New Insights on the Seismic Response of

Rigid Inclusions considering Strength and

Stiffness Degradation of Grout and
Improved Numerical Boundary Conditions
WS 5
RECENT ADVANCES Armin W. Stuedlein, PhD, P.E., [Link] | Oregon State University
IN RIGID INCLUSIONS Matthew D. Gibson, PhD, P.E. | Clarity Engineering, LLC

ORGANIZATION
TC211 | GROUND IMPROVEMENT
 Acknowledgments
• Funding source: Deep Foundations Institute (DFI)
• Phase I: concluded
– Improve understanding of basic seismic mechanisms
– Conduct studies using traditional unit cell conditions
– Effect of steel reinforcements, layer thicknesses, ground motion variability
– Focus on kinematic soil-rigid inclusion-interaction
• Phase II: on-going
– Expand domain to consider realistic boundary conditions
– Effect of steel reinforcements, etc.
– Study post-shaking load bearing performance
– Propose simplified procedure for flexural design
• Special thanks to Menard for sponsoring this talk
2
 Overview of this Presentation
• Motivation for this work and review of prior art
• Critical aspects of the 3D numerical simulations
• Comparison of RI performance considering the Mohr-T and Concrete
constitutive models under monotonic and cyclic loading
• 3D numerical simulations of rigid inclusion (RI) ground improvement
– Ground motions, free-field simulations and “simple” unit cell results
– Comparative performance of RIs: Mohr-T vs. Concrete Model
– Expanded “unit” cell simulations: RI Groups under realistic boundary conditions
– Post-shaking footing performance under RI Groups: virtual footing loading tests
• Concluding remarks and research needs
3
 Motivation: To Advance our Understanding
• Seismic soil-rigid inclusion-interaction
– Fundamental questions about kinematic interaction remain
– Seismic load transfer mechanisms (static is well understood)
• How do RI’s “mitigate” liquefaction & poor soils?
– Increase (or decrease) liquefaction factor of safety?
– Reduce intolerable settlement of supported structures?
– Modify surface site response and seismic loads to structures?
• Seismic design guidelines for rigid inclusions:
– Maintain vertical support of structures before, during, & after EQ
– Design for flexural demand

4
 Prior Art and Interpretation
of Rigid Inclusion Performance

5
 Prior Art and Interpretation of Improvement
Cyclic Shear Stress Redistribution (CSSR)
• Baez (1995):
– Introduced the concept of cyclic shear stress
redistribution for stone columns
– Composite section with shear strain compatibility
(SSC) assumption
– Cyclic shear stress redistribution: ground
improvement attracts dynamic shear stresses away
from liquefiable soils to reduce demand
– Ratio of shear modulus of stone column and soil
controls CSSR
• Extended to high modulus columns and rigid
6
inclusions Baez (1995)
 Prior Art and Interpretation of Improvement
Cyclic Shear Stress Redistribution (CSSR)
• Rayamajhi et al. (2014), work
conducted at OSU & UC Davis:
– Unit cell numerical framework now
widely adapted in many studies
– Confirmed that pure SSC cannot
occur
– Focused on quantifying CSSR for
assessments of liquefaction
mitigation (within the Simplified
Method)

7 Rayamajhi et al. (2016)

 Prior Art and Interpretation of Improvement
Framework for Interpretation of CSSR
• Rayamajhi et al. (2014; 2016):
– Interpretation within the Simplified Method
– Cyclic stress ratio (loading) of unimproved (U) ground:
 v amax,U
CSRU  0.65    rd ,U  = tcyc K
MSF /’v0
v g
– Cyclic stress ratio of improved (I) ground: Note that the effect
 v amax, I of Arching is not
CSRI  0.65    rd , I  MSF  K  considered here
 v g
– CSRI / CSRU = the CSR reduction factor due to reinforcement:
amax, I  rd , I
RCSR   Ra max  Rrd
amax,U  rd ,U
8
Prior Art and Interpretation of Improvement
Framework for Interpretation of CSSR Rayamajhi et al. (2014)

• RCSR: Ratio of soil shear stress (calculated)

• Ramax: Ratio of soil acc. (calculated)
• Rrd = RCSR / Ramax: a catch all parameter that
– Captures all SSI and site response effects
– Correlates to gr, Gr, and ar
Gr = ratio of shear modulus in RI / shear modulus of soil, GRI / Gsoil
gr = ratio of shear strain in RI / shear strain in soil, gRI / gsoil = 1 under SSC
ar = area replacement ratio, CG = 1.0 for circular elements ( = Ar)

𝑟𝑑
𝑟 𝑟 𝑟 𝐺 𝑟
𝑟

• Rrd trends confirmed by Demir & Ozener (2020) and

Baez (1995)
9 Wang, Stuedlein, et al. (2021)
 Prior Art and Interpretation of Improvement
Physical Model and Full-Scale Observations Soil-cement columns (typ.)

