ADVA
A ANCE ED COU
C RSE
E
ON
O CCOMP
PUTAATIO
ONALL GE
EOTE
ECHN
NICS
S
SINGA
S APO
ORE
Ven
nue National University o
of Singaporee
Datte 23 to 25 N
November 2 2011
Asssociate Proffessor Tan SSiew Ann
Natioonal Universiity of Singapore
(Co
ourse leaderr)
Dr Shen Rui Fu
u Natioonal Universiity of Singapore
Dr William Cheeang Plaxiis AsiaPac, SSingapore
Org
ganised by
1 of 607
CONTENTS
Lectures
CG1 Concepts of Plasticity and The Mohr-Coulomb Model 5
CG2 Density and Shear Hardening 26
CG3 Hardening Soil Model 49
CG4 Determination of Soil Parameters 49
CG5 Soft Soil Model 73
CG6 Soft Soil Creep Model 108
CG8 Drained and Undrained Analysis 171
CG9 Structural Elements & Modelling Excavations in Plaxis 196
CG10 Modelling of Groundwater 210
CG11 Slope Stability Analysis and Rain-fall induce failures 225
CG12 Phi-C reduction analysis in Plaxis 267
CG13 Introduction of Plaxis 3D 279
CG14 Geometry, Calculation and Output Mode 295
CG15 Modelling of Piled Foundations in 3D 321
CG16 Modelling of Tunnels and Tunnelling in 3D 367
CG17 Embedded Elements in Plaxis 3D 409
CG18 Modelling of Excavations in 3D 426
Exercises
SGE1 Simulation of laboratory triaxial test 468
SGE2 Simulation of laboratory oedometer test 474
SGE3 2D Excavation : Anchored Excavation 513
SGE4 2D Slope Stability: Stabilisation of an Embankment Using Geotextile 540
SGE5 3D Piled Foundation: Fleiden Piled Raft Analysis 559
SGE6 3D Excavation: 3D Effects in Strutted Retaining Wall Analysis 576
2 of 607
DAY 1
DAY 2
3 of 607
DAY 3
4 of 607
CG1 - Concepts of Plasticity and
Mohr-Coulomb Model
Ronald Brinkgreve
Plaxis BV /
Delft University of Technology
Contents
• Aspects of real soil behaviour
• Stresses and strains
• Stress paths in standard soil tests
• Standard drained triaxial test (CD-test)
• Oedometer test
• Consolidated undrained triaxial test (CU-test)
• Basic concepts of the Mohr-Coulomb model
• Elastic strains, plastic strains
• Yield function, plastic potential
• Parameters
• Possibilities and limitations of the M-C model
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES
p-axis p-axis
q-axis
1
Rendulic plane 2
q-axis
-3
2= 3 -32
-2
Principal stress space Rendulic plane
p-axis
Deviator plane
-3
-2 -3
-2
Principal stress space Deviator plane (-plane)
(p = constant)
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES
v
1
-1
3 3 v
-1
Axial loading
-3
n
-3 -’3
v 0
1
-1
3 3
pw
-1
-’1
Axial loading
-3
-3 -’3
1
-1
1
ln 1
-1
-’1
Axial loading
-’3
xy
dxy
dxy
xy
v
xy
-’1
Shearing
-’3
d d e d p (strain increments)
d xx
e
1 0 0 0 d ' xx
d e 1 0 0 0 d ' yy
yy
e 0 d ' zz
d zz 1
1 0 0
e E
d xy 0 0 0 2 2 0 0 d ' xy
e
d yz 0 0 0 0 2 2 0 d ' yz
d e 0 0 0 0 0 2 2 d ' zx
zx
3
2
t* = ½(’3 - ’1) t * 4
1 ' '
xx yy2 2
xy
s* = ½(’3+’1) s* 2
1 ' '
xx yy
1
2 '3 '1 c' cos ' 12 '3 '1 sin '
f 1
2
'3 '1 12 '3 '1 sin ' c' cos '
Note: Compression is negative and ’1 ’2 ’3
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES
f 1
2
'3 '1 12 '3 '1 sin ' c' cos '
= 30°
= 0°
-2 -3
d3
- 1
d1
E
d1 E
d3 1
- 1
d1
1
3
xy
p
d yy yy
tan
p
d xy
xy
dilatancy
-1 1
1
1- Eoed
-1
-3
NON-COHESIVE GRANULAR
MATERIAL
r
shear stress
reality
f=0 f ≈ r – As0.9
s
mean stress r
Models
f = r –As -B
A = sinφ
B = c cosφ s
εy
COHESIVE GRANULAR
MATERIAL : CLAY
We • wp < w < wl r
consider : • w = water
f=0
content
• wp = plastic
limit tension cut-off
• wl = liquid s
limit
test data : f ≈ r - As - B
• cohesion = (B/cosφ) depends on water content
• effective cohesion very small for w ≈ wl
• effective cohesion significant for w ≈ wp
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES
tension cut-off
s s
Questions ?
