Simplified

Critical-State Soil Mechanics

Paul W. Mayne
Georgia Institute of Technology
 PROLOGUE
 Critical-state soil mechanics is an
effective stress framework describing
mechanical soil response
 In its simple form here, we consider
only shear loading and compression-
swelling.
 We merely tie together two well-known
concepts: (1) one-dimensional
consolidation behavior (via e-logsv’
curves); and (2) shear stress-vs.
normal stress (t-sv’) plots from direct
 Critical State Soil Mechanics
(CSSM)
 Experimental evidence
 1936 by Hvorslev (1960, ASCE)
 Henkel (1960, ASCE Boulder)
 Parry (1961)
 Kulhawy & Mayne (1990): Summary of
200+ soils
 Mathematics presented elsewhere
 Schofield & Wroth (1968)
 Roscoe & Burland (1968)
 Wood (1990)
 Jefferies & Been (2006)
 Basic form: 3 material constants (f', Cc,
Cs) plus initial state parameter (e0, svo',
 Critical State Soil Mechanics
(CSSM)
 Constitutive Models in FEM packages:
 Original Cam-Clay (1968)
 Modified Cam Clay (1969)
 NorSand (Jefferies 1993)
 Bounding Surface (Dafalias)
 MIT-E3 (Whittle, 1993)
 MIT-S1 (Pestana, 1999; 2001)
 Cap Model
 “Ber-Klay” (Univ. California)
 others (Adachi, Oka, Ohta, Dafalias, Nova, Wood, Huerkel)
 "Undrained" is just one specific stress path
 Yet !!! CSSM is missing from most textbooks and
undergrad & grad curricula.
 One-Dimensional Consolidation
Sandy Clay (CL), Surry, VA: Depth = 27 m
1.0
svo'=300 kPa
sp'=900
0.9
Cr = 0.04 kPa
Void Ratio, e

0.8 Overconsolidation Ratio, OCR

=3
Cs = swelling index (= Cr)
0.7
cv = coef. of consolidation
D' = constrained modulus
0.6
Cae = coef. secondary compression
Cc = 0.38
k ≈ hydraulic conductivity
0.5
1 10 100 1000 10000

Eff ective Vertical Stress, s ’ (kPa)

svo'
v
 Direct Shear Test Results
Slow Direct Shear Tests on Triassic Clay,NC Slow Direct Shear Tests on Triassic Clay, Raleigh, NC
140 140
sn' Strength Parameters:
f ' = 26.1 o

(kPa)
Shear Stress, t (kPa)

120 120
(kPa)= c' = 0;
Peak
214.5
100 100

t
80 80

Shear Stress,
Peak

60 135.0 60
0.491 = tan f '
40 40
Peak
20 45.1 20

0 0
0 1 2 3 4 5 6 7 8 9 10 0 50 100 150 200 250

Displacement, d (mm) Eff ective Normal Stress, sn' (kPa)

t t sv’ t t sv’
d

gs

Direct Shear Box (DSB) Direct Simple Shear (DSS)

 CSSM for Dummies
CC

Void Ratio, e
Void Ratio, e

sCSL’  ½sNC’
e00 NC NC

CSL CSL

sCSL’ sNC’ Log sv' Effective stress sv'

CSL

Shear stress t tanf'

CSSM Premise:
“All stress paths
fail on the critical
state line (CSL)” f
c=0
Effective stress sv'
 CSSM for Dummies
CC
Void Ratio, e

Void Ratio, e
e0
De NC
NC
ef
CSL CSL

svo
Log sv' Effective stress sv'
CSL

Shear stress t
tmax = c + s tanf tanf'
STRESS PATH No.1
NC Drained Soil
Given: e0, svo’, NC
(OCR=1)
Drained Path: Du = 0
Volume Change is c’=0
svo
Contractive: evol =
Effective stress sv'
 CSSM for Dummies
CC
Void Ratio, e

Void Ratio, e
e0
NC
NC
CSL
CSL
svf svo Effective stress sv'
Log sv' CSL
tanf'
STRESS PATH No.2 Shear stress t
NC Undrained Soil Du
tmax = cu=su
Given: e0, svo’, NC
(OCR=1)
Undrained Path: DV/V0 = 0
+Du = Positive Excess svf
Porewater Pressures svo
Effective stress sv'
 CSSM for Dummies
CC

Void Ratio, e
Void Ratio, e

NC NC

CSL CSL

Log sv' Effective stress sv'

CSL

Shear stress t
tanf'
Note: All NC
undrained
stress paths are
parallel
to each
DSS: other,
s /s ’ =thus:
u vo NC Effective stress sv'
su/svo’ = constant
 CSSM for Dummies
CC

Void Ratio, e
Void Ratio, e

CS OC
NC NC

CSL CSL

Effective stress sv'