• Centrifuge tests (Rayamajhi et al. 2015)

– Liquefaction not prevented
– Settlement of columns << soil
• Blast-liquefaction tests (Gianella & Stuedlein 2017)
– Excess pore pressures indicative of shear
strains >> under SSC assumption
– Piles bearing into competent layer settled
1/4 that of floating piles
• Liquefaction not significantly reduced,
but RI is strongly beneficial to structure
10 Rayamajhi et al. (2015)
 Prior Art and Interpretation of Improvement
Free-field Modeling to Estimate Shear & Moment Demand

• Gingery et al. (2018) present a simplified approach to estimate kinematic

and inertial contributions to flexure
– Site response or simplified method (Gingery 2007) used
to estimate kinematic displacement field, y(z)
– Incremental application of displacement field +
inertial load components in L-Pile to estimate
shear & flexure
– Which p-y curves to use? (!)
– p-y curves for “liquefied sand” represent flow
failure, not cyclic mobility

11 Gingery et al. (2018)

 Our Work Aims to Move our Understanding
Forward
• Soil-structure interaction:
Static/dynamic arching and load
transfer mechanisms • System Response:
• Installation conditions: Applied – Ground surface response spectra
foundation stresses, liquefiable – Liquefaction behavior
layer thickness, RI embedment, – Dynamic axial load transfer
ground motion variability – Dynamic moment and shear demands
• RI structural behavior: Nonlinear – Reinforcement demands
RI behavior: cracking, inelastic
moment-curvature, hinging
12
 Overview of Modeling Requirements &
Improvements

See:
Gibson, M.D. and Stuedlein, A.W. (2022). “Observations on the Seismic Loading of Rigid Inclusions based on 3D Numerical Simulations.”
13 Journal of the Deep Foundations Institute, 16(3), 1-22.
 Critical Modeling Needs
3D Numerical Simulations
• Accurate bending dynamics; selected sufficiently fine
mesh, w/ short elements and square meshing
arrangement*
• Appropriate SSI mechanisms; selected interface
elements to capture changes in effective stress and
relative displacements* P2PSand Model (FLAC3D)

• Sufficient 3D nonlinear constitutive models:

– Liquefiable soil: P2PSand model* (Cheng & Detournay 2021)
– Non-liquefiable soil: elastic-plastic Mohr-Coulomb with
hysteretic model*
– Steel reinforcements: elastic-plastic cable elements
– Grout: Focus of next several slides

14 *see Gibson & Stuedlein (2022) Cheng & Detournay (2021)

 Critical Modeling Needs
Required Nonlinear Bending Behavior

• Rigid Inclusion: continuum with discrete reinforcements

• Section characteristics & abilities:
– Unconfined grout
– Confined grout: Modify internal parameters per
Mander et al. (1988)
– Vertical reinforcement: Explicitly modeled
– Tensile cracking: Constitutive model w/ reversible tensile
plastic strain
– Moment-Curvature varies with axial: emergent
– Plastic moment varies with axial: emergent
– Progressive flexural failure: emergent
• Fine lateral mesh for moment-curvature resolution
15
 Constitutive Models for RI Grout
• Phase I Study (Gibson & Stuedlein 2022): Mohr-T Model
– Produced nonlinear bending stiffness, EI
– Stiffness degradation from stress cycles 35 5075

– No reduction in compressive strength 30 Compressive

Strength Degraded
4350

Compressive Stress (MPa)

Compresive Stress (psi)

Strength
– Additional damping required 25 Degraded 3625
Strength
• Phase II study: Concrete Model 20
following
Tensile 2900
Stresses
– Nonlinear bending stiffness 15 2175
– Strain softening 10 1450
– Improved stiffness degradation from 5  725
stress cycles
0 0
– Strength degradation from tensile stresses
-5 -725
– Realistic damping 0 0.1 0.2 0.3 0.4 0.5
• “Damage” accumulates: % reduction Compressive Strain (%)

in Young’s modulus
16
 Constitutive Models for RI Grout
Calibration & Validation
• Calibration of FLAC3D model:
– Full-scale observations of moment-curvature
(drilled shaft, Li et al. 2019)
– 1.1 m dia., 2% Steel
• Validation: against L-Pile
• With both Grout models, we capture:
– Excellent agreement with full-scale
measurements
– Emergent moment-curvature from nonlinear
responses along the section (center, edge)
• Mohr-T performs slightly better @ initial
cracking under monotonic loading
Li, Q., Stuedlein, A.W., and Marinucci, A. (2019). "Effect of Casing and High-Strength Reinforcement on the Lateral Load Transfer Characteristics of Drilled Shaft
17 Foundations." Journal of Geotechnical and Geoenvironmental Engineering, 145(9), 04019056.
 LPILE (0 kN) LPILE (500 kN)
FLAC3D Edge (MohrT, 0 kN) FLAC3D Center (MohrT, 0 kN)
FLAC3D Edge (MohrT, 500 kN) FLAC3D Center (MohrT, 500 kN)