1. Introduction
2. Hyperbolic stress-strain curve
3. Shear strain contours
4. Unloading – reloading
5. Density hardening
6. Double hardening
Singapore 2011
26 of 607
Introduction
Universität Stuttgart
Stress path in oedometer loading - unloading
Introduction
Universität Stuttgart
Settlements for strip footing
0,2
s / smax [-]
0,4
Singapore 2011
27 of 607
Introduction
Universität Stuttgart
Mohr Coulomb
100
vertical displacements [mm]
80
Hardening Soil
60
40
20
0
0 5 10 15 20
-20
-40
Universität Stuttgart
Data from triaxial tests will be considered
Triaxial samples
The traditional triaxial sample is relatively long sample (H/D = 2) without lubricated endplates.
A high quality research sample is relatively short (H/D = 1) and it is tested with lubricated end plates.
Singapore 2011
28 of 607
Part 2 Hyperbolic stress-strain curve
Universität Stuttgart
Stress-strain curve from a standard drained triaxial test
1 σ 1
3 = constant
3
1
1 isotropic loading
σ 3
q σ1 σ 3
dense soil
q f = failure value
loose
qf q cs
q cs = critical state value
1
q σ 1 σ 3
Asymptote
q a = qasymptotive
Hyperbola
1
end of isotropic compression is taken as
undeformed reference state
qa q
ε1
2 E 50 q a q
Singapore 2011
29 of 607
Part 2 Hyperbolic stress-strain curve
Universität Stuttgart
Definition of secant modulus E50
q σ 1 σ 3 E50
1
50%
Hyperbola
50%
1
m
σ 3 a
E 50 E ref
50
p ref a
E ref
50
= reference modulus for primary loading at 50% of strength
q σ 1 σ 3
Asymptote
qf
Hyperbola qf qa R f 0.9
Rf
1
qf 2 sin
qa with q f 3 a a c cot
Rf 1 sin
Duncan & Chang (1970) Non-linear analysis of stress and strain in soils. Journal of the Soil Mechanics
and Foundation Division (ASCE), Vol. 96 No. SM5.
Singapore 2011
30 of 607
Part 2 Hyperbolic stress-strain curve
Universität Stuttgart
Parameters for hyperbolic law
effective cohesion : c
friction angle :
sin
sin a wi th R f 0.9
sin R f 1 sin
300 kPa 3
3 300 kPa c 10kPa
1 3
19
200
3 ’ = 200 E ref
50 15MPa
m 0.7
3 ’ = 100
100
m
measured σ 3 c cot a
hyperbolic E 50 E ref
50
p ref c cot a
0 ε 1 [%]
0 1 2 3 4
Data from: Amann, P.; Breth, H. & Stroh, D. (1975), Verformungsverhalten des Baugrundes beim
Baugrubenaushub und anschließendem Hochhausbau am Beispiel des Frankfurter Tons, Report 15 of
Geotechnical Institute of the University of Darmstadt.
Singapore 2011
31 of 607
Universität Stuttgart
Part 3
Universität Stuttgart
q σ 1 σ 3
Asymptote
qa
Hyperbola
qa q
1
2 E 50 q a q
3
γ shear strain ε 1 ε 3 ε1
2
3 qa q
γ
4 E 50 q a q
2 sin a
q a q asymptotic a asymptotic
qa σ 3 a
1 sin a
Singapore 2011
32 of 607
Part 3 Shear strain contours
q q
0 .05
0 .01
p´ p´
m
3 qa q σ 3 a 2 sinφ a
E 50 E ref
qa (σ 3 a)
4E 50 q qa 50 1 sin φ a
p ref a
Universität Stuttgart
Data for Fuji river sand after Ishihara et al. (1975)
1
3
p´
´
Singapore 2011
33 of 607
Part 3 Shear strain contours
3
2 3 1
2
´ ´
Calculated shear strain contours are shown by dotted lines. They are calculated by using,
m
3 qa q σ 3 a 2 sinφ a
E 50 E ref qa (σ 3 a)
4E 50 q qa 50
p ref a 1 sin φ a
a 0, φ a 38, 50 30 MPa,
Eref m 0.5
p'
´
Ref. : Tatsuoka & Ishihara (1974). “Yielding of sand in triaxial compression“. Soils and Foundations, Vol. 14, No. 2.
