

 g
u drift = – D gc ----------
g

Using a turbulence model that solves for the turbulent kinetic energy, k, together with
a gas concentration that is not assumed to be low, the stress tensor contains an extra
contribution, and the momentum equations becomes

u l
l l +  l  l u l  u l =
t
(6-103)
= – p +    l   l +  T   u l + u l – ---    u l I – ---  l  l kI +  l  l g + F
T 2 2
 3  3

662 | CHAPTER 6: FLUID FLOW INTERFACES

References for the Bubbly Flow Interfaces
1. A. Sokolichin, G. Eigenberger, and A. Lapin, “Simulations of Buoyancy Driven
Bubbly Flow: Established Simplifications and Open Questions”, AIChE Journal,
vol. 50, no. 1, pp. 24–49, 2004.

2. D. Kuzmin and S. Turek, Efficient Numerical Techniques for Flow Simulation in

Bubble Column Reactors, Institute of Applied Mathematics, University of Dortmund,
2000.

3. D. Kuzmin, S. Turek, and H. Haario, Finite Element Simulation of Turbulent

Bubbly Flows in Gas-liquid Reactors, Ergebnisberichte Angew, Math, 298, University
of Dortmund, 2005.

4. M.P. Schwarz and W.J. Turner, “Applicability of the Standard k- Turbulence Model
to Gas-stirred Baths”, Applied Mathematical Modelling, vol. 12, pp. 273–279, 1988.

5. C. Crowe, M. Sommerfeld, and Y. Tsuji, Multiphase Flows with Droplets and

Particles, CRC Press, 1998.

THEORY FOR THE BUBBLY FLOW INTERFACE | 663

 T he o r y f o r t h e Mi xt u re Mod el
Interface
The mixture model is a macroscopic two-phase flow model, in many ways similar to
the bubbly flow model. It tracks the averaged phase concentration, or volume fraction,
and solves a single momentum equation for the mixture velocity. It is suitable for
mixtures consisting of solid particles or liquid droplets immersed in a liquid.

The Mixture Model interface theory is described in this section:

• The Mixture Model Equations

• Dispersed Phase Boundary Condition Equations
• Turbulence Modeling in Mixture Models
• Slip Velocity Models
• References for the Mixture Model Interfaces

The Mixture Model Equations

In the Mixture Model interfaces the particle-fluid combination is regarded as a single
flowing continuum with macroscopic properties such as density and viscosity. The two
phases consist of one dispersed phase and one continuous phase. The mixture model
is valid if the continuous phase is a liquid, and the dispersed phase consists of solid
particles, liquid droplets, or gas bubbles. For gas bubbles in a liquid, however, the
bubbly flow model is preferable. The mixture model relies on the following
assumptions:

• The density of each phase is approximately constant.

• Both phases share the same pressure field.
• The particle relaxation time is short compared to the time-scales of the macroscopic
flow.

The mixture density is given by

 = c c + d d

664 | CHAPTER 6: FLUID FLOW INTERFACES

where

• c and d denote the volume fractions of the continuous phase and the dispersed
phase (SI unit: m3/m3), respectively
• c is the continuous phase density (SI unit: kg/m3), and
• d is the dispersed phase density (SI unit: kg/m3).

The volume flux for each phase is

jd = d ud

jc = c uc

where uc and ud (SI unit: m/s) are the continuous and the dispersed phase velocity
vectors, respectively.

In previous versions of COMSOL Multiphysics (prior to version 5.4, see

Mass-Averaged Mixture Velocity), the mixture velocity used was the mass-averaged
mixture velocity u (SI unit: m/s), defined as

c c uc + d d ud
u = -------------------------------------------
- (6-104)


The mixture velocity used here is the volume-averaged flux density, or

volume-averaged mixture velocity j (SI unit: m/s), defined as

j = jd + jc

The continuity equation for the mixture is

 t +    u  = 0 (6-105)

In the Mixture Model interfaces it is assumed that the densities of both phases, c and
d, are constant, and therefore the following alternative form of the continuity
equation for the mixture is used

  j = m dc  ---- 1-
1- – ----- (6-106)
   c d

The momentum equation for the mixture is

THEORY FOR THE MIXTURE MODEL INTERFACE | 665

 j t +   j   j +  c   j slip   j = –  p +    Gm + g + F
T
–     c  1 +  c  u slip j slip 

–  c   j   j slip +     D md  d  j + j slip m dc  ---- 1-

1- – -----
 c  d

where:

• j is the velocity vector (SI unit: m/s)

•  is the density (SI unit: kg/m3)
• p is the pressure (SI unit: Pa)
•  is the reduced density difference (SI unit: kg/kg)
• uslip is the slip velocity vector between the two phases (SI unit: m/s)
• jslip is the slip flux (SI unit: m/s)
• Gm is the sum of the viscous and turbulent stresses (SI unit: kg/(m·s2))
• Dmd is a turbulent dispersion coefficient (SI unit: m2/s)
• mdc is the mass transfer rate from the dispersed to the continuous phase (SI unit:
kg/(m3·s))
• g is the gravity vector (SI unit: m/s2), and
• F is any additional volume force (SI unit: N/m3)

The slip flux is defined for convenience as,

j slip =  d  c u slip (6-107)

Here, uslip (SI unit: m/s) denotes the relative velocity between the two phases. For
different available models for the slip velocity, see Slip Velocity Models.

The reduced density difference  is given by

d – c
 = -----------------
c

The sum of the viscous and turbulent stresses is

T 2 2
 Gm =   +  T   j + j  – ---   +  T     j I – --- kI (6-108)
3 3

666 | CHAPTER 6: FLUID FLOW INTERFACES

where (SI unit: Pa·s) is the mixture viscosity, T (SI unit: Pa·s) the turbulent viscosity,
and k (SI unit: m2/s2) is the turbulent kinetic energy, when available. If no turbulence
model is used, T and k equal zero

The transport equation for d, the dispersed phase volume fraction, is

 
   +     d  d u d  =     d D md  d  – m dc (6-109)
t d d

where mdc (SI unit: kg/(m3·s)) is the mass transfer rate from the dispersed to the
continuous phase and Dmd is a turbulent dispersion coefficient (SI unit: m2/s) (see
Turbulence Modeling in Mixture Models), accounting for extra diffusion due to
turbulent eddies. When a turbulence model is not used, Dmd is zero. Assuming
constant density for the dispersed phase and using Equation 6-106, Equation 6-109
can be rewritten as

  m dc 
 d  + j   d +    j slip  =    D md  d  – -------------
- (6-110)
t d c

The continuous phase volume fraction c is

c = 1 – d

MASS TRANSFER AND INTERFACIAL AREA

It is possible to account for mass transfer between the two phases by specifying an
expression for the mass transfer rate from the dispersed phase to the continuous mdc
(SI unit: kg/(m3·s)).

The mass transfer rate typically depends on the interfacial area between the two phases.
An example is when gas dissolves into a liquid. In order to determine the interfacial
area, it is necessary to solve for the dispersed phase number density (that is, the number
of particles per volume) in addition to the phase volume fraction. The Mixture Model
interface assumes that the particles can increase or decrease in size but not completely
vanish, merge, or split. The conservation of the number density n (SI unit: 1/m3) then
gives

n j  n  n
+   c u slip  =    D md n  – nm dc  ---- 1-
1- – -----
+
t   
c d

The number density and the volume fraction dispersed particles give the interfacial area
per unit volume (SI unit: m2/m3):

THEORY FOR THE MIXTURE MODEL INTERFACE | 667

 13 23
a =  4n   3 d 

MASS-AVERAGED MIXTURE VELOCITY

In previous versions of COMSOL Multiphysics (prior to version 5.4), the mixture
velocity used was the mass-averaged mixture velocity u (SI unit: m/s), see
Equation 6-104. The continuity equation was

m dc
  c –  d      d  1 – c d u slip – D md  d  + ---------
- + c    u  = 0 (6-111)
d

and the transport equation for d was

  m dc
 d  +     d u d  =    D md  d  – ---------
- (6-112)
t d

The dispersed phase velocity vector ud and the mass fraction of the dispersed phase cd
are given by

u d = u +  1 – c d u slip

d d
c d = -----------
-


The continuity equation Equation 6-111 is incompatible with the no-penetration

condition of u at the walls, and u is not divergence-free. The volume-averaged velocity
j is solenoidal and Equation 6-106 is compatible with the no-penetration condition at
the walls (as long as there is no mass transfer or the densities of the two phases are
equal).

Dispersed Phase Boundary Condition Equations

In addition to the boundary conditions for the mixture, specify boundary conditions
for the dispersed phase on Wall, Inlet, and Outlet nodes. For these boundary types, the
boundary condition’s Settings window contains a Dispersed Phase Boundary
Condition section.

DISPERSED PHASE CONCENTRATION

This is the default for the Inlet node. Specify the dispersed phase volume fraction:

 d =  d0

668 | CHAPTER 6: FLUID FLOW INTERFACES

Enter the Dispersed phase volume fraction (dimensionless) in the d0 field.

If Solve for interfacial area is selected under the Physical Model section, the dispersed
phase number density is also specified. The dispersed phase number density can be
specified directly,

n = n0

or, when Diameter of particles/droplets is selected, computed from the diameter of

particles and droplets dd:

 d0
n = 6 ---------3-
d d

DISPERSED PHASE OUTLET

This is the default for the Outlet node. This boundary condition is appropriate for
boundaries where the dispersed phase leaves the domain with the dispersed phase
velocity ud. No condition is imposed on the volume fractions at the boundary.

A corresponding flux condition is also applied for the number density equation (when
Solve for interfacial area is selected under the Physical Model section).

DISPERSED PHASE FLUX

Using this boundary condition, specify the dispersed phase flux through the boundary:

–n   d ud  = Nd

Enter the Dispersed phase flux (SI unit: m/s) in the N  d field.

If Solve for interfacial area is selected under the Physical Model section, the number
density flux through the boundary is also specified:

– n   nu d  = N n d

SYMMETRY
This boundary condition, which is useful on boundaries that represent a symmetry line
for the dispersed phase, sets the dispersed phase flux through the boundary to zero:

n   d ud  = 0

A corresponding flux condition is also applied for the number density equation (when
Solve for interfacial area is selected under the Physical Model section).

THEORY FOR THE MIXTURE MODEL INTERFACE | 669

 NO DISPERSED FLUX
This is the default for the Wall node. This boundary condition represents boundaries
where the dispersed phase flux through the boundary is zero:

n   d ud  = 0

In the case of moving and , the dispersed flux through the boundary is:

n   d ud  = n   d j 

This implies that the flux, relative to the boundary motion, is zero. A corresponding
flux condition is also applied for the number density equation (when Solve for
interfacial area is selected under the Physical Model section).

Turbulence Modeling in Mixture Models

Turbulence modeling is particularly relevant for dilute flows, that is, for flows with a
low dispersed phase volume fraction. For dense flows, the mixture viscosity usually
becomes high. In such cases, the flow is laminar and no turbulence modeling is
necessary.

The Mixture Model, Turbulent Flow interfaces include the turbulence models
described in Theory for the Turbulent Flow Interfaces.

The turbulence must be accounted for in the calculation of the dispersed phase volume
fraction. This is accomplished by introducing a turbulent dispersion coefficient Dmd
(SI unit: m2/s) in Equation 6-110 as

T
D md = ---------
-
 T

where T is the turbulent particle Schmidt number (dimensionless). The particle

Schmidt number is usually suggested as a value ranging from 0.35 to 0.7. In the
physics interface the default value is 0.35. The so-called drift velocity is included in
Equation 6-110 as a diffusion term

    + j   +    j  =    D   – m dc 
-------------
-
t d d slip md d
d c

Similarly, the number density equation (used to compute the interfacial area) for
turbulent flow corresponds to

670 | CHAPTER 6: FLUID FLOW INTERFACES

 n j  n  n
+   c u slip  =    D md n  – nm dc  ---- 1-
1- – -----
+
t   
c d

The inclusion of the drift velocity in Equation 6-110 modifies the flux of the dispersed
phase. The effective dispersed phase flux jd,eff and velocity ud,eff are given by

j d, eff = j d – D md  d

D md
u d, eff = u d – ------------  d
d

jd,eff and ud,eff are available in the physics interface as [Link] and [Link].

Slip Velocity Models

The Mixture Model interfaces contain three predefined models for the relative velocity
between the two phases uslip (SI unit: m/s):

• The Schiller–Naumann model

• The Haider–Levenspiel model
• The Hadamard–Rybczynski model

All three models use the following relation for the slip velocity:

C   – d  
----
3 - – j –  j   j + g + F
--- ------d- u slip u slip = – ------------------- (6-113)
4 dd c  t 

where Cd (dimensionless) is the particle drag coefficient. Essentially, interpret the

relation as a balance between viscous drag and buoyancy forces acting on the dispersed
phase.

The Schiller–Naumann model models the drag coefficient according to

 24- 0,687
 ---------  1 + 0,15Re p  Re p  1000
Cd =  Re p

 0,44 Re p  1000

where Rep is the particle Reynolds number

d d  c u slip
Re p = ---------------------------


THEORY FOR THE MIXTURE MODEL INTERFACE | 671

 Note that this definition of the particle Reynolds number uses the viscosity of the
mixture and not the viscosity of the continuous phase, as would be customary for a
single particle in a pure fluid with viscosity c. This choice incorporates the hindrance
effect of the other particles on the slip velocity, see Ref. 17.

Because the particle Reynolds number depends on the slip velocity, an implicit
equation must be solved to obtain the slip velocity. Therefore, the Mixture Model
interfaces add an additional equation for

2
u slip

when the Schiller–Naumann slip model is used. The Schiller–Naumann model is

particularly well-suited for solid particles in a liquid.

The Haider–Levenspiel model is applicable to nonspherical particles. It models the

drag coefficient according to

B  Sp  C  Sp 
C d = ----------  1 + A  S p Re p  + ----------------------------------------
24
-
Re p  1 + D  S   Re
p p

where A, B, C, and D are empirical correlations of the particle sphericity. The sphericity
is defined as the ratio of the surface area of a volume equivalent sphere to the surface
area of the considered nonspherical particle

A sphere
0  S p = ------------------  1
A particle

The correlation coefficients are given by

2.3288-6.4581S p + 2.4486S p2
A  Sp  = e
B  S p  = 0.0964+0.5565S p
4.905-13.8944S p + 18.4222S p2 – 10.2599S p3
C  Sp  = e
1.4681 + 12.2584S p – 20.7322S p2 + 15.8855S p3
D  Sp  = e

The diameter used in the particle Reynolds number is that of the volume equivalent
sphere. The equation for the squared slip velocity is also added when the Haider–
Levenspiel slip model is used.

672 | CHAPTER 6: FLUID FLOW INTERFACES

The Hadamard–Rybczynski drag law is valid for particle Reynolds numbers less than 1,
for particles, bubbles, and droplets. The drag coefficient for liquid droplets or bubbles
is

 2 c 
 1 + --3- ----- d
-
24 
C d = ----------  --------------------
Re p 
 1 + -----c- 
 d 

which yields the following explicit expression for the slip velocity

 c 
2 1 + ------ 
  –  d d d  d  
u slip ----
= – --------------------------- -------------------- – j t –  j   j + g + F
18  c   
1 + 2 --- ------ 
 3  d

For solid particles, the slip velocity is given by

2
  –  d d d 
----
- – j –  j   j + g + F
u slip = – --------------------------
18  t 

when Rep  1. For very small gas bubbles, the drag coefficient is observed to be closer
to the solid-particle value. This is believed to be caused by surface-active impurities
collecting on the bubble surface.

References for the Mixture Model Interfaces

1. M. Manninen, V. Taivassalo, and S. Kallio, On the Mixture Model for Multiphase
Flow, VTT Publications, 288, VTT Energy, Nuclear Energy, Technical Research
Center of Finland (VTT), 1996.

2. C. Crowe, M. Sommerfeld, and Y. Tsuji, Multiphase Flows with Droplets and

Particles, CRC Press, 1998.

3. M.S. Nigam, “Numerical Simulation of Buoyant Mixture Flows,” Int. Jour. of

Multi. Flow., vol. 29, pp. 983–1015, 2003.

4. M. Ungarish, Hydrodynamics of suspensions, Springer-Verlag, 1993.

THEORY FOR THE MIXTURE MODEL INTERFACE | 673

 Theory for the Euler-Euler Interface
The Euler–Euler Model interfaces are based on averaging the Navier–Stokes equations
for each present phase over a volume that is small compared to the computational
domain but large compared to the dispersed phase (particles, droplets, or bubbles).
The two phases, the continuous and the dispersed phase, are assumed to behave as two
continuous and interpenetrating fluids, and the physics interface solves one set of
momentum equations for each phase.

In this section:

• The Euler–Euler Model Equations

• Some Notes on Boundary Conditions
• Turbulent Two-Phase Flow Modeling
• References for the Euler–Euler Model Interfaces

The Euler–Euler Model Equations

MASS BALANCE
The following continuity relations hold for the continuous and dispersed phases
(Ref. 3):

    +     u  = m (6-114)
t c c c c c dc

 
   +     d  d u d  = – m dc (6-115)
t d d

Here  (dimensionless) denotes the phase volume fraction,  (SI unit: kg/m3) is the
density, u (SI unit: m/s) the velocity of each phase, and mdc is the mass transfer rate
from the dispersed to the continuous phase (SI unit: kg/(m3·s)). The subscripts c and
d denote quantities relating to the continuous and the dispersed and phase,
respectively. The following relation between the volume fractions must hold

c = 1 – d

Both phases are considered to be incompressible, in which case Equation 6-114 and
Equation 6-115 can be simplified as:

674 | CHAPTER 6: FLUID FLOW INTERFACES

  c m dc
+     c u c  = ---------
- (6-116)
t c

 d m dc
+     d u d  = – ---------
- (6-117)
t d

If Equation 6-117 and Equation 6-116 are added together, a continuity equation for
the mixture is obtained:

    d u d + u c  1 –  d   = m dc  ---- 1-
1- – ----- (6-118)
  
c d

In order to control the mass balance of the two phases, the Euler–Euler Model
interfaces solves Equation 6-117 together with Equation 6-118. Equation 6-117 is
used to compute the volume fraction of the dispersed phase, and Equation 6-118 is
used to compute the mixture pressure.

MOMENTUM BALANCE
The momentum equations for the continuous and dispersed phases, using the
nonconservative forms (see for example Ref. 3 or Ref. 9), are:


c c  u  + uc    uc  =
t c (6-119)
–  c p +  c    c +  c  c g + F m,c +  c F c + m dc  u int – u c 


d d  u  + ud    ud  =
t d (6-120)
–  d p +  d    d +  d  d g + F m,d +  d F d – m dc  u int – u d 

Here p (SI unit: Pa) is the mixture pressure, which is assumed to be equal for the two
phases. In the momentum equations the viscous stress tensor for each phase is denoted
by (SI unit: Pa), g (SI unit: m/s2) is the vector of gravitational acceleration, Fm
(SI unit: N/m3) is the interphase momentum transfer term (that is, the volume force
exerted on each phase by the other phase), F (SI unit: N/m3) is any other volume
force term, and uint is the interphase velocity (SI unit: m/s).

In these equations, the influence of surface tension in the case of liquid phases has been
neglected, and the potential size distribution of the dispersed particles or droplets is
not considered.

THEORY FOR THE EULER-EULER INTERFACE | 675

 For fluid-solid mixtures where d  c, Equation 6-120 is modified in the manner of
Enwald (Ref. 5)


d d  u  + ud    ud  =
t d (6-121)
–  d p – p s +  d  d g + F m,d +  d F d – m dc  u int – u d 

where ps (SI unit: Pa) is the solid pressure.

The fluid phases in the above equations are assumed to be Newtonian and the viscous
stress tensors are defined as

 c =  cm  u c +  u c  – ---    u c I
T 2
 3 

 d =  dm  u d +  u d  – ---    u d I
T 2
3

where  cm and  dm (SI unit: Pa·s) are the dynamic viscosity model of the respective
phase. Observe how the viscous terms in equation Equation 6-119 and
Equation 6-120 appear with their volume fractions outside the divergence operators.
The equations can also be derived so that the terms read   cc and   dd
instead (see for example Ref. 4 or Ref. 5). It depends on exactly how the derivation is
carried out. formulation in Equation 6-119 and Equation 6-120 are so that the
momentum equation can be divided by c and d, respectively as described below.

In order to avoid singular solutions when the volume fractions tend to zero, the
governing equations above are divided by the corresponding volume fraction. The
implemented momentum equation for the continuous phase is:

 F m,c m dc
 c (u c) +  c u c    u c  = – p +    c +  c g + ----------- + F c + ----------  u int – u c 
t c c

The implemented momentum equations for the dispersed phase in the case of
dispersed liquid droplets or bubbles is:


d  u  + ud    ud  =
t d
(6-122)
F m,d m dc
– p +    d +  d g + -----------
- + F d – ----------  u int – u d 
d d

676 | CHAPTER 6: FLUID FLOW INTERFACES

and in the case of dispersed solid particles:


d  u  + ud    ud  =
t d
(6-123)
p F m,d m dc
– p – ---------s +  d g + -----------
- + F d – ----------  u int – u d 
d d d

VISCOSITY MODELS
The Newtonian viscosities of interpenetrating media are not readily available. Instead
empirical and analytical models for the dynamic viscosity of the two-phase mixture
have been developed by various researchers, usually as a function of the dispersed
volume fraction. Using an expression for the mixture viscosity, the default values for
the dynamic viscosities of the two interpenetrating phases are taken to be:

 cm =  dm =  mix (6-124)

A simple mixture viscosity covering the entire range of particle concentrations is the
Krieger type model (Ref. 5):

 d  – 2.5 d,max
 mix =  c  1 – --------------
- (6-125)
  
d,max

Here d,max is the maximum packing limit, by default 0.62 for solid particles.
Equation 6-125 can be applied when c  d. An extension of Equation 6-125 can
be applied for liquid droplets/bubbles:

 d + 0.4 c
– 2.5 d,max -------------------------
-
d  d + c
 mix =  c  1 – --------------
- (6-126)
  
d,max

For liquid droplets/bubbles the default value of d,max is 1.

For fluid-fluid systems, the simple relation

 mix =  c  c +  d  d (6-127)

can sometimes be applied or the viscosity for each individual phase can be applied to
the respective momentum equation.

An alternative to mixture viscosity for dispersed solid particles, that is particles in gas
flow, is to model the particle interface with a solid pressure instead (see Solid Pressure).

THEORY FOR THE EULER-EULER INTERFACE | 677

 A small viscosity is however necessary for numerical robustness and Gidaspow (Ref. 6)
suggested to use

 dm = 0.5max   d  d,res  (6-128)

This is consistent with the requirement that  dm  0 when d  0 (Ref. 9), but in
practice it can suffice with any small enough, nonzero value.

INTERPHASE MOMENTUM TRANSFER

In all the equations, Fm denotes the interphase momentum transfer, that is the force
imposed on one phase by the other phase. Considering a particle, droplet, or bubble
in a fluid flow, it is affected by a number of forces, for example, the drag force, the
added mass force, the Basset force, and the lift force. The most important force is
usually the drag force, especially in fluids with a high concentration of dispersed solids,
and hence this is the predefined force included in the Euler–Euler model. The drag
force added to the momentum equation is defined as:

ꞏ
F drag,c = – F drag,d = u slip (6-129)

where  is a drag force coefficient and the slip velocity is defined as

u slip = u d – u c

The drag force on the dispersed phase is equal to the one on the
continuous phase but is working in the opposite direction.

The drag force coefficient, , is for solid particles often written as (Ref. 2)

3
 = ----------  d C D  c u slip (6-130)
4d d

For fluid-fluid system, Equation 6-130 is regularized to read (Ref. 12)

3
 = ----------  d  c C D  c u slip (6-131)
4d d

which is consistent with a vanishing  if c  0. All drag force models except Gidaspow
are based on Equation 6-130 for solid particles and on Equation 6-131 for bubbles
and droplets.

678 | CHAPTER 6: FLUID FLOW INTERFACES

Gidaspow
For dense particle flows with a high concentration of the dispersed phase — for
example, in fluidized bed models — the Gidaspow model (Ref. 6) for the drag
coefficient can be used. It combines the Wen and Yu (Ref. 7) fluidized state expression:

For c > 0.8

3 c  d  c C d – 2.65
 = ---------------------------- u slip  c (6-132)
4d d

with the Ergun (Ref. 8) packed bed expression:

For c < 0.8

2
c d d c
 = 150 -----------2- + 1.75 ----------- u slip (6-133)
c dd dd

In the above equations, dd (SI unit: m) is the dispersed particle diameter, and Cd is the
drag coefficient for a single dispersed particle. The drag coefficient is in general a
function of the particle Reynolds number

d d  c u slip
Re p = -------------------------
-
c

No universally valid expression for the drag coefficient exists. Using the implemented
Gidaspow model, Cd is computed using the Schiller–Naumann relation.

Schiller–Naumann
The Schiller–Naumann model describes the drag coefficient in Equation 6-130 and
Equation 6-131 for a single rigid sphere (see for example Ref. 3)

 24- 0.687
 ---------  1 + 0.15Re p  Re p  1000
Cd =  Re p (6-134)

 0.44 Re p  1000

It is valid for particle Reynolds numbers up to some critical Reynolds number,

approximately equal to 2.5·105 where drag crisis occurs.

The Schiller–Naumann model is appropriate for diluted flows since the correlation in
Equation 6-134 is based on a single particle and does not take particle-particle
interaction into account.

THEORY FOR THE EULER-EULER INTERFACE | 679

 The Schiller–Naumann model can be applied to small fluid particles since they do not
deform significantly.

Hadamard–Rybczynski
The Hadamard–Rybczynski drag law is valid for dilute flows with particle Reynolds
number less than one for particles, bubbles, and droplets and is defined as:

 2 c 
 1 + --3- ----- d
-
C d = ----------  --------------------
24
Re p 
 1 + -----c- 
 d 

for bubbles and droplets and as

24-
C d = ---------
Re p

for solid particles (the Stokes limit). For very small gas bubbles, the drag coefficient is
observed to be closer to the solid-particle value. This is believed to be caused by
surface-active impurities collecting on the bubble surface.