Constitutive Models for RI Grout

FLAC3D Edge (Conc, 0 kN) FLAC3D Center (Conc, 0 kN)
FLAC3D Edge (Conc, 500 kN) FLAC3D Center (Conc, 500 kN)
100
UNREINFORCED
Monotonic moment-curvature responses of RIs

Bending Moment (kN-m)

75 Unreinforced RIs
(only): Nominal
• Unreinforced RIs: moment capacity
calculated using axial
thrust force, section
– ACI (2014), L-Pile: elastic-perfectly plastic responses. 50 modulus, and modulus
of rupture following
ACI (2014)
– No strain hardening when axially-loaded
25
– FLAC3D, Mohr-T: significant addition to bending
capacity when loaded 0
– FLAC3D, Concrete: slightly softer center response, 0 0.01 0.02 0.03 0.04 0.05

slightly stiffer and stronger edge response 250

STEEL CAGE
6 - #8 Gr 60
• Reinforced RIs:

Bending Moment (kN-m)

200
– L-Pile captures confined grout, but returns an ultimate
150
plastic response
– FLAC3D, Mohr-T: confinement + progressive hardening 100

– FLAC3D, Concrete: identical center response, notably 50

stiffer and stronger edge response

0
0 0.01 0.02 0.03 0.04 0.05
18 Curvature (1/m)
 Constitutive Models for RI Grout
Cyclic Moment-Curvature Responses of Unreinforced RIs
• Lateral quasi-static, cyclic loading test simulations of rigid inclusions (12 cycles)
• Mohr-T Model: slow rate of ultimate moment reduction; stiffness degradation decoupled
from (constant) compressive strength
• Concrete Model: increased rate of moment reduction; effect of post-peak softening,
tension excursions, damage accumulation
• Concrete Model: realistic energy dissipation and hinging behavior
Lateral Displacement MohrT Model Concrete Model

25 100 100 100

20 80 80 80
Lateral Displacement (mm)

15 60 60 60
10 40 40 40
Moment (kN-m)

Moment (kN-m)
Moment (kN-m)

5 20 20 20
0 0 0 0
-5 -20 -20 -20
-10 -40 -40 -40
-15 -60 -60 -60
-20 -80 -80 -80
-25 -100 -100 -100
0 2 4 6 8 10 12 -0.100 -0.075 -0.050 -0.025 0.000 0.025 0.050 0.075 0.100 -25 -20 -15 -10 -5 0 5 10 15 20 25
19 Loading Cycles Curvature (1/m) Lateral Displacement (mm)
 Selected Prior Results using the Mohr-T
Constitutive Model for RI Grout

20
 3D Numerical Simulations of Rigid Inclusions
Base Case & Variations Foundation w/ 190 kPa
Soil Load Transfer Platform 0.3 m
Med. Stiff Clay Crust 0.3 m
• Unit cell model w/periodic BCs
• Consists of (base case in blue): Interfaces Around RI

– Foundation w/ 190 kPa (4 ksf) loading Liquefiable Layer 2.5 m

(Dr = 45%)
– 0.3 m thick load transfer platform
– Medium stiff clay crust, Hcr: 0.3 m, 2.5 m 0.67D embedment
– Liquefiable layer, Hliq: 2.5, 5.0, 7.5, 10 m AR = 5.8%, 1.7 m spacing
– 5 m thick stiff clay bearing layer
– 0.3 m thick elastic, compliant base (Vs = 300 m/s)
• 0.46 m diameter rigid inclusions: Stiff Clay Bearing Layer
5.0 m
– Unreinforced, single center bar, steel cage
– Embedment, ERI: 0.67D, 6.0D
– Area replacement ratios, AR: 2.8, 5.8, 9.8, 14.4% Stiff Compliant Elastic
21 Base (Vs = 300 m/s)
 3D Numerical Simulations of Rigid Inclusions
Base Case & Variations
Crust, Liq. Layer thickness, Embedment
Embedment
Base Case

0.3 m

2.5 m
10 m
2.5 m
5.0 m
10 m
0.67D 2.5 m

6D 6D
6D
0.67D

22
 3D Numerical Simulations of Rigid Inclusions
Ground Motion (GM) Selection & Free Field (FF) Responses
RSN1495 approximates the
median GM; used as the base
• Necessary to consider ground motion case GM in our simulations

variability at any site Individual

Pseudo-Spectral Acceleration, RotD50 (g)

motions with
1.00
• Motions selected from PEER GM dataset
RSN1495
(in red)

(Set #1A; Baker et al. 2013):