Singapore 2011
34 of 607
Part 3 Shear strain contours
Universität Stuttgart
Part 4
Unloading - Reloading
Singapore 2011
35 of 607
Part 4 Unloading - Reloading
Unloading – reloading in standard drained triaxial tests Universität Stuttgart
qf failure line
E50
deviatoric stress |1 - 3|
m
σ 3 a
E ur E ref
ur
p ref a
1
axial strain
loose : E ur ( 3 - 5) E 50 dense : E ur ( 2 - 3) E 50
Singapore 2011
36 of 607
Part 4 Unloading - Reloading
1
Δε 1e Δ σ 1 ν ur Δ σ 2 ν ur Δ σ 3
E ur
1
Δε e2 ν ur Δ σ 1 Δ σ 2 ν ur Δ σ 3
E ur
1
Δε 3e ν ur Δ σ 1 ν ur Δ σ 2 Δ σ 3
E ur
m
a
ν ur Poisson' s ratio 0 .2 E ur E urref 3 a c´cot ´
p ref a
Universität Stuttgart
Part 5
Density hardening
Singapore 2011
37 of 607
Part 5 Density hardening
Typical results of triaxial tests Universität Stuttgart
qf qr = qcrit
1
n n loose
loose
ncritical
dense
dense
clay sand
1 1
n = porosity
Henkel (1960) : The shear strength of saturated remoulded clays. Research Conf. on Shear Strength of
Cohesive Soils, Boulder, Colorado (ASCE).
Singapore 2011
38 of 607
Part 5 Density hardening
Measured contour lines for a kaolin clay Universität Stuttgart
q [MN/m²]
vol 0.25
p [MN/m²]
Biarez, J. & Hicher, P.-Y. (1994), Elementary Mechanics of Soil Behaviour, Balkema - Publishers.
Die Proben bestanden aus einem gesättigten Kaolin-Ton und wurde allseitig bis auf unterschiedliche Werte von p’ konsolidiert.
Danach wurden sie bis zum Bruch unter undrainierten Bedingungen einaxial belastet.
q [MN/m²]
vol 0.25
vol 0.20
0.14
0.07
p’ [MN/m²]
Biarez, J. & Hicher, P.-Y. (1994), Elementary Mechanics of Soil Behaviour, Balkema - Publishers.
Singapore 2011
39 of 607
Part 5 Density hardening
Ellipse as volumetric strain contour Universität Stuttgart
q
1
q2
M p pP
M 2 p
peak point
6 sin '
with : M
pp p 3 sin '
1
pp
2
Ellipse:
q2
p p p
Kref = reference bulk modulus M 2 p
Ishihara, Tatsuoka and Yasuda (1975). “Undrained deformation and liquefaction of sand under cyclic stresses“,
J. Soils and Foundations, Vol. 15, No. 1.
Singapore 2011
40 of 607
Part 5 Density hardening
Loading and unloading / reloading Universität Stuttgart
q 6sin cs q
M 6sin
3 sin cs
1 3 sin
1
β•pp
elastic elastic
pp p´ pp p´
NC
K 0 is controlled by β. On inputting realistic
no control of K NC value of one needs
K NC relatively large β-
0 0
values and one gets a relatively steep yield
cap
Singapore 2011
41 of 607
Universität Stuttgart
Part 6
Double hardening
q
q
elastic elastic
q HS-model
double hardening
pp p´
Singapore 2011
42 of 607
Part 6 Double hardening
Universität Stuttgart
Four stiffness zones
q
E50 & Eoed
combined hardening
E50
Eoed
shear hardening
cap hardening
Eur
elastic
pp
p
Universität Stuttgart
Appendix
Singapore 2011
43 of 607
Density Hardening of Soils Part 3: Contour lines in p-q-plane
Universität Stuttgart
The expression for vol is relatively simple
1 p
1 ε vol ref p
1 m K ref p ref
Ellipse:
Kref = reference bulk modulus q2
p p p
M 2 p
OCR POP
p p
Singapore 2011
44 of 607
Part 2 Hyperbolic stress-strain curve
ref
E 50
can be estimated from E ref
oed
Universität Stuttgart
50 E oed
E ref ref
50 2 E oed
E ref ref
NC-Clays:
E ref
sec
m m
1 a 3v a
E oed E ref E ur E ref
oed p a ur p a
ref ref
Singapore 2011
45 of 607
Part 3 Unloading - Reloading
HS-model does not include hysteresis Universität Stuttgart
q = 1 - 3
1
The unloading-reloading modulus Eur is the secant stiffness in large unloading-reloading loops. The modulus
E0 at small strain is a manifold of Eur.
σ 3v 100 kPa
e 0 0.54
Data for ε 1 3%
: E ur (1 .9 - 3.3) E 50
Singapore 2011
46 of 607
Complex reloading provides new yield point Universität Stuttgart
q q
p'
1
The last yield point and the new yield point are on a so-called yield locus. Soil
behaves essentially elastic as long as the stress remains below the yield
surface.