Log sv' sp'
CSL

Overconsolidated States: Shear stress t tanf'

e0, svo’, and OCR = sp’/svo’

where sp’ = svmax’ = Pc’ =
preconsolidation stress;
OCR = overconsolidation ratio
Effective stress sv' sp'
 CSSM for Dummies
CC

Void Ratio, e
Void Ratio, e

e0
OC
CS NC
NC

CSL CSL

svo' svf' Log sv' Effective stress sv'

CSL
Stress Path No. 3 Shear stress t tanf'
Undrained OC Soil: Du
e0, svo’, and OCR
Stress Path: DV/V0 = 0
Negative Excess Du
svo' Effective stress sv'
 CSSM for Dummies
CC

Void Ratio, e
Void Ratio, e

e0
OC
CS NC
NC

CSL CSL

svo' Log sv' Effective stress sv'

CSL
Stress Path No. 4 Shear stress t tanf'
Drained OC Soil:
e0, svo’, and OCR
Stress Path: Du =
0Dilatancy: DV/V > 0
0
svo' Effective stress sv'
 Critical state soil mechanics
• Initial state: e0, svo’, and OCR = sp’/svo’
• Soil constants: f’, Cc, and Cs (L = 1-Cs/Cc)
• For NC soil (OCR =1):
 Undrained (evol = 0): +Du and tmax = su = cu
 Drained (Du = 0) and contractive (decrease evol)
• For OC soil:
 Undrained (evol = 0): -Du and tmax = su = cu
 Drained (Du = 0) and dilative (Increase evol)
There’s more ! Semi-drained, Partly undrained, Cyclic response…
 Equivalent Stress Concept
CC

Void Ratio, e
NC
Void Ratio, e

e0
OC NC
De CS
sp' sp'
ep

CSL CSL

svo' svf' se' Log sv' Effective stress sv'

1. OC State (eo, svo’, sp’) CSL

2. Project OC state to NC Shear stress t tanf'

line for equivalent stress, su

’ = Cs log(sp’/svo’)
seDe at se’
De = Cc log(se’/sp’) suOC = suNC

3. se’ = svo’ OCR[1-Cs/Cc] svo' se ' Stress sv'

 Critical state soil mechanics
• Previously: su/svo’ = constant for NC soil
• On the virgin compression line: svo’ = se’
• Thus: su/se’ = constant for all soil (NC &
OC)
• For simple shear: ssue/s
’=e’ =
svo½sin f’
’ OCR[1-Cs/Cc]
• Equivalent stress:
Normalized Undrained Shear
Strength:

su/svo’ = ½ sinf’ OCRL

where L = (1-C /C )
 Undrained Shear Strength from CSSM
0.4
AGS Plastic
Amherst
Ariake
Bootlegger
su/svo' NC (DSS)

0.3 Bothkennar
Boston Blue
Cowden
Hackensack
0.2 James Bay
Mexico City
Onsoy
Porto Tolle
Portsmouth
0.1 Rissa
San Francisco
Silty Holocene
su/ svo' NC (DSS) =½sinf'
Wroth (1984)
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
sinf'
 Undrained Shear Strength from CSSM

10 Amherst CVVC
DSS Undrained Strength, su/ svo'

Atchafalaya
Bangkok
Bootlegger Cove
L Boston Blue
f' =40 o
Intact Cowden
30o Drammen
Clays
20o Hackensack
Haga
1 Lower Chek Lok
Maine
McManus
Paria
L Portland
su/svo' = ½ sinf' OCR Portsmouth
Silty Holocene
Note: L = 1 - Cs/Cc  0.8 Upper Chek Lok
0.1 40
30
1 10 100 20
Overconsolidation Ratio, OCR
 Porewater Pressure Response from CSSM
1 Amherst CVVC
L
Dus/svo' = 1 - ½cosf'OCR Atchafalaya
Bangkok
0
Normalized Porewater, Du/ svo'

Bootlegger Cove
Boston Blue
-1 Cowden
Drammen
Hackensack
-2
Haga
Lower Chek Lok
-3 Maine
Intact McManus
Clays Paria
-4
Portland
f' = 20o 30o 40o Portsmouth
-5 Silty Holocene
L = 0.9 0.8 0.7
Upper Chek Lok
20
-6
30
1 10 100
40
Overconsolidation Ratio, OCR
 Yield Surfaces
NC NC
CSL

Void Ratio, e
OC
OC
Void Ratio, e

sp'

CSL

sp' Normal stress sv'