– Strike-slip, Mw = 7 events
– Source-to-site distance of 10 km 0.10

– Sites with Vs,30 ranging from 200 to 400 m/s

(average 300 m/s) Median of 11 Selected GMs
16th / 84th Quantile (Selected GMs)
• Selected 11 of 40 GMs to match median Median (Suite of 40 GMs; Baker et al. 2011)
16th / 84th Quantile (Suite of 40 GMs)

and St. Dev. of dataset 0.01

0.01 0.1 1 10
Period (s)
23
 Seismic Response of Rigid Inclusions
Effect of Ground Motion Variability: Amplification @ Ground Surface
Ratio of surface response spectra: RI / FF
3.0
• BASE CASE: Unreinf. RIs, AR = 5.8%, RSN1495
(Base Case)

Amplification Ratio: Spectral Acceleration

Hliq = 2.5 m, ERI = 0.67D, Hcr = 0.3 m
2.5 10 Other
• Both elastic and nonlinear RIs

with Rigid Inclusion / Free-field

Motions

• Effect of ground motion variability 2.0 Median of 11

Selected GMs
apparent: different motions amplify (Elastic RIs)
Median of 11
different periods 1.5 Selected GMs

• All PGAs decrease. Overall (median) 16th / 84th

Quantile
de-amplification; due to: 1.0

• Increased stresses from surface

0.5
loading, complicated by arching
phenomena (free-field represents PGAs
0.0
unloaded condition) 0.01 0.1 1 10
24 Period (s)
 Seismic Response of Rigid Inclusions
Effect of GM Variability: Cyclic Shear Stress Reduction (CSSR)

• No shear strain compatibility (SSC), as expected

• Relative to the free-field, cyclic stress ratios reduce significantly (to ~10 - 20%)
in presence of RI RCSR Rrd

• Reduction in CSR is not Base Case

(RSN1495)
fully-due to reduction in PGA 10 other
Motions

• Effect of foundation loading, (a) Median of 11

Selected GMs (b)
complicated by arching
phenomenon No
• Affects the shear stress SSC:
reduction coefficient, Rrd gr << 1

25
 Seismic Response of Rigid Inclusions
Effect of Arching: Static and Cyclic StressesFree-Field
Ratio of Vertical RI w/ 190
Total and kPa
Effective Stress,  v0 /  ′v0
1.0 1.2 (4 ksf)
1.4 1.6 1.8
0
• Initial stress state strongly influences
triggering of liquefaction
4
• Cyclic stress ratio (Simplified Method):
Effect of
8 Arching
Effect of Pa
Surface

Depth (m)
Loading
12 Unloaded
• Surface loading reduces the ratio Free-Field

of stresses
16
• Arching reverses some of this reduction Free-Field (0 KSF), 2.5 m thick Liq. Layer
Free-Field (0 KSF), 5.0 m thick Liq. Layer
• CSSR, if considered, must account for 20
Free-Field (0 KSF), 10 m thick Liq. Layer
Free-Field (4 KSF), 2.5 m thick Liq. Layer
these phenomena RI (4 KSF), 2.5 m thick Liq. Layer
RI (4 KSF), 5.0 m thick Liq. Layer
RI (4 KSF), 10 m thick Liq. Layer
26 24
 Seismic Response of Rigid Inclusions
Effect of GM Variability: Structural Responses (Unreinforced RIs)
Maximum structural responses during ground motion
• Maximum shear stress ratio, Moment capacity: 31 kN-m (0 Axial Load) to 41 kN-m (200 kN AL)
moment, shear force, and axial Initial
Static
load during GM Load
• Shear stresses picked up by At end of
RI much larger than that of shaking
surrounding soil (dashed)

• Moment distributions exhibit

variability, but all sections
remain elastic
• 0.67D embedment, results in
pinned-pinned response
• Some axial load is shed due to
EPWPs and partial loss of
arching
27
 Seismic Response of Rigid Inclusions
Temporary Loss of Arching & Effect on EPWPs
• As EPWPs increase, stresses within the LTP Unreinforced RI with 5 m thick Liq. Layer
and soils that had arched onto the RI begin to 250 350

transfer (unload) to pore water Total Stress, v

Stress and Pore Pressure (kPa)

300

• Manifests as an increase in total stress of the 200

v

Axial Load in RI, QRI (kN)

250
surrounding soils, which reduces the axial load 150
v0
v0

on the RI 200

Axial Load, QRI 150

• A general definition of the excess pore pressure 100

ratio is necessary: 100

50
𝒆 𝒆,𝑺 𝒘𝒂𝒗𝒆𝒔 𝒆,𝒖𝒏𝒍𝒐𝒂𝒅𝒊𝒏𝒈 Effective Stress ′v 50
𝒖 𝒖 Pore Pressure, u
𝒗𝟎 𝒗𝟎 𝒗,𝒖𝒏𝒍𝒐𝒂𝒅𝒊𝒏𝒈 𝟎 0 0
0 10 20 30 40 50
• Mechanism occurs during shaking, post-shaking Time (s)

dissipation leads to re-acquisition of arching and

axial load + drag loads (needs study!)
28
 Revisiting some Phase I Results with the
Concrete Constitutive Model for RI Grout