Yield loci of sand after Tatsuoka & Ishihara (1974) Universität Stuttgart
p'
Many samples of Fuji River sand were tested to obtain pairs of yield points and thus segments of yield loci by
Tatsuoka & Ishihara (1974): “Yielding of sand in triaxial comprtession”, Soils and Foundations 14 (2)
Singapore 2011
47 of 607
Data for a loose sand after Ishihara et al. (1975)
Universität Stuttgart
p ( kg / cm 2 )
1 kg / cm 2 100 kPa
1
3
9 I 1a I a2 / I a3
sin 2 m
2 1 I 1a I a2 / I a3
Singapore 2011
48 of 607
Hardening Soil & Hardening Soil with
Small Strain Overlay
William Cheang
Notes partly by: Dr Ronald Brinkgreve
Plaxis b.v. and TU Delft
Characteristics:
49 of 607
Hardening Soil model
Hyperbolic stress-strain relationship in (tri)axial loading:
(Duncan-Chang model)
q
1 qult
q
1 / E0 1 / qult E0
Rf qult
q / E0 Eur
1
1 q / qult
E0 = initial stiffness 1
qult = asymptotic value of q (related to strength)
Rf = ‘failure ratio’ (standard value 0.9)
50 of 607
Shear hardening in the HS model
Elastoplastic formulation of hyperbolic q-1 relationship:
1 q 2q
f* p
2 1p vp
E50 1 q qa Eur
m m
c cot ' ' 3 c cot ' ' 3
E 50 E ref
Eur E ref
c cot ' p c cot ' p
50 ref ur ref
1 q 2q
f* p
2 1p vp
E50 1 q qa Eur
m m
c cos ' '3 sin ' c cos ' '3 sin '
E50 E ref
Eur E ref
c cos ' p sin ' c cos ' p sin '
50 ref ur ref
51 of 607
Shear hardening in the HS model
Elastoplastic formulation of hyperbolic q-1 relationship:
Elastic
q MC failure line q MC failure line
plastic
m 3p,fric
2p,fric
1p,fric
p’
p’
52 of 607
Shear hardening in the HS model
Flow rule: d vp , fric d p , fric
sin m with:
sin m sin cv
sin m
q MC failure line 1 sin m sin cv
m>0 sin ' sin
m
cv sin cv
1 sin ' sin
m<0 (in principle)
Note:
m < 0 is not taken
into account
p’
Compaction hardening
q
in the HS model
MC failure line
Yield function (associated):
q2
f cap p 2 pc2
2
Cap
Hardening rule: pc
1 m
pc fc = 0
vp ,cap
1 m p ref pc p’
is determined by K0 nc c 1
is determined by Eoed
m
ref c cot ' '1
Eoed Eoed ref
c cot ' p v
53 of 607
Compaction and Shear hardening
in the HS model
Cap
Cone
54 of 607
Small-strain stiffness in the HS model
(HSsmall)
Strain(path)-dependent elastic overlay model:
G0
Gs
1 0.385 / 0.7
G0
1 0.385 / 0.7 2
Gt Gur
G starts again at G0
Gur after full strain reversal
Gt Cyclic loading
G0 leads to Hysteresis
Gs
G0
G0
55 of 607
Small-strain stiffness in the HS model
(HSsmall)
(Excel sheet)
G0 0.7
Gt Gs
Gur
56 of 607
Parameters of the HS(small) model
Parameters:
E50ref Secant stiffness from triaxial test at reference pressure
Eoedref Tangent stiffness from oedometer test at pref
Eurref Reference stiffness in unloading / reloading
G0ref Reference shear stiffness at small strains (HSsmall only)
0.7 Shear strain at which G has reduced to 70% (HSsmall only)
m Rate of stress dependency in stiffness behaviour
pref Reference pressure (100 kPa)
ur Poisson’s ratio in unloading / reloading
c’ Cohesion
’ Friction angle
Dilatancy angle
Rf Failure ratio qf /qa like in Duncan-Chang model (0.9)
K0nc Stress ratio ’xx/’yy in 1D primary compression
q
3=pref c 1=pref 1
qult
(, c)
E50ref qf=Rf qult
Eurref
0.5 qf 1
Eoedref
1 v
Triaxial test Oedometer test
57 of 607
Parameters of the HS model
Eoed [MPa] for NC-soils and ´ = 100 kPa
105
rock After Janbu (1963)
104
Janbu :
m
103
Eoed Eref
oed
sandy gravel pref
102
sand
10
more general:
m
Norwegian a
Eoed E ref
pref a
clays oed
1
ref
Eoed 1
2
ref
E50 Order of magnitude (very rough)
50000 kPa
ref
Eoed Correlation with Ip for pref=100 kPa
Ip
500 kPa
ref
Eoed Correlation by Vermeer
wL 0.1
ref
Eoed p ref * Relationship with Soft Soil model
58 of 607
Parameters of the HS model