Log sv'
CSL

 Yield surface Shear stress t

represents 3-d
preconsolidation

 Quasi-elastic
behavior within the
yield surface
Normal stress sv'
 Critical state soil mechanics
• This powerpoint: geosystems.ce.gatech.edu
• Classic book: Critical -State Soil Mechanics
by Schofield & Wroth (1968):
http://www.geotechnique.info
• Schofield (2005) Disturbed Soil Properties
and Geotechnical Design Thomas Telford
• Wood (1990): Soil Behaviour and CSSM
• Jefferies & Been (2006): Soil liquefaction: a
critical-state approach
www.informaworld.com
 ESA versus TSA
• Effective stress analysis (ESA) rules:
 c' = effective cohesion intercept (c' = 0 for OCR < 2 and c' ≈
0.02 sp' for short term loading)
 f' = effective stress friction angle
 t = c' + s' tan f' = Mohr-Coulomb strength criterion
 sv' = sv - u0 - Du = effective stress
• Total stress analysis (TSA) is (overly) simplistic for clay
with strength represented by a single parameter, i.e.
"f = 0" and tmax = c = cu = su = undrained shear
strength (implying "Du = 0")
Explaining the myth that "f = 0"
The effective friction angle (f') is usually
between 20 to 45 degrees for most soils.
However, for clays, we here of "f = 0"
analysis which applies to total stress analysis
(TSA). In TSA, there is no knowledge of
porewater pressures (PWP). Thus, by
ignoring PWP (i.e., Du = 0), there is an
illusional effect that can be explained by
CSSM. See the following slides.
 0.8 0.8
f' = 30 °
Cc = 0.50
Cr = Cs = 0.05

Void Ratio, e
Void Ratio, e

0.7 0.7

0.6 0.6

0.5 0.5
10 100 1000 0 100 200 300 400 500
Log Effective stress, s v' sv' (kPa)
300

(Undrained) Total Stress

t = Shear Stress (kPa)

Analysis - Consolidated
200

Undrained Triaxial Tests 100

Three specimens initially consolidated

to svc' = 100, 200, and 400 kPa 0
0 100 200 300 400 500

sv ' (kPa)
 (Undrained) Total Stress Analysis
In TSA, however, Du not known, so plot stress paths for "Du = 0"
Obtains the illusion that " f ≈ 0° "
300
t = Shear Stress (kPa)

200
su400
su200
su100
100

0
0 100 200 300 400 500

Effective stress, sv' (kPa)

 0.8 0.8

Void Ratio, e
Void Ratio, e

0.7 0.7

VCL
0.6 CSL 0.6
Cs from Pc' = 400 kPa
Cs from Pc' = 500 kPa
Cs from Pc' = 600 kPa

0.5 0.5
10 1000 0 100 200 300 400 500 600
s v' (kPa) s v' (kPa)

300

t = Shear Stress (kPa)

Another set of undrained Total
Stress Analyses (TSA) for UU tests 200

on clays: 100

UU = Unconsolidated Undrained
0
0 100 200 300 400 500 600

s v ' (kPa)
 (Undrained) Total Stress Analysis
Again, Du not known in TSA, so plot for stress paths for "Du = 0"
Obtains the illusion that " f = 0° "
300
t = Shear Stress (kPa)

200

su
100

0
0 100 200 300 400 500

Effective stress, sv' (kPa)

Explaining the myth that "f = 0"
 Effective Stress Analyses (ESA)
• Drained Loading (Du = 0)
• Undrained Loading (DV/V0 = 0)

 Total Stress Analyses (TSA)

 Drained Loading (Du = 0)
 Undrained Loading with "f = 0"
analysis: DV/V0 = 0 and "Du = 0"
Cambridge University q-p' space
s1'
SL
C 6 sin  '
q = (s1 - s3)

s2' = s3' Triaxial Mc 

Compression 3  sin  '

Undrained OC
Stress Path

M c  OCR 
Undrained NC ( su / p0 ' )TC   
Stress Path 2  2 
Drained
Stress Path
3V : 1H

svo' = P0'
P' = (s1' + s2' + s3')/3
 Port of Anchorage, Alaska

0.8
10
Deviatoric Stress = q* = (s 1-s 3)/s p'

Critical State Soil Mechanics

0.7
(Modified Cam Clay)
Bootlegger f ' = 27.7o
0.6
Cove Clay

Strength Ratio, su/s vo'

L = 0.75

0.5

0.4
1
0.3 M c = (q/p')f = 1.10
M = 6sinf '/(3-sinf ')
c DSS Data
0.2
f ' = 27.7o CIUC Data
MCC Pred CIUC
0.1
MCC Pred DSS
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.1
1 10 100
Effective Stress, p'* = ( s 1'+s 2'+s 3')/(3s p') Overconsolidation Ratio, OCR
 Cavity Expansion – Critical State Model for Evaluating
OCR in Clays from Piezocone Tests
1/ 
 1  q T  ub  
OCR 2   
 1.9 5 M  1   vo '  
where M = 6 sinf’/(3- Overconsolidation Ratio, OCR
sinf’) 0 1 2 3 4 5 6
0

and L = 1 – Cs/Cc  0.8 2

4
Depth (meters) 6

fs 8
Bothkennar, UK
10

12
CPTU
ub 14
CRS
IL Oed
16 RF

qc  qT
18

20
Cambridge University q-p' space
L 6 sin  '
CS Mc 
3  sin  '
q = (s1 - s3)

Yield Surface
Original Cam Clay
Modified Cam Clay

Bounding Surface

Cap
Cap Model
Model

Pc'