29
 Previous Findings from Unit Cell Analyses
Gibson & Stuedlein (2022)

• Effect of Surface Crusts & Embedment (0.67D vs. 6D):

– Thin Crusts: No effect of RI toe embedment on surface amplification
– Thick Crusts: Amplifies short and medium periods, embedment can increase amplification
Fixed (crust) – Free (toe): moment and shear demand increases, hinges possible
Fixed (crust) – Fixed (toe): greater increases in moment and shear
• Effect of Liquefiable Layer Thickness (2.5 m to 10 m thick):
– Free-field: PGAs increase with layer thickness due to dilation spiking
– With RIs: PGA decreases relative to free-field, amplified periods lengthen
– RIs alter the ground motion due to kinematic SSI, prevents or reduces dilation spiking
– Little sensitivity of spectral acceleration to reinforcement (provided hinges do not develop)
BASED ON MOHR-T MODEL
30
 Seismic Response of Rigid Inclusions
Effect of Reinforcement (Hliq = 10 m): Surface Accelerations & EPWPs
0.6 1.0
PGARI = 0.28g
• Free-field:

Surface Accleration (g)

Excess Pore Pressure

0.4
100 PGAFF = 0.79g 0.8
0.2
80
– Strong60 dilation spikes

Ratio, ru
0.0 0.6
-0.2 EPWPs 0.4
– Large20PGA (0.79g) 40
Moment (kN-m)

-0.4
0.2
-0.6
Unreinforced
UnreinforcedRI
• Rigid inclusion w/Mohr-T Model:
0 RI
-0.8 0.0
-20 0 10 20 30 40 50 60
0.6 Time (s) 1.0
– Delays
-40
onset of EPWP generation PGARI = 0.32g

Surface Accleration (g)

Excess Pore Pressure

-60 0.4
PGAFF = 0.79g 0.8
0.2
– Mitigates dilation spiking, shaking amplitudes
-80

Ratio, ru
0.0 0.6
-100

– can reduce excess pore pressure (in thin layers)

-0.100 -0.075 -0.050 -0.025 0.000 0.025 0.050 0.075 0.100
Curvature (1/m)
-0.2
-0.4
0.4

0.2
• Increasing RI Stiffness: -0.6
-0.8
#32 (#10) Center
Single Bar
Steel Center Bar
0.0
0 10 20 30 40 50 60
– Attracts greater flexural demand 0.6
PGARI = 0.33g
1.0

Surface Accleration (g)

Excess Pore Pressure

0.4 PGAFF = 0.79g Steel Cage 0.8
– No real change in dynamic system SSI 0.2

Ratio, ru
0.0 0.6

– …but this is not the whole story, as cyclic -0.2

Acceleration: Steel Cage-Reinforced RI
0.4

degradation of grout not captured very well..! -0.4

-0.6
Acceleration: Free-Field
Excess Pore Pressure Ratio: Steel Cage-Reinforced RI
0.2
Excess Pore Pressure Ratio: Free-Field
-0.8 0.0
0 10 20 30 40 50 60
31 Time (s)
 Seismic Response of Rigid Inclusions
Effect of Reinforcement (Hliq = 10 m): Surface Accelerations & EPWPs
150
0.6 1.0
PGARI = 0.43g

Surface Accleration (g)

• Rigid inclusion w/Concrete Model: Unreinforced

Excess Pore Pressure

0.4
PGAFF = 0.79g 0.8
0.2 Center Bar
100

Ratio, ru
– Captures appropriate grout degradation 0.0 Steel Cage 0.6
-0.2 0.4
– Acceleration time histories hold the clues -0.4
50
0.2

(kN-m)
-0.6
Unreinforced RI
Unreinforced RI
– Damage accumulates, plastic hinging develops -0.8
0 0
0.0
10 20 30 40 50 60

Surface AcclerationMoment
– Unit cell degrades into simple SDOF system, 0.6
PGARI = 1.20g
1.0

(g)

Excess Pore Pressure

0.4
PGAFF = 0.79g 0.8
oscillates with soil, leading to dilative -50
0.2

Ratio, ru
0.6
amplification  this is not realistic! -0.2
0.0

-100 0.4
100 -0.4
0.2
80 -0.6
-150
#32 (#10)
Single Center
Steel CenterBar
Bar
60 -0.8 0.0
0
0 10
10 20
20 30
30 40
40 50
50 60
60
40
Moment (kN-m)