ref
Eoed E50
ref
Order of magnitude by Schanz
Correlation by Lengkeek
ref
Eoed RD 60 MPa for pref=100 kPa
Schanz (1998)
59 of 607
Parameters of the HS model
Eurref (3 to 5) E50
ref
Eurref
G ref
(2.5 to10)G ref
where ref
G
2(1 ur )
0 ur ur
0.7 (1 to 2) 10 4
60 of 607
Initial conditions for the HS model
’yy0 c ’yy0 c
Initial stresses:
’yy
’yy0 follows from soil weight
Prestress and pore pressure
’c
Initial
CAP ’xx0 = K0 ’yy0
POP 1
ur
’yy0 1 ur
Initial stress
K 0nc '0yy POP ur
1 ur
POP
1 K0
K0nc
'0yy
ur
’xx0 ’xx K 0 OCR K 0nc OCR 1
1 ur
61 of 607
Initial conditions for the HS model
Initial stresses:
q
MC failure line
K0nc line
Output:
pc Cap
pc
p’0, q0 ' OCR ' OCRiso
p eq
62 of 607
Comparison HS model and MC model
Isotropic compression test:
Custom
1000 MC
HS.vlt
900
800
700
p' [kN/m²]
600
500
400
300
200
100
0
0 0.01 0.02 0.03 0.04 0.05 0.06
v
MC
HS.vlt
200
| 1 - | [kN/m²]
3
100
0
0 -0.01 -0.02 -0.03 -0.04 -0.05 -0.06 -0.07 -0.08
1
63 of 607
Comparison HS model and MC model
Drained triaxial test at 3=100 kPa :
Custom
0.009 MC
HS.vlt
0.006
0.003
v
-0.003
400
| 1 - | [kN/m²]
300
3
200
100
0
0 -0.01 -0.02 -0.03 -0.04 -0.05 -0.06 -0.07 -0.08
1
64 of 607
Comparison HS model and MC model
Drained / undrained triaxial test at 3=100 kPa :
E1DS
500
MC(u)
HS(u).vlt
MC.vlt
400 HS.vlt
| 1 - | [kN/m²]
300
3
200
100
0
0 -0.01 -0.02 -0.03 -0.04 -0.05 -0.06 -0.07 -0.08
1
300
q [kN/m²]
200
100
0
0 -100 -200 -300
p' [kN/m²]
65 of 607
Comparison HS model and MC model
One-dimensional compression test (oedometer):
Custom
1000 MC
HS.vlt
900
800
700
' 1 [kN/m²]
600
500
400
300
200
100
0
0 0.005 0.01 0.015 0.02 0.025 0.03
1
One-dimensional
MC
HS.vlt
-1000
(oedometer):
-800
-700
' 1 [kN/m²]
-600
-500
-400
-300
-200
-100
0
0 -200 -400
' 3 [kN/m²]
66 of 607
Comparison HS model and MC model
One-dimensional compression test (oedometer):
Stress state after unloading
HS MC
67 of 607
Hardening Soil model
68 of 607
Examples of parameter selection
200
150
dense 40
100
50
pref a
E50ref E50
0 'x a
0 1 2 3 4 5
100kPa
Axial strain [%] 20000kPa 32MPa
40kPa
Dense 40
-3 1-sin 1.2
-2
2sin Dense 40 2 sin 4.2
sin 0.29
-1
0 1 2 3 4 5
0
1 17
69 of 607
Examples of parameter selection
-0,4
-0,6
pa ' a
-0,8 ref
Eoed Eoed
-1 1 ' a
-1,2
850kPa 100kPa
-1,4 53MPa
0 200 400 600 800 1000 1200 0.008 400kPa
Vertical effective stress [kPa]
70 of 607
Examples of parameter selection
200
180
Deviatoric stress, q [kPa]
160
140
120
Plaxis 40
100 dense 40
80
60
40
20
0
0,00 1,00 2,00 3,00 4,00 5,00
Axial strain [%]
-3
-2,5
-2 from PLAXIS 40
Dense 40
-1,5
-1
-0,50,00 1,00 2,00 3,00 4,00 5,00
0
0,5
1
71 of 607
Examples of parameter selection
0
Test data
-0,2
Vertical strain [%]
-0,4 Plaxis
-0,6
-0,8
-1
-1,2
-1,4
0 200 400 600 800 1000 1200
Vertical effective stress [kPa]
72 of 607
CG5 Soft Soil Model with creep
1
Cs swelling
index
NC-Line = normal consolidation line
compression index Cc
1
log
ln
A
1
1
Cc
Cs 2.3 1 e0
strain ε
A
2.3 1 e0
e
0 existing stress
p preconsolidation stress
0
p
NC-Line
log
NC-Linie
creep
Cs
1
0 p
log
t
0
ec
1
Cα = secondary compression index
log t
tc
tc = end of consolidation
t + t
= c + CB log = c + CB log tc c = consolidation strain
tc tc tc = end of consolidation
+ t
e ec - C log with C (1+ e0 ) CB
Buisman (1936): Results of long duration settlement tests, Proceedings 1st International Conference on
Soil Mechanics and Foundation Engineering, Vol. 1, pp. 103-107
Garlanger (1972): The consolidation of soils exhibiting creep under constant effective stress,
Géotechnique 22 (1), pp. 71-78
e e
ee Cs log
Δe
ec Cc log
24 hours
(NCL)
1 week
log t log
tc 30´ τ = 24 h
Strain rate
NCL
e
N.B.