P' = (s1' + s2' + s3')/3

 Anisotropic Yield Surface
Mc = 6sinf’/(3-sinf’)

SL
C
q = (s1 - s3)

Yield Surface fctn(K0NC)

sp’
OC Y2

e0 Y1
NC
svo’ Y3 = Limit State
K0
G0
P’ = (s1’ + s2’ + s3’)/3
Cambridge University q-p' space Mc 
6 sin  '
3  sin  '
Yield Surface SL
C
q = (s1 - s3)

Apparent
fctn(K0NC)
Mc

sp’

Y3 = Limit State

P' = (s1' + s2' + s3')/3

MIT q-p' space
q = ½(s1 - s3) tan  c sin  '

fctn(K0NC)
Yield Surface
sp’

Diaz-Rodriguez, Leroueil,
OC and Aleman (1992, JGE)

P' = ½(s1' + s3')

 Yield Surfaces
of Natural Clays

Diaz-Rodriguez,
Leroueil, & Aleman
(ASCE Journal
Geotechnical
Engineering July 1992)
Friction Angle of Clean Quartz
Sands
(Bolton, 1986 Geotechnique)
 State Parameter for Sands, y
(Been & Jefferies, 1985; Jefferies & Been 2006)

l10

void l10 Wet of Critical

ecsl
ratio (Contractive)
e y = e0 - ecsl
e0 VCL
Dry of Critical
(Dilative)
CSL
p0'
p' = ⅓ (s1'+s2'+s3')
log p'
 State Parameter for Sands, y
(Simplified Critical State Soil Mechanics)

y = (Cs - Cc )∙log[ ½ cos f' OCR ]

l10 Du = (1 - ½ cos f' OCRL ]∙svo'
l10
void then CSL = OCR = 2/cosf'
ecsl
ratio
e y = e0 - ecsl
e0 VCL
CSL
p0' p' = ⅓ (s1'+s2'+s3')

log p'
 State Parameter for Sands, y
(Been, Crooks,L& Jefferies, 1988)
log OCRp = log2 + Y/(k-l)
where OCRpp = R = overconsolidation ratio in
Cambridge q-p' space, L = 1-k/l, l = Ccc/ln(10) =
compression index, and k  Css/ln(10) = swelling
index

Georgia Tech
 MIT Constitutive Models
 Whittle et al. 1994: JGE Vol. 120 (1)
"Model prediction of anisotropic behavior
of Boston Blue Clay"
 MIT-E3: 15 parameters for clay
 Pestana & Whittle (1999) "Formulation of
unified constitutive model for clays and
sands" Intl. J. for Analytical & Numerical
Methods in Geomechanics, Vol. 23
 MIT S1: 13 parameters for clay
 MIT S1: 14 parameters for sand
MIT E-3 Constitutive Model

Whittle (2005)
MIT S-1 Constitutive Model
Pestana and Whittle (1999)
 MIT S-1 Constitutive Model
Predictions for
Berlin Sands
(Whittle, 2005)
 Critical state soil mechanics
• Initial state: e0, svo’, and OCR = sp’/svo’
• Soil constants: f’, Cc, and Cs
• Link between Consolidation and Shear Tests
• CSSM addresses:
 NC and OC behavior
 Undrained vs. Drained (and other paths)
 Positive vs. negative porewater pressures
 Volume changes (contractive vs. dilative)
 su/svo’ = ½ sinf’ OCRL where L = 1-Cs/Cc
• Yield surface represents 3-d preconsolidation
• State parameter: y = e00 - ecsl
csl
 Simplified Critical State Soil Mechanics
NC
CC

Void Ratio, e
dilative
Void Ratio, e

NC con
sol
eOC ida
t
CS -Du swelling ion
OC
eNC +Du
CS contractive
L C SL

Log sv' s p' Effective stress sv' f'

tmax = stanf
Four Basic Stress S L
Shear stress C
Paths: tmax = su NC
2 1

1. Drained NC (decrease
su OC
DV/Vo) Yield
3 Surface
2. Undrained NC (positive Du)
t

3. Undrained OC (negative
4 sCS½sp sp '
tmax = c+stanf c'
Du)
Effective stress sv'
4. Drained OC (increase