Ishii–Zuber
The Ishii–Zuber model (Ref. 10) can for solid particles be regarded as a generalization
of the Schiller–Naumann model. For solid particles it can be formulated as

 24 1 + 17.67f   d  6 / 7 
C D = max  -----------  1 + 0.1Re m
0.75  0.45 ----------------------------------------------  (6-135)
 Re m 18.67f   d  

where

f  d  = c  c  m  (6-136)

and Rem is a Reynolds number based on the mixture viscosity

d d  c u slip
Re m = -------------------------
- (6-137)
m

The Ishii–Zuber model is formulated with the assumption that m is described by a

Krieger type model (Equation 6-125 for solids or Equation 6-126 for droplets and
bubbles). The dependence on m means that the Ishii–Zuber model is valid also for
dense suspensions.

680 | CHAPTER 6: FLUID FLOW INTERFACES

For bubbles and droplets, Ishii and Zuber identify three regimes. At low Reynolds
numbers, bubbles and droplets behave in the same way as solid particles and the drag
coefficient can be described by

24
C D visc = -----------  1 + 0.1Re m
0.75  (6-138)
Re m

When the Reynolds number becomes higher, bubbles and droplets start to deform in
such a way that the drag coefficient increases linearly with diameter but is independent
of viscosity. Ishii and Zuber refer to this as the distorted particle regime and suggest
the following correlation for this regime:

2 1 + 17.67f   d  6 / 7
C D dist = --- Eo ---------------------------------------------- (6-139)
3 18.67f   d 

where Eo is the Eötvös number

g  c –  d d d2
Eo = ------------------------------
- (6-140)


where in turn g is the gravitational constant and  is the surface tension coefficient.
The Eötvös number relates surface tension forces to gravitational forces. The second,
“distorted” regime ends when hydrodynamic instabilities prevents the drag from
increasing further and the distorted regime is capped by the following value:

8
C D cap = ---  c2 (6-141)
3

The complete model for bubbles and droplets can hence be written:

 C D visc if C D visc  C D dist

CD =  (6-142)
 min  C
 D dist C D cap  if C D visc  C D dist

Haider–Levenspiel
The Haider–Levenspiel model is applicable to dilute flows with nonspherical solid
particles. It models the drag coefficient according to

B  Sp  C  Sp 
C d = ----------  1 + A  S p Re p  + ----------------------------------------
24
-
Re p   1 + D  S   Re
p p

THEORY FOR THE EULER-EULER INTERFACE | 681

 where A, B, C, and D are empirical correlations of the particle sphericity (see for
further details).

Tomiyama and Others

Tomiyama and others (Ref. 11) provided three correlations for diluted bubbles and
droplets. The correlation for purified fluids read

 16 48 8 Eo 
C D = max  min ---------  1 + 0.15Re p0.687  ---------  --- -----------------  (6-143)
 Re p Re p 3 Eo + 4 

Fluid pure enough for Equation 6-143 to be valid is typically only achievable in
laboratory environments. Even small amounts of surface-active impurities can collect
at the droplet/bubble interface and cause a surface tension gradient which resists
surface movement. The correlation for slightly contaminated fluids therefore prescribe
a slightly higher drag coefficient:

 24 72 8 Eo 
C D = max  min ---------  1 + 0.15Re p0.687  ---------  --- -----------------  (6-144)
 Re p Re p 3 Eo + 4 

This correlation is the default one and applies to droplets and bubbles in fluids of
contamination levels comparable to ordinary tap water.

When the contamination level is high, the impurities effectively prevent all motion
within the droplet or bubble and the following correlation is suggested:

 24 8 Eo 
C D = max  min ---------  1 + 0.15Re p0.687  --- -----------------  (6-145)
 Re p 3 Eo + 4 

SOLID PRESSURE
For fluid-solid mixtures where d  c, for example solid particles in gas flow, a model
for the solid pressure, ps in Equation 6-123, is needed. The solid pressure models the
particle interaction due to collisions and friction between the particles. The solid
pressure model implemented uses a gradient diffusion based assumption:

p s = – G   c  c

where the empirical function G can be thought of as a modulus of elasticity or powder

modulus. A common form for G is (Ref. 5)

B1 c + B2
G   c  = 10 (6-146)

682 | CHAPTER 6: FLUID FLOW INTERFACES

The available predefined models (all defined in Ref. 5) are those of Gidaspow and
Ettehadieh,

-8.76 c + 5.43
G   c  = 10 (6-147)

Ettehadieh,

-10.46 c + 6.577
G   c  = 10 (6-148)

and Gidaspow,

-10.5 c + 9.0
G   c  = 10 (6-149)

Observe that the solid pressure term theoretically replaces the stress term in the
dispersed phase momentum equation. A small viscosity is however necessary for
numerical robustness. The Gidaspow formulation or a small constant value is
appropriate. The viscosity for the continuous phase,  cm , can be taken as the single
phase value, that is  cm =  c .

NOTES ON THE IMPLEMENTATION

There are several equations with potentially singular terms and expressions such as the
term

F m,d
------------
d

in Equation 6-122. Also, d is a degree of freedom and can therefore, during

computations, obtain unphysical values in small areas. This can, in turn, lead to
unphysical values of material properties such as viscosity and density. To avoid these
problems, the implementation uses the following regularizations:

• 1k, k  c,d is replaced by1k,pos where

 k pos = max   k  k,res 

The residue volume fractions are per default set to 104 but can be changed in the
section of the interface settings.
• When appropriate, for example when evaluating drag models and mixture
viscosities, k, k  c,d are replaced by

THEORY FOR THE EULER-EULER INTERFACE | 683

  k reg = min  max   k 0  1 

Note however that the continuity equation, Equation 6-118, uses k without
regularization in order to guarantee mass conservation.

Some Notes on Boundary Conditions

An effect of dividing the momentum equations with c and d, respectively is that for
example ud obtains a value even if d  0. In many cases ud will tend to uc in lack of a
dispersed phase. So on inlets where d  0, it is appropriate to set uduc.

A fixed value for d should only be prescribed if there is no transport of d toward the
inlet. If, for example, a cloud of dispersed phase with higher d than the prescribed
value somehow reaches the inlet, d will actually be “consumed”.

When specifying a dispersed phase flux, make sure to specify an appropriate velocity to
go with the flux to avoid unphysical concentrations of d close to the boundary.

At walls, where it is possible to prescribe a wall leakage, the continuous phase normal
velocity is set to

ꞏ
mc
u c  n = – -------------------
-
 c  c pos

For a dispersed phase leakage:

ꞏ
md
u d  n = – --------------------
-
 d  d pos

Wall leakage for one phase works best if combined with slip condition for the other
phase so that the mixture is given maximum freedom to react to the leakage.

Turbulent Two-Phase Flow Modeling

When the characteristic Reynolds number of the two-phase mixture under
investigation becomes high, the flow transitions and becomes turbulent. The influence
of the turbulence on the flow characteristics (mixing, particle dispersion, pressure
drop, and so on) are usually significant and warrants the use of a turbulence model.

In order to model fluid flow turbulence, the Euler–Euler Model, Turbulent Flow
interface uses the k- turbulence model. This is realized by solving transport equations

684 | CHAPTER 6: FLUID FLOW INTERFACES

for the kinetic energy k (SI unit: m2/s2) and the dissipation rate of turbulent kinetic
energy  (SI unit: m2/s3).

The interface includes the possibility to model the turbulent flow of the two-phase
mixture, or to solve for turbulent flow of each phase separately. In the former case, one
set of k- equations are solved, while in the latter two sets of k- equations are solved
for, one for each phase.

MIXTURE TURBULENCE
Mixture turbulence assumes that the turbulence effects on both the continuous and
dispersed phase can be modeled by solving for the turbulence of the resulting
two-phase mixture. Here this is done using a two-equation k- model where the
transport of turbulence quantities is based on the mass-averaged mixture velocity

c c um + d d um
u m = ----------------------------------------------
-,


and a volume-averaged mixture viscosity

 m =  c  cm +  d  dm

This implies that the model is appropriate for flows where the relaxation time of the
dispersed particles (the time scale on which particles react to changes in the carrier fluid
velocity) is not significantly different to the time scale of the turbulence. It is also
appropriate for stratified flows, where the mixture mainly consists of one of the phases.

Setting the Two-phase turbulence interface property to Mixture, the turbulence of the
two-phase flow is modeled by solving the following k and equations:

k  T 
 ------ +  u m  k =      m + ------ k + P k –  (6-150)
t   k 

  T   2
 ----- +  u m   =      m + ------   + C 1 --- P – C 2  ----- (6-151)
t    k k k

The equations correspond to the standard two-equation k- model including

realizability constraints.

The production term is defined accordingly as

THEORY FOR THE EULER-EULER INTERFACE | 685

 P k =  T  u m :  u m +  u m  T  – ---    u m  2 – --- k  u m
2 2
3 3

and the resulting turbulent viscosity is defined as

k2
 T = C  ------


where C is a model constant. The viscous stress tensors for the phases are hence
defined as

 c =   cm +  T   u c +  u c  – ---    u c I – ---  c kI ,

T 2 2
 3  3

 d =   dm +  T   u d +  u d  – ---    u d I – ---  d kI .

T 2 2
3 3

Assuming mixture turbulence, the transport equation for the volume fraction is:


   +     d u d  =    D md  d  (6-152)
t d

Here the dispersion of the particulate phase by the turbulent fluctuations is modeled
using a gradient based hypothesis. The turbulent dispersion coefficient is defined from
the turbulent viscosity of the two-phase mixture in the manner of

T
D md = ---------
-
 T

where T is a turbulent particle Schmidt number (dimensionless).

The default values of the dimensionless parameters using the Mixture Two-phase
turbulence model are:

CONSTANT VALUE

C 0.09
C1 1.44
C2 1.92
k 1.0
 1.3
Kv 0.41

686 | CHAPTER 6: FLUID FLOW INTERFACES

CONSTANT VALUE

B 5.2
 0.35

PHASE-SPECIFIC TURBULENCE
The phase-specific Two-phase turbulence model assumes that the turbulent flow of the
continuous and dispersed phase can be modeled by solving for the turbulence of each
phase separately by using two sets of k- equations. The model implies that the time
scales of the turbulent flow of each phase can differ, but it is also computationally more
expensive than assuming solving one set of k- equations for the mixture.

Setting the Two-phase turbulence interface property to Phase specific, the turbulent flow
of the two phases is modeled by solving two sets of k and equations, one for each
phase. For the continuous phase, the transport equations for k and are

k c  T,c 
 c -------- +  c u c  k c =      cm + ---------- k  + P k c –  c  c (6-153)
t  k c c

 c  T,c c  d2
 c ------- +  c u c   c =      cm + ---------
-  c + C 1,c ----- P k c – C 2,c  ------ (6-154)
t   c kc kd

while for the dispersed phase the corresponding equations are

k d  T,d 
 d --------- +  d u d  k d =      dm + ----------
- k  + P k  d –  d  d (6-155)
t   k d d

 d  T,d d  d2
 d -------- +  d u d   d =      dm + ----------  d + C 1,d ------ P k d – C 2,d  ------ (6-156)
t     d   kd kd

The equations for each phase correspond to the standard two-equation k- model
including realizability constraints.

The production terms are defined as

P k c =  T,c  u c :  u c +  u c  T  – ---    u c  2 – ---  c k d   u c

2 2
3 3

P k d =  T,d  u d :  u d +  u d  T  – ---    u d  2 – ---  d k d   u d

2 2
 3  3

THEORY FOR THE EULER-EULER INTERFACE | 687

 The resulting turbulent viscosity, applied individually in the momentum equations of
the continuous and the dispersed phases, are

k c2 k d2
 T c =  c C  c ------ ,  T d =  d C  d ------ (6-157)
c d

where Cc and Cd are model constants. The viscous stress tensors for the phases are
hence defined as

 c =   cm +  T c   u c +  u c  – ---    u c I – ---  c k c I ,

T 2 2
3 3

 d =   dm +  T d   u d +  u d  – ---    u d I – ---  d k d I .

T 2 2
 3  3

As in the case of mixture turbulence, the dispersion of the particulate phase by the
turbulent fluctuations is modeled using a gradient based hypothesis (Equation 6-152).
For phase-specific turbulence the dispersion coefficient is modeled as a volume average
of the respective turbulent diffusivity of each phase

 T c  T d
D md = -----------   c ----------- +  d -----------
1
 T d  c d 

using a turbulent particle Schmidt number T,d (dimensionless).

The default values of the turbulence model parameters for phase specific turbulence are

CONTINUOUS VALUE DISPERSED VALUE

PHASE PHASE
CONSTANTS CONSTANTS

Cc 0.09 Cd 0.09

C1c 1.44 C1d 1.44
C2c 1.92 C2d 1.92
kc 1.0 kd 1.0
c 1.3 d 1.3
Kv,c 0.41 Kvd 0.41
Bc 5.2 Bd 5.2
d 0.35

688 | CHAPTER 6: FLUID FLOW INTERFACES

References for the Euler–Euler Model Interfaces
1. M.J.V. Goldschmidt, B.P.B. Hoomans, and J.A.M. Kuipers, “Recent Progress
Towards Hydrodynamic Modelling of Dense Gas-Particle Flows”, Recent Research
Developments in Chemical Engineering, Transworld Research Network, India, pp.
273–292, 2000.

2. M. Ishii and T. Hibiki, “Thermo-Fluid Dynamics of Two-Phase Flow”, Springer

US, 2006.

3. C. Crowe, M. Sommerfeld, and Y Tsuji, Multiphase Flows with Droplets and

Particles, CRC Press, Boca Raton, 1998.

4. B.G.M. van Wachem, J.C. Schouten, C.M. van den Bleek, R. Krishna, and J.L.
Sinclair, “Comparative Analysis of CFD Models of Dense Gas-Solid Systems”, AIChE
Journal, vol. 47, no. 5, pp. 1035–1051, 2001.

5. H. Enwald, E. Peirano, and A.-E. Almstedt, “Eulerian Two-Phase Flow Theory

Applied to Fluidization”, Int. J. Multiphase Flow, vol. 22, pp. 21–66, 1996.

6. D. Gidaspow, Multiphase Flow and Fluidization, Academic Press, San Diego,

1994.

7. C.Y. Wen and Y.H. Yu, “Mechanics of Fluidization”, Chemical Engineering

Progress Symposium Series, vol. 62, pp. 100–110, 1966.

8. S. Ergun, “Fluid Flow Through Packed Columns”, Chemical Engineering

Progress, vol. 48, pp. 89–94, 1952.

9. M. Ungarish, Hydrodynamics of Suspensions, Springer-Verlag, 1993.

10. M. Ishii and N. Zuber, “Drag Coefficient and Relative Velocity in Bubbly, Droplet
or Particulate Flows”, AIChE Journal, vol. 25, no. 5, 1979.

11. A. Tomiyama, I. Kataoka, I. Zun and T. Sakaguchi, “Drag coefficients of single

bubbles under normal and micro gravity conditions”, JSME International Journal,
series B. Vol. 41, pp. 472–479, 1998.

12. P.J. Oliveira and R.I. Issa, “Numerical aspects of an algorithm for the Eulerian
simulation of two-phase flows”, Int. J. Numer. Meth. Fluids, vol. 43, pp. 1177–1198,
2003.

THEORY FOR THE EULER-EULER INTERFACE | 689

 Theory for the Coupling of Fluid
Flow to Electrochemical Reactions
Also see Theory for the Coupling of Mass Transport to Electrochemical Reactions for
how to calculate molar and mass fluxes, sources, and sinks.

Momentum Sources and Sinks

The flow velocity, u (SI unit: m/s), at a boundary is coupled to the total mass flux of
the species at the electrode surface in the following way

1
u  n = ---
  ni  n (6-158)
i

where (SI unit: kg/m3) is the density of the flowing media, n the surface normal and
ni is the mass flux of species i.

For flow interfaces solving for the velocity field, no-slip conditions are applied at the
electrode surface, resulting in

1 
u = –  ---
  Mi Ni n (6-159)
i

where Mi (SI unit: kg/mol) is the species mass and Ni (SI unit: mol/(m2·s)) the molar
flux calculated by Faraday’s law.

For a porous electrode the sum of all mass fluxes are added to the conservation
equation according to:


------ +    u  =
t  Ri mass (6-160)
i

Note that because the mass is usually not conserved within the species transporting
phase (the right-hand side above being nonzero), the velocity field is not divergence
free.

690 | CHAPTER 6: FLUID FLOW INTERFACES

 7

Heat Transfer Interfaces

As with all other physical descriptions simulated by COMSOL Multiphysics, any

description of heat transfer can be directly coupled to any other physical process.
This is particularly relevant for systems based on chemical reactions and mass
transfer along with fluid flow. These physics interfaces are found under the Heat
Transfer branch ( ).

The basic functionality for simulating heat transfer in homogeneous media is

described in The Heat Transfer Interfaces in the COMSOL Multiphysics Reference
Manual. The Fuel Cell & Electrolyzer Module has extra functionality for
simulating heat transfer in porous media. The documentation of The Heat Transfer
in Porous Media Interface is found in the Heat Transfer Module User’s Guide,
which is supplied with the Fuel Cell & Electrolyzer Module irrespective of whether
a separate Heat Transfer Module license is purchased.

Most electrochemistry interfaces automatically define heat source variables that

may be coupled to the heat transfer interfaces either manually or by the use of a
Electrochemical Heating multiphysics node. See the Theory for Electrochemical
Heat Sources section in the Electrochemistry Interfaces chapter for more
information on how these heat source variables are computed.

691
692 | CHAPTER 7: HEAT TRANSFER INTERFACES
| 693
694 | CHAPTER 7: HEAT TRANSFER INTERFACES
 8

Thermodynamics

This chapter describes how you can use the thermodynamics functionality to
define thermo-physical and transport properties. These properties can be used
when simulating chemical reaction systems or any type of transport model
involving mass transfer, fluid flow, or heat transfer.

In this chapter:

• Using Thermodynamic Properties

• Thermodynamic Models and Theory

695
 Using Thermodynamic Properties
In this section:

• Workflow for Thermodynamics Property Calculations

• Thermodynamics
• Thermodynamic System
• External Thermodynamic Packages
• External Thermodynamic System
• Predefined System
• Exporting and Importing Thermodynamic Systems
• Species Property
• Mixture Property
• Equilibrium Calculation
• Generate Chemistry
• Generate Material
• Coupling with the Reaction Engineering and the Chemistry Interfaces
• Evaluating a Property Function in a Physics Interface
• User-Defined Species
• References

Workflow for Thermodynamics Property Calculations

The thermodynamic properties database, together with the thermodynamic calculator,
make it possible to evaluate physical properties of fluids such as heat capacity, thermal
conductivity, density, and diffusivity, as well as equilibrium compositions of two phases
at equilibrium. The properties can be calculated for pure fluids, mixtures, and for
two-phase fluid systems consisting of pure species as well as mixtures.

Using the thermodynamics functionality involves the following main steps:

696 | CHAPTER 8: THERMODYNAMICS

1 Add the Thermodynamics node and create a thermodynamic system to add properties
for a selected set of chemical compounds, referred to as chemical species in this
manual. A thermodynamic system can be created in the following ways:
- A built-in Thermodynamic System uses the built-in, or user-defined, property
database with native functionality for thermodynamic property calculations.
- An External Thermodynamic System uses properties from an installed external
package provider. For more information on how to enable the use of external
packages see the Installing External Thermodynamic Packages — An Example.
- A Predefined System uses the built-in database with native functionality to set up
one of the following common systems: Dry air, Moist air, or Water-steam.
2 Once a thermodynamic system is set up, it can be used to define a multitude of
property functions. The Generate Material option provides a fast route to defining
the material properties most commonly required for fluid flow, heat transfer and
mass transfer modeling. By specifying the composition and phase of the mixture, a
Material node can automatically be set up. The material node can include properties
such as density, heat capacity, heat capacity ratio, thermal conductivity and viscosity.
3 Users of the Reaction Engineering, or Chemistry interface can couple chemical
species in these interfaces with the chemical species in the thermodynamic system.
When Coupling with the Reaction Engineering and the Chemistry Interfaces the
required property parameters and functions are automatically added and visualized
as nodes under the corresponding thermodynamic system node.
4 As an alternative to the functions automatically set up when using Generate Material
or coupling to, for example, Chemistry, functions or constants can be set up
manually for any thermodynamic system. This is accomplished using the
thermodynamic features Species Property, Mixture Property or Equilibrium
Calculation. You can, for example, create a Species Property function describing the
density of a fluid, and use this function in a fluid-flow interface. With Mixture
Property you can define mixture functions, that is, functions that depend on the
composition of a mixture, describing for example density, enthalpy, or heat capacity.

The functionalities described in Step 2– 4 generate functions and constants, defined by

a thermodynamic system, that can be used in any physics interface in COMSOL
Multiphysics. The functions are used to evaluate fluid properties that depend on
variables such as temperature, pressure, and mixture composition. For more
information, see Evaluating a Property Function in a Physics Interface.

USING THERMODYNAMIC PROPERTIES | 697

 Thermodynamics
To access the functionality for thermodynamic calculations, right-click the Global
Definitions node in the Model Builder tree and select Thermodynamics ( ). When the
model includes a Component, Thermodynamics is also available in the Physics toolbar.

Using the Reaction Engineering or Chemistry interfaces, Thermodynamics is also available

on the Reaction Engineering or Chemistry toolbar.

Figure 8-1: The Thermodynamics is always available under Global Definitions.

Thermodynamic properties given as parameters or functions are collected in a so called

thermodynamic system. Each thermodynamic system includes a set of available species
and phases, as well as different models used for the thermodynamic property
calculations, for example an equation of state such as Peng–Robinson.

Select Thermodynamic System in the context menu to add a thermodynamic system

that uses the built-in database included with COMSOL Multiphysics.

As an alternative, selecting External Thermodynamic System can make use of an

installed thermodynamics software to make the corresponding calculations.

The option Predefined System, allows you to set up one of the following systems: Dry
air, Moist air, or Water-steam.

698 | CHAPTER 8: THERMODYNAMICS

It is also possible to import previously created and stored thermodynamic systems by
selecting Import Thermodynamic System (see Exporting and Importing
Thermodynamic Systems).

In addition to creating thermodynamic systems, you can also define User-Defined

Species in Thermodynamics. Use User-Defined Species to add new species that are not
available in the COMSOL database. You can also edit available species in the database.

Figure 8-2: Adding Thermodynamic Systems to the Thermodynamics node. Note that
external package providers have been installed.

When a thermodynamic system has been added, thermodynamic and transport

properties are defined by functions. These can be used to describe properties for pure
species, mixtures, and for gas–liquid (or liquid–liquid) systems at equilibrium.

Thermodynamic System
A thermodynamic system is used to describe properties of pure species and mixtures of
chemical compounds for liquids, gases, liquid–vapor equilibria, and liquid–liquid
equilibria. It specifies the available species and the phases (states of aggregation) that
are present in the modeled system. It also defines and evaluates the functions for
thermodynamic and transport properties of the chemical system, that is, the species
and mixture properties for liquids, gases, and phase equilibria.

USING THERMODYNAMIC PROPERTIES | 699

 To create a Thermodynamic System, start the Thermodynamic System Wizard in one of
the following ways:

• Right-click the Global Definitions node, select Thermodynamics and choose

Thermodynamic System.
• Using the Reaction Engineering interface, choose Thermodynamic System from the
Thermodynamics section of the Reaction Engineering toolbar.

• Correspondingly, using the Chemistry interface, choose Thermodynamic System

from the Thermodynamics section of the Chemistry toolbar.

The Thermodynamic System Wizard contains the following steps:

1 Select System
2 Select Species
3 Select Thermodynamic Model

SELECT SYSTEM
Use the Select System step in the wizard to define the phases in the modeled system.
You can select Gas, Liquid, Vapor-liquid, Vapor-liquid-liquid, or Liquid-liquid. The names
of the phases in the Selected system table can be changed by editing the element in the
Name column. Click the Next button ( ) to proceed to the next step in the wizard.

Figure 8-3: Thermodynamic System Wizard: Select System.

SELECT SPECIES
Use the Species Type filter text field to search among the available species in the COMSOL
database or the User-Defined Species. Species can be searched for by typing the name,
CAS number, or the chemical formula. Using a formula, all species whose chemical
formula includes a combination of the given atoms are listed. For example, when
searching for ammonia NH3 or H3N can be used.

Select one or more species in the list and click the Add Selected button ( ) to add
species to the Selected species table. You can also press the Enter key to add either the

700 | CHAPTER 8: THERMODYNAMICS

first species in the list, or the species that have been selected. To remove species from
the Selected species table, use either the Remove Selected button ( ) or the Remove
All button ( ), both located below the table. Click the Next button ( ) to proceed
to the next step in the wizard.

Figure 8-4: Thermodynamic System Wizard: Select Species.

SELECT THERMODYNAMIC MODEL

The Thermodynamic model controls the relations between the intensive
thermodynamic properties of the system such as pressure, specific volume, and
temperature. It also controls how extensive properties like enthalpy, entropy, or the
heat capacity of the system are computed. Furthermore, the Thermodynamic model
specifies the default models used to compute transport properties.

The following Thermodynamic Models are available for all phases:

• Peng–Robinson
• Peng–Robinson (Twu)
• Soave–Redlich–Kwong
• Soave–Redlich–Kwong (Graboski–Daubert)

For a system including only gas phase, or for liquid phase systems with thermodynamic
models requiring an equation of state, the Ideal Gas Law is added to the available
models in the Gas phase model list or Equation of state list.

USING THERMODYNAMIC PROPERTIES | 701

 When the system includes one or more liquid phases, the following additional models
are also available (in the Liquid phase model list):

• Chao–Seader (Grayson–Streed)
• Wilson
• NRTL
• UNIFAC VLE
• UNIQUAC
• Regular Solution
• Extended Regular Solution
• Ideal Solution

These models correspond to activity models.

To calculate properties of water and steam, use the Water (IAPWS) model. This is
available for systems that include a single water species.

For recommendations regarding the appropriate thermodynamic model, based on the

available phases and the system pressure level, see the section Selecting the Right
Thermodynamic Model.

Select the Advanced options checkbox in order to manually control the models used for
thermodynamic properties, transport properties, and surface tension. When this
checkbox is selected, all available property models are shown in the property model
table. The available property models are dependent on the phases available in the
thermodynamic system.