1 e0
creep
unloading
p p 0
0
log log
de c s c 1 c
e e e e c
dt ln10 ln10 p
Cc Cs
Typical soil data: Cs Cc / 10 and C Cc / 30 27
C
27
1
It follows that the creep rate is proportional to
OCR27
p
It follows that the creep rate is negligibly small for OCR >
1.4
Cc Cs
27
C ´ C
0.005 ´ 0.005 1
creep rate: e c
ln10 day day OCR27
p p
e
OCR ec/day (%) ec/year
1 0.5 -
1.1 0.04 0.146
1.2 0.0035 0.013
1.3 0.0004 0.0015
NC-line
1.4 0.00006 0.0002 with p
1.5 0.000009 0.00003
log
oedometer general
Cc
logp ec / Cc Cs lnppe ec /
ln10
27
C pe
27
C
e c e
c
ln10 pe
p p ln10
Definition of pe and constants λ and comes from Modified Cam Clay model
1 q
q 2
3 p 3 2
2 3 3 2 p
1
mean stress: p 1 2 3
3
1
deviatoric stress: q 1 2 2 2 3 2 3 1 2
2
q2
q pe p
CS - Line M2p
pe ppe
6 sinCS
M
3 sinCS
cap
CS Critical State
ec pp
e
Both with moving cap (density hardening):
q
MCC: cap is moved by primary loading
e c
q
c pe
ε c vol
σ
27
σ
ε c e c * pe
cvol
1 e0 ppe
p pe / p
pe ppe
q q
drained drained
failure
undrained undrained
p´ vertical strain 1
Undrained triaxial tests after Vaid et al. (1977) and simulation with SSC
model
undrained q
q
CS-Line
fast shearing
slow
p´
Vaid & Campanella (1977): Time-dependent behaviour of undisturbed clay, J. Geotech. Engng. Div., ASCE;
Vol. 103(7), pp. 693-709
Kulhawy & Mayne (1990): Manual on Estimating Soil Properties for Foundation Design
cap
p´
ppe
c´cot ´
6 sin
Mmc
3 sin
6 sincs
Mcs Value of cs can be selected such that K0nc = 1 – sin
3 sincs
´
The more difference between Mmc and Mcs the steeper the cap. The above picture would
suggest the possibility of tensile stresses, but these can be omitted by using a „tension cut-
off“.
Cc
deformation parameters: * modified compression index
1 e0 (1 e0 ) ln10
special parameters:
p
p ( or POP p 0 or OCR )
K 0 Knc
0 OCR 0 1 sin
Knc
POP p 0
y
0 p 0 p
y
p
Knc
0
1- vur POP
1
vur
y0
x
x0
ur ur
E ref
oed
* pref / E ref
oed
c ´, ´, c ´, ´,
p 0 p 0 ( input with OCR or POP )
E ref
50 -
- *
m always m = 1
q MC q MC
p´ p`
1278
5.5
1178
centre of gravity 58 m
22.5 m
wP = 28 % wL = 38 % w = 28 %
wP = 30 % wL = 70 % w = 52 %
wP = 13 % wL = 43 % w = 24 %
wP = 25 % wL = 51 % w = 38 %
Laval – Cc1
Laval – Cc2
Lancellotta e Pepe (1990)
Calabresi et al. (1993)
first analysis
last analysis
Cc w 0.1
* L
2.31 e 5
3 1 ur Cs
* * / 10
2.3 1 ur 1 e
C
* * / 30
2.31 e
z kPa
z0
1278
maximum excess
pore pressure 59
kPa
1370
maximum excess
pore pressure 3
kPa
1993
relative shear stresses
I MC
I MC
I MC
34
35
36
37
39
41
60
ult MCC MC
40
20 K0 Compression
0
-20 Extension
Measured qult
-40
-60
Mf=6sin()/3+sin(
-80
)
MC
-100
MCC Mf=6sin()/3-
-120 Soft Soil
-140 sin()
0 10 20 30 40 50 60 70 80 90 100 110
p' (kPa) =1/3×(′1+′2+′3)
42
Modified
Measur Mohr
Cam Soft Soil
ement* Coulomb
Clay
Tx.
33 40 30 33
Comp.
Tx.
21 28 30 24
Ext.
* For tx. compression cu/′v0 ≈ 0.33, for tx. extension cu/′v0 ≈ 0.21
43
60
40
MCC
20
0
K0
-20 Mf=6sin()/3+sin(
Measured qult
-40 )
-60
-80
MC
-100
MCC
-120 Soft Soil Extension Mf=6sin()/3-
-140
sin()
0 10 20 30 40 50 60 70 80 90 100 110
p' (kPa) =1/3×(′1+′2+′3)
44
Modified
Measur Mohr
Cam Soft Soil
ement* Coulomb
Clay
Tx.
33 61 35 42
Comp.
Tx.
21 43 35 30
Ext.