0.6 1.0
PGARI = 1.05g

Surface Accleration (g)

Excess Pore Pressure

20
0.4 PGAFF = 0.79g Steel Cage 0.8
0 0.2

Ratio, ru
-20 0.0 0.6
-40 -0.2 0.4
-60 -0.4
Acceleration: Steel Cage-Reinforced RI
Acceleration: Free-Field
-80 0.2
-0.6 Excess Pore Pressure Ratio: Steel Cage-Reinforced RI
-100 Excess Pore Pressure Ratio: Free-Field
-0.8 0.0
-0.100 -0.075 -0.050 -0.025 0.000 0.025 0.050 0.075 0.100 0 10 20 30 40 50 60
32 Curvature (1/m) Time (s)
 Selected Phase II Results
-
Concrete Constitutive Model for RI Grout
Realistic Boundary Conditions

33
 Phase II: 3D Simulations of Isolated Footings
M
• Embedded, isolated footing within very large unit cell
• Footing within center of tributary “structural bay,”
intended to capture improved boundary conditions
(but without structural resistance) Embedded Footing
Load transfer platform
• Four (4) rigid inclusions per footing,
190 kPa (4 ksf) applied bearing pressure
• Post-shaking footing loading test 
Liquefiable Layer
maintenance of vertical support
• Future efforts: include effect of inertial loading
with SDOF lumped mass
• Current efforts: kinematic loading, only
Bearing Layer

34
 Phase II: 3D Simulations of Isolated Footings
10.0 m

3.0 m
• 10 m-wide unit cell Footing
Interface
Footing
• 3 m-wide square footing Stone
Fill
0.9 m

Unit 1 0.3 m
• Interface elements at all soil-
structure interfaces Cage Reinforced
RI’s w/ Interfaces

• Embedded footing captures

5m
passive resistance to
Unit 2
kinematic loading Liquefiable

• Four RIs per footing,

embedded 0.67D and 6D
• Unreinforced, single #32
Unit 3 5m
(#10) bar, and steel cage Non-Liquefiable

(6 - #25 bars)
35
GM: RSN1495 𝒖̈ 𝒈
 Group Rigid Inclusion Responses
Dynamic Load Transfer Characteristics of Surrounding Soil
With Steel Cage
• Between “footings” in unit cell,
soil liquefies Passive
resistance

• SSI maintains some non-zero Pa

effective stress below footing
ru > 0.98 ru > 0.98
• Static stresses below footing
ru ≈ 0.90
transferred to
bearing layer
during shaking Load transfer to
bearing layer
as RIs &
footing settle
• Passive
resistance is
mobilized
36
 Group Rigid Inclusion Responses
Vertical Footing Displacement Response
With Steel Cage
• Loss of effective stress below
footing  loss of some shaft
resistance
• Footing settles and toe of rigid meters

inclusion displaces downward to

maintain vertical support
• Vertical footing displacement:
93 to 96 mm
Toe of RI
settles to
maintain load
transfer

37
 Group Rigid Inclusion Responses
SSI Mechanisms: Unreinforced RIs (Embedment 0.67D)
0.3
All time histories at depths of maximum moment Depth = 5.97 m

Cyclic Stress Ratio,

0.2
Depth = 2.49 m
0.1

• Cyclic shear stresses trigger EPWPs; but

CSR
0.0
-0.1

no transfer of total stress to pore water -0.2

-0.3
(a)

• Axial load increases initially as soils

150

Total Stress and Pore

125

Poressure (kPa)
below footing lose effective stress 100
75
Total Stress, v
Pore Pressure, u
50
• Kinematic SSI loads RI in flexure 25
(b)
0
• 1stcracking @ 2.49 m, t ≈ 16 sec 400

Footing Settlement (mm)

0

RI (kN)
• 2nd cracking occurs @ 5.97 m, t ≈ 21 sec 300 Settlement 50

Axial Load, QRI

Axial Load
• Moment resistance degrades as RI 200 100

accumulates damage 100

Time of initial cracking @ 2.49 m
Time of initial cracking @ 2.49 m (c)
150

0 200

• Axial load is shed to surrounding soil as it Moment, M (kN-m)

75
50
Time of initial cracking @ 5.97 m

dilates: local load reversals timed with tcyc 25

0

• Footing settlement rate increases after

-25
-50 Mic = 51 kN-m (350 kN)

hinging  180 mm total

(d)
-75
0 10 20 30 40 50 60
38 Time (s)
 Group Rigid Inclusion Responses
SSI Mechanisms: Single #32 (#10) Center Bar-Reinforced (Emb. 0.67D)
0.3
All time histories at depths of maximum moment Depth = 5.97 m