When only a gas phase is present, models are available for the following properties:

• Gas diffusivity
• Gas thermal conductivity
• Gas viscosity

For a single-phase liquid, or for a two-phase liquid–liquid system, there are models for
the following properties:

• Liquid diffusivity
• Liquid diffusivity at infinite dilution
• Liquid thermal conductivity

702 | CHAPTER 8: THERMODYNAMICS

• Liquid viscosity
• Liquid volume

For systems with both vapor and liquid phases, all of the above properties are available,
as well as gas–liquid surface tension.

Figure 8-5: Thermodynamic System Wizard: Advanced thermodynamic model options

available for a Vapor–liquid system.

Click the Finish button ( ) to exit the Thermodynamic System Wizard and add the
corresponding system under the Thermodynamics node. Note that the default node
label reflects the available phase in the system. For example, when creating a vapor–
liquid system, a node labeled Vapor-Liquid System is added. The name of the system can
be changed using the Label text field in the Settings window for the system.

SETTINGS
Selecting a Thermodynamic System or Predefined System node shows its Settings
window with a description of the package content and the available properties. Editing

USING THERMODYNAMIC PROPERTIES | 703

 the text in the Label text field changes the name of the system. Below follows a
description of the thermodynamic system settings.

Figure 8-6: Thermodynamic System Settings.

Species
Lists the species included in the thermodynamic system. You can change the list of
species by right-clicking the Thermodynamic System node and selecting Add or Remove
Species.

Phases
Lists the phases included in the thermodynamic system. You can add or remove phases,
such as gas, liquid, or liquid–vapor, by right-clicking the Thermodynamic System node
and selecting Define System.

Species Property
Lists the parameters and functions that describe thermodynamics and transport
properties for pure species. Such functions may describe density, heat capacity, thermal
conductivity, or other thermodynamic and transport properties. For more
information, see Species Property.

704 | CHAPTER 8: THERMODYNAMICS

Mixture Property
Lists the available mixture property functions. Note that you have to have a mixture
model defined in order to use these functions. You can define a mixture property by
right-clicking the Thermodynamic System node and selecting Mixture Property.

Thermodynamic Model
When a liquid is present, use the Liquid phase model list to select the thermodynamic
model for this phase.

When a gas is present, use the Gas phase model list to select the thermodynamic model
for this phase. If both gas and liquid phases are available, and the Liquid phase model is
set to an equation of state, then the Gas phase model is set to the same model
automatically.

Property Models
Displays the settings for the available individual property models. The property models
can be changed by selecting different values from the combo box in the Model
column. For example, for Gas thermal conductivity you can select Kinetic theory or Ideal
from the corresponding combo box. You can also select the property model in the last
step of the thermodynamic system wizard by selecting the Advanced option checkbox,
see Select Thermodynamic Model.

Binary Interaction Parameters

This section can be used to inspect and edit binary interaction parameters. Select a pair
of species (first and second column, Species 1 and Species 2) as shown in Figure 8-7
to edit the corresponding binary interaction parameter currently in use. In cases that
the thermodynamic model has more than one type of binary interaction parameter,
each type of the interaction parameter will be shown in a column in the table.

Figure 8-7: Entering the Binary Interaction Parameters for the Soave–Redlich–Kwong
model.

When a binary interaction parameter is not available in the database for a pair of
species, the parameter value is set to zero (default value).

USING THERMODYNAMIC PROPERTIES | 705

 MODIFYING AN EXISTING SYSTEM
It is possible to modify an existing thermodynamic system, for example by changing
the included species or phases. To do so, right-click the Thermodynamic System node
and select Add or Remove Species, or Define System from the context menu. Both of
these options are available for systems created using Thermodynamic System. For a
Predefined System, the option Add or Remove Species is available for Dry Air and Moist
Air. For a Water-steam system, neither of the two options are available.

When the properties of a User-Defined Species have been changed, select Update
From Species for the changes to take effect in the thermodynamic system.

Figure 8-8: Modifying the species and phases of a thermodynamic system.

Add or Remove Species

Selecting Add or Remove Species takes you to the Select Species step in the
Thermodynamic System Wizard. Use the Add Species button ( ) to add a species to the
Selected Species list. To remove a species, select it in the list and use the Remove Species
button ( ).

Make sure to click the Finish button ( ) in order for the changes to take effect.

Note that adding or removing species in a thermodynamic system that is currently

coupled to a Reaction Engineering interface or a Chemistry interface breaks this
coupling. See the Coupling with the Reaction Engineering and the Chemistry
Interfaces for how to update this coupling.

706 | CHAPTER 8: THERMODYNAMICS

Define System
Selecting Define System takes you to the Select System step in the Thermodynamic
System Wizard. You can select the desired system from the list.

Make sure to click the Finish button ( ) button in order for the changes to take
effect.

Note that changing the phase(s) in a thermodynamic system that is currently coupled
to a Reaction Engineering interface or a Chemistry interface breaks this coupling. See
the Coupling with the Reaction Engineering and the Chemistry Interfaces section for
how this coupling can be updated.

Update From Species

This option is available in a Thermodynamic System that includes a User-Defined
Species. Use it to update the system following a change in the species properties.

WARNING INFORMATION
A sanity check is always performed when a Thermodynamic System is created. If any
problems are found, a Warning Information node listing the problems is added under
the package. One example when this occurs is when a parameter required for the
thermodynamic model, typically a binary interaction parameter, is not available in the
database.

External Thermodynamic Packages

The thermodynamics functionality includes support for creating properties from
external thermodynamic packages. In order to use external packages, the external
software has to be CAPE-OPEN compliant. Thermodynamics in COMSOL
Multiphysics supports the CAPE-OPEN Thermo 1.1 and the CAPE-OPEN Thermo
1.0 standards.

CAPE-OPEN is an open interface standard for chemical process simulation software

(Ref. 1). This standard defines rules and interfaces that allows linking between
compliant software. Some examples of CAPE-OPEN compliant software are:

• COCO/TEA (Free of charge package provider; see Ref. 2).

• Multiflash™ (KBC Advanced Technologies plc).
• VMGThermo™ (Virtual Materials Group, Inc.).
• Simulis® Thermodynamics (ProsSim SA).

USING THERMODYNAMIC PROPERTIES | 707

 • Aspen Properties, aspenONE® (Aspen Technology, Inc.).
• COMThermo® packages (comes with Aspen HYSYS®).

Accessing external thermodynamic systems allows species properties, mixture

properties, as well as phases equilibrium calculations (gas–vapor, liquid–liquid) to be
obtained from external software, such as the ones listed above.

INSTALLING EXTERNAL THERMODYNAMIC PACKAGES — AN EXAMPLE

TEA is a thermodynamics thermodynamic system that handles the physical and
thermodynamic property calculations for the simulation environment COCO. The
property data bank contains more than 190 commonly used chemicals, and the
package exhibits more than 100 property calculation routines with their analytical or
numerical derivatives. COCO is maintained by AmsterCHEM and is free to download
from the Internet (Ref. 2).

To use the external physical and thermodynamic property calculations from

COCO/TEA in COMSOL Multiphysics models, you need to go through the
following steps:

1 Download and install COCO, which includes the TEA thermodynamic system
manager. The software is available from
[Link]/index_download.html.
2 Create and configure a thermodynamic system template that handles physical and
thermodynamics calculations needed for your model. If you have already created a
package template earlier, or if an adequate thermodynamic system already exists in
the installation, this step is not needed.
3 Create an External Thermodynamic System node as detailed in the next section.
The installed packages are available in the Select Property Package step of the
Thermodynamic System Wizard.

External Thermodynamic System

The following assumes that you have installed external thermodynamic packages from
a CAPE-OPEN compliant software in the manner outlined in Installing External
Thermodynamic Packages — An Example.

Locate the Thermodynamics node and right-click to add an External Thermodynamic

System (Figure 8-2). Using a Reaction Engineering or Chemistry interface, you can also
add an external package by choosing External Thermodynamic System from the
Thermodynamics section of the Reaction Engineering or Chemistry toolbar.

708 | CHAPTER 8: THERMODYNAMICS

SELECT PROPERTY PACKAGE
All installed packages, from providers such as COCO/TEA, Aspen Properties, and so
on, are available in the Thermodynamic System Wizard (Figure 8-9). Select one of the
packages and click the Finish button ( ) in the upper right corner of the wizard in
order to add the package as an External Thermodynamic System node under the
Thermodynamics node.

When adding the package its default label reflects the included phases. For example, a
node labeled Vapor-Liquid System (External) is created when adding an external
package containing a vapor and a liquid phase. The name of the system can be changed
using the Label text field in the Settings window for the system.

Figure 8-9: Available external thermodynamic systems. Use the Thermodynamic System
Wizard to browse the contents of the installed external thermodynamic systems on your
system. The example shows the packages shipped with the COCO provider.

SETTINGS
Selecting the External Thermodynamic System node shows the thermodynamic system
settings with sections for available species, phases, species properties, and mixture

USING THERMODYNAMIC PROPERTIES | 709

 properties. Editing the text in the Label text field changes the name of the system.
Below follows a description of the thermodynamic system settings.

Provider
Contains information about the external thermodynamic system provider such as
version, CAPE-OPEN compliance, and a short description of the external software
that provides the thermodynamic system.

Species
Lists the species included in the selected thermodynamic system.

Phases
Lists the phases included in the selected thermodynamic system for example gas,
liquid, gas–vapor, or liquid–liquid.

Species Properties
Lists the parameter values or the functions that describe species properties, for example
molar mass, and properties available for pure compounds, such as density as a function
of temperature.

710 | CHAPTER 8: THERMODYNAMICS

Mixture Properties
Lists the available functions that describe mixture properties. One example is the
density of a nonideal mixture as a function of composition.

Figure 8-10: Settings window for an External Thermodynamic System.

The thermodynamic functionality supports package persistence. This

means that an MPH-file for a model that uses a thermodynamic system
stores that package definition in the file. You can use the mph-file to
access a thermodynamic system defined in the corresponding model
without having to store the thermodynamic system as a separate file.

USING THERMODYNAMIC PROPERTIES | 711

 Predefined System
A Predefined System uses the built-in database with native functionality to set up one
of the following common systems; dry air, moist air, or water and steam.

To create a Predefined System, start the Thermodynamic System Wizard in one of the
following ways:

• Right-click the Global Definitions node, select Thermodynamics and choose Predefined
System.
• Using the Reaction Engineering interface, choose Predefined System from the
Thermodynamics section of the Reaction Engineering toolbar.

• Correspondingly, using the Chemistry interface, choose Predefined System from the
Thermodynamics section of the Chemistry toolbar.

The Thermodynamic System Wizard that opens is identical to the wizard used to create
a Thermodynamic System, but with the important difference that all settings have been
entered for the chosen predefined system. The wizard contains the following steps:

1 Select System
2 Select Species
3 Select Thermodynamic Model

SELECT SYSTEM
In the Select System window, choose one of the predefined systems in the list. Choose
among Dry air, Moist air, and Water-steam. Click the Next button ( ) to proceed to
the next step in the wizard.

Figure 8-11: Thermodynamic System Wizard: Select Predefined System.

SELECT SPECIES
If Dry or Moist air was chosen in the Select System step, the Selected species table is
populated with the species relevant for the chosen system. For Dry air the species are:
nitrogen, oxygen, argon, carbon dioxide, neon, and helium. For Moist air the list also

712 | CHAPTER 8: THERMODYNAMICS

contains water. If instead Water-steam was chosen in the Select System step, the Next
button in that step guides you directly to the Select Thermodynamic Model pane. No
additions of species can be done in the wizard.

To add more species to the Dry air and Moist air systems, use the Species Type filter
text field to search among the available species in the COMSOL database or the
User-Defined Species. Species can be searched for by typing the name, CAS number,
or the chemical formula. Select one or more species in the list and click the Add Selected
button ( ) to add species to the Selected species table. To remove species from the
Selected species table, use either the Remove Selected button ( ) or the Remove All
button ( ), both located below the table. Click the Next button ( ) to proceed to
the next step in the wizard.

Figure 8-12: Thermodynamic System Wizard: Select Species.

SELECT THERMODYNAMIC MODEL

This last step of the Thermodynamic System Wizard allows you to choose what
thermodynamic model to use. For the system Dry air and Moist air, the Ideal Gas law
is chosen by default. For Water-steam, the model Water (IAPWS) is used. For more
information about this step in the Thermodynamic System Wizard, see Select
Thermodynamic Model for Thermodynamic System.

USING THERMODYNAMIC PROPERTIES | 713

 Click the Finish button ( ) to exit the Thermodynamic System Wizard and add the
corresponding Predefined System under the Thermodynamics node. Note that the
default node label reflects the chosen predefined system.

SETTINGS
Selecting a Predefined System node shows its Settings window with a description of the
package content and the available properties. For information about the content in the
Settings window, see Settings for Thermodynamic System and Predefined System.

MODIFYING A PREDEFINED SYSTEM

Please see Modifying an Existing System for information about how to modify an
existing thermodynamic system.

Exporting and Importing Thermodynamic Systems

EXPORT THERMODYNAMIC SYSTEM

Right-click the thermodynamic system node for the thermodynamic system that you
want to export (Figure 8-8) and select Export Thermodynamic System. Use the Select
File window to store the thermodynamic system.

Note that only the thermodynamic system definition is exported, for example the
underlying species data (from the database), the included phases, and the applied
thermodynamic models. Property values and functions created using the
thermodynamic system are not stored.

IMPORT THERMODYNAMIC SYSTEM

Right-click the Thermodynamics node and select Import Thermodynamic System
(Figure 8-2). Use the Select File window to navigate to a thermodynamic system file
and click Open.

It is possible to manually modify the species properties, such as parameters and

temperature-dependent functions, in the created Thermodynamic System by using
Export Thermodynamic System and Import Thermodynamic System. In the saved xml file
from Export thermodynamic system, there are all parameters and
temperature-dependent properties for species, which are saved inside a <Compound
CompID="name"></Compound> block. For example, to edit the acentric factor of
methane from 0.01141 to 0.2 perform the following steps

• Open the saved xml file for the thermodynamic system

714 | CHAPTER 8: THERMODYNAMICS

• Find the compound methane which starts with the line:
<Compound CompID="methane">
• Find the line with the acentric factor and change it to
<AcentricFactor >0.2</AcentricFactor>.

To use the system with the updated species, right-click the Thermodynamics node and
select Import System.

For temperature-dependent functions, for example for the (saturated) liquid density,
the database input is of the form:

<Density>
<Phase>Liquid</Phase>
<Coefficients>Tlb;a0;a1;a2;a3;Tub</Coefficients>
<Data>Tlb;f(Tlb);Tub;f(Tub)</Data>
<Comment></Comment>
</Density>

Here Tlb and Tub defines the lower and upper bound for a temperature range, and a1
to a4 are the corresponding coefficients for a cubic polynomial as f(T) = a0 + a1T+
a2T2+ a3T3 fitted for that range. Multiple sequential ranges can be added by
appending temperature ranges and coefficient sets. The <Data></Data> block can be
used to input data points directly.

The unit for temperature is K. The unit used for the temperature dependent properties
(fitted polynomials functions) are defined in the table below:
TABLE 8-1: UNIT FOR TEMPERATURE-DEPENDENT PROPERTIES.

PROPERTY UNIT

Liquid density mol/m3

Ideal heat capacity J/mol/K
Ln vapor pressure 1
Heat of vaporization J/mol
Vapor viscosity Pa·s
Ln liquid viscosity 1
Vapor thermal conductivity W/m/K
Liquid thermal conductivity W/m/K
Surface tension N/m

USING THERMODYNAMIC PROPERTIES | 715

 Species Property
A Species Property is used to define and compute a pure species property. The available
properties consist of both parameters and functions. Some examples of available
parameters are molar mass, Lennard Jones diameter, and dipole moment. Some
examples of available functions are density, enthalpy, heat capacity, and viscosity. The
property functions created are either dependent on temperature alone, or both on
temperature and pressure. For all property functions, the first order derivative with
respect to temperature and, when applicable, with respect to pressure are automatically
defined. The second-order derivatives with respect to temperature and pressure are
available for property functions of density and molar volume.

Right-click the relevant Thermodynamic System node (see Figure 8-8), Predefined
System node, or the relevant External Thermodynamic System node, and select Species
Property ( ) to start the Species Property Wizard.

The Species Property Wizard contains the following steps:

1 Select Properties
2 Select Phase
3 Select Species
4 Species Property Overview

SELECT PROPERTIES
First use the Amount base unit list to define the base unit. Select mol or kg.

Use the filter to search among the available properties. Select one or more properties
in the list and click the Add Selected button ( ) to add them to the Selected properties
list.

716 | CHAPTER 8: THERMODYNAMICS

Click the Next button ( ) to proceed to the next step, selecting the phase.

Figure 8-13: Species Property Wizard; Select Properties pane.

SELECT PHASE
Use the list to specify the phase for the species property. The available phases
correspond to the ones defined for the thermodynamic system.

Figure 8-14: Species Property Wizard; Select Phase pane.

Note that for parameters (constants) this step is skipped.

Click the Next button ( ) to proceed to the next step, selecting the species.

SELECT SPECIES
Select one or more of the species available in the thermodynamic system and use the
Add Selected button ( ) to add them to the Selected species table. One property
function is created for each of the selected species.

USING THERMODYNAMIC PROPERTIES | 717

 Click the Next button ( ) to proceed to the Species Property Overview in the wizard.

Figure 8-15: Species Property Wizard; the Select Species step.

SPECIES PROPERTY OVERVIEW

The final step of the wizard presents an overview of the parameters and functions
defined for the corresponding property. Click the Finish button ( ) to exit the wizard
and add the species property to the current thermodynamic system.

THE SPECIES NODE UNDER A SYSTEM

When creating a species properties, the resulting parameters and functions are
collected under a node with the species name. You can create new functions from an

718 | CHAPTER 8: THERMODYNAMICS

existing species node under the system. Right-click the species and select Species
Property to start the Species Property Wizard.

Figure 8-16: Create a Species Property from an existing species node.

USING THERMODYNAMIC PROPERTIES | 719

 SETTINGS
Selecting a Species Property node shows the settings for such property.

Figure 8-17: Species Property settings.

Definition:
Shows the definition of a property that is defined as a parameter or a function, for
example the name of the parameter or function.

You can use the Parameter name or Function name fields to specify or change the name
of a parameter or a function. The section also provides information about the property
type and the species it is defined for.

For a parameter, this section also displays the numerical value.

720 | CHAPTER 8: THERMODYNAMICS

For a function, this sections displays the arguments for the functions and the first order
derivatives with respect to these arguments.

Plot Parameters
Available for property functions in order to plot a selected function for a given set of
argument values.

Apply a Lower limit and an Upper limit for each argument, and click the Plot button
( ) to plot the function using the given argument range. You can also click the Create
Plot button ( ) in order to create a plot group, under the Results node.

Properties Window
To see the reference for constant or temperature dependent functions, right-click on
the function and select Properties. This opens the Properties window.

Figure 8-18: Show constant and temperature dependent properties references for species
functions.

USING THERMODYNAMIC PROPERTIES | 721

 The reference for the constant or temperature dependent property is available in the
Comments field in the Node Properties section.

Figure 8-19: Properties window for species function.

Mixture Property
A Mixture Property is used to compute a property function that depends on the
concentration of the species in a thermodynamic system. Some examples of the
available property functions are density, enthalpy, heat capacity, and surface tension.

Apart from the composition, the mixture property functions are also dependent on
temperature and pressure. The first-order derivatives with respect to temperature and
pressure are automatically defined. The second-order derivatives with respect to
temperature and pressure are available for property functions of density and molar
volume.

Right-click the relevant Thermodynamic System node (see Figure 8-8), Predefined
System node, or the relevant External Thermodynamic System node, and select Mixture
Property to start the Mixture Property Wizard.

The Mixture Property Wizard contains the following steps:

1 Select Properties
2 Select Phase

722 | CHAPTER 8: THERMODYNAMICS

3 Select Species
4 Mixture Property Overview

SELECT PROPERTIES
The settings in this step corresponds to those for Select Properties in the Species
Property Wizard. Click the Next button ( ) to proceed to the next step.

SELECT PHASE
Use this list to specify the phase, among the ones available in the system, for the
selected mixture property.

Some properties require that the system consists of two phases, for example surface
tension. A two-phase system may consist of a combination of liquid–vapor or liquid–
liquid phases.

SELECT SPECIES
First select the Species composition base unit to be used for function arguments. Select
Mole fraction or Mass fraction.

Select the species to be included in the list. Use the Add All button ( ) to add all
species in the thermodynamic system. It is also possible to select a subset of the
available species. In that case use the Add Selected button ( ) to add species. The
Selected species table is updated as you add species.

Click the Next button ( ) to proceed to the next step in the wizard.

MIXTURE PROPERTY OVERVIEW

The final step of the wizard displays an overview of the functions that the wizard has
added. Click the Finish button ( ) to exit the wizard and add the properties to the
current thermodynamic system.

SETTINGS
Selecting a desired Mixture Property node to display its settings window.

THE MIXTURE NODE UNDER A SYSTEM

When creating mixture properties, the resulting functions are collected under the
Mixture node. You can create new functions from an existing Mixture node. Right-click

USING THERMODYNAMIC PROPERTIES | 723

 the node and select Mixture Property from the context menu to start the Mixture
Property Wizard.

Figure 8-20: Mixture properties can be created by right-clicking a Mixture node under a
thermodynamic system.

Equilibrium Calculation
The Equilibrium Calculation functionality is used to compute the resulting equilibrium
conditions for a mixture of a set of species and phases.

Equilibrium calculations are often used for processes with vapor–liquid equilibrium
(VLE), so-called flash calculations. A typical process that requires flash calculations is
a distillation process, where a multiphase feed stream is separated into a vapor and a
liquid product, and where the concentrations of the species in each phase are required.
Equilibrium calculations involve combining the VLE-equations with material balances
for the chemical species, and in some cases with an energy balance. Flash calculations
are typically used to obtain estimates of:

• Bubble point at given temperature.

• Bubble point at given pressure.
• Dew point at given temperature.
• Dew point at given pressure.
• Equilibrium conditions at given pressure and temperature.
• Equilibrium conditions at given pressure and enthalpy.
• Equilibrium conditions at given pressure and entropy.
• Equilibrium conditions at given internal energy and specific volume.

The last three examples are often considered more difficult, since they require energy
balances and relations for computing enthalpy and entropy. Thermodynamics in

724 | CHAPTER 8: THERMODYNAMICS

COMSOL Multiphysics can handle all of the above cases using the equilibrium
calculation functionality. Phase envelopes, bubble point, and dew point can be
calculated for any number of species.

Right-click the relevant Thermodynamic System node (see Figure 8-8), Predefined
System node, or the relevant External Thermodynamic System node, and select
Equilibrium Calculation to start the Equilibrium Calculation Wizard.

The Equilibrium Calculation Wizard consists of the following steps:

1 Select species
2 Equilibrium Specifications
3 Equilibrium Function Overview

The Equilibrium Calculation settings include three types of equilibrium functions.

• The first function type is used to detect whether a phase is present in the system and
includes “exist” in its name.
• The amount function computes the total amount of material in each phase.
• The phase composition functions compute the mass or mole fraction of each species
in each phase, depending on the selected base unit for the equilibrium calculation.

SELECT SPECIES
Select one or more of the species available in the thermodynamic system and use the
Add Selected button ( ) to add them to the Selected species table. Click the Next
button ( ) to proceed to the next step in the wizard.

EQUILIBRIUM SPECIFICATIONS
Use the Amount base unit list to use either kg or mol as base unit.

Select two Equilibrium conditions that define the current equilibrium, for example a
given pressure and a temperature. These equilibrium conditions are used as arguments
in the equilibrium functions, in addition to the composition (overall fractions of
species).

The available equilibrium conditions are: Temperature, Pressure, Phase fraction, Energy
(or Internal energy of formation), Enthalpy (or Enthalpy of formation), Specific volume,
Density, and Entropy (or Entropy of formation). For chemical reactions, it is

USING THERMODYNAMIC PROPERTIES | 725

 recommended to use Enthalpy of formation, Entropy of formation, or Internal energy of
formation, since they account for heat of reactions.

Figure 8-21: Equilibrium Calculation Wizard; Equilibrium Specifications pane.

Selecting Phase fraction as one of the equilibrium conditions activates the Solution type
input field. This can be used to indicate the direction of the desired solution, which is
of great use especially near critical points. The options Undefined, Normal, or Retrograde
define different directions for the search of the solution to the equilibrium equations.
Using Normal means that the derivative of the vapor phase fraction with respect to
temperature (at constant pressure and composition) is kept positive and the derivative
of the vapor phase fraction with respect to pressure (at constant temperature and
composition) is kept negative. Using Retrograde means that the opposite sign of the
previous mentioned derivatives are enforced.

For a single species, the critical point is the highest pressure and temperature at which
two phases (liquid and vapor) are distinguishable. However, for some multispecies
systems, the critical point is a point between the dew point and the bubble point. In
this case, the critical point does not represent the maximum pressure or the maximum
temperature of vapor–liquid coexistence. This phenomenon is known as retrograde
condensation. This means that under isothermal conditions, when the pressure
decreases, some of the vapor condenses into liquid instead of expanding or vaporizing.
An example of such system is formation of liquid hydrocarbons in a gas reservoir as the
pressure decreases below the dew point pressure. In this case, setting Solution type to
Normal or Retrograde may not be sufficient to distinguish between the two solutions.

726 | CHAPTER 8: THERMODYNAMICS

Note that the Solution type setting is only available for a built-in thermodynamic
systems. For external thermodynamic systems, the corresponding functionality needs
to be supplied by the thermodynamic software provider. For instance, the
COCO/TEA provider does not support the Normal or Retrograde options. In those
cases, the Solution type should be Undefined.

Click the Next button ( ) to proceed to the next step.

EQUILIBRIUM FUNCTION OVERVIEW

In this step, you can review all the functions including units and arguments. Click the
Finish button ( ) to exit the wizard and add functions at equilibrium state to the
current thermodynamic system.

THE EQUILIBRIUM CALCULATION NODE

When creating an Equilibrium Calculation, the resulting functions are collected under
the Mixture node. You can create new functions from an existing Mixture node.
Right-click it and select Equilibrium Calculation to start the Equilibrium Calculation
Wizard (see Figure 8-20).

SETTINGS
Selecting an Equilibrium Calculation node displays the settings including the property
functions, see Figure 8-22.