* For tx. compression cu/′v0 ≈ 0.33, for tx. extension cu/′v0 ≈ 0.21
45
46
47
cu by Method A
Direct input cu
c = cu, ′ = 0°
ESP by Mohr Coulomb in
undrained shearing
P′
ESP for real soft clays
48
49
Failure envelope
Elastic behaviour
50
p 1 / K 0 p
q 0 1 / 3G q
Note: p = v
51
u B 3 A1 3
u
Failure line
M.C. ESP
53
Yield locus
ESP
54
-40 0.6
A 0.5
-30 0.4 A = 0.33
-20 0.3
Mohr Coulomb
0.2
-10 Mohr Coulomb
0.1
0 0.0
0.00 -0.02 -0.04 -0.06 -0.08 -0.10 0.00 -0.02 -0.04 -0.06 -0.08 -0.10
a [-] a [-]
-ve u: compressive excess pwp
HK clay isotropically consolidated tx. comp.
55
Note:
q/p' ratio at peak
deviatoric stress < q/p'
ratio at critical state
56
Isotropic Anisotropic
model model
Yield locus
Yield locus
Modified Cam-Clay
Sekiguchi-Ohta (1977)
type
57
1 6 sin
cu
pini
2 3 sin
1 6 sin
cu
pini
2 3 sin
58
′ sin ′ K0 cu/’v0
59
p0 ini
M f pini
cu
2
2 pini
60
* *
M
M f pini 2
p0 ini *
cu 2
cap
2 M f M cap pini
2
61
Mcap
Mf
62
63
=24° =24°
64
65
66
67
Settlement [m]
0.00 Settlement
OCR0=2.0
of 10m
-0.05
thick soil
-0.10
OCR0=1.4 * = 0.10
-0.15 * = 0.02
* = 0.005
-0.20
ur = 0.15
OCR0=1.1 c′ = 0.0 kPa
-0.25
′ = 25°
-0.30 OCR0=1.0 = 0°
-0.35
K0nc = 0.577
0 200 400 600
Time (day)
800 1000 (1-sin ′)
68
69
Contents
• Modified Cam-Clay model and
critical state
• Soft Soil model
• Soft Soil Creep model
• Parameters
• Creep and undrained analysis
• SSC compare with MC Model
108 of 607
Modified Cam-Clay
q Modified Cam-Clay:
Yield contour is an ellipse
M
1
plastic
elastic
pp p’
q2
Yield function: f p' p' p p
M2
Modified Cam-Clay
Elastic deformation is generated according to:
p'
ee e0 e ln 0 unloading/reloading
p
Total deformation is generated according to:
p'
e e0 ln primary compression
p0
e = void ratio
= swelling index
= compression index
109 of 607
Modified Cam-Clay
Generally we prefer notation in strains:
p'
ve ve 0 * ln 0 , *
p 1 e
p'
vp vp 0 ( * * ) ln , *
0
p 1 e
εv = volumetric strain
* = modified swelling index
* = modified compression index
Modified Cam-Clay
q
q2
εs f 2 p' p' p p
M
dεs dε
M dεv
1
p’
pp εv
The Cam-Clay model has an associated flow rule, hence
the yield function f equals the plastic potential function g;
That is: f=g
110 of 607
Modified Cam-Clay
Plastic volumetric strains
p' p
vp vp 0 ( * * ) ln 0
( * * ) ln 0p
p pp
dp p
d vp ( * * )
pp
Hardening rule
Modified Cam-Clay
Right from the M-line (“wet side”): q < M p’
d vp 0, d sp 0 (compaction, hardening)
d vp 0, d sp Failure!
CSL f=0
111 of 607
Soft Soil
q
M
1
M K0nc
1
p’
pp
Soft Soil
Oedometer test
d xx d zz K0NC d yy , d xy 0
d xxe d zze , d xye 0
d xxp d zzp , d xyp 0
Gives a relation between K0nc and M:
*
(1 K0 )(1 2 ) * 1
M 3
(1 K 0 ) 2
(1 2 K0 ) 2 *
(1 2 K 0 )(1 2 ) (1 K 0 )(1 )
*
M 3.0 2.8K 0
(Brinkgreve, 1994)
112 of 607
Soft Soil
q
M K0NC
1
α
p’
pp
M determines both the strength and the amount of
plastic strains generated when following the K0NC path
0.4 K 0NC 0.6 1.3 M 1.9
6sin( )
tan( ) M 32 46
3 sin( )
15 30 0.6 K0NC 0.9
Soft Soil
q M
1 MC-line
K0NC
α
p’
pp
113 of 607
Soft Soil Creep
Model based on soil behaviour in oedometer test
• Creep decreases logarithmic with time
• Creep increases with load step, hence it increases with
effective stress level.
114 of 607
One-dimensional creep
modelling
One-dimensional compression:
B/C
d '1 C '1
d1 d1e d1c A
'1 p
with:
c
p p 0 exp (development of preconsolidation stress)
B
A = 1D swelling index (elasticity)
A+B = 1D compression index (total strain)
C = 1D creep index
15
One-dimensional creep
modelling
One-dimensional compression:
16
115 of 607
Soft Soil Creep model
3D model: Division of strains:
d d e d c d p
c dg c
d d1 Creep strains (viscoplastic, time-dependent)
d '
dg f
d p d2 Plastic strains according to MC (failure)
d '
17
Two problems:
• What is the “instantaneous plastic volume strain” in case of
undrained behaviour?
