1

Chapter 2 FLOW OF FLUIDS THROUGH

SOLID POROUS

I- INTRODUCTION

Porous materials are generally the site of transfer processes.

matter, heat, quantity of motion and sometimes chemical reactions
with the fluids that pass through them.

These phenomena can cause structural changes and

textural of porous solids.

Due to the structural complexity of porous solids, the flow of a

fluid through a bed of particles has mainly been the subject of experimental studies that
We arrived at empirical correlations; these correlations are significant for the engineer.
valuable tools for reasonably estimating the pressure drops experienced by the
fluid and the transfer coefficients between fluids and solid particles.

II-FLOW OF A SINGLE FLUID THROUGH A MEDIUM

POREUX WITH INVARIABLE TEXTURE

II-1- REMINDER ON DARCY'S AND HAGEN'S EXPERIMENTAL LAWS

POISEUILLE

2
 Darcy states that the average speed of a fluid with viscosity (η) flowing in
laminar flow along a uniform straight pipe section is proportional
they kept pressure.

0∆
Um=
η

0
Or: proportionality coefficient
η

B0represents the specific permeability of the porous medium.

1cm

∆ = 2− 1= 1

P1 P2
- The Darcy represents the unit of permeability. This unit (Darcy) is the
permeability of a 1 cm cube side traversed by a fluid with viscosity
η = 1 cp, with a flow rate Q = 1 cm3and under a differential pressure∆ =
1 , the sectionΩ = 1cm²
m= 1cm/s
{
B 0=
Δ
Where B01 Darcy = 0.987 10-8cm²

Furthermore, the law of Poiseuille gives the pressure drop of a fluid.

viscosity (η) also flowing in laminar regime along a pipe
of uniform straight section.

Δ = 32 2

D: diameter of the pipe

Y: circulation factor of the straight section

3
II-2- KOZENY MODEL

II-2-1- statement of the model

Kozeny assimilates the porous volume to a bundle of cylindrical pores.

diameter (dp) and of length (ZpThese pores are supposed to be identical and
independent of each other, that is to say, not interconnected.

Zp= Z.T

Cos = Z

1
T=
Cos

Let A: total internal surface of the solid contained in the volume V.

Porous solid⟹number of pores of height Zp

A = NP.
↑
Internal surface of a cylinder

Pore volume Vp
2
Vp= Np.πZp= V
4
Porosity

Let's define the volumetric surface area of the layer.

ac =

total internal surface

V: volume of the barrel (column)

4
Volume of air occupied by the grains

ag=
V(1− )

4
= 2
=
4

4
⟹ =

1 4
= where =
(1− ) (1− )

Number of pores Np
4
= = 2

2
Knowing that V = Ω.Z =
4
Section of the trunk

2
4 4
2
= 2
= 16g²
2 (1− )²

2 (1− )² ²
=
16
Pore speed

Q=Ω =
Fluid flow through the barrel

2 2 (1− )² ² 2
= =
4 16 4
2( )²
1− 16 ²
= 2 ( 1− ⟹ = ⟹ =
16 )²
T: tortuosity factor

5
II-2-2 case of the laminar regime

Let's apply the Poiseuille equation to the pores (since they are)
cylindrical conductors

1
∆A = 32 2

Circular driving⟹ = 1( )

1
∆A = 32 16g²
2 (1− )²

2 (1− )²
=2 ² 3

∆ (1−
2
)²
=2 ² 3

∆ 2(1− )² Kozeny-Carman equation

=ℎ 3

Withℎ = 2 ²Kozeny constant

ℎ = 4,5 ± 15

For stackings of isometric grains (of the same dimensions) in a

cylindrical drive, usually given ℎ =5

Porosity increases with grain size.

Porosity increases if the grains are of different shapes.
If the observed values for the Kozeny constant are markedly
greater than 5, therefore the model ceases to be valid

II-2-2-1-Expression of the Kozeny-Carman equation in expression form

undimensional

6
Reminder:

Coefficient of friction

f= 2

τ: work of the shear tangential constraint

U2mkinetic energy
Reynolds number

pressure drop
4
∆ =

It is the friction that causes the drop in pressure.

4 2
∆ =

Let's apply this to the pores:

And

= Δ 2
4
4
=
(1− )
4
=
(1− )
Δ 4 1
=
(1− )4 2 ²
²
Δ 3
= Kozeny - Carman equation
(1− ) 3 2

Δ ( 1− )²
=2 ² 2
3

7
 Δ 3

2 (1−
=2 ²
)²
Δ 3 1 1
3 2
=2
(1− )² (1− )
4
=8
(1− )
. = 8y
= ( ) = ( )
Modified Reynolds
′ = ⟹ ′ =
4 (1− )

Δ 3
′=
(1− ) 2

′
4T. ′ =8y
3

′ ′= 2= ′ ′=
Kozeny-Carman equation
∗ 2 ℎ ⇒ ∗ ℎ still modified

II-2-2-2-Expression of the Kozeny-Carman equation as a function of the diameter of the

grain

When the porous layer is made up of a stack of grains, one can

Express the Kozeny-Carman equation in terms of the grain diameter.