USING THERMODYNAMIC PROPERTIES | 727

 Plot Parameters:
Can be used to evaluate and plot the equilibrium function for a range of argument
values. First specify which of the included functions to plot. Then apply a Lower limit
and Upper limit for each argument and click the Plot button ( ).

Figure 8-22: Settings window; Equilibrium Calculation.

Generate Chemistry
The Chemistry interface includes functionality to automatically define and use mixture
properties by interfacing to a Thermodynamic System node. How to manually set this

728 | CHAPTER 8: THERMODYNAMICS

coupling up is described in Coupling with the Reaction Engineering and the
Chemistry Interfaces.

The Generate Chemistry Wizard can be used to directly create a Chemistry interface
from an existing Thermodynamics System node. The resulting Chemistry interface is
automatically populated with the selected species, and is set up to be fully coupled to
the thermodynamics. This means for example that functions for mixture properties,
like the density enthalpy, are automatically defined and used in the generated
Chemistry interface.

Right-click the relevant Thermodynamic System node, Predefined System node, or

External Thermodynamic System node, and select Generate Chemistry to start the
Generate Chemistry Wizard.

The Generate Chemistry Wizard contains the following steps:

1 Select Species
2 Chemistry Settings

SELECT SPECIES
Select the species to be included in the list. Use the Add All button ( ) to add all
species in the thermodynamic system. It is also possible to select a subset of the
available species. In that case use the Add Selected button ( ) to add species. The
Selected species table is updated as you add species.

Click the Next button ( ) to proceed to the next step in the wizard.

CHEMISTRY SETTINGS
The table at the top shows the selected species and their respective chemical formula.
For each species, a corresponding Species feature will be added in the Chemistry
interface. As an option, use the Name in Chemistry columns to specify another name of
the species in Chemistry. This means that the species name could be different in
Thermodynamics and Chemistry for a species. By default the chemical formula is used.

Component
Use the Component list to select the model component to which the Chemistry
interface will be added. The already existing components are shown in the top of the
list. A new component can be created by selecting the desired space dimension.

Mass Transfer
The Mass transfer list controls the intended mixture type for the Chemistry interface.
Select Diluted species to model a mixture where the species are assumed diluted in a

USING THERMODYNAMIC PROPERTIES | 729

 high concentration solvent. For this choice, also select Solvent. Optionally, select
Concentrated species to model a mixture where all species are assumed to contribute to
the mixture properties. In the resulting Chemistry interface the Type setting in the
Mixture Properties section is defined accordingly.

Phase
Select the Phase of the mixture from the Phase list. Included in the list are the phases
available in the Thermodynamic System from which the Chemistry is generated.

Click the Finish button ( ) to create the desired Chemistry interface and exit the
wizard.

Generate Material
Once a thermodynamic system is defined, it can be used to define a multitude of
property functions. The Generate Material option provides a fast route to defining the
material properties most commonly required for fluid flow, heat transfer and mass
transfer. By specifying the composition and the phase of the mixture, a Material node
can automatically be set up. The material node can include properties such as the
density, heat capacity, heat capacity ratio, thermal conductivity, viscosity, and the
diffusivity. Two types of property functions can be used, either functions from the
underlying thermodynamics, or interpolation functions created from these. The
interpolation functions are computationally more efficient since it is cheap in terms of
CPU-time to evaluate polynomials, whereas the thermodynamics functions are more
complex and potentially solve implicit systems iteratively. Interpolation functions are
appropriate when the temperature and pressure is not expected to vary wildly, and a
sensible number of nodes is needed to reproduce the behavior, then the cost is paid
upfront during the creation of the material. The computational savings are typically
most pronounced for larger space-dependent models of higher dimensions. To validate
that the interpolation error is acceptable, create two materials, one with, and one
without interpolation functions, and compare the material properties in the relevant
region of (T, p)-space.

Right-click the relevant Thermodynamic System node (or an External Thermodynamic

System node or Predefined System node), and select Generate Material to start the
Generate Material Wizard.

The Generate Material Wizard contains the following steps:

1 Select Phase
2 Select Species

730 | CHAPTER 8: THERMODYNAMICS

3 Select Properties
4 Define Material

SELECT PHASE
Use this list to specify the phase of the material. Select among the ones available in the
system. This step of the Generate Material Wizard is not available for the Predefined
Systems Dry Air and Moist Air.

Click the Next button ( ) to proceed to the next step in the wizard.

SELECT SPECIES
Select the species to be included in the list. Use the Add All button ( ) to add all
species in the thermodynamic system. It is also possible to select a subset of the
available species. In that case use the Add Selected button ( ) to add species. The
Selected species table is updated as you add species.

Notice that when using Generate Material for the Predefined System Water-steam this
step is not included in the wizard.

Material Composition
Specify the mixture composition in terms of the Mole fractions or Mass fractions of all
species. The fractions should sum to one. This setting is only available for systems with
more than one component, and for materials with more than one selected species.

Click the Next button ( ) to proceed to the next step in the wizard.

SELECT PROPERTIES
Use this step to select the material properties to be included.

For a liquid phase material the following properties are available:

• Density, ()
• Diffusion coefficient at infinite dilution
• Heat capacity (Cp)
• Thermal conductivity (k)
• Viscosity ()

For a gas phase material, the Heat capacity ratio ( = Cp/Cv) is also available.

To add a property to the Selected properties list click the Add Selected button ( ).
To remove properties, use the Remove Selected button ( ) or the Remove All
button ( )

USING THERMODYNAMIC PROPERTIES | 731

 When adding the Diffusion coefficient at infinite dilution, also select which of the species
is the solvent. The material will include one diffusion coefficient for each of the
remaining (solute) species.

Figure 8-23: Generate Material Wizard; Select Properties pane with Diffusion coefficient
at infinite dilution added.

Additional Properties
The additional properties below are defined from the selected properties and added to
the output properties of the material.

For a liquid phase material the isobaric coefficient of thermal expansion is defined as

1 
 iso = – --- ------- (8-1)
 T

For a gas phase material, the speed of sound c, and the isentropic compressibility f are
also defined as

p 1  (8-2)
c=  ------  f = ------ ------
  p

Click the Next button ( ) to proceed to the next step.

732 | CHAPTER 8: THERMODYNAMICS

DEFINE MATERIAL

Component
Use the Component list to select where the Material will be added. The already existing
components are shown in the top of the list. A new component can be created by
selecting the desired space dimension. Select Global at the end of the list to add the
material under Global Definitions. A global material can be utilized in any component,
through the use of a Material Link. It can also be used to define phase specific
properties when linked to a subnode of a Porous Material.

Figure 8-24: Generate Material Wizard; Define Material pane.

Function Type
The material can be set up to use functions defined under the thermodynamics system,
or interpolation functions defined from these. The interpolation functions are
significantly faster to evaluate, especially for space dependent models, but requires that
the temperature range and the pressure range are specified.

Select Thermodynamics to base the material properties on functions in the current

thermodynamic system. The required functions are created, if not already present, and

USING THERMODYNAMIC PROPERTIES | 733

 added to the system. When evaluating material properties, the underlying
thermodynamics functions are called.

Select Interpolation to generate interpolation functions for the material properties. In

this case the required functions are first created and added to the system. Next, the
functions are evaluated in a number of sampling points over a given range, for both
temperature and pressure, leading to a grid of (T, p) pairs. The result of the evaluation
is stored and used to define interpolation functions for all material properties.

For the Interpolation points, select Same for all functions to use the same number of
sampling points for all functions. By selecting Individual, the number of sampling
points can be given per function. The latter can for example be used to increase the
number of points for a function that is known to vary faster than the rest of the
functions.

Select one of the preset levels for the Maximum number of interpolation points. This is
used to provide a maximum size of the underlying data for the interpolation functions.
The Number of points for temperature multiplied by the Number of points for pressure
should not exceed the Maximum number of interpolation points. If the total number of
sampling points for any function, varying temperature and pressure, is exceeded, a
warning is shown.

There are two types (or forms) of input values for pressure, Pa and log10(Pa). The later
is often used in cases where the pressure changes in a very large range, for example, for
a system from atmosphere to a vacuum system. The log10(Pa) is expressed as
log10(p/1[Pa]), here p is pressure ([Pa]).

When the selection from Pressure form changes, the corresponding input fields for
pressure will also follow the change, for both Same for all functions and Individual,
respectively.

Define the temperature range and pressure range for the temperature and pressure by
providing the Low and High end values. To neglect the variation in either temperature
or pressure, give the same value for the low and high end value. This can for example
be used when the pressure variation is known to be negligible. For each range, also
define the number of evaluation points in the Number of points field. The underlying
functions are evaluated at uniformly distributed positions across the temperature and
pressure (uniform logarithmic values for log10(Pa)) range.

When Individual has been selected for the Interpolation points, the Ranges and
interpolation points can be specified per function to be created.

Click the Finish button ( ) to create the specified Material and exit the wizard.

734 | CHAPTER 8: THERMODYNAMICS

Coupling with the Reaction Engineering and the Chemistry
Interfaces
Using the Reaction Engineering or Chemistry interfaces, all species property parameters
and property functions required by these interfaces can be created automatically by
coupling to an existing Thermodynamic System, Predefined System, or External
Thermodynamic System under Thermodynamics. Examples of species properties that can
be created automatically are molar mass, heat capacity, enthalpy, and entropy for each
species. Parameters and functions for these properties are created by the
thermodynamic system. The Reaction Engineering and Chemistry interfaces can also be
used to define transport properties for the resulting mixture (all species in the
interface). When coupled, the following mixture properties can be automatically
created: heat capacity, density, thermal conductivity, and dynamic viscosity.

The Reaction Engineering interface and the Chemistry interface are only
available in a limited set of add-on products. See
[Link] for more details on
availability.

Note that using a thermodynamic system significantly increases the modeling

capabilities in the Reaction Engineering and Chemistry interfaces. All ideal and nonideal
thermodynamic models, for gases and liquids, are directly available and also
automatically updated by editing the settings for the thermodynamic system in use.
The Chemistry interface can furthermore be used to make the mixture properties
readily available in space-dependent models for modeling of mass transport, heat
transfer, or fluid flow.

REACTION ENGINEERING

Mixture Properties
You can couple a Reaction Engineering interface with an existing thermodynamic
system in the Reaction Engineering interface’s settings window.

You need to have at least one species defined in the Reaction Engineering interface in
order to couple it to a thermodynamic system. You can make this coupling in the
Settings window for the Reaction Engineering interface by selecting the Thermodynamics
checkbox in the Mixture Properties section.

USING THERMODYNAMIC PROPERTIES | 735

 Select a thermodynamic system from the Thermodynamic system list. Use the Phase list
to select the phase to be used in the Reaction Engineering interface.

Figure 8-25: Coupling Reaction Engineering Interface with Thermodynamics.

Species Matching
The Species Matching section is activated when the Thermodynamics checkbox is selected
in the Mixture Properties section; see above. The species in the Reaction Engineering
interface can be matched to a species in the thermodynamic system. This step ensures
that the arguments in the thermodynamic system functions are correctly defined.

Use the lists in the From Thermodynamics column to match each species in the interface
to a species in the coupled thermodynamic system.

For each species matched, the required property parameters and functions are added
under to the corresponding thermodynamic system.

When all species are matched, the interface is considered fully coupled and functions
representing mixture properties, such as density, are also added automatically under
the corresponding thermodynamic system.

Figure 8-26: Matching the species in Reaction Engineering to those in the corresponding
thermodynamic system.

Calculate Transport Properties

When the interface is fully coupled to a thermodynamic system, property functions for
the mixture transport properties are created automatically when the Calculate mixture

736 | CHAPTER 8: THERMODYNAMICS

properties checkbox is selected. The properties calculated by the thermodynamic
system display Thermodynamics in the corresponding list; see below.

Figure 8-27: Select the Calculate mixture properties checkbox when coupled to a
thermodynamic system.

CHEMISTRY

Mixture Properties
You can couple a Chemistry interface with an existing thermodynamic system in the
Chemistry interface’s settings window. You can make this coupling in the Settings
window for the Chemistry interface by selecting the Thermodynamics checkbox in the
Mixture Properties section. You need to have at least one species defined in the
Chemistry interface in order to couple it to a thermodynamic system.

Select a thermodynamic system from the Thermodynamic system list. Use the Phase list
to select the phase to be used in the Chemistry interface.

Species Matching
The Species Matching section is activated when the Thermodynamics checkbox is selected
in the Mixture Properties section. Here you can match the variables for the
concentrations, and by this calculate mixture properties (transport and thermodynamic
properties). For information on how to specify the dependent variables to be used, see
Species Matching in The Chemistry Interface documentation.

You can match the species in the Chemistry interface with those in the corresponding
thermodynamic system in the Species matching section’s table. Use the lists in the
column with the title From Thermodynamics to match each species in the Chemistry
interface to a species in the corresponding thermodynamic system. This ensures that
the composition function arguments in the thermodynamic system are correctly
defined.

For each species matched, the species property parameters and functions required by
the Chemistry interface are automatically created and added under the corresponding
thermodynamic system.

USING THERMODYNAMIC PROPERTIES | 737

 When all species are matched, the interface is fully coupled to Thermodynamics and
functions representing mixture properties, such as the density, are added automatically
under the thermodynamic system that the interface is coupled to.

Figure 8-28: Matching concentration variables in a Chemistry interface, as well as species

in a coupled Thermodynamic System.

Calculate Transport Properties

When the interface is fully coupled to a thermodynamic system, property functions for
the mixture transport properties are created automatically when the Calculate mixture
properties checkbox is selected. The properties calculated by the thermodynamic
system display Thermodynamics in the corresponding list; see Figure 8-27.

MIXTURE PROPERTY DEFINITIONS WHEN COUPLING TO

THERMODYNAMICS
When all species in a physics interface (Reaction Engineering or Chemistry) are matched
to the corresponding species in a thermodynamic system, the mixture properties are
calculated based on the composition of the mixture. For example, consider Zmix(T, P,
n1, …, nm), which denotes a generic extensive mixture property for a mixture of m
number of species. The property function’s arguments are the temperature T, the
pressure P, and the number of moles, n, for each species.

When the system is fully matched, the mixture property Zmix is defined as

m
Z mix =  ni Zi (8-3)
i=1

where Z i is the partial molar of species i:

738 | CHAPTER 8: THERMODYNAMICS

 Z mix  T P n 1 n 2  n m 
Z i =  ----------------------------------------------------------------------- (8-4)
ni T P n j  i

The definition of partial molar properties can be rewritten using mole fraction
derivatives as:

Z i  T P n 1  n m  = Z i  T P x 1  x m  = (8-5)


Z  -----------------------------------------------------------
Z mix  T P x 1  x m 
 mix  T P x 1  x m  + 
-
xi  T P x
 ij

m Z mix  T P x 1  x m  
–   xi  -----------------------------------------------------------
xi
-
 T P x

 
i=1
ij


When the system is partially coupled, which means that some but not all species have
been coupled, the mixture property is instead calculated assuming ideal mixing:

m
Z mix =  ni Zi  T P  (8-6)
i=1

using the pure species properties Zi  ZmixTPxi =10…0.

Evaluating a Property Function in a Physics Interface

The functions set up using a thermodynamic system in Thermodynamics can be used
directly in any physics interfaces in COMSOL Multiphysics, in the same way as for
standard functions. Simply type in the function name with the proper arguments in the
appropriate field of the model to use the function. Figure 8-29 shows an example of
this from the Reaction Engineering interface. In this example, the Species
Thermodynamic Expressions section gets input for the molar entropy, the enthalpy and
the heat capacity of toluene. These functions, generated using Species Property for a

USING THERMODYNAMIC PROPERTIES | 739

 thermodynamic system containing toluene, have been entered in the corresponding
input fields using the appropriate arguments.

Figure 8-29: The enthalpy and heat capacity of toluene are evaluated using property
calculations through a thermodynamics system. In this case, the functions are temperature
and pressure dependent properties for a pure component.

User-Defined Species
Use this feature to specify the properties of a new species not available in the included
COMSOL database. This feature can also be used to edit a database species, by using
an existing species as a template, to create a new one. A User-Defined Species can be
included in a Thermodynamic System feature, or Predefined System feature, together
with predefined database species and other user-defined species.

The data needed for a user-defined species includes both material properties, such as
the molar mass and the vapor pressure, and properties for specific thermodynamic
models or transport models (see Thermodynamic Models and Theory). For instance,
in order to use the UNIFAC thermodynamic model, you need to define UNIFAC
groups for the new species.

Note that the thermodynamics calculator includes measures to handle missing species
properties, sometimes by applying approximations. For instance, if the Wilson volume
is not available, the liquid volume at normal boiling point is used instead. If data for
this is also missing, it is estimated from the saturated liquid density correlation.

When creating a user-defined species it is recommended to add the following common

material properties:

• Molecular mass
• Critical properties
• Acentric factor
• Absolute entropy
• Standard enthalpy of formation
• Ideal gas heat capacity

740 | CHAPTER 8: THERMODYNAMICS

• Vapor pressure
• Heat of vaporization
• Saturated liquid density
• Transport properties

EXAMPLE MODEL
For an example of using a User-Defined Species see this application example:

Dissociation in a Tubular Reactor: Application Library path

Chemical_Reaction_Engineering_Module/Thermodynamics/dissociation

THE USER-DEFINED SPECIES WIZARD

To create a User-Defined Species, start the User-Defined Species Wizard in one of the
following ways:

• Right-click the Global Definitions node, select Thermodynamics and choose

User-Defined Species.
• If the model already includes the Thermodynamics node, right-click it and choose
User-Defined Species.
• Using the Reaction Engineering or Chemistry interface, choose User-Defined Species
from the Thermodynamics section of the Reaction Engineering or Chemistry toolbar.

The User-Defined Species Wizard contains the following steps:

1 Enter Name and Formula

2 Enter Parameters
3 Define Properties

USING THERMODYNAMIC PROPERTIES | 741

 ENTER NAME AND FORMULA
Enter the Name, CAS number, and the Chemical formula for the new species.

Figure 8-30: User-Defined Species Wizard; Enter Name and Formula.

If you want to edit a species in the COMSOL database, select the Use species from
database as template checkbox. Then locate the species to edit in the list. The filter text
field can be used to search among the available species.

Figure 8-31: User-Defined Species Wizard; edit an available species by selecting

Use species from database as template.

Click the Next button ( ) to proceed to the next step.

ENTER PARAMETERS
Define the necessary parameter for the new species.

742 | CHAPTER 8: THERMODYNAMICS

Constants
Use this table to define material constants such as molar mass, critical temperature, and
standard enthalpy of formation.

Figure 8-32: User-Defined Species Wizard; enter material constants.

Structure Information
If you want to use the UNIFAC model, specify the structure formula, the UNIFAC
groups and their number occurrences.

Figure 8-33: User-Defined Species Wizard; specify the structure information in the Enter
Parameters step.

USING THERMODYNAMIC PROPERTIES | 743

 Model Parameters
Specify the parameters for the thermodynamic models and transport models in use.

Figure 8-34: User-Defined Species Wizard; specify model parameters.

Click the Next button ( ) to proceed to the next step.

DEFINE PROPERTIES
The last step in the wizard is to add temperature dependent properties for the new
species. All temperature dependent properties are defined using cubic polynomials on
the form

2 3
f  T  = a0 + a1 T + a2 T + a3 T (8-7)

Each property can consist of an arbitrary number of temperature intervals, each using
the above form. Click the Add button ( ) under the table for a specific property to
add an interval.

The following temperature dependent properties are necessary in order to compute

the basic set of mixture properties when including the User-Defined Species in a
thermodynamic system:

• Ideal gas heat capacity

• Liquid density
• Vapor pressure
• Heat of vaporization

744 | CHAPTER 8: THERMODYNAMICS

Note that some thermodynamic properties, such as the enthalpy and entropy, of a
species or mixture, are estimated from the ideal gas heat capacity and depends on the
thermodynamic model applied for the system (see Thermodynamic Properties
Definitions).

Click the Finish button ( ) in the top of the Settings window to finalize the species
and add it to the User-Defined Species node under Thermodynamics.

Figure 8-35: User-Defined Species Wizard; specify temperature-dependent properties.

SETTINGS
Selecting a species node under User-Defined Species shows the Settings window
including the definitions of all species properties. Properties are categorized into

USING THERMODYNAMIC PROPERTIES | 745

 sections as seen in Figure 8-36 below. All properties can be edited by expanding the
corresponding section.

When a species properties has been changed, any Thermodynamic System

that includes the species needs to be updated for the changes to take effect
(see Update From Species).

Figure 8-36: The Settings window for a User-Defined Species.

References
1. [Link]

2. Cape Open to Cape Open Simulation Environment,

[Link]/index_download.html

746 | CHAPTER 8: THERMODYNAMICS

Thermodynamic Models and Theory
In this section:

• Introduction
• Thermodynamic Models
• Selecting the Right Thermodynamic Model
• Species Property References
• Thermodynamic Properties Definitions
• Standard Enthalpy of Formation and Absolute Entropy Terms
• Reference State
• Transport Properties
• Surface Tension
• References

Introduction
In this chapter, we review the theory behind the thermodynamic properties database
and its functions. The thermodynamic models in the database are available for single
phase, gas or liquid, and phase equilibrium systems for two or more phases such as
vapor–liquid equilibrium (VLE), vapor-liquid–liquid equilibrium (VLLE) and liquid–
liquid equilibrium (LLE).

Thermodynamic Models
In the following sections, the available thermodynamic models are described:

• Equation of State Models

• Liquid Phase Models

EQUATION OF STATE MODELS

An equation of state (EOS) model is an algebraic relation between the absolute
pressure (P), the molar volume (V), and the absolute temperature (T). An equation of
state that is at least cubic in volume and in the form of

PV
Z  -------- = f v  V T  (8-8)
RT

THERMODYNAMIC MODELS AND THEORY | 747

 where Z is the compressibility factor and R the universal gas constant, can be used to
describe both gases and liquids. The cubic equations of state are widely used to
describe the gaseous and liquid volumetric behavior of pure fluids and also all other
properties with extensions to mixtures. A comprehensive comparison of different cubic
EOS models can be found in Ref. 2 and Ref. 3.

The equations of state models available in the thermodynamic properties database are:

• Ideal Gas Law

• Peng–Robinson
• Peng–Robinson (Twu)
• Soave–Redlich–Kwong
• Soave–Redlich–Kwong (Graboski–Daubert)
• Water (IAPWS)

Ideal Gas Law

The ideal gas law equation of state is

P = RT
-------- (8-9)
V

As the name suggests, the ideal gas law is only applicable to gases. In fact, its use is
limited to gases at low to moderate pressures.

Peng–Robinson
The classical Peng–Robinson (PR) equation of state Ref. 5 is given by

RT - – ------------------------------------
a
P = ------------ - (8-10)
V – b V 2 + 2bV – b 2

where for species i

2 2
R Tc
a i =  A -------------- (8-11)
Pc

RT c
b i =  B ----------- (8-12)
Pc

1
A = 1
--- + ---  B  4 + 10 B  (8-13)
3 3

748 | CHAPTER 8: THERMODYNAMICS

  B =   – ----------------
63 - – 0.03125 (8-14)
 1024 

1
---
27 -  3
 =  ---------------
351 - + ------------ 2 (8-15)
 32768 2048 

The alpha function, i, is given by

2
 i =  1 +  0.37464 +1.54226 i – 0.26992 i2   1 – T r i   (8-16)

For mixtures

  xi xj  1 – kPR,i,j 
a =  a  i  a  j (8-17)
i j

 xi bi
b = (8-18)
i

The binary interaction parameters (BIPs), kPR, are symmetric with zeros in the
diagonal:

k PR = k PR,i,j (8-19)

k PR,i,j = 0 (8-20)

When binary interaction parameters are missing in the database for a set of species, the
value is set to zero (a warning node is created). The values for critical temperature, Tc,
critical pressure, Pc, and acentric factor, i must be specified for all species.

Peng–Robinson (Twu)
For the Twu modification (Ref. 6) of the Peng–Robinson model, the alpha function,
i, is replaced by

 Ni  Mi – 1   NM
i = T + R exp  L i  1 – T r ii i   (8-21)

The binary interaction parameters kPR, are used for the Twu modification. Acentric
factor, i, is not used in this model but critical temperature and critical pressure must
be specified for all species. The species specific fit parameters Li, Mi, Ni can be
determined by fitting the pure species phase equilibrium to the vapor pressure curve.

THERMODYNAMIC MODELS AND THEORY | 749

 Soave–Redlich–Kwong
The classical Soave–Redlich–Kwong equation of state Ref. 7 is given by

RT - – ----------------------
a -
P = ------------ (8-22)
V – b VV + b

with

1-
 = ------- (8-23)
T

and for the pure species i,

2 2.5
R Tc
a i =  A ------------------ (8-24)
Pc

RT c
b i =  B ----------- (8-25)
Pc

1
A = 1
--- + ---  B  3 + 3 B  (8-26)
3 3

1 2
--- ---
1 3 3
B = ------ 2 27 – 1
--- (8-27)
27 3

For mixtures

  xi xj  1 – kSRK,i,j 
a = ai aj (8-28)
i j

 xi bi
b = (8-29)
i

The binary interaction parameters, kSRK, are symmetric with zero in the diagonal:

k SRK ,i,j = k SRK ,j,i (8-30)

k SRK ,i,j = 0 (8-31)

If a value is missing for kSRK,i,j in the database, it is set to zero (a warning node is
created). The values for critical temperature, Tc, critical pressure, Pc, and acentric
factor, i, must be specified for all species. The Soave–Redlich–Kwong equation of

750 | CHAPTER 8: THERMODYNAMICS

state is a version of Equation 8-22 modified by Soave Ref. 8, where for pure species i,
the alpha function is modified to

2
 i =  1 +  0.480 +1.574 i – 0.176 i2   1 – T r i   (8-32)

Soave–Redlich–Kwong (Graboski–Daubert)
The Graboski and Daubert Ref. 9 modification of the Soave–Redlich–Kwong equation
of state replaces alpha function by

2
 i =  1 +  0.48508 +1.55174 i – 0.1561 i2   1 – T r i   (8-33)

except for H2, where it is replaced by

 H 2 = 1.202 exp  – 0.30228T r H 2  (8-34)

Water (IAPWS)
The International Association for the Properties of Water and Steam (IAPWS)
provides a set of correlations to compute properties of water in different states. The
correlations available in COMSOL correspond to the version named Industrial
Formulation 1997 (IF-97) Ref. 10 - Ref. 12. The correlations are valid in the
following ranges:

273.15 K  T  1073.15 K p  100 MPa (8-35)

1073.15 K  T  2273.15 K p  50 MPa (8-36)

The following water and steam properties, available by creating a Species Property, are
computed using the IAPWS correlations: density, volume, enthalpy, internal energy,
Helmholtz energy, heat capacity at constant pressure, and heat capacity at constant
volume. The amounts present in the vapor and liquid phase can be computed using an
Equilibrium Calculation.