Cyclic Stress Ratio,

0.2
Depth = 2.49 m
0.1
• Same ground motion; similar CSRs until

CSR
0.0
-0.1
reduction towards end of GM

Cyclic
-0.2
(a)
-0.3

• Axial load increases as soils below footing

150

Total Stress and Pore

125

Poressure (kPa)
lose effective stress – but never fall below 100
75
Total Stress, v
Pore Pressure, u
static load magnitude 50

Total
25
(b)
• Kinematic SSI loads RI in flexure 0

(mm)
Footing Settlement (mm)
400 0
Settlement
• Initial cracking occurs @ the same locations

RI (kN)
300 50

Load, QRI
Axial Load
• Plastic hinging of section does not occur 200 100

Axial Load,
100 Time of initial 150

• Axial load is maintained: surrounding soil

Axial
cracking @ 2.49 m (c)
0 200

doesn’t work (dilate) as hard

100 Mc = 93 kN-m (350 kN)
(kN-m)
Moment, M (kN-m) 75
50
25
• No change in footing settlement rate, total 0
Moment,

-25

settlement is smaller  120 mm total -50

-75
Mic = 51 kN-m (350 kN)
(d)
-100
0 10 20 30 40 50 60
39 Time (s)
 Group Rigid Inclusion Responses
SSI Mechanisms: Steel Cage-reinforced RI (Embedment 0.67D) 0.3
All time histories at depths of maximum moment Depth = 5.97 m

Stress Ratio,
0.2
Depth = 2.49 m
0.1

Cyclic Stress
• Now w/ steel cage (6 - #25 bars, Gr 60)

CSR
CSR
0.0
-0.1

Cyclic
-0.2
• Axial load increases as soils below footing -0.3
150
(a)

lose effective stress, maintains higher axial

Pore
Stress and Pore
125

(kPa)
Poressure (kPa)
load relative to static load 100 Total Stress, v

Poressure
75 Pore Pressure, u

Total Stress
50
• Kinematic SSI loads RI in flexure

Total
25
(b)
0
• Initial cracking occurs @ the same locations,

(mm)
Settlement(mm)
400 0
Settlement
2nd cracking delayed

(kN)
RI (kN)

Footing Settlement
300 50

QRI
Axial Load
• FS on plastic hinging > 1.8

Load, Q
200 100

Axial Load,
100 Time of initial cracking @ 2.49 m 150
• Stiff RI does most of the work to pick up

Footing
Axial
(c)
0 200

footing load; little need for soil dilation (kN-m)

M (kN-m)
150
150
100
100 Time of initial cracking @ 5.97 m
50
50
• Gradual rate of settlement rate, total
Moment, M

00
Moment,

settlement is smallest  96 mm -100

-100
-50
-50 Mic = 51 kN-m (350 kN)
Mc = 227 kN-m (350 kN) (d)
-150
-150
00 10
10 20
20 30
30 40
40 50
50 60
60
40 Time
Time (s)
(s)
 Group Rigid Inclusion Responses
SSI Mechanisms: Effect of RI Embedment in Bearing Layer
Depth = 5.97 m
Depth = 2.49 m
500

Axial Load, QRI (kN)

400

• Considering RI reinforced with a #32 (#10)

300

200

center bar: 100

Single Center Bar: 0.67D
6D
0
• Kinematic loading of RI with low embedment (0.67D) 100
75

Moment, M (kN-m)
leads to soil dilation and load sharing 50
25

• RI with greater embedment (6D) provides greater

0
-25

moment resistance at contact w/ bearing layer

-50
-75
-100
• Reduces soil work, leads to greater axial loads 0 10 20 30
Time (s)
40 50 60

700

• Steel cage-reinforced RI (6 - #25 [#8] bars) 600 Steel Cage: 6D

0.67D

Axial Load, QRI (kN)

500
400
• 0.67D: Low embedment leads to low moment 300

demand at bearing layer contact 200

100
0
• 6D: greater embedment leads to increased moment 150
100

Moment, M (kN-m)
at bearing layer contact 50

• Load drop at depth is local, affected by numerical

0
-50

interface element between steel and grout continuum -100

-150

41
– needs further investigation 0 10 20 30
Time (s)
40 50 60
 Group Rigid Inclusion Responses
Footing Settlement: Effect of RI Embedment in Bearing Layer

• Shallow embedment: settlement time 0

histories identical until ~24 sec: 25
• Unreinforced RI: rate of settlement increases

Settlement (mm)
50
dramatically 1.5 cycles after 2nd hinge develops 75
• Reinforced RIs: no hinging, load transfer 100
Unreinforced (0.67D)
maintained, reduced settlement 125 Single Bar (0.67D)
150 Steel Cage (0.67D)
• Deeply-embedded RIs: improved load Single Bar (6D)
175
transfer to bearing layer leads to reduced 200
Steel Cage (6D)
displacement: 0 10 20 30 40 50 60
• Single Bar: 120 mm  82 mm Time (sec)