2 6
= = =
9
6

Δ ( 1− )²
= 36ℎ 2

8
 64 1−
= =
4 1−

= 6(1− ) ′

′ =
4

Δ 3
= 2
6(1− ) 3

′= 3 ⟹

′∗ = 6ℎ (1− )

II-2-2-3- Expression of Darcy's permeability

1 1
0= Δ ⁄
= (1− )2
ℎ 2
3

3
0=
ℎ 2 (1− )2

II-2-3 cases of turbulent regime

It is known from the Reynolds expression that in a straight section of a

cylindrical flow, the laminar regime begins to be disturbed when Re reached
2000 and we also know that the intermediate regime is established at = 4000, can-
Shall we continue the analogy and apply it to the pores of the Kozeny model?

Knowing that the pores of this model actually make up a whole distribution
small flaws, large fissures, big openings and waves of singularities
In reality, these passages of the fluid are the seat of whirlpools and eddies even in
low flow and in a non-uniform manner.

9
 As the flow rate increases, an increasingly larger part of the porous volume
operates in a turbulent regime, and experimentation shows that the transition from
the transition from laminar flow to turbulent flow is not abrupt but rather very

progressive. LEVA in tracing ′ = ( ′ ) shows that the turbulent regime is established

from ′ ≥ 0,2.

So at ′ ≥ 0,2we have a turbulent regime by compiling the results

experimental results observed by various authors in turbulent regimes, Carman proposes the
following correlation:

′ −0,1
′= 5R ′ −1 + 0,4

Relation of Carman⟹ turbulent regime

II-3- MODEL OF THE TORTUOUS CYLINDRICAL PORE BEAM

II-3-1 description of the model

≥1

In the Kozeny model, we called ⁄ = tortuosity factor

but in reality this term is very poorly chosen since it was assumed that the pore was
straight and did not have any real twists (elbows, singularities) in
which of the kinetic energy would be lost in the pores.

10
 A more advanced model assumes that the axis of each pore
Cylindrical is a broken line whose segment length is on the order of
the length of the grain diameter.

Each elbow is the seat of a swirl in which part of the energy

the fluid's kinetics is converted into heat, in addition to the pressure drop by
viscous flow on the walls of the pore.

II-3-2 Burke Plummer equation

1 2
∆ = ∗ ∗2

Pressure loss⟹ due to a loss of kinetic energy

Number of bends (energy loss at the bends)

average fluid velocity through the pore

B: coefficient of proportionality

Each elbow is a segment of length

1 2
Δ = ∗ ∗
2

. . 2 ²
Δ = 4 2 ²
(1− )

3 (1− )
Δ = 2
8λ 3

3
With
8
=ℎ = −

11
 Δ (1− ) 2
=ℎ 3

II-3-3 Ergun equation

∆ (1− )²
2
(1− ) 2
=ℎ 3
+ℎ 3

Applied in laminar flow and even for an intermediate regime

II-3-3-1-Expression of the Ergun equation in dimensionless form

∆ 3 (1− )
2
=ℎ +ℎ
(1− )

′= ℎ
⁄ ′ +ℎ

In general, we take

′= 5R ′ −1 + ℎ

II-4-MODELS OF REGULAR ARRAYS OF IDENTICAL SPHERES

Consider a spherical obstacle placed in a uniform flow, this

the obstacle is spherical and is subjected to a force called drag which in
laminar flow is equal to:

F = 3πηU Stock Relations

1
= 2
= 2

Or work of constraint

12
 1 3πhU 1 12
= 2
= 2 2
=
4

12
=

→
T

∆
= ∗

Ncnumber of obstacles per unit volume

∆
pressure drop per unit volume

F: force per unit volume

1(1− ) Volume occupied by the grains

= 3
6

V porous volume

V(1- volume occupied by the grains

1(1− ) 6(1− )
= 3 = 3 =
6
↓
Number of obstacles

∆ 6(1− )
= 3
∗ 3πηU

∆ 1
= 18ηU (1− ) 2
⟶ Laminar flow

This relationship is valid only in laminar flow and is valid only if the
spheres are sufficiently distant from each other.

13
II-5- CONCLUSION ON THE MODELS

II-5-1 Cases of laminar flow regime

the tortuous pore beam model (Ergun) leads to a remarkable

in accordance with the expression in the case of a stacking of isometric grains of
sphericity close to 1 and porosity close to 0.7.