Other parameters and thermodynamic properties available for the system are provided
from the COMSOL database.

LIQUID PHASE MODELS

Activity Coefficient Models

Activity coefficient models are used to describe liquid mixtures at low to moderate
pressures and temperatures where the equations of state are inadequate. When using
activity coefficient models, the partial fugacity for species i in the liquid phase is defined
as

THERMODYNAMIC MODELS AND THEORY | 751

 f̂ i l = x i  i  i sat P i sat F i (8-37)

which yields

P i sat
̂ i l =  i  i sat ------------- F i (8-38)
P

where the activity coefficient, i, describes the nonideal liquid phase and ̂ i sat P i sat is
the fugacity at the vapor–liquid phase boundary at equilibrium for the pure species i.
The Poynting correction, Fi, describes the pure species fugacity deviation from the
boiling curve and can be expressed as

 P V i l 
F i = exp 
 P i sat
--------- dp
RT 
(8-39)

For an incompressible liquid, an approximation of the Poynting correction can be

done according to

V i l
F i  exp  ---------  P – P i sat  (8-40)
RT

The Poynting correction can often be ignored for moderate pressure. Hence,
Equation 8-38 can be expressed as

P i sat
̂ i l =  i  i sat ------------- (8-41)
P

If the vapor phase is considered ideal, then  i sat = 1 and the above equation reduces
to

P i sat
̂ i l =  i ------------- (8-42)
P

This reduction can be selected explicitly in case the vapor phase is not ideal.

The logarithmic transformation of Equation 8-38 is:

ln ̂ i l = ln  i + ln  i sat + ln P i sat – ln P + ln F i (8-43)

Ideal Solution
For an ideal solution the activity coefficient is equal to one, which gives:

ln  i = 0 (8-44)

752 | CHAPTER 8: THERMODYNAMICS

Regular Solution
The Scatchard–Hildebrand equation Ref. 14 for a nonpolar mixture is

2
V i   i –  av 
ln  i = -------------------------------
- (8-45)
RT

where Vi is species molar volume and i is species solubility parameter, and av is

  xi Vi i  (8-46)
 av = ----------------------------
i

  xi Vi 
i

The volume parameter, Vi, is set equal the liquid volume, Vi,l,b at normal boiling point
which must be specified for all species. The solubility parameter, i must be specified
for all species and can be estimated from the normal heat of vaporization, Hvap,iand
the liquid volume at normal boiling point as

H vap i
 i  ------------------- (8-47)
V i l b

Extended Regular Solution

The extended Scatchard–Hildebrand equation adds the Flory and Huggins correction
to the regular solution model Ref. 16:

2
V i   i –  av 
- + ln   i  + 1 –  i
ln  i = ------------------------------- (8-48)
RT

where

Vi
 i = ----------------------
- (8-49)
  xi Vi 
i

Wilson
Wilson Ref. 17 derived his activity coefficient model from a consideration of
probabilities of neighboring molecules in a liquid

THERMODYNAMIC MODELS AND THEORY | 753

  
 x   
 -----------------------
j j i 
ln  i = 1 –  
- – ln 
   xj i j (8-50)
j
 
 x k  j k

j
k

V w j  i j
 i j = ----------- exp  – -------- (8-51)
V w i  T

where the Wilson volume parameter, Vw,i, is a species-specific parameter describing

volume. If the volume parameter is not available, the liquid volume at normal boiling
point is used. The Wilson binary interaction parameters i,j are specified in terms of
absolute temperature. The binary interaction parameter matrix is nonsymmetric and
with zeros in its diagonal. All off-diagonal values must be specified

 i i = 0 (8-52)

 i i = 1 (8-53)

NRTL
Renon and Prausnitz (Ref. 18) formulated a three parameter activity coefficient model
that is able to describe liquid–liquid equilibrium; the nonrandom two-liquid (NRTL)
model:

  
 xj j i Gj i 
 x j G j i 
  x k  k j G k j 
 (8-54)
ln  i = j
------------------------------
- +   -------------------------   i j – ----------------------------------
 x G 
k


 xj Gj i j   k k  j   x k G k  j 
j  k  k 

A i j
 i j = --------- (8-55)
T

G i j = exp  –  i j  i j  (8-56)

The three parameters are Ai,j, Aj,i, and i,j. A more general form is implemented here:

A i j
 i j = --------- + B i j (8-57)
T

G i j = exp   i j –  i j  i j  (8-58)

754 | CHAPTER 8: THERMODYNAMICS

The binary interaction parameters, Ai,j, are specified in terms of absolute temperature.
The diagonal values are zero and the matrix is nonsymmetric. All off-diagonal values
must be specified.

The binary interaction parameters, Bi,j, have values of zero on the diagonal and the
matrix is nonsymmetric. For each pair of species, at least Ai,j or Bi,j should be specified.

The randomness parameters, i,j, have values of zero on the diagonal and the matrix
is symmetric. All off-diagonal values must be specified. Alternatively one can set the
more generic form directly specifying parameter i,j for which diagonal values are zero
and the matrix is nonsymmetric. For each pair of species at least i,j or i,j should be
specified.

If any value for these parameters is missing in the database, it is set to zero (warning
node is created).

A i i = 0 (8-59)

B i i = 0 (8-60)

 i i = 0 (8-61)

 j i =  i j (8-62)

 i i = 0 (8-63)

G i i = 1 (8-64)

 i i = 0 (8-65)

UNIQUAC
Abrams and Prausnitz followed up with another two-liquid model known as Universal
Quasi Chemical equation (UNIQUAC) (see Ref. 19), which is formulated in terms of
two activity coefficients:

ln  i = ln  i comb + ln  i res (8-66)

The first term is the combinatorial part contributes to the Gibbs free energy
originating from size and shape effects as

z   i 
ln  i comb = 1 –  i + ln  i – --- q i  1 + ----i + ln  ----  (8-67)
2  i   i 

THERMODYNAMIC MODELS AND THEORY | 755

 and the second term is the residual part from chemical interactions between the
molecules,

   

  j

 x j q j  j i
 x j q j  j i 

ln  i res = q i  1 – ln  -------------------------
 
- –
 --------------------------- (8-68)
x j q j  x j q j  k j



 j 
j  
k

where

ri
 i = ----------------
- (8-69)
 xj rj
j

qi
 i = ----------------- (8-70)
 xj qj
j

–  E i j
 i j = exp  ----------------- (8-71)
T

The coordination number is taken equal to z = 10. The binary interaction energy
parameters, Ei,j, are specified in terms of absolute temperature (K). The diagonal
values are zero and the matrix is nonsymmetric. All off-diagonal values must be
specified.

E i i = 0 (8-72)

 i i = 1 (8-73)

The volume parameters, ri, and surface area parameters, qi, are model-specific
parameters for each species. If the parameters are not specified, they can be derived
from the van der Waals volume, VVDW,i, and area, AVDW,i, respectively

V VDW,i
r i = -------------------------------------
- (8-74)
–3
0.01517  10

A VDW,i
q i = ------------------------ (8-75)
–5
2.5  10

756 | CHAPTER 8: THERMODYNAMICS

For all species ri or VVDW,i and qi or AVDW,i must be specified.

UNIFAC
The UNIQUAC Functional-group Activity Coefficients (UNIFAC; see Ref. 20) uses
the same equations as UNIQUAC but the parameters are constructed from group
contributions. The model can be used if UNIQUAC parameters are not available for
all species. The activity coefficients are calculated from Equation 8-66. The
combinatorial part follows from equation Equation 8-67, where

ri =   k i r k (8-76)
k

qi =   k i q k (8-77)
k

where rk and qk are the values for group k in species i, andk,i is the number of
occurrences of group k in molecule. The residual term in Equation 8-66 is calculated
from a summation over functional groups:

ln  i res =  k i  ln k res – ln i k res  (8-78)

k

The values for ln(k,res) are calculated from the mixture containing all species at a
specified composition x. The values for ln(i,k,res) are calculated for a mixture of group
k considering only pure species i. Both are defined, for functional group k, by

   

  l

 x l q l  l k
 x l q l  k l 

ln  k res = q k  1 – ln  --------------------------- –
   ----------------------------- (8-79)
x l q l 
 x l q l  m l



 l 
j  
m

where xl and xm are the compositions of functional group l and m in the mixture

 x i  m i
i
x m = ----------------------------------- (8-80)
 
 
 x i  k i
 
i k

For the calculation of a pure species’ i residual activity we get:

THERMODYNAMIC MODELS AND THEORY | 757

  m i
x m = ----------------
- (8-81)
  k i
k

The volume parameters rk and surface area parameters qk are model-specific

parameters for each group. The binary interaction between groups k and m is

A k m
 k m = exp  – ------------
- (8-82)
T 

The binary interaction parameters, Ak,m, are specified in terms of absolute

temperature. The diagonal values are zero, the matrix is nonsymmetric. All
off-diagonal values must be specified.

The default group and interaction parameters are those published by the UNIFAC
consortium (Ref. 21 through Ref. 26), with added groups from Balslev and Abildskov
(Ref. 27) but can be modified per package or database. The groups must be specified
for all species. Note that the interaction parameter matrix is sparse, and a package can
only be used if all interaction parameters for all used groups are specified.

Chao–Seader (Grayson–Streed)
The Chao–Seader model Ref. 13 correlates liquid phase partial coefficients for pure
species, for use of hydrogen and hydrocarbon mixtures at elevated pressure and
temperatures. It is expressed by

ln  i l = ln  i + ln  i l 0 (8-83)

The activity is based on the Scatchard–Hildebrand equation Ref. 14 and presented in

Equation 8-45 to Equation 8-47. Chao–Seader specific values for liquid volume, Vi,
and solubility parameter, i, are used.

If Vi is unspecified, it can be estimated by

a molar volume at normal boiling point

b van der Waals volume
c from the Rackett model, Equation 8-80, at normal boiling point temperature.
If the Chao–Seader specific solubility parameter, i, is not specified, the generic
solubility parameter is used.

The fugacity coefficient for pure species i is correlated as

log  i l 0 = log  i l 1 +  i log  i l 2 (8-84)

758 | CHAPTER 8: THERMODYNAMICS

 A i 1
log  i l 1 = A i 0 + ---------- + A i 2 T r i + A i 3 T r2 i + A i 4 T r3 i (8-85)
T r i

+  A i 7 T r2 i + A i 6 T r i + A i 5 P r i +  A i 9 T r i + A i 8 Pr2 i – log P r i

A i 12
log  i l 2 = A i 10 + ------------
- + A i 11 T r i + A i 13 T r3 i + A i 14  P r i – 0.6  (8-86)
T r i

A Chao–Seader specific acentric factor i is used. If it is unknown, it can be set equal

to the generic acentric factor. The parameter values are taken from the later publication
of Grayson and Streed Ref. 15 and given in Table 8-2.
TABLE 8-2: CHAO–SEADER GRAYSON–STREED FUGACITY COEFFICIENT PARAMETERS.

H2 CH4 OTHER ALL

Ai,0 1.50709 1.36822 2.05135 Ai,10 -4.23893

Ai,1 2.74283 -1.54831 -2.10899 Ai,11 8.65808
Ai,2 -0.02110 Ai,12 -1.22060
Ai,3 0.00011 0.02889 -0.19396 Ai,13 -3.15224
Ai,4 -0.01076 0.02282 Ai,14 -0.02500
Ai,5 0.008585 0.10486 0.08852
Ai,6 -0.02529
Ai,7 -0.00872
Ai,8 -0.00353
Ai,9 0.00203

The corresponding vapor phase model is Redlich–Kwong equation of state. The

Chao–Seader (Grayson–Streed) model is valid when

- 255 K < T < 533 K

- P < 6.89 MPa
- Pr < 0.8
- 0.5 < Tr,i < 1.3, for all hydrocarbons except CH4
- xCH4 < 0.3
- For systems that include CH4 or H2S:

 xi Tr i  0.93
i

THERMODYNAMIC MODELS AND THEORY | 759

 The enthalpy, entropy, and Gibbs free energy can be calculated from Equation 8-109
to Equation 8-118.

LIQUID VOLUME MODELS

When an activity coefficient model or the Chao–Seader (Grayson–Streed) model is
used, a liquid volume model must be explicitly selected. It is also possible to assign a
liquid volume model when an equation of state is used.

For liquids the density is defined as the reciprocal of the liquid volume:

1-
 l = ----- (8-87)
Vl

Using a cubic equation of state, the solution has 1, 2, or 3 different roots for the
volume. The liquid density is defined as the root producing the highest density, and
the vapor density as the root producing the lowest density.

Equation of State
When an equation of state is selected as the liquid phase model, the liquid volume is
by default set to be calculated using the same equation of state model. The other liquid
volume models; Ideal Mixture, COSTALD, and Rackett are also available.

Note, the vapor phase model and the liquid volume cannot use different equation of
state models.

Ideal Mixture
For an ideal mixture the liquid volume is computed from the pure species densities
(corresponding to the saturated liquid density):

xi
V l m =  --------
 i l
(8-88)
i

In this case the pure species densities corresponds to the saturated liquid density, which
is available as a temperature dependent correlation for all species in the built-in
database.

COSTALD
Hankinson and Thomson Ref. 28 presented the Corresponding States Liquid Density
(COSTALD) equation as

V - = V
------------ r ref  1 – V r   (8-89)
V mix

760 | CHAPTER 8: THERMODYNAMICS

 V r ref = 1 + a  1 – T r  1 / 3 + b  1 – T r  2 / 3 + c  1 – T r  + d  1 – T r  4 / 3 (8-90)

e + fT r + gT r2 + hT r3
V r  = ---------------------------------------------------
- (8-91)
T r – 1.00001

1   
V mix = --- 
4  xi Vi + 3   xi Vi2 / 3   xi Vi1 / 3  (8-92)
i i i

 2
 
 x i T c i V i

i
T c = ------------------------------------------ (8-93)
V mix

 -----
T
 - T  Tc
Tr =  Tc (8-94)
1 T  Tc


 =  xi i (8-95)
i

where the volume, Vi, and acentric factor, ifor species i are model specific
parameters. If the COSTALD volume parameter is unspecified, it is estimated from the
van der Waals volume when VVDW,i0.3103 m3/mol

2
V i = 5.385 V VDW i – 5.1022VVDW 3
+ 79.524VVDW (8-96)
i i
4
– 99.316VVDW 5
+ 100.88VVDW 6
– 1152.7VVDW
i i i

otherwise, it can be set equal to the critical volume as

V i = V c i (8-97)

If the COSTALD acentric factor, i is not specified, it can be set equal to the generic
acentric factor for species i. A critical temperature, Tc,i must be specified for all species.

The correlation parameter values are:

TABLE 8-3: REDUCED VOLUME CORRELATION PARAMETERS IN COSTALD.

a -1.52816 e -0.296123
b 1.43907 f 0.386914

THERMODYNAMIC MODELS AND THEORY | 761

 TABLE 8-3: REDUCED VOLUME CORRELATION PARAMETERS IN COSTALD.

c -0.81446 g -0.0427258
d 0.190454 h -0.0480645
Rackett
The Rackett equation Ref. 29 computes the liquid density at the saturation point, and
can be used to describe liquid density at any pressure using the assumption that the
liquid is incompressible. The equation and its condition can be expressed as:

 x i T c i     1 +  1 – Tr 2  7   
2

Vl =  R
  - 
---------------------------
 MW  i P c i   Z r i


  xi  MW i (8-98)
i i i

Tc =  xi Tc i (8-99)
i

 -----
T
 T- T  Tc
Tr =  c (8-100)
1 T  Tc


Critical temperatures, Tc,i, critical pressures, Pc,i, and molecular weights, Mi, must be
specified for all species. The model parameter, Zr,i, must be specified for all species. If
the value is not available it can be set to the critical compressibility factor:

Z r i = Z c i (8-101)

Selecting the Right Thermodynamic Model

Which thermodynamic model is the most appropriate depends on a number of things,
for example the nature of the properties to be used, the operational pressure and
temperature, and not least the availability of the required model parameters. One
general introduction to how to select thermodynamic models, as well as estimating
missing parameters and validating properties can be found in Ref. 77.

762 | CHAPTER 8: THERMODYNAMICS

Below is a decision tree that can be used as a guide for choosing the thermodynamic
model:

Figure 8-37: Decision tree to select thermodynamics model.

Species Property References

All references used for the included constants and temperature-dependent properties
are available in the database file. The reference comments can be inspected in the
settings window by right-clicking a species property node (under a Thermodynamic
System) and selecting Properties.

Thermodynamic Properties Definitions

The thermodynamic properties provided are listed as species and mixture property in
a Property package.

THERMODYNAMIC MODELS AND THEORY | 763

 Ideal Gas
The ideal gas law is independent of composition and determines V at given T and P.
Density can be calculated from

1-
 = --- (8-102)
V

The partial fugacity coefficients

ln ̂ i = 0 (8-103)

The ideal gas enthalpy for mixture

 T 
 xi  Hi ig T T
H ig = + C P i ig dT (8-104)
ref
ref 
i

where Hi,ig,Tref relates the enthalpy of an ideal gas to the enthalpy at the selected
reference state for species i.

The ideal gas entropy for species i

 T C P i ig 
- dT – R ln  --------
P -
S ig =  x i  S i ig T ref – R ln x i +
 T ref
-----------------
T   P ref
 (8-105)
i

where Si,ig,Tref is the entropy of an ideal gas to the entropy of species at the selected
reference state.

The Gibbs free energy follows from

G ig = H ig – TS ig (8-106)

Equation of State
The equation of state determines V at given x, T, and P. Density can be expressed as
Equation 8-102. The partial fugacity coefficients are derived from

P
ˆ
0  Vi – --------
1 RT
ln ̂ i = -------- (8-107)
RT P

where Vi is the partial molar volume. The enthalpy, entropy, and Gibbs free energy
follow from the partial fugacity coefficients and the ideal gas contributions as:

764 | CHAPTER 8: THERMODYNAMICS

  ln ̂ i
H = H ig – RT 2  xi  -------------
T 
- (8-108)
i

 ln ̂ i
S = S ig – R  xi  ln ̂i + T -------------
T 
- (8-109)
i

G = G ig – RT  xi ln ̂i (8-110)
i

Heat Capacity

The heat capacity at constant pressure is calculated from

H
C P =  -------- (8-111)
 T  P,x

It is available when the enthalpy is available. The heat capacity at constant volume is
defined by

H
C v =  -------- (8-112)
 T  v,x

The relationship between the heat capacity at constant pressure and constant volume
can be expressed as

V P
C p – C v = T  -------  ------- (8-113)
 T P,x  T v,x

were v is a function of T and P. A Taylor expansion of vat constant composition gives

V V
v =  ------- T +  ------- P = 0 (8-114)
 T P,x  P T,x

Rearranging equations above gives

V 2
T  -------
 T P,x
C v = C P + ------------------------ (8-115)
 V
-------
 P T,x

Cv is available if Cp and volume are available and fluid is compressible by means of

volume is pressure dependent. For ideal gas Equation 8-113 express as

THERMODYNAMIC MODELS AND THEORY | 765

 Cp – Cv = R (8-116)

Specific heat capacity ratio is defined as

C
 = ------p- (8-117)
Cv

Activity Coefficient Models

The enthalpy, entropy, and Gibbs free energy follow from the activity coefficients and
the ideal gas contributions in Equation 8-104 – Equation 8-106. The activity
coefficients describe the deviation of chemical potentials from the ideal liquid phase,
so heats of vaporization need to be accounted for

 2  ln  i
T
H =  xi  Hi ig T ref
+ T ref
C P i ig dT – H i vap – RT --------------
T
(8-118)
i
2  ln ̂ i sat
– R T -----------------------
T 

 T C P i ig P i sat
- dT – R ln  -------------
S =  xi  Si ig T ref
– R ln x i + T ref
-----------------
T  P ref 
- (8-119)
i
 ln  i  ln ̂ i sat H i vap
+ ln  i + T -------------- + ln ̂ i sat + T ----------------------- + R --------------------
T T T 

 T 
G =  x i  H i ig T ref +
 T ref
C P i ig dT – T  S i ig T ref +

(8-120)
i
T C P i ig  Pi sat  
------------------ dT + RT  ln x i + ln  i + ln ̂ i sat + ln -------------
T ref
T   P ref  
- 

Note that if the vapor phase is ideal, then the saturated fugacity, ̂ i sat , contribution
can be ignored.

Other Properties
Partial fugacity is calculated from

fˆi = x i ̂ i P (8-121)

Internal energy is calculated from

766 | CHAPTER 8: THERMODYNAMICS

 U = H – PV +  xi Ui ig ref (8-122)
i

where Ui,ig,ref is the enthalpy of an ideal gas to the species enthalpy at the selected
reference state.

Helmholtz energy is calculated from

A = U – TS = G – PV (8-123)

The K-values for phases p and q are taken from

̂ i q
K i p q = --------- (8-124)
̂ i p

If only liquid phases are defined, the K-value calculation is reduced to

 i q
K i p q = --------
- (8-125)
 i p

M mix =  xi Mi (8-126)
i

Standard Enthalpy of Formation and Absolute Entropy Terms

Formation terms are not included in the enthalpy, entropy, or internal energy. If the
enthalpy of formation is specified for all species, EnthalpyF (HF) represent the enthalpy
including the formation terms. If the absolute entropy of all species are specified then
the EntropyF (SF) includes the absolute entropy.

At reference conditions, the heat of any reaction relates to the heat of formation as

H reac,ref =  i  Hf ref  (8-127)

i

Note that for reacting flow or a heat balance in a reactor when the heat of reaction is
explicitly taken into account, the enthalpy should not include the heat of formation.

The following enthalpy expression should be used:

HF = H +  xi Hf i ref (8-128)

i

THERMODYNAMIC MODELS AND THEORY | 767

 Hf,i,ref is calculated such that it matches the reference state of the package, even if the
species have a different reference state. Similarly, the absolute entropy of each species
is taken into account according to

SF = S +  xi Sabs i (8-129)
i

The entropy balance over a process that includes reactions should include either the
entropy of reaction and use S, or use SF without entropy of reaction.

In database, a set of parameters introduced to define Absolute entropy and Standard

enthalpy of formation for ideal gas and different phases (Vapor, liquid, Solid, and
Aqueous infinite dilution). One can use these parameters to calculate Gibbs free energy
of reaction when different phases are available (See Electrode reaction).

SPECIES GIBBS FREE ENERGY OF FORMATION

As the COMSOL ThermoDB does not include entropy of formation of species,
entropy Equation 8-129 and in consequence Gibbs free energy Equation 8-110 are
estimated from absolute entropy’s value. Therefore, the absolute value of entropy of
formation and Gibbs free energy of formation of the species are not correct. However,
for our applications these absolute values are not required while the value of entropy
and Gibbs free energy change of reaction (the difference value) are correct.

If user needs the absolute value of entropy of formation, it is possible to estimate it by:

S i = S f i =  j Sabsj (8-130)
j

where the entropy of formation of species i is calculated from its elemental constituent
j. For example, entropy of formation of ammonia at 298 K is

N 2  g  + 3H 2  g   2NH 3 (8-131)

S f NH3 = 2S abs NH3 –  S abs N2 + 3S abs H 2  (8-132)

The values from RHS of Equation 8-132 can be extracted from

COMSOLThermoDB. This value should be used in Equation 8-128 to estimate the
absolute value of Gibbs free energy of species.

768 | CHAPTER 8: THERMODYNAMICS

Reference State
The reference temperature, pressure, and phase can be specified for each package and
for species within a package. The reference phase can be a real phase or an ideal gas. If
a reference temperature, pressure, or phase is unspecified for a particular species, the
package reference temperature, pressure, or phase is used for that species.

The reference values for enthalpy, Hi,ref, entropy, Si,ref, and internal energy, Ui,ref, are
calculated so that the pure species enthalpy, entropy, and internal energy are equal to
zero at reference conditions.

The reference values for enthalpy, including formation terms, and entropy, including
absolute terms, are calculated such that the corresponding property for pure species i
has a value equal to the specified formation term at the species reference conditions.

Transport Properties
This section includes definitions of the models available in for thermal conductivity,
viscosity, and diffusivity:

• Thermal Conductivity
• Viscosity
• Diffusivity

THERMAL CONDUCTIVITY

Vapor

Ideal

The thermal conductivity correlations is according to:

v =  xi i v + v P (8-133)

i

The pressure correction v,P is calculated from the method of Stiel and Thodos, see
Ref. 30, which is applicable for r  3, but is less accurate for H2, strongly polar gases,
and gases with a high degree of hydrogen bonding, such as H2O and NH3

P c2 / 3
 v P = ----------------------------- A  exp   B r  + C   (8-134)
MT c1 / 6 Z c5

THERMODYNAMIC MODELS AND THEORY | 769

 -
 r = ---- (8-135)
c

1-
 c = ----- (8-136)
Vc

The mixing rules are as suggested by Yorizane, see Ref. 31:

 x i  x j V c i j
Vc = (8-137)
i j

 x i  x j V c i j T c i j (8-138)
i j
T c = -----------------------------------------------------
-
Vc

 xi i
 = (8-139)
i

Z c = 0.291 – 0.08 (8-140)

RT c Z c
P c = ----------------- (8-141)
Vc

 xi Mi
M = (8-142)
i

where the binary constants are

1
V c i j = ---  V c1/i3 + V c1 /j3  3 (8-143)
8

T c i j = T c i T c j (8-144)

The Stiel and Thodos coefficients are

TABLE 8-4: PRESSURE CORRECTION PARAMETERS.