• Steel Cage: 96 mm  30 mm

42
 Group Rigid Inclusion Responses
Post-Shaking System Stability: Footing Loading Test, Unreinf. RIs
Lateral Displacement (mm) Vertical Displacement (mm) Moment (kN-m) Axial (kN)

• Profiles of displacement,
-200 -100 0 100 200 300 0 50 100 150 200 -50 -25 0 25 50 0 100 200 300
0 0

moment, and shear during 1 1

shaking (24.6s)
2 2

• After shaking, at start of

loading test, axial load is 3 3

initially low

Depth (m)
Depth (m)

4 4

• Load picks up with footing 5 5

displacement… 6 6

• Until RI loses stability at 7 In-Shaking (t = 24.6s) 500 7 In-Shaking (t = 24.6 s)

highest plastic hinge Footing Test: Just after

Load Drop 400
Footing Loading Test: Just
after load drop

(w/ zero moment transfer) 8 8

Axial Load (kN)

300

• Footing and RI plunges, 200

but axial load is greater 100

Starting Load ≈ 120 kN
Depth = 2.49 m

than initial static load

Depth = 5.97 m
0
0 10 20 30 40 50 60 70 80 90 100
Settlement (mm)
43
 Group Rigid Inclusion Responses
Post-Shaking System Stability: Footing Loading Test, #32 Center Bar
Lateral Displacement (mm) Vertical Displacement (mm) Moment (kN-m) Axial (kN)
-200 -100 0 100 200 300 0 50 100 150 200 -75 -50 -25 0 25 50 75 0 100 200 300 400 500
0 0

• Post-seismic status of RIs with single 1 1

center bar
2 2

• No sudden loss of integrity of RI 3 3

• Load is transferred through shaft

Depth (m)
4 4

resistance as soil dilates at interface

until entire RI plunges at ~50 mm 5 5

• Details of load transfer needs further

6 6

study, but significant post-seismic 7 In-Shaking (t = 24.6s)

Footing Test: 15 mm
600 7 In-Shaking (t = 24.6 s)

Footing Test: 15 mm

support of vertical load Displacement 500 Displacement

8 8

Axial Load (kN)

400

• Better responses for RI with Cage 300

200
Depth = 2.49 m
100
Depth = 5.97 m
0
0 10 20 30 40 50 60 70 80 90 100
44 Settlement (mm)
 Some Conclusions
• We continue to improve our modeling approaches to study basic seismic SSI
mechanisms for rigid inclusions
– Previous unit cell simulations using Mohr-T model failed to capture appropriate dynamic grout
degradation
– Basic unit cell models will not be able to capture restraint provided by the “free field,” and cannot
capture reality
• Depending on demand, rigid inclusions may experience significant nonlinear
inelasticity during shaking
• This may be alright, if code-based life-safety is the performance target
• Post-shaking footing loading tests show that initial static axial loads in RIs can be
maintained
– Unreinforced RIs: little excess post-shaking axial capacity due to degradation of grout
– Reinforced RIs: significantly greater axial capacity reserves
45
 Our Work Continues
• Assess moment and shear demand with varying
liquefiable layer thicknesses
• Inertial effects
• Post-shaking consolidation and reacquisition of arching
• Simplified design procedure for moment demands: 1D
site response  section design
Thank you!

46
 [ Extra Slides ]

47
 3D Numerical Simulations of Rigid Inclusions
Ground Motion (GM) Selection & Free Field (FF) Responses
Base Model Case, 11 Motions Variations on the base case

Cyclic Stress Ratio, CSR Excess Pore Pressure Cyclic Stress Ratio, CSR Excess Pore Pressure
Ratio, ru Ratio, ru
0.0 0.5 1.0 1.5 2.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.5 1.0 1.5 2.0 0.0 0.2 0.4 0.6 0.8 1.0
0 0

1 2

2 4
Legend for (c) and (d)
3 Base Case
6 (RSN1495)
Depth (m)

Median of 11
4 Selected GMs
8 Base with 5.0 m
Liq. Layer
5 Base with 7.5 m
Legend for (a) and (b) 10 Liq. Layer
Base with 10 m
6 Base Case Liq. Layer
(RSN1495)
12 Base with 2.5 m
thick Crust
10 other
7 Motions Base with 4 ksf
14 Surcharge
Median of 11
8 Selected GMs
(a) (b) (c) (d)
48 16
 Group Rigid Inclusion Responses
Rigid Inclusion Displacement Response
93 mm 96 mm

• Vertical RI movements appear

as “rigid body” translation
• But in reality, local meters
displacements indicate Volumetric
otherwise Internal strains:
Displacement yielding 5 to 12%
• Strains in grout are significant Exaggerated 3x
Shear strain:
5%

92 mm 94 mm
49