On the other hand, the layers of fibers that are often more or less intertwined
porosities between 0.7 and 0.99, these layers can be represented by the
latest swarm model that also gives good results in accordance with
the experience.

II-5-2 Cases of turbulent regime

Only the pore beam model that leads to a semi-empirical relationship

that of Burke Plummer and an empirical relationship, that of Carman.

III- FILTRATION

III-1- INTRODUCTION

.. . . … . . ..
P1 ⟶ Suspension
… . . .. …. ..
∆ ⟶ Cake (canvas, filter)

P2 ⟶ Filtrate

∆ =∆ +∆
∆ Pressure drop due to the cake
∆ Pressure drop due to the support

Filtration is a mechanical separation technique that allows for the removal.

solid particles in suspension or in a fluid constitute the continuous phase and
the solid constitutes the dispersed medium.

The main element of this process consists either of a porous medium and
in this case we will talk about mass or depth filtration, either through a filter, such as
what cloth, membrane and we will talk in this case about cake filtration, through which
the fluid can flow, while the solid particles of equivalent diameter

14
higher than the opening of porous media constitute the cake, so the cake
will serve as 2themefilter.

III-2- DESCRIPTION OF THE CAKE FILTRATION PROCESS

Suspension

Filter Framework

Once the thickness of the cake = the cakes

2
come together, the liquid

cannot pass, so the pressure increases, so we will say that we have reached the maximum
From filtration, we remove the cake and restart the filtration.

We can determine the filtration time of the suspension based on the

filtering surface (Σ )based on density, on the specific mass of the
suspension, depending on temperature and pressure.

III-3- DEFINITION AND REMINDERS

III-3-1 Permeability

Permeability 0 is defined as the ease of flow of a fluid

through a porous medium of length Z

0=

dR: being the resistance of the layer to flow per unit area

Specific resistance of the cake

The resistance dR of the layer dZ to the flow of a fluid per unit of

surface is proportional to the mass of the cake per unit area.

=
Ω
MGcake mass

Ω Filtering surface

15
 Coefficient of proportionality

=
⁄Ω

: actually represents the resistance of the cake of thickness Z (or resistivity by

unit of mass.

III-4-EXPRESSION OF THE PRESSURE DROP FROM THE EQUATION OF

DARCY

=−
0 = Ω=

=− =− =
Ωdt
Ω Ω
and

=−
Ω² =

If we call the mass of the cake deposited per unit volume of filtrate

= ⁄
{ ⟹ =−
= Ω²

And as = and
Ω
= ⟹

Δ =− ⟹− Δ =
Ω²
By analogy for the support

−Δ =
With Rs : support resistance to flow per unit area

−Δ = − (Δ +Δ =) ( + )

III-4-1 Constant pressure filtration

Δ =

16
 1
=
Ω
1
−Δ = 1− 2= ( + )
Ω

We replace = Ω

1
Δ =( + ) = +
Ω Ω Ω² Ω

= +
Δ Ω² ΩΔ

= ²+
2Ω²Δ ΩΔ

⁄ = +
ΩΔ 2Ω²Δ

By plotting based on the volume of the filtrate V, we should end up with a straight line.

experimental from which we could deduce values of:

Rsfrom the y-intercept

- from the slope of the line hence RG

III-4-2 Constant flow filtration

= ⟹Δ =(+ )
Ω

Like= =Ω

Δ = 2 +

If the cake is compressible the factor varies withΔ .

III-5- EXPRESSION OF PRESSURE DROP FROM THE MODEL OF

KOZENY

Δ = Δ 1+ Δ 2 Δ 1 Δ 2 ℎ

17
 à cake

According to Kozeny-Carman

( 1− 1²) ′
Δ 1= 36ℎ 3 2 1= = 1
1

( 1− 2²)
Δ 2= 36ℎ ′ 3 2 2= = ( 2= )
2 Ω

′
Δ = 1+
Ω

1= ∶

∶
.
= = { ∶
mass volume of cake = 1− + () )

′ .
Δ = +
Ω Ω
′
= ′ ⟹ ′ =
1= =
Ω

.
Δ = +
Ω Ω

1
Δ = +
Ω Ω²
III-5-1A constant pressure

= ²+
2Ω2 Δ ΩΔ

⁄ = +
2Ω2 Δ ΩΔ
III-5-2 Filtration at constant flow rate

Δ = + .
Ω Ω²

18
Q = constant⟹ Um constant speed (= ; = )
Ω Ω

Δ = +

Δ = + 2

= ⟹ Δ ( )=

If the cake is compressible⟹ = (Δ )

19
20