A B C
r0.5 A1=2.702E8 B1=0.535 C1=-1
0.5r2.0 A2=2.528E8 B2=0.670 C2=-1.069
r2.0 A3=0.574E8 B3=1.155 C3= 2.016

770 | CHAPTER 8: THERMODYNAMICS

However, in order to ensure 0th-order continuity at r  0.5 and r  2.0, the
following coefficients for   r  2.0 are recalculated from

A 1 C 1 + A 1 exp  0.5 B 1  – A 3 C 3 – A 3 exp  2.0 B 3 

A 2 = – ------------------------------------------------------------------------------------------------------------------------------
- (8-145)
exp  2.0 B 2  – exp  0.5 B 1 

A 1 C 1 exp  2.0 B 2  + A 1 exp  0.5 B 1  exp  2.0 B 2 

C 2 = – ------------------------------------------------------------------------------------------------------------------------------
- (8-146)
A 1 C 1 + A 1 exp  0.5 B 1  – A 3 C 3 – A 3 exp  2.0 B 3 
A 3 C 3 exp  2.0 B 2  + A 3 exp  0.5 B 2  exp  2.0 B 3 
– ------------------------------------------------------------------------------------------------------------------------------
-
A 1 C 1 + A 1 exp  0.5 B 1  – A 3 C 3 – A 3 exp  2.0 B 3 

The vapor thermal conductivity correlation must be available for all species. Also
critical volumes, Vc,i, critical temperatures, Tc,i, molecular weights Mi, and acentric
factors i must be specified for all species.

Kinetic Theory

Lindsay and Bromley (see Ref. 32) provided an equation for the interaction parameters
of the method of Wassiljewa (see Ref. 33) based on the kinetic theory, to provide
mixture thermal conductivity from pure species values

 
 x 
 --------------------- + 


i i v
v = (8-147)
  v P
i
 
 x j  i j

j

2
 T + 3   
 --- T b i   9
 T +  --4- T b i T b j
 M 3 / 4  2  
= ---  1 + ----------  -------  ------------------------------   ----------------------------------------------
1 i v j
 i j (8-148)
4  j v M i T + 3  
--- T    T +  --- T b i 
3
 
  2 b j    2  

where the pressure correction v,P is calculated from Equation 8-134. Both vapor
thermal conductivity correlation i,v and the vapor viscosity correlation i,v must be
available for all species. In addition, all normal boiling points Ti,b, molecular weights
Mi, critical volumes Vc,i, critical temperatures Tc,i, and acentric factors i must be
specified.

Water (IAPWS)

The International Association of the Properties of Water and Steam recommend an

equation Ref. 34 which is valid in the following range:

THERMODYNAMIC MODELS AND THEORY | 771

 273.15 K  T  773.15 K p  100 MPa (8-149)

773.15 K  T  923.15 K p  70 MPa (8-150)

923.15 K  T  1073.15 K p  40 MPa (8-151)

Liquid
The following mixture models are available for liquid thermal conductivity

Ideal

To calculate the mixture liquid thermal conductivity, l,m, the values of pure liquid
thermal conductivity correlations are mixed ideally

 xi i l
 l m = (8-152)
i

The pressure dependence is based on the work of Missenard (Ref. 51) where

 l m P
----------------- = 1 + QP r0.7 (8-153)
 l m

where Q is correlated as

Q = -0.013693477 + 0.057427673T r + 0.001617831  T r   ln P 1  (8-154)

and the following mixing rules are used

T-
T r = ----- (8-155)
Tc

P-
P r = ----- (8-156)
Pc

RT c Z c
P c = ----------------- (8-157)
Vc

 x i V c i
Vc = (8-158)
i

 x i T c i
Tc = (8-159)
i

772 | CHAPTER 8: THERMODYNAMICS

  x i Z c i
Zc = (8-160)
i

All liquid thermal conductivity correlations must be specified. All values for critical
temperatures, Tc,i, critical volumes, Pc,i and critical compressibility factors, Zc,i must
be specified for all species i.

Power Law

The values of pure liquid vapor thermal conductivity correlations are mixed according
to the following power law

1 = xi
-----------
 l2 m  --------
 l i
(8-161)
i

All liquid thermal conductivity correlations must be specified. The model is valid for
pure compound thermal conductivity values that are no more apart than a factor of 2
(Ref. 52 and Ref. 53). The pressure dependence is introduced using Equation 8-153
through Equation 8-160. All values for critical temperatures, Tc,i, critical volumes, Pc,i
and critical compressibility factors, Zc,i must be specified for all spices i.

Local Composition

The local composition model by Rowley (Ref. 53) uses an ideal and excess
contribution

 l m =  l ideal +  l excess (8-162)

The ideal part is based on mass fractions

 l ideal =  i i l (8-163)

i

xi Mi
 i = -------------------
- (8-164)
 xj Mj
j

The excess term is based on NRTL local concentrations

THERMODYNAMIC MODELS AND THEORY | 773

   i j Gj i  j i – i l 
 l excess = ------------------------------------------------------------------
i j - (8-165)
 j G j i
j

where Gj,i follows from Equation 8-58. The binary interaction terms follow from

 i  i i  i l +  j  j j  j l
 i j =  j i = -------------------------------------------------------- (8-166)
 i  i i +  j  j j

which is symmetric, and on the diagonal,

 i i =  i l (8-167)

and

 i i =  i   i +  j G j i  (8-168)

with i is the composition in the binary mixture of species i and j and the local
composition is equi-molar

M i G j i
 i = -------------------------------------------------- (8-169)
M i G j i + M j G i j

All liquid thermal conductivity correlations must be specified. The pressure

dependence is introduced using Equation 8-153 – Equation 8-160. All values for
critical temperatures Tc,i, critical volumes Pc,i, critical compressibility factors Zc,i, and
molecular weights Mi must be specified for all compounds i. In addition, all NRTL
binary interaction parameters Ai,j must be specified. Unspecified values for NRTL
interaction parameters Bi,j are set to zero. The randomness parameters i,j have values
of zero on the diagonal and the matrix is symmetric. All off-diagonal values must be
specified. NRTL model is presented in Equation 8-54 to Equation 8-65.

Local Composition (Modified)

Rowley (Ref. 52) adapted the local composition model by replacing the mixing rule in
Equation 8-166 by the following

M i   i  i i  i l + M j   j  j j  j l
 i j =  j i = ---------------------------------------------------------------------------------
- (8-170)
M i   i  i i  + M j   j  j j 

774 | CHAPTER 8: THERMODYNAMICS

which he found to produce better model predictions in most cases where both the
Local Composition model and Power Law model have trouble. However, the model
is not as generally applicable; for instance, systems containing H2O are not well
described by this model due to the low molecular weight of H2O.

Water (IAPWS)

The International Association of the Properties of Water and Steam recommend an

equation Ref. 34 which is valid in the following range:

273.15 K  T  773.15 K p  100 MPa (8-171)

773.15 K  T  923.15 K p  70 MPa (8-172)

923.15 K  T  1073.15 K p  40 MPa (8-173)

VISCOSITY

Vapor

Wilke

Wilke, see Ref. 35, based his method for mixture viscosity of the vapor phase on kinetic
theory:

 
 x 
 ---------------------
i i v 
v =  
-

(8-174)
i
 
 x j  i j

j

2
  i v M j 1 / 4
 1 + ----------  -------- 
  j v Mi 
 i j = ------------------------------------------------------ (8-175)
M
8  1 + -------i
 M j

The vapor viscosity correlation i,v must be available for all species. In addition, all
molecular weights Mi must be specified.

Brokaw

Brokaw (see Ref. 36) uses the same basic equation as Wilke (Equation 8-174).
However, Equation 8-175 is replaced by

THERMODYNAMIC MODELS AND THEORY | 775

  i v
 i j = S i j A i j ---------
- (8-176)
 j v

and the interaction parameter is defined as

 
 i j  RM i j – RM i0.45 j


A i j = --------------------  1 + ----------------------------------------------------------------------------------------- (8-177)
RM i j  1 + RM i0.45 
2  1 + RM i j  + ------------------------------------------------ j 
   i j   1 + RM i j 

where

 4M i M j  1 / 4 M
 i j =  ----------------------------2 , RM i j = -------i (8-178)
  Mi + M   Mj
j

The vapor viscosity correlation, i,v must be available for all species i. In addition, all
molecular weights Mi must be specified. If Lennard–Jones energy i (see Ref. 37)
Stockmayer’s polar parameter s,i (Ref. 38 and Ref. 39) are specified for both species
i and j then

T 2- + 1 ---  
1 + --------
 i  j 4 s i s j
S i j = ---------------------------------------------------------------------------
- (8-179)
T 1 T 1
1 + ---- + ---  s i 1 + ---- + ---  s j
i 4 j 4

Otherwise,

S i j = 1 (8-180)

Davidson

The Davidson method, see Ref. 40, requires fewer compound specific parameters than
Brokaw, while reported accuracy is almost as good, and in the case of H2, even
surpasses it. The Davidson model only requires molar masses and the viscosities of the
pure gases. The model is based on fluidity, which is defined to be the reciprocal
viscosity.

f = --1- (8-181)


The fluidity of the mixture is then calculated as

776 | CHAPTER 8: THERMODYNAMICS

  yi yj A 
f =   -------------------  E i j
 i j 
(8-182)
i j

where yi is the momentum fraction of species i; Ei,j is the momentum transfer

coefficient of the species pair i, j; and A is an empirical species-independent parameter
set to 1/3. The momentum fraction is given by

xi Mi
y i = ----------------------------
- (8-183)
 xj Mj 
j

and the momentum transfer coefficient is taken as

2 Mi Mj
E i j = --------------------------
- (8-184)
Mi + Mj

High Pressure Modification

To account for the effect of pressure on vapor viscosity, a pressure correction can be
applied. The pressure dependence is based on kinetic gas theory, which adds the
following term to the vapor viscosity:

 M t P c2 / 3  10 – 7
- --------------------------- 
 v P =  ----------------------- (8-185)
 T c1 / 6   101325 2 / 3 

where  is calculated from the correlation of Jossi (Ref. 41), which is applicable for
r  3.0. It is less accurate for H2, strongly polar gases and gases with a high degree of
hydrogen bonding such as H2O and NH3.

The correction factor is due to using pressure, atm, and viscosity, cP, units in Jossi’s
correlation. It is expressed as:

1
---
4
  + 1  = 1.0230 + 0.23364 r + 0.58533  r2 (8-186)

– 0.40758  r3 + 0.093324  r4

-
 r = ---- (8-187)
c

THERMODYNAMIC MODELS AND THEORY | 777

 1-
 c = ----- (8-188)
Vc

The following mixture rules are used

RT c Z c
P c = -----------------
Vc
, Vc =  x i V c i , T c =  x i T c i (8-189)
i i

Zc =  x i Z c i , M t =  x i M i (8-190)
i i

The values for critical volumes, Vc,i, critical temperatures, Tc,i, critical compressibility
factors, Zc,i and molecular weights, Mi must be specified for all species i.

The high pressure correction is available for the Wilke, Brokaw, and Davidson mixture
models. The vapor viscosity follows from

 v =  v Wilke +  v P (8-191)

where v,Wilke is calculated from Equation 8-174.

Pedersen Corresponding States Model

The corresponding states viscosity model of Pedersen (Ref. 42 and Ref. 43) applies to
both vapor and liquid phases of hydrocarbon mixtures. The selected reference species
is CH4.

The CH4 viscosity is calculated from Ref. 44, modified by Pedersen and Fredenslund
(Ref. 45) to avoid issues below 91 K where CH4 becomes solid

 CH 4 =  CH 4 0 +  CH 4  CH  1 + F 1  CH  2 + F 2  CH  3 (8-192)
4 4 4

where

F1 = H + 1-
------------- (8-193)
2

F1 = 1 – H-
------------- (8-194)
2

exp  T  – exp  – T  -
H = ------------------------------------------------------------ (8-195)
exp   T  + exp  – T  

T = T – 91 K T  K (8-196)

778 | CHAPTER 8: THERMODYNAMICS

Here, CH4 is used in g/cm3; for the mass-mole conversion of CH4, a molecular
weight of MCH4 = 16.042568 g/mol is used.

The dilute gas part is given by

–L L2 L3
 CH4 0 = ---------1- + ----------- – ----------- (8-197)
T T 2 / 3 T1 / 3
+ L4 –L5 T 1 / 3 + L6 T 2  3 – L7 T + L8 T 4  3 – L9 T 5 / 3

The first density correction for the moderately dense gas is given by

2
 CH 4 1 = L 10 – L 11  1.4 – ln ----------------
T
(8-198)
168 K

The remainder is given by the empirical correlation

L 12   L 15 
 CH 4 2 = exp  --------
- – L 13  exp   10  CH 4   L 14 – ----------- + (8-199)
 T    T 3 / 2
  CH –  c CH 4 L 17 L 18  
4 -   CH   L 16 + --------
 ------------------------------------ - + ---------  – 1
 4  T 
 c CH 4  T  2 

The correction term for solid CH4

L 19   L 22 
 CH 4 3 = exp  --------
- – L 20  exp   10  CH 4   L 21 – ----------- (8-200)
T   T 3 / 2
  CH 4 –  c CH 4 L 24 L 25  
+  -----------------------------------   CH4   L 23 + --------- + --------
-  – 1
  c CH 4  T T  2 

with the values of the parameters L1 through L25 are listed in Table 8-5 below:
TABLE 8-5: METHANE VISCOSITY NUMERICAL COEFFICIENTS.

L1 2.090975·105 L10 1.696985927 L19 9.74602

L2 2.647269·105 L11 0.133372346 L20 44.6055
L3 1.472818·105 L12 188.73011594 L21 18.0834
L4 47167.40 L13 10.35060586 L22 4126.66
L5 9491.827 L14 17.571599671 L23 0.976544
L6 1219.979 L15 3019.3918656 L24 81.8134
L7 96.27993 L16 0.042903609488 L25 15649.9

THERMODYNAMIC MODELS AND THEORY | 779

 TABLE 8-5: METHANE VISCOSITY NUMERICAL COEFFICIENTS.

L8 4.274152 L17 145.29023444

L9 0.08141531 L18 6127.6818706
Here, CH4 is used in g/cm ; the critical density is given by c,CH4  0.16284 g/cm3.
3

The following equation by McCarty (Ref. 46) is solved for the density of CH4

 N 5 N 6
P = N 1  CH T +  CH
2  – N 2 T + N 3 T – N 4 + ------- – ------2- (8-201)
4 4
 T T 

 N 9 N 10  – N T + N + –------------ N 13

+  CH
3  N 7 T – N 8 + ------- – ---------- +  CH 4
4
- + N 14  CH 5
4
 T T 
2 11 12
T  4

 N 15 N 16  CH7
 N 18 N 19  CH9
+  CH - + N 17 ------------- +  CH
4 4
6  – ---------- + --------- 8  ---------- + ---------- + N 20 -------------
4
 T T 
2 T 4
 T T 
2 T

 3  N 22 N 23  N 24 N 25  N 26 N 27

+ exp  – N 21  CH
2    CH - – ---------- +  CH
 --------- 5 - + ---------- +  CH
 --------- 7
 ---------
- + ----------
4
 4
T 2
T 
3 4
T 2
T 
4 4
 T2 T 
3

 N 28 N 29  N 30 N 31  N 32 N 33 4 
+  CH
9 - – ---------- +  CH
 – --------- - + ---------- +  CH
11  – --------- 13  ---------
- – ---------- + N 34 T  
4
 T 2
T 
4 4
 T 2
T 
3 4
T 2
T
3 

where CH4 is used in mol/l.

TABLE 8-6: COEFFICIENTS IN THE FUNCTIONAL FORM OF THE MCCARTY EOS.

N1 0.08205616 N13 2.8685285973 N25 1.6428375992·106

-3
N2 0.018439486666 N14 0.11906973942·10 N26 0.21325387196
N3 1.0510162064 N15 0.0085315715699 N27 37.791273422
N4 16.057820303 N16 3.8365063841 N28 0.1185701681·10-4
N5 848.44027562 N17 0.24986828379·10-4 N29 31.630780767
N6 42738.409106 N18 0.57974531455·10-5 N30 0.4100678294·10-5
N7 0.76565285254·10-3 N19 0.0071648329297 N31 0.0014870043284
N8 0.48360724197 N20 0.12577853784·10-3 N32 3.151226153·10-9
N9 85.195473835 N21 0.0096 N33 0.2167077474·10-5
N10 16607.434721 N22 22240.102466 N34 0.2400055107·10-4
N11 0.37521074532·10-4 N23 1.4800512328·106
N12 0.028616309259 N24 50.498054887

With the viscosity and density of CH4 defined, the viscosity of any mixture, mcan be
calculated from the corresponding states principle

780 | CHAPTER 8: THERMODYNAMICS

 –1
Tc ------ Pc 2 / 3 Mm 1 / 2 
 m =  ----------------- 6  -----------------  ---------------  -------------  CH 4 P 0 T 0 (8-202)
 T c CH   P c CH   M CH    CH 
4 4 4 4

where the CH4 viscosity CH4,P0,T0 is calculated at temperature T0 and pressure P0:

T c CH  CH
T 0 =  ------------------4 T  -------------4 (8-203)
Tc 

P c CH  CH
P 0 =  -----------------4 P  -------------4 (8-204)
 Pc    

The following mixing rules are used for the critical properties, see Ref. 47:

  x i x j  i j T c i T c j
(8-205)
Tc = i j
-----------------------------------------------------------
-
  xi xj i j
i j

8   x i x j  i j T c i T c j
(8-206)
Pc = i j
--------------------------------------------------------------
-
 2

   xi xj i j
i j

T c i 1 / 3  T c j 1 / 3 3
 i j =   ---------- + ---------- (8-207)
  P c i  P c j 

The parameter  is

–3 0.5173
 = 1 + 7.378 10  r1.847 M m (8-208)

 CH4 P 0 T 0
 r = --------------------------
- (8-209)
 c CH 4

where

T c CH
T 0 =  -----------------4 T (8-210)
Tc

THERMODYNAMIC MODELS AND THEORY | 781

 P c CH
P 0 =  -----------------4 P (8-211)
Pc

with c,CH4  0.16284 g/cm3. For CH4

–3 0.5173
 CH 4 = 1 + 7.378 10  r1.847 M CH 4 (8-212)

where Equation 8-209 is used. The mixture molecular weight is a function of the
weight-averaged molecular weight and the number-averaged molecular weight

–4 2.303 2.303
M m = 1.304 10  M W – MN  + MN (8-213)

 xi Mi
2

i (8-214)
M W = --------------------
-
MN

 xi Mi
MN = (8-215)
i

where the power in Equation 8-213 is determined by fitting to experimental viscosity

data.

Note that pure species vapor viscosity correlations i,v are not required. However, for
each species i, molecular weight Mi, critical temperature, Tc,i, and critical pressure, Pc,i
must be specified.

Water (IAPWS)

The International Association of the Properties of Water and Steam recommend an

equation Ref. 48–Ref. 49 for industrial application which is valid in the following
range:

273.16 K  T  1173.15 K 0  p  pt (8-216)

T m  p   T  1173.15 K p t  p  300 MPa (8-217)

T m  p   T  873.15 K 300 MPa  p  350 MPa (8-218)

T m  p   T  433.15 K 350 MPa  p  500 MPa (8-219)

T m  p   T  373.15 K 500 MPa  p  1000 MPa (8-220)

782 | CHAPTER 8: THERMODYNAMICS

where Tm is the pressure dependent melting temperature and pt is the triple-point
pressure. In accordance with industrial application recommendations, critical region
correction is not applied.

Liquid
The following mixture models are available for liquid viscosity.

Molar Logarithmic Mixing

The values of pure species log liquid viscosity, ln i,l are mixed ideally using mole
fractions xi

ln  l m =  xi ln i l (8-221)
i

where m,l is the mixture viscosity of liquids.

Mass Logarithmic Mixing

The values of pure species log liquid viscosity correlation are mixed ideally using the
weight fractions i

ln  l m =  i ln i l (8-222)
i

Pedersen Corresponding States Model

The Pedersen Corresponding States Model described above for the gas phase viscosity
also applies to the liquid phase. Pure species liquid viscosity correlations are not
required. However, for each species i, molecular weight Mi, critical temperature Tc,i,
and critical pressure Pc,i, must be specified.

Cubic mixing

The mixture viscosity is defined using the cubic root average in terms of the mole
fractions xi

 3
 l m = 
  x i  i1 /l3

(8-223)
i

The model is noted in Ref. 50 to provide reasonable results for hydrocarbon mixtures
of similar components.

THERMODYNAMIC MODELS AND THEORY | 783

 The model requires that the log liquid viscosity correlation is available for all species i.

Cubic mass mixing

The mass fraction equivalent of the previous model is

 3
 l m = 
   i  i1 /l3

(8-224)
i

The model requires that the log liquid viscosity correlation is available for all species i.

Water (IAPWS)

The International Association of the Properties of Water and Steam recommend an

equation Ref. 48–Ref. 49 for industrial application which is valid in the following
range:

273.16K  T  1173.15K 0  p  pt (8-225)

T m  p   T  1173.15 K p t  p  300 MPa (8-226)

T m  p   T  873.15 K 300 MPa  p  350 MPa (8-227)

T m  p   T  433.15 K 350 MPa  p  500 MPa (8-228)

T m  p   T  373.15 K 500 MPa  p  1000 MPa (8-229)

where Tm is the pressure dependent melting temperature and pt is the triple-point

pressure. In accordance with industrial application recommendations, critical region
correction is not applied.

DIFFUSIVITY
Two types of diffusion coefficients are supported. Diffusion coefficients in infinitely
diluted systems, and Maxwell–Stefan diffusion coefficients.

For dilute systems, the binary diffusion coefficient D0i,j represent the diffusivity of
species i in a medium consisting of pure species j. This corresponds to the Fickian
diffusion coefficient.

For any mixture, the binary Maxwell–Stefan diffusion coefficient D i j , represents the
inverse drag coefficient of species i moving past species j (Ref. 54–Ref. 57). This

784 | CHAPTER 8: THERMODYNAMICS

property is referred to as the Maxwell–Stefan diffusivity. The Maxwell–Stefan
diffusivity is symmetric, D i j = D j i , and the diagonal elements D i i are not used.

Gas Phase Diffusion Coefficient at Infinite Dilution

The following models are available for the diffusion coefficients at infinite dilution in
the vapor phase:

• Fuller–Schettler–Giddings
• Wilke–Lee

Automatic When the Gas diffusivity property model is set to Automatic, the Fuller–
Schettler–Giddings model is used, provided that the Fuller diffusion volume is known
for both species (i and j), otherwise the Wilke–Lee model is used.

Fuller–Schettler–Giddings

Fuller and others (Ref. 58) modified the Chapman–Enskog relation to correlate binary
diffusion coefficient for species i and j in the vapor phase according to the Fuller–
Schettler–Giddings (FGS) model:

–2 1.75 1- + ------ 1-
1.01325  10 T ------
0 Mi Mj
D i j = -------------------------------------------------------------------------------- (8-230)
1 1 2
 --- ---
 3   3
P  v +  v 

 
  
 i  
j

where T denotes the temperature (K), Mi the molecular weight of species i (g/mol)
and P is the pressure (Pa). vi are the atomic diffusion volumes (Fuller diffusion
volume, cm3), which are estimated using group contribution for each species
(Ref. 59):
TABLE 8-7: ATOMIC AND STRUCTURAL DIFFUSION VOLUME INCREMENTS.

GROUP CONTRIBUTION

C 15.9
H 2.31
O 6.11
N 4.54
F 14.7
Cl 21
Br 21.9

THERMODYNAMIC MODELS AND THEORY | 785

 TABLE 8-7: ATOMIC AND STRUCTURAL DIFFUSION VOLUME INCREMENTS.

GROUP CONTRIBUTION

I 29.8
S 22.9
Aromatic Ring -18.3
Heterocyclic Ring -18.3

For some simple molecules the values below, determined from regression, are used:
TABLE 8-8: DIFFUSION VOLUMES OF ATOM AND SIMPLE MOLECULES.

SPECIES FULLER DIFFUSION

VOLUME

He 2.67
Ne 5.98
Ar 16.2
Kr 24.5
Xe 32.7
H2 6.12
D2 6.84
N2 18.5
O2 16.3
CO 18
CO2 26.9
N2O 35.9
NH3 20.7
H2O 13.1
SF6 71.3
Cl2 38.4
Br2 69
SO2 41.8
Air 19.7

Wilke–Lee

Wilke and Lee (Ref. 60) also modified the Chapman–Enskog relation to correlate
binary diffusion coefficient in vapor phase according to:

786 | CHAPTER 8: THERMODYNAMICS

 3.03 –  --------------------- T
0.98 1.5

0 –2 MW i j
D i j = 10  -------------------------------------------------------
2
- (8-231)
P i j  D MW i j

where

2 -
MW i j = -------------------- (8-232)
1
------- + ------1-
Mi Mj

The length scale i,j for the interaction is taken from the Lennard–Jones diameter
parameters of species i and j:

LJ LJ
 i + j 
 i j = ------------------------------
- (8-233)
2
LJ
If  i is not specified in the database, it is instead estimated from:

1
---
LJ –8 3
i = 1.18  10 V i l b (8-234)

where Vi,l,b is the molar volume of species at normal boiling point.

The collision integral D is evaluated from (Ref. 61):

-0.15610
 D = 1.06036 T + 0.19300 exp  -0.47635 T  (8-235)
+ 1.03587 exp  -1.52996 T  + 1.76474 exp  -3.89411 T 

with

T-
T = ------- (8-236)
 i j

The energy scale i,j for the interaction is taken from the Lennard–Jones energy
parameters of species i and j:

LJ LJ
 i j = i j (8-237)
LJ
If  i is not specified in the database, it is instead estimated from:

LJ
i
--------
- = 1.15T i l b (8-238)
k

THERMODYNAMIC MODELS AND THEORY | 787

 where k is the Boltzmann constant and Ti,l,b is the normal boiling point temperature.

Gas Phase Maxwell–Stefan Diffusivity

For gas phase diffusion the Maxwell–Stefan diffusivities are defined from the models
for gas phase diffusivity at infinite dilution

0
D i j = D i j i  j (8-239)

All models for gas phase diffusion at infinite dilution are symmetric, the diffusivity of
species i in species j equals that of species j in species i. In addition, under the ideal gas
assumption, the Maxwell–Stefan diffusion coefficient matches the Fick diffusion
coefficient. This implies that the Maxwell–Stefan gas diffusivities provided are
independent of composition.

Liquid Phase Diffusion Coefficients at Infinite Dilution

The following models are available for the diffusion coefficients at infinite dilution in
liquid phase:

• Wilke–Chang
• Tyn–Calus
• Hayduk–Minhas
• Siddiqi–Lucas
• Erkey–Rodden–Akgerman

Automatic When the Liquid diffusivity at infinite dilution property model is set to
Automatic, a selection for each solute i in solvent j will be made from the models
according to the following rules. Except in the case of a temperature correlation and
the Erkey–Rodden–Akgerman model, all of these rules require that the log liquid
viscosity correlation, ln j is available in the database for species j and that the liquid
volume at normal boiling point, Vi,l,b is available for the species i.

• If a temperature correlation is available for species i and j, it is used.

• If the solvent is water, the Siddiqi–Lucas correlation (for aqueous systems) is used.
• If the solute and solvent are both normal paraffins, the Hayduk–Minhas correlation
(for normal paraffins) is used.
• If the solvent is a normal paraffin, and the liquid density correlation for the solvent
is available, and the solute is hydrogen, carbon-monoxide or carbon-dioxide, the
Erkey–Rodden–Akgerman correlation (for normal paraffins) is used.

788 | CHAPTER 8: THERMODYNAMICS

• If parachors Pi and Pj are both available, and the liquid volume at normal boiling
point, Vi,l,b is available for the solvent, the Tyn–Calus correlation is used.
• If parachors Pi and Pj are both available, the Hayduk–Minhas correlation is used.
• If the solute and solvent are both organic molecules and the liquid volume at normal
boiling point, Vi,l,b is available for the solvent, the Siddiqi–Lucas method is used.
• If the molecular weight Mj is available for the solvent, the Wilke–Chang correlation
is used.
• If the solvent is a normal paraffin, and the liquid density correlation for the solvent
is available, and the solute is a normal paraffin, the Erkey–Rodden–Akgerman
correlation (for normal paraffins) is used.
• If the solvent is a normal paraffin, and the liquid density correlation for the solvent
is available, and the Lennard–Jones diameter for the solute is available, the Erkey–
Rodden–Akgerman correlation (for normal paraffins) is used.

For the Siddiqi–Lucas method, a molecule is considered organic if it has at least one C
atom bound to anything other than O or C atoms. This is determined by the SMILES
formula, if available. If, for any pair of species, the required input data for none of the
above models is available, the entire property liquid diffusion coefficient at infinite
dilution is not available.

Wilke–Chang

The correlation by Wilke and Chang (Ref. 62) for liquid phase diffusion coefficients at
infinite dilution is:

0 – 18 T j Mj
D i j = 1.858795959  10 --------------------
0.6
- (8-240)
 j V i l b

Molecular weight, Mj, and log liquid viscosity correlation, ln j, for species j and liquid
molar volume at normal boiling point for species i, Vi,l,b, is required. The Wilke–
Chang association parameter j, if unavailable, is set to:
TABLE 8-9: ASSOCIATION PARAMETERS FOR SOLVENT.

SOLVENT ASSOCIATION
PARAMETER

Water 2.26 (Ref. 63)

Methanol 1.9
Ethanol 1.5
Others 1

THERMODYNAMIC MODELS AND THEORY | 789

 Species are identified by their CAS number or SMILES formula. The Wilke–Chang
correlation is not suitable for diffusion of water. If water is the solute, the correction
suggested by Kooijman (Ref. 64) is applied where liquid molar volume of water at
normal boiling point, Vwater,l,b is multiplied by 4.5.

Tyn–Calus

The Correlation by Tyn and Calus (Ref. 65) for liquid diffusion coefficients at infinite
dilution reads:

1
 --6-  0.6
T  V i l b P j
0 – 16  
D i j = 8.93  10 ------------------------------------- (8-241)
1-
--
 3  0.6
 j  V j l b P i
 

The log liquid viscosity correlation ln j should be available for species j, The liquid
volume at normal boiling point, Vi,l,b and parachor, Pi should be available for both
species i and j.

If the solvent is nonpolar (dipole moment is zero), and the solvent is methanol, ethanol
or1-butanol or if the solvent is a mono-hydroxy alcohol, both the liquid volume at
normal boiling point, Vj,l,b and the parachor, Pj are corrected by a factor of 8 × 103j.

If the solute is water, both the liquid volume at normal boiling point, Vi,l,b and
parachor, Pi for the solute are corrected by a factor 2. This factor also applies if it is
detected from the SMILE formula that the species is an organic acid (a carboxyl group
is found), except in the cases where the solvent is water, methanol, or n-butanol.

Hayduk–Minhas

Hayduk and Minhas (Ref. 66) suggested three different correlations for liquid
diffusion coefficients at infinite dilution.

Aqueous Solutions: in case the solvent is water (derived from CAS number or SMILES
formula), the correlation reads:

790 | CHAPTER 8: THERMODYNAMICS

 0 – 12 1.52  0.07244359601 
D i j = 1.25  10 T  ----------------------------------------- - – 0.292 (8-242)
 0.19
V i l b 
 10
–6
 9.58
---------------------------- – 1.12
 V i l b 
3
   j  10 

Normal paraffin solutions: In case both the solute and solvent are normal paraffins
(derived from SMILES formula), the correlation reads:

10.2  10
–6
 ---------------------------- – 0.791
  
    10 3  Vi l b 
– 16 1.47  j
---------------------------------------------------------------------
0
D i j = 7.30889362  10 T (8-243)
 0.71 
 V i l b 
 

For all other systems, the correlation is:

1.29
0 – 16 T Pj
D i j = 3.8934239  10 ---------------------------------------------
0.92 0.23 0.42
- (8-244)
 j  V j l b P i

If the solvent is nonpolar (dipole moment is zero), and the solvent is methanol, ethanol
or1-butanol or if the solvent is a mono-hydroxy alcohol, both the liquid volume at
normal boiling point, Vj,l,b and the parachor, Pj are corrected by a factor of 8  103j.

If the solute is water, both the liquid volume at normal boiling point, Vi,l,b and
parachor, Pi for solute are corrected by a factor 2. This factor also applies if it is
detected from the SMILES formula that the species is an organic acid (a carboxyl
group is found), except in the cases where the solvent is water, methanol, or n-butanol.

Siddiqi–Lucas

Siddiqi and Lucas (Ref. 67) suggested correlations for liquid diffusion coefficients at
infinite dilution for aqueous system (including gases) and for organic solutions. For
normal paraffins systems, they recommended the Hayduk–Minhas model.

For aqueous solutions:

0 – 17 T
D i j = 1.295436901  10 --------------------------------------
1.026 0.5473
- (8-245)
j  V i l b 

For all other systems, the equation for organic solution is used as:

THERMODYNAMIC MODELS AND THEORY | 791

 0.265
0 – 15 T  V j l b 
D i j = 1.459473761  10 ----------------------------------
0.907 0.45
- (8-246)
j  V i l b 

Where T is the temperature, Vi,l,b is the liquid volume at normal boiling point and j
is the viscosity of the solvent.

Erkey–Rodden–Akgerman

The correlation by Erkey and others (Ref. 68) for liquid diffusion coefficients at
infinite dilution in normal paraffins is:

ref 6
0 –9 T  V – V j  10
D i j = 94.5  10 ------------------------------------------------------------------------------------ (8-247)
0.239 0.781 – 20 1.134
Mi Mj   i  j  10 

where the reference volume is

ref 1 3
Vj = ------- b j N Av  j (8-248)
2

NAv is Avogadro’s number, and the deviation from closest packing volume is given by

i
b j = 1.206 + 0.0632  ----- (8-249)
  j

The correlation is fitted to normal paraffins, hydrogen, carbon monoxide and carbon
dioxide diffusing in normal paraffins. The molecular weight Mi should be available for
both solute and solvent. The Lennard–Jones diameter, i is estimated from Bondi
group contribution method (Ref. 69), and for some species are taken from Ref. 68 and
Ref. 70 as:
TABLE 8-10: MOLECULAR DIAMETERS FOR SPECIES.
10
SPECIES   10 (M)

H2 2.92
CO 3.72
CO2 3.97
1/3
n-CjH2j+2 (21.82+32.44*j)

To prevent the diffusion coefficients from becoming negative, the minimum difference
of (V  Vjref) is considered to be 1012 mol/m3.

792 | CHAPTER 8: THERMODYNAMICS

Liquid Phase Maxwell–Stefan Diffusivity
The Maxwell–Stefan liquid diffusion coefficients are calculated from the liquid
diffusion coefficients at infinite dilution. The diagonal values should be ignored and
are set to zero. The Vignes (Ref. 71) model for diffusion in binary solutions can be
extended to multicomponent systems (Ref. 72):

 Di j x  1 
xk
D i j = k
(8-250)
k=1

where D i j x k  1 denotes the Maxwell–Stefan liquid diffusion coefficient for species i

and j in the limited of pure species k. If k = j, it represents a binary system of species i
and j where i is infinitely diluted. In addition, at infinite dilution the thermodynamics
factor (activity) (Ref. 57) becomes unity and the Maxwell–Stefan diffusivity equals to
Fick diffusivity:

0 0
D i j xj  1 = D i j = D i j (8-251)

Similarly,

0 0
D i j x i  1 = D j i = D j i (8-252)

Models for the D i j x k  1 for k  i and k  j remain to be defined in such a way that
symmetry is ensured:

D i j = D j i (8-253)

and to ensure continuity if both species i and j vanish:

lim D i j xk  1 = lim D i j x k  1 (8-254)

xi  0 xj  0

For binary systems, Equation 8-250 reduces to the Vignes interpolation formula.

Wesselingh–Krishna

Wesselingh and Krishna (Ref. 72) proposed:

0 0
D i j x k  1 = D i j D j i (8-255)

leading to:

THERMODYNAMIC MODELS AND THEORY | 793

 0  1 + xj – xi   2 0  1 + xi – xj   2
D i j =  D i j   D j i  (8-256)

Kooijman–Taylor

Kooijman and Taylor (Ref. 73) found on a limited number of systems that this
provides better results:

0 0
D i j xk  1 = D i k D j k  k  i j  (8-257)

leading to:

n
xj xi xk

0 0 0 0
D i j =  D i j   D j i  D i k D j k  (8-258)
k = 1
k  i j

Wesselingh and Bollen (Ref. 74) asserted that this is a reasonable estimate.

Krishna–van Baten

Krishna and van Baten (Ref. 75), on the basis of data obtained from molecular
dynamic simulations, proposed the following Vignes-based (Ref. 71) interpolation:

 --------------
xi 
-  --------------
xj 
-
 x i + x j  x i + x j
0 0
D i j x k  1 =  D i k   D j k   k  i j  (8-259)

The model reduces to the Kooijman–Taylor model for xi = xj. The value of D i j x k  1
is undefined in the limit of both xi  0 and xj  0. From a physical point of view this
is inconsequential as the value cancels out in the expressions obtained for the fluxes
using the Maxwell–Stefan equations. Nevertheless, in order to obtain well-defined
values and composition derivatives of D i j xk  1 itself, the equation is modified to:

xi + x 
 ----------------------------- xj + x 
 -----------------------------
 x i + x j + 2 x  x i + x j + 2 x
0 0
D i j x k  1 =  D i k   D j k   k  i j  (8-260)

where x  1010 is taken as a small composition. The limiting case for both xi  0 and
xj  0 also reduces to the Kooijman–Taylor model. This equation leads to:

794 | CHAPTER 8: THERMODYNAMICS

 n xi + x
 ----------------------------- xj + x
 ----------------------------- xk
  x + x + 2 x  x i + x j + 2 x 
0 xj 0 xi   D0  i j 

0
D i j =  D i j   D j i   D j k  (8-261)
 i k 
k = 1 
k  i j

Surface Tension
The available surface tension models are presented below.

GAS–LIQUID SURFACE TENSION

Ideal
The gas–liquid surface tension is predicted by ideally mixing the pure species
correlations. It is independent of pressure, vapor temperature or composition.

 xi l i vl
 vl = (8-262)
i

where the vapor–liquid surface tension correlations, i,vl, must be specified for all
species i, and are evaluated at the temperature of the liquid phase.

Winterfeld
Following Winterfeld (Ref. 76), the vapor–liquid surface tension is predicted by
mixing the pure species correlations according to

x i l x j l  i vl  j vl
  ---------------------------------------------
 i l  j l
-
 vl = -------------------------------------------------------------
i j - (8-263)
 x i l  2

 --------
  i l
i

It is independent of pressure, vapor temperature and composition. The vapor–liquid

surface tension correlations and liquid density correlations must be specified for all
species i, and are evaluated at the temperature of the liquid phase.

References
1. A. Akerberg, CFD analyses of the gas flow inside the vessel of a hot isostatic press,
Master of Science Thesis, KTH School of Industrial Engineering and Management,
Stockholm, Sweden, 2012.

THERMODYNAMIC MODELS AND THEORY | 795

 2. M.A. Trebble and P.R. Bishnoi, “Accuracy and consistency comparisons of ten cubic
equations of state for polar and non-polar compounds”, Fluid Phase Equilibria,
vol. 29, pp. 465–474, 1986.

3. B.E. Poling, J.M. Prausnitz, and J.P. O’Connell, The Properties of Gases and
Liquids, McGraw Hill, international edition, 2007.

4. M.C. Clapeyron, “Mémoire sur la puissance motrice de la chaleur,” Journal de

l’École polytechnique, vol. 23, pp. 153–190, 1834 (in French).

5. D.Y. Peng and D. Robinson, “A new two-constant equation of state,” Industrial

and Engineering Chemistry: Fundamentals, vol. 15, pp. 59–64, 1976.

6. C.H. Twu, J.E. Coon, and J.R. Cunningham, “A new generalized alpha function
for a cubic equation of state, Part 1, Peng–Robinson equation,” Fluid Phase
Equilibria, vol. 105, pp. 49–59, 1995.

7. O. Redlich and J.N.S. Kwong, “On the thermodynamics of solutions an equation

of state, fugacities of gaseous solutions,” Chem. Rev., vol. 44, no. 1, pp. 233–244,
1949.

8. G. Soave, “Equilibrium constants from a modified Redlich–Kwong equation of

state,” Chem. Eng. Sci., vol. 27, no. 6, pp. 1197–1203, 1972.

9. M.S. Graboski and T.E. Daubert, “A modified Soave equation of state for phase
equilibrium calculations, 3, systems containing hydrogen,” Ind. Eng. Chem. Process.
Des. Dev., vol. 18, no. 2, pp. 300–306, 1979.

10. J.R. Cooper, “Revised Release on the IAPWS Industrial Formulation 1997 for the
Thermodynamic Properties of Water and Steam”, The International Association for
the Properties of Water and Steam, IAPWS R7-97, 2012.

11. K. Watanabe, “Revised Supplementary Release on Backward Equations for the

Functions T(p,h), v(p,h) and T(p,s), v(p,s) for Region 3 of the IAPWS Industrial
Formulation 1997 for the Thermodynamic Properties of Water and Steam,” The
International Association for the Properties of Water and Steam, Kyoto, Japan, Sep.
2004.

12. F. Marsik, “Supplementary Release on Backward Equations for Specific Volume as

a Function of Pressure and Temperature v(p,T) for Region 3 of the IAPWS Industrial
Formulation 1997 for the Thermodynamics Properties of Water and Steam,” The
International Association for the Properties of Water and Steam, Santorini, Greece,
July 2005.

796 | CHAPTER 8: THERMODYNAMICS

13. K.C. Chao and J.D. Seader, “A general correlation of vapor-liquid equilibria in
hydrocarbon mixtures,” AIChE J., vol. 7, no. 4, pp. 598–605, 1961.

14. J.H. Hildebrand and R.L. Scott, “The solubility of non-electrolytes,” J. Phys.
Chem., vol. 55, no. 4, pp. 619–620, 1951.

15. H.G. Grayson and C.W. Streed, “Vapor-liquid equilibria for high temperature,
high pressure hydrogen-hydrocarbon systems,” 6th World Petroleum Congress, 19–
26 June, Frankfurt am Main, Germany, IV, pp. 169–180, 1963.

16. J.H. Hildebrand, J.M. Prausnitz, and R.L. Scott. Regular and Related Solutions,
Van Nostrand Reinhold Co., 1970.

17. G.M. Wilson, “Vapor-liquid equilibrium. xi. a new expression for the excess free
energy of mixing,” J. Am. Chem. Soc., vol. 86, pp. 127–130, 1964.

18. H. Renon and J.M. Prausnitz, “Local compositions in thermodynamic excess

functions for liquid mixtures,” AIChE J., vol. 14, no. 1, pp. 135–144, 1968.

19. D.S. Abrams and J.M. Prausnitz, “Statistical thermodynamics of liquid mixtures:
A new expression for the excess Gibbs energy of partly or completely miscible
systems,” AIChE J., vol. 21, no. 1, pp. 116–128, 1975.

20. A. Fredenslund, R.L. Jones, and J.M. Prausnitz, “Group-contribution estimation

of activity coefficients in nonideal liquid mixtures,” AIChE J., vol. 21, no. 6,
pp. 1086–1099, 1975.

21. S. Skjold-Joergensen, B. Kolbe, J. Gmehling, and P. Rasmussen, “Vapor-liquid

equilibria by unifac group contribution. revision and extension,” Ind. Eng. Chem.
Process. Des. Dev., vol. 18, no. 4, pp. 714–722, 1979.

22. J. Gmehling, P. Rasmussen, and A. Fredenslund, “Vapor-liquid equilibria by unifac

group contribution. revision and extension. 2,” Ind. Eng. Chem. Process. Des. Dev.,
vol. 21, no. 1, pp. 118–127, 1982.

23. E.A. Macedo, U. Weidlich, J. Gmehling, and P. Rasmussen, “Vapor-liquid

equilibria by unifac group contribution. 3. Revision and extension,” Ind. Eng. Chem.
Process. Des. Dev., vol. 22, no. 4, pp. 676–678, 1983.

24. D. Tiegs, J. Gmehling, P. Rasmussen P, and A. Fredenslund, “Vapor-liquid

equilibria by unifac group contribution. 4. Revision and extension,” Ind. Eng. Chem.
Process. Des. Dev., vol. 26, no. 1, pp. 159–161, 1987.

THERMODYNAMIC MODELS AND THEORY | 797

 25. H.K. Hansen, P. Rasmussen, A. Fredenslund, M. Schiller, and J. Gmehling,
“Vapor-liquid equilibria by unifac group contribution. 5. Revision and extension,”
Ind. Eng. Chem. Process. Des. Dev., vol. 30, no. 10, pp. 2352–2355, 1991.

26. R. Wittig, J. Lohmann, and J. Gmehling, “Vapor-liquid equilibria by unifac group

contribution. 6. Revision and extension,” Ind. Eng. Chem. Process. Des. Dev., vol. 42,
no. 1, pp. 183–188, 2003.

27. K. Balslev and J. Abildskov, “Unifac parameters for four new groups,” Ind. Eng.
Chem. Res., vol. 41, pp. 2047–2057, 2002.

28. R.W. Hankinson and G.H. Thomson, “A new correlation for saturated densities
of liquid and their mixtures,” AIChE J., vol. 25, no. 4, pp. 653–663, 1979.

29. H.G. Rackett, “Equation of state for saturated liquids,” J. Chem. Eng. Data,
vol. 15, no. 4, pp. 514–517, 1970.

30. L.I. Stiel and G. Thodos, “The thermal conductivity of nonpolar substances in the
dense gaseous and liquid regions,” AIChE J., vol. 10, no. 1, pp. 26–30, 1964.

31. M. Yorizane, S. Yoshimura, H. Masuoka, and H. Yoshida, “Thermal conductivities

of binary gas mixtures at high pressures: nitrogen-oxygen, nitrogen-argon, carbon
dioxide-argon, and carbon dioxide-methane,” Industrial & Engineering Chemistry
Fundamentals, vol. 22, no. 4, pp. 458–463, 1983.

32. A.L. Lindsay and L.A. Bromley, “Thermal conductivity of gas mixtures,”
Industrial and Engineering Chemistry, vol. 42, no. 8, pp. 1508–1511, 1950.

33. A. Wassiljewa, “Heat conduction in gas mixtures,” Physikalische Zeitschrift,

vol. 5, no. 22, pp. 737–742, 1904.

34. J.R. Cooper, “Release on the IAPS Formulation 1985 for the Thermal
Conductivity of Ordinary Water Substance,” The International Association for the
Properties of Water and Steam, Berlin, Germany, Sep. 2008.

35. C.R. Wilke, “A viscosity equation for gas mixtures,” J. Chem. Phys., vol. 18, no. 4,
pp. 517– 520, 1950.

36. R.S. Brokaw, “Approximate formulas for the viscosity and thermal conductivity of
gas mixtures. ii,” J. Chem. Phys., vol. 42, no. 4, pp. 1140–1147, 1965.

37. J.E. Lennard–Jones, “On the determination of molecular fields 1. from the
variation of the viscosity of a gas with temperature,” Proceedings of the Royal Society
of London. Series A, Containing Papers of a Mathematical and Physical Character,
vol. 106, pp. 441–462, 1924.

798 | CHAPTER 8: THERMODYNAMICS

38. W.H. Stockmayer, “Second virial coefficients of polar gases,” J. Chem. Phys.,
vol. 9, pp. 398–402, 1941.

39. F.M. Mourits and F.H.A. Rummens, “A critical evaluation of Lennard–Jones and
Stockmayer potential parameters and of some correlation methods,” Can. J. Chem.
Eng., vol. 55, pp. 3007–3020, 1977.

40. T.A. Davidson, “A Simple and Accurate Method for Calculating Viscosity of
Gaseous Mixtures”, Report of Investigations 9456, United States Department of the
Interior, Bureau of Mines, 1993.

41. J.A. Jossi, L.I. Stiel, and G. Thodos, “The viscosity of pure substances in the dense
gaseous and liquid phases,” AIChE J., vol. 8, no. 1, pp. 59–63, 1962.

42. K.S. Pedersen and P.L. Christensen, “Phase Behavior of Petroleum Reservoir
Fluids,” CRC Press/Taylor & Francis Group, 2007.

43. K.S. Pedersen, A. Fredenslund, P.L. Christensen, and P. Thomassen, “Viscosity of

crude oils,” Chem. Eng. Sci., vol. 39, no. 6, pp. 1011–1016, 1984.

44. H.J.M. Hanley, W.M. Haynes, and R.D. McCarty, “The viscosity and thermal
conductivity coefficients for dense gaseous and liquid methane,” J. Chem. Phys.,
vol. 6, no. 2, pp. 597–609, 1977.

45. K.S. Pedersen and A. Fredenslund, “An improved corresponding states model for
the prediction of oil and gas viscosities and thermal conductivities,” Chem. Eng. Sci.,
vol. 42, no. 1, pp. 182–186, 1987.

46. R.D. McCarty, “A modified Benedict-Webb-Rubin equation of state for methane

using recent experimental data,” Cryogenics, pp. 276–280, May 1974.

47. S. Murad and K.E. Gubbins, “Corresponding states correlation for thermal
conductivity of dense fluids,” Chem. Eng. Sci., vol. 32, no. 5, pp. 499–505, 1977.

48. J.R. Cooper, “Revised Release on the IAPWS Formulation 2008 for the Viscosity
of Ordinary Water Substance”, The International Association for the Properties of
Water and Steam, Berlin, Germany, Sep. 2008.

49. K. Daucik, “Revised Release on the Pressure along the Melting and Sublimation
Curves of Ordinary Water Substance”, The International Association for the
Properties of Water and Steam, Plzen, Czech Republic, Sep. 2011.

50. J. Gabitto and M. Barrufet, “Experimental and theoretical determination of heavy

oil viscosity under reservoir conditions,” Technical reports, Office of Fossil Energy,
USDOE Office of Fossil Energy, United States, 2002.

THERMODYNAMIC MODELS AND THEORY | 799

 51. A. Missenard, “Conductivité thermique des liquides organiques d’une serie ou
d’un ‘groupe de liquides’,” Revue Générale de Thermique, vol. 101, no. 5, pp. 649–
660, 1970 (in French).

52. R.L. Rowley, G.L. White, and M. Chiu, “Ternary liquid mixture thermal
conductivities,” Chem. Eng. Sci., vol. 43, no. 2, pp. 361–371, 1988.

53. R.L. Rowley, “A local composition model for multicomponent liquid mixture
thermal conductivities,” Chem. Eng. Sci., vol. 37, no. 6, pp. 897–904, 1982.

54. W.D. Niven, “The Scientific Papers of James Clerk Maxwell”, Cambridge
University Press, 1890.

55. J.C. Maxwell, “On the dynamical theory of gases,” Philos. Trans. R. Soc.,
vol. 157, pp. 49–88, 1867.

56. J. Stefan, “Uber das Gleichgewicht und die Bewegung, insbesondere die Diffusion
von Gasgemengen,” Sitzungsbrichte der Kaiserlichen Akademie der Wissenschaften
Wien, 2te Abteilung a, vol. 63, pp. 63–124, 1871.

57. R. Taylor and R. Krishna, “Multicomponent Mass Transfer,” Wiley-Interscience,

1993.

58. E.N. Fuller, P.D. Schettler and J.C. Giddings, “A new method for prediction of
binary gas-phase diffusion coefficients,” Industrial and Engineering Chemistry,
vol. 5, pp. 19–27, 1966.

59. E.N. Fuller, K. Ensley and J.C. Giddings, “Diffusion of halogenated hydrocarbons
in helium. The effect of structure on collision cross sections,” J. Phys. Chem., vol. 73,
no. 11, pp. 3679–3685, 1969.

60. C.R. Wilke and C.Y. Lee, “Estimation of diffusion coefficients for gases and
vapors,” Industrial and Engineering Chemistry, vol. 47, no. 6, pp. 1253–1257,
1955.

61. P.D. Neufeld, A.R. Janzen, and R.A. Aziz, “Empirical equations to calculate 16 of
 l s 
the transport collision integrals  for the Lennard–Jones (12-6) potential,”
J. Chem. Phys., vol. 57, pp. 1100–1102, 1972.

62. C.R. Wilke and P. Chang, “Correlation of diffusion coefficients in dilute

solutions,” AICHE J., vol. 1, no. 2, pp. 264–270, 1972.

63. W. Hayduk and H. Laudie, “Prediction of diffusion coefficients for nonelectrolytes

in dilute aqueous solutions,” AICHE J., vol. 20, no. 3, pp. 611–615, 1974.

800 | CHAPTER 8: THERMODYNAMICS

64. H.A. Kooijman, “A modification of the Stokes-Einstein equation for diffusivities
in dilute binary mixture,” Ind. Eng. Chem., vol. 41, pp. 3326–3328, 2002.

65. M.T. Tyn and W.F Calus, “Diffusion coefficients in dilute binary mixtures,”
J. Chem. Eng. Data, vol. 20, no. 1, pp. 106–109, 1975.

66. [Link] and [Link], “Correlation for prediction of molecular diffusivities

in liquids,” Can. J. Chem. Eng., vol. 60, no. 2, pp. 295–299, 1983.

67. M.A. Siddiqi and K. Lucas, “Correlations for prediction of diffusion in liquids”,
Can. J. Chem. Eng., vol. 64, pp. 839–843, 1986.

68. C. Erkey, J.B. Rodden and A. Akgerman, “A correlation for predicting diffusion
coefficients in alkanes”, Can. J. Chem. Eng., vol. 68, pp. 661–665, 1990.

69. A. Bondi, “van der Waals Volumes and Radii”, J. Phys. Chem., vol. 68, no. 3,
pp. 441-451, 1964.

70. J.O. Hirschfelder, C.F. Curtiss and R.B. Bird, “The Molecular Theory of Gases
and Liquids”, Wiley-Interscience, 1964.

71. A. Vignes, “Diffusion in Binary Solutions,” Ind. Eng. Chem. Fundam., vol. 5,
pp. 189–199, 1966.

72. J.A. Wesselingh and R. Krishna, “Mass Transfer”, Ellis Horwood Ltd, 1990,
ISBN: 0-13-553165-9.

73. H.A. Kooijman and R. Taylor, “Estimation of diffusion coefficients in

multicomponent liquid systems”, Ind. Eng. Chem., vol. 30, pp. 1217–1222, 1991.

74. J.A. Wesselingh and A.M. Bollen, “Multicomponent diffusivities from the free
volume theory”, Chem. Eng. Res. Des., vol. 75, no. 6, pp. 590–602, 1997.

75. R. Krishna and J.M. van Baten, “The Darken relation for multicomponent
diffusion in liquid mixtures of linear alkanes: An investigation using molecular
dynamics (MD) simulations,” Ind. Eng. Chem., vol. 44, pp. 6939–6947, 2005.

76. P.H. Winterfeld, L.E. Scriven, and H.T. Davis, “An approximate theory of
interfacial tension of multicomponent systems: Applications binary liquid-vapor
tensions,” AICHE J., vol. 24, no. 6, pp. 1010–1014, 1978

77. E.C. Carlson, “Don’t Gamble With Physical Properties For Simulation”, Chem.
Eng. Prog., vol. 92, no. 10, pp. 35–46, 1996.

THERMODYNAMIC MODELS AND THEORY | 801

802 | CHAPTER 8: THERMODYNAMICS
 9

Multiphysics Coupling Nodes

The Fuel Cell & Electrolyzer Module has multiphysics couplings available under
certain conditions.

When a predefined multiphysics interface is added from the Model Wizard or Add
Physics windows, it adds the constituent interfaces and the Multiphysics Couplings
node, which automatically includes one or more multiphysics couplings.

If the constituent physics interfaces are added one at a time, then it adds an empty
Multiphysics Couplings node. When you right-click this node, you can choose from
the available multiphysics couplings.

The Multiphysics Branch in the COMSOL Multiphysics Reference

Manual.

In this chapter, the following multiphysics couplings nodes are described:

• Electrochemical Heating • Space Charge Density Coupling

• Potential Coupling • Reacting Flow, H2/O2 Gas Phase

803
 Electrochemical Heating
Use the Electrochemical Heating multiphysics coupling ( ) to define domain and
boundary heat sources in a heat transfer interface, based on the sum of irreversible
(Joule heating and activation losses) and reversible heat in an electrochemistry
interface. The node also defines the temperature in the electrochemistry interface to
be equal to that of the heat transfer interface.

The settings of this node are similar to the Electromagnetic Heating node, described
in the COMSOL Multiphysics Reference Manual.

Theory for Electrochemical Heat Sources

Potential Coupling
The Potential Coupling multiphysics coupling ( ) applies the electrolyte potential
variable from the source interface into the model inputs of the destination interface.
The potential variable is typically used to compute the migration flux in the destination
interface.

COUPLED INTERFACES
Specify the physics interface that provides the potential (Electrostatics, for instance) in
the Source list and the Chemical Species Transport interface in the Destination list.

Space Charge Density Coupling

The Space Charge Density Coupling multiphysics coupling ( ) computes the local
space charge based on local concentrations and species charges in a Chemical Species
Transport interface. The computed space charge is then added as a contribution to the
Poisson equation in an Electrostatics interface. The charge is added in the same way as
a Space Charge Density feature does.

COUPLED INTERFACES
Specify the Chemical Species Transport interface in the Source list and the
Electrostatics interface in the Destination list.

804 | CHAPTER 9: MULTIPHYSICS COUPLING NODES

Reacting Flow, H2/O2 Gas Phase
The Reacting Flow, H2 Gas Phase and the Reacting Flow, O2 Gas Phase nodes couple a
fluid flow interface to a Hydrogen Fuel Cell (fc) or a Water Electrolyzer (we) interface.

The node sets up a two-way bidirectional coupling so that the velocity field and
pressure distribution, calculated by the fluid flow interface, are used by the fc/we
interface, whereas the density and dynamic viscosity, calculated by the fc/we interface,
are used by the fluid flow interface.

| 805
806 | CHAPTER 9: MULTIPHYSICS COUPLING NODES
 10

Glossary

This Glossary of Terms contains application-specific terms used in the Fuel Cell &
Electrolyzer Module software and documentation. For finite element modeling
terms, mathematical terms, and geometry and CAD terms, see the glossary in the
COMSOL Multiphysics Reference Manual. For references to more information
about a term, see the index.

807
 Glossary of Terms
anode An electrode where oxidation occurs. Anodic charge transfer current densities
and overpotentials are positive by definition.

cathode An electrode where reduction occurs. Cathodic charge transfer current

densities and overpotentials are negative by definition.

cell voltage The difference in potential over an electrochemical cell.

charge transfer reaction A reaction during which charge is transferred from an

electron conducting phase (electrode) to an ion conducting phase (electrolyte).

charge transfer current density The current density at an electrolyte-electrode

interface associated with a charge transfer reaction.

concentrated electrolyte An electrolyte where the concentration of charged species is

so high that the interactions (friction forces) between the ions needs to be considered.
See also diluted electrolyte.

current collector A part of an electrochemical device with high electric conductivity

for conducting current out of the cell. See also current feeder.

current feeder A part of an electrochemical device with high electric conductivity for
conducting current into the cell. See also current collector.

diluted electrolyte An electrolyte where the charged species are diluted in a solvent so
that the interaction between the ions can be neglected. See also concentrated
electrolyte.

electric potential The potential in the electrode (electron conducting) phase, denoted
s (phis).

electrode An electron conductor.

electrode reaction See charge transfer reaction.

electrokinetic flow Transport of fluid or charged particles within a fluid by means of

electric fields.

808 | CHAPTER 10: GLOSSARY

electrolyte An ion conductor.

electrolyte potential The potential in the electrolyte (ionic) phase, denoted l (phil).

electroneutrality A situation where the net charge of any small control volume of a
domain is zero. Electroneutrality can usually be assumed for electrolytes, with an
exception for the very proximity of the electrolyte-electrode interface, which is due to
the charge separation within the electrochemical double layer.

equilibrium potential The potential difference between the electrode and electrolyte
at which the net current density of an electrode reaction is zero.

Faraday’s Law A mathematical expression relating the flux of a species over an

electrode-electrolyte interphase to be proportional to the charge transfer current
density.

gas diffusion electrode (GDE) A porous electrode that also includes gas pores in the
porous matrix.

gas diffusion layer (GDL) A porous matrix containing an electron conducting media
and gas pores. A gas diffusion layer contains no electrolyte phase.

GDL See gas diffusion layer (GDL).

GDE See gas diffusion electrode (GDE).

Gibbs free energy A thermodynamic potential that depends on the enthalpy and
entropy of the reacting species.

intrinsic volume averages The physical properties of the fluid, such as density,
viscosity, and pressure.

intercalation The process of ions reacting on the surface of a host particle to form a
solid, intercalated, species within the host. The intercalated species can then be
transported further into the particle by diffusion processes.

migration The transport of charged species in an electrolyte due to the electric force
imposed by the electric field.

morphology correction parameter A correction parameter that is used to account for

morphology effects of a porous structure.

GLOSSARY OF TERMS | 809

 Nernst–Planck equation A commonly used equation for describing the transport of
charged species in a diluted electrolyte. The equation includes transport due to
diffusion, migration and convection. The Nernst–Planck equation can be modeled in
COMSOL Multiphysics using the Tertiary Current Distribution, Nernst–Planck
interface (and the Transport of Diluted Species interface). The Nernst–Planck
equation is often combined with an electroneutrality condition.

Nernst–Planck–Poisson equation In certain situations, for instance when striving to

resolve the electrochemical double layer at an electrode, and electroneutrality cannot
be assumed, the Nernst–Planck equation is combined with the Poisson’s equation for
describing the charge density.

overpotential A deviation from the equilibrium potential.

oxidation A charge transfer reaction where electrons are produced.

perfectly mixed electrolyte. An electrolyte where no concentration gradients are

present, for instance due to stirring. See also secondary current distribution.

primary current distribution The current distribution that is established when

overpotentials can be neglected — that is, when the current distribution is governed
by the electrolyte and electrode conductivities only. See also secondary current
distribution and tertiary current distribution.

Poisson’s equation An equation for relating the electric field to the space charge
density. Can be modeled in COMSOL using the Electrostatics interface.

pore electrolyte An electrolyte in the pores of a porous matrix. See porous electrode.

porous electrode A porous matrix containing both electron conducting and ion
conducting media.

potentiostat A device that monitors and controls the potential of an electrode versus
a reference electrode.

reduction A charge transfer reaction where electrons are consumed.

reference electrode An electrode used for reference when defining the potentials of
the electrodes in a cell. A good reference electrode is stable and does not vary in

810 | CHAPTER 10: GLOSSARY

potential. Typically no net reactions occur at a reference electrode surface, and the net
current density is zero.

resistive film A film with ohmic resistivity, typically formed as a result of electrode
reactions.

secondary current distribution The current distribution that is established when

concentration overpotentials can be neglected — that is, when the current distribution
is governed by the activation overpotentials and the electrolyte conductivity. See also
tertiary current distribution and primary current distribution. Note that COMSOL
models using the Secondary Current Distribution interface can in fact be modeling a
tertiary current distribution because mass transport activation losses can be present in
the electrode reaction current density expressions.

solid-electrolyte interphase (SEI) A resistive and passivating layer formed on the

graphite electrode particles in the negative electrode of a lithium-ion battery.

stoichiometric coefficients The number of species reacting in a reaction formula. In a

redox reaction the oxidized species have negative sign and the reduced species have
positive sign by convention.

superficial volume averages The flow velocities, which correspond to a unit volume of
the medium including both pores and matrix. These are sometimes called
Darcy velocities, defined as volume flow rates per unit cross section of the medium.

supporting electrolyte An electrolyte with an excess of charge carriers so that the

conductivity can be assumed not to be affected by the currents flowing in the cell.

surface molar flux The tangential flux in the surface dimension as governed by
diffusion according to Fick’s law.

tertiary current distribution The current distribution that is established when

concentration overpotentials cannot be neglected — that is, when the current
distribution is governed by the activation and concentration overpotentials, as well as
the electrolyte conductivity. The electrolyte conductivity can also be nonconstant due
to the currents flowing in the cell. See also secondary current distribution and primary
current distribution.

GLOSSARY OF TERMS | 811

812 | CHAPTER 10: GLOSSARY
I n d e x
A absolute pressure 437, 533, 550, 570, 589 surface reactions 349
activation overpotential 181, 188, 220 transport of concentrated species 287
added mass force 678 transport of diluted species 234, 341
adsorption 262, 273 boundary stress (node) 450
amperometric sensing 217 Brinkman equations 648
analyte 217 Brinkman equations interface 522
anodes 40 theory 648
anodic Tafel equation 191 bubble number density 660
anodic transfer coefficient 182 Butler–Volmer and Tafel expressions
Application Libraries window 31 tertiary current distribution 127
application library examples Butler–Volmer equation 49
convection and diffusion 239 electroanalysis 222
Darcy’s law 460 electrode kinetics 188
free and porous media flow 509 secondary current distribution 181
laminar flow 435
C CAPE-OPEN compliant packages 707
migration in electric field 239
cathodes 40
mixture model, laminar flow 568
cathodic charge transfer coefficient 182
storage model 503
cathodic Tafel equation 191
transport of concentrated species 286
CFL number
transport of diluted species 232
settings 434, 525, 547, 567, 586
Arrhenius parameters 319
charge transfer reaction 181
atmosphere/gauge (node) 497
charge transfer reactions 148
average linear velocity 640
checklist, for modeling 58
B Basset force 678 chemical equilibrium 316
boundary current source (node) 174 chronoamperometry 217
boundary electrolyte potential 88, 155, common settings 29
345 continuity equation, multiphase flow 665
boundary nodes contributing velocity (node)
Brinkman equations 525 Darcy’s law 481
bubbly flow 548 convection 383
Darcy’s law interface 460 coulometry 217
electrochemistry interfaces 138 creeping flow (spf) interface 429
electrode, shell 171 critical point 726
Euler–Euler interface 587 current (node) 344
free and porous media flow 509 current density (node) 344
mm interfaces 568 current source (node) 173

INDEX| 813
 cyclic voltammetry 217 electrode, shell 175
electric reference potential (node) 162
D Darcy velocity 383, 490
electroanalysis interface 176
Darcy’s law interface 458
theory 217
Darcy’s law interface theory 637
electroanalysis, definition 217
Debye length 44
electrochemical cells 40
dense flows 679
electrochemical impedance spectrosco-
diffusion models 282
py (EIS) 217
dispersed liquid droplets 676
electrochemical sensing 217
dispersed phase boundary conditions
electrode 180
668
electrode (node)
dispersed phase number density 667
electrochemistry 140
dispersed phase particles 674
electrode, shell 172
dispersed solid particles 677
electrode current (node) 90, 158
documentation 30
electrode current density (node) 90, 158
domain nodes
electrode current source (node) 140
Brinkman equations 525
electrode line current source (node) 124
bubbly flow 548
electrode points current source (node)
Darcy’s law 460
125
electrochemistry interfaces 138
electrode potential (node) 160
electrophoretic transport interface
electrode power (node) 91, 159
341
electrode reaction (node)
Euler–Euler interface 587
electrochemistry 148
free and porous media flow 509
electrode surface (node) 143
mm interfaces 568
electrode symmetry axis current source
transport of concentrated species 287
(node) 124
transport of diluted species 234
electrode, shell interface 170
double layer capacitance (node) 153
theory 215
drag force 678
electrode-electrolyte boundary interface
drag law, Hadamard–Rybczynski 671
(node)
E edge electrode (node) 123
current distribution 156
edge nodes
electrode-electrolyte interface coupling
Darcy’s law interface 460
(node)
electrochemistry interfaces 138
free and porous media flow 486, 516
electrode, shell 171
transport of diluted species 254, 296
surface reactions 349
electrode-electrolyte interfaces 40
electric insulation (node)
electrodes 39
electrode, shell 174
electrolysis 39
electric potential (node)

814 | I N D E X
electrolyte 180 Euler–Euler model, laminar flow (ee) in-
electrolyte (node) terface 584
primary and secondary current distri- theory 674
bution 119 external short circuit (node) 161
tertiary current distribution 132 external thermodynamic packages 707
electrolyte current (node) 155
F faradaic current 53
electrolyte current density (node) 89,
Faraday’s constant 182
156
Faraday’s law 379
electrolyte current source (node) 141
Faraday’s laws of electrolysis 221
electrolyte line current source (node)
Fick diffusion laws 219
124
Fick equations and diffusivities 395
electrolyte points current source (node)
Fick’s law approximation diffusion 401
125
Fick’s law diffusion model 283
electrolyte potential 183
flash calculation
electrolyte potential (node) 88, 155
retrograde solution type 726
electrolyte symmetry axis current
solution type 726–727
source (node) 125
flash calculations 724
electrolyte-electrode domain interface
flow continuity (node) 454, 562, 599
(node) 154
flow rate in SCCMs 445
electrolytes 39
fluid flow
electromagnetic heat source (multiphys-
Brinkman equations theory 648
ics coupling) 804
fluid properties (node) 510
electroneutrality 44
bf interfaces 549
electrophoretic transport interface 335
fluid-solid mixtures 676
elevation 484
flux (node) 276, 346
emailing COMSOL 32
transport of concentrated species 301
EnthalpyF 726
transport of diluted species 276, 346
EntropyF 726
flux discontinuity (node) 246
Eötvös number 661
Darcy’s law 493
equilibrium potential 181
transport of concentrated species 306
Equilibrium Reaction
transport of diluted species 246
theory for 373
fracture flow (node) 500
equilibrium reaction (node) 251, 307
free and porous media flow interface 507
equilibrium reaction group (node) 330
theory 644
Ergun packed bed expression 679
Freundlich exponent 389
Ettehadieh solid pressure model 683
fully developed flow 445
Euler–Euler equations 683
G galvanic cells 39
bubbly flow 657
gas boundary conditions 562

INDEX| 815
 gas constant 423 primary and secondary current distri-
gas diffusion 395 bution 120
gas reacting fluids 311, 325 single-phase, laminar flow 438
general stress (boundary stress condi- spf interfaces 438
tion) 451 surface reactions 351
Gibbs free energy 321 tertiary current distribution 135
Gidaspow and Ettehadieh solid pressure transport of concentrated species 299
model 683 transport of diluted species 240
Gidaspow model 679 inlet (node) 442
Gidaspow solid pressure model 683 bf interfaces 556
gravity 457 ee interface 596
gravity (node) mm interfaces 578
bf interfaces 553 single-phase flow 442
ee interfaces 593 insulation (node) 88, 142, 345
mm interfaces 573 interior wall (node)
ground (node) 174 bf interfaces 560
ee interface 598
H H2 inlet (node)
mm interfaces 581
hydrogen fuel cell 92
internet resources 30
Hadamard–Rybczynski drag law 671
interphase momentum transfer 678
Haider–Levenspiel slip model 671
intrinsic volume averages 648
harmonic perturbation (node) 160
irreversible reaction 316
Henry’s law 552
heterogeneous rate constant 222 K knowledge base, COMSOL 33
hydraulic head (node) 497 Knudsen diffusion 295
Hygroscopic Swelling 265 Krieger type model 677
Krieger type viscosity model 572
I implementing, Euler–Euler equations 683
inflow (node) 243, 347 L laminar bubbly flow (bf) interface 545
transport of concentrated species 302 theory 657
initial values (node) laminar flow interface 430
bf interfaces 554 laminar mixture model (mm) interface
Brinkman equations 532 theory 664
corroding electrode surface node 161 Langmuir constant 389
Darcy’s law 482 leaking wall, wall boundary condition
depositing electrode surface node 161 440, 517
ee interface 592 lift force 678
electrode, shell 173 line mass source (node)
free and porous media flow 515 fluid flow 456
mm interfaces 574 species transport 249

816 | I N D E X
 line source Nernst–Planck equations 43, 178
species transport 377 Nernst–Planck–Poisson equations 44
liquid reacting fluids 311, 325 Nernst–Planck–Poisson Equations inter-
local face 333
CFL number 434, 525, 547, 567, 586 no flow (node) 492
no flux (node) 242
M mass action law, laminar flow 422
hydrogen fuel cell 91
mass based concentrations (node) 240
transport of concentrated species 303
mass fraction (node) 300
no slip, wall boundary condition 439, 517
mass source (node)
no viscous stress (open boundary) 450
Brinkman equations 531
nodes, common settings 29
Darcy’s law 480
nonconservative formulations 375
mass transfer (node)
non-faradaic reactions (node) 162
bf interfaces 552
normal current density (node) 174
ee interfaces 593
normal stress, normal flow (boundary
mm interfaces 572
stress condition) 451
mass transport 394
Maxwell–Stefan diffusion model 282 O O2 inlet (node)
Maxwell–Stefan diffusivity matrix 398 hydrogen fuel cell 92
mixture continuity (node) 583 Ohm’s law 116, 180, 380
mixture model, laminar flow (mm) inter- ohmic drop 218
face 564 open boundary (node)
mixture properties (node) 569 single-phase flow 449
mixture viscosity 571 spf interfaces 449
mixture-averaged diffusion model 283, transport of concentrated species 307
399 transport of diluted species 250
modeling checklist 58 outflow (node)
momentum balance equations 675 electrophoretic transport 347
monolayer adsorption 413 transport of concentrated species 305
MPH-files 31 transport of diluted species 243
multicomponent diffusion 397 outlet (node) 446
multiphase flash calculations 708 bf interfaces 558
multiphase flow theory 657, 664 hydrogen fuel cell 93
multiphysics couplings mm interfaces 579
electromagnetic heat source (node) single-phase flow 446
804 spf interfaces 446
overpotential 48, 181
N Nernst equation 222
Nernst–Einstein relation 238, 257, 261, P pair nodes
379 Brinkman equations 525

INDEX| 817
 Darcy’s law interface 460 free and porous media flow 485, 515
electrochemistry interfaces 138 porous electrode reaction (node) 141
electrode, shell 171 porous electrodes 119, 181
electrophoretic transport interface porous matrix double layer capacitance
341 (node) 142
Euler–Euler interface 587 porous media and subsurface flow
free and porous media flow 509 Brinkman equations interface 522
surface reactions 349 Darcy’s law interface 458
transport of concentrated species 287 free and porous media flow interface
transport of diluted species 234 507
partially saturated porous media (node) theory, Brinkman equations 648
258 theory, free and porous media flow
periodic condition (node) 97, 122 644
transport of diluted species 247 porous media transport properties
periodic flow condition (node) 452, 559, (node) 255, 258, 292
581 Porous Medium (node)
permeability models 467, 477, 505 Brinkman equations 511, 526
persistence potential (node) 345
of thermodynamic packages 711 potentiometry 217
pervious layer (node) 498 potentiostat 225
phase properties (node) 588 pressure (node) 487, 491, 494
physics interfaces, common settings 29 pressure head (node) 496
point mass source (node) pressure point constraint (node) 455
fluid flow 456 primary current distribution 41
species transport 249 primary current distribution interface
point nodes 116
Brinkman equations 525 theory 177
Darcy’s law interface 460 pseudo time step 420
electrochemistry interfaces 138 pseudo time stepping
electrode, shell 171 settings 434, 525, 547, 567, 586
Euler–Euler interface 587
R reacting fluids, gases and liquids 311, 325
free and porous media flow 509
Reacting Volume 241, 278
surface reactions 349
reaction (node) 315
transport of diluted species 234
reaction coefficients (node) 254
point source
reaction thermodynamics (node) 331
species transport 376
reactions (node)
porous electrode (node) 120, 133
surface reactions 351
porous electrode coupling (node) 253
tertiary current distribution 134

818 | I N D E X
 transport of concentrated species 299 445
transport of diluted species 241 standard flow rate 445
reference electrode (node) 162 standard settings 29
resistive film 187 stoichiometric coefficients 321
retardation factor 389 Stokes equations 429
reversible reaction 316 Stokes flow 429
reversible reaction group (node) 328 stratified porous media 387
Reynolds number 358 superficial volume average, porous me-
slip velocity models 671 dia 648
superficial volume averages, porous me-
S SCCM, flow rate in 445
dia 383
Schiller–Naumann slip model 671
supporting electrolyte 217
Schmidt number 670, 686, 688
Supporting Electrolytes 379
secondary current distribution 41
surface concentration (node) 352
secondary current distribution interface
surface equilibrium reaction (node) 252,
116
308
theory 177
surface properties (node) 350
SEMI standard E12-0303 444
surface reactions interface 348
separator (node)
theory 412
current distribution 134
symmetry (node) 449
single-phase flow interface
bf interfaces 559
laminar flow 430
Darcy’s law 492
sliding wall 441, 577, 595
electrochemistry 143
slip model
mm interfaces 580
Hadamard–Rybczynski 671
spf interfaces 449
Schiller–Naumann 671
transport of concentrated species 306
slip model theory 660
transport of diluted species 246
Haider–Levenspiel 671
Reynolds number 671 T Tafel equation 191
slip, wall boundary condition 440, 517 Tafel law 49
solid pressure and particle interaction technical support, COMSOL 32
682 tertiary current distribution 41
Soret effect 402 tertiary current distribution interface
species (node) 323 theory 177
species activity (node) 331 tertiary current distribution, Nernst–
species group (node) 331 Planck interface 127
species source (node) 345 theory
species thermodynamics (node) 332 bf interfaces 657
standard cubic centimeters per minute Brinkman equations 648

INDEX| 819
 electroanalysis interface 217 228, 270
electrode, shell interface 215 theory 371
Euler–Euler model, laminar flow 674 turbulent bubbly flow (bf) interface
free and porous media flow 644 theory 657
mm interfaces 664 turbulent mixture model (mm) interface
primary current distribution interface theory 664
177 turbulent Prandtl number 358
secondary current distribution inter- two-fluid Euler–Euler model
face 177 bubbly flow 657
surface reactions 412
U Ungarish 673
tertiary current distribution interface
V viscous drag, coefficient 660
177
voltammetry, cyclic 217
transport of concentrated species in-
volume averages 648
terface 394
volume force (node) 438, 531
transport of diluted species in porous
free and porous media flow 514
media interface 372
spf interfaces 438
transport of diluted species interface
371 W wall (node)
thermal diffusion 402 bf interfaces 554
thermodynamics package 707 ee interface 594
adding 708 free and porous media flow 516
thickness mm interfaces 575
fracture 268 single-phase flow 439
out-of-plane 485 websites, COMSOL 33
thin diffusion barrier (node) 250 Wen and Yu fluidized state expression
Thin Impermeable Barrier 251 679
tortuosity factors 385
traction boundary conditions 450
transport mechanisms 289, 295
transport of concentrated species in po-
rous media interface 286
transport of concentrated species inter-
face 280
theory 394
transport of diluted species in porous
media interface 232
theory 372
transport of diluted species interface

820 | I N D E X