• Creep is also plastic behaviour – why the difference?
116 of 607
Soft Soil Creep
Therefore the Soft Soil creep model assumes:
* *
p ' * p eq * vc
v
e c *
eq , with p p e
eq eq * *
p'
v v p p0
pp
v * eq
e c *
, with p p e
eq eq * *
v v
p' p p p0
p
q M
1
p eq p eq p’
p
p p
117 of 607
Soft Soil Creep
Under constant load the cap expands in time and so
generates plastic volume strains: creep!
• Hence, the preconsolidation pressure Pp increases
in time and therefore so does the OCR.
118 of 607
Parameters
Parameters *, * and μ* are obtained from oedometer tests
μ* is the creep rate when OCR = 1
ε1=εv ε1=εv
ln p’ ln t
1
*
μ* 1
* 1
Parameters
Parameters *, * and μ* can be linked to internationally
normalized parameters:
Cc 2 Cr C
* * *
2.3(1 e) 2.3 1 e 2.3(1 e)
119 of 607
Parameters
* * primary compression plasticity
Creep ratio:
* creep plasticity
CR < 5
Very creep-sensitive materials; fresh deposits, soft sediment
CR > 25
Materials with little creep.
One may consider using the Soft Soil model instead
Parameters
Overconsolidation pressure pp
• Important parameter: it determines the initial creep rate
• No direct input in Plaxis, but input by means of OCR or POP
120 of 607
Parameters
OCR - Simulation of history
1. Perform K0 procedure with OCR=1 or
Gravity Loading in 1 day
2. Define a next calculation phase, plastic nil-step,
with a time interval equal to the estimated geological time in da
that the material has been at the site since deposition, that is, the
last primary loading step.
OCR - Estimation
Perform K0 procedure with OCR according to:
* *
ln( t )
*
OCR e
121 of 607
Creep and undrained
analysis
Creep and consolidation
Creep
Increases excess pore pressures and consolidation
time; BUT Decreases effective stresses and so the
creep rate
Consolidation
Increases effective stresses and so the creep rate; BUT
Decreases excess pore pressures
x y
B * = 0.025
* = 0.005
c = 1 kPa
φ = 28º
=0º
1m
A
3 alternatives:
μ*= 0.0040 (CR = 5)
μ*= 0.0020 (CR = 10)
μ*= 0.0001 (CR = 200)
122 of 607
Creep and undrained
analysis
Excess PP [kN/m2]
-120
CR = 5
CR = 10
CR = 200
-90
0
1 10 100 1e3 1e4 1e5
Time [day]
0.03 26 mm
CR = 5
CR = 10
CR = 200
0.06
77 mm
0.09
97 mm
0.12
1 10 100 1e3 1e4 1e5
Time [day]
123 of 607
Creep and undrained
analysis
Uy [m]
0
-0.02
-0.04
CR=10
-0.06
OCR = 1.5
-0.08 OCR = 1.2
OCR = 1.0
-0.1
1 10 100 1e3 1e4 1e5
Time [day]
Cam-Clay * * -
1 e 1 e
1 2 1
Dutch literature * * *
C p' C p ( ontl .) 2.3 C's
Den Haan * A B * 2A * C
Cc Cs C
International lit. * * *
2.3(1 e) 1 e 2.3 (1 e)
34
124 of 607
Comparison of SSC and MC
model
Stress-strain development in various stress paths:
* 0.10
* 0.02 E 2000 kPa
* 0.005 0.30
ur 0.15 c’ 0.0 kPa
c’ 0.0 kPa ’ 25°
’ 25° 0°
0°
K0nc 0.677 (1-sin ’)
35
1000
p' [kN/m2] MC SSC
800
600
400
200
0
0 0.2 0.4 0.6 0.8
eps-v
36
125 of 607
SSC model in standard tests
Isotropic compression:
1000
p' [kN/m2]
100 1 1
* *
10
1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
eps-v 37
38
126 of 607
SSC model in standard tests
CID tri-axial test at different loading rates:
160
q' [kN/m2]
120
900 kPa/d
80
9 kPa/d
40
0
0 0.05 0.10 0.15 0.20 0.25 0.30
eps-1
39
20
0
0 0.01 0.02 0.03 0.04 0.05
eps-1
40
127 of 607
SSC model in standard tests
CIU-test at different loading rates (only creep):
80
q' [kN/m2] 900 kPa/d
90 kPa/d
60
9 kPa/d
40
20
0
0 20 40 60 80 100
p' [kN/m2]
41
160
9 kPa/d
q' [kN/m2]
120
90 kPa/d
900 kPa/d
80
9000 kPa/d
40
00 40 80 120 160
p' [kN/m2]
42
128 of 607
Comparison of SSC and MC
model
The role of OCR in self-weight loading and creep:
Settlement
Settlement [m]
0.00 of 10m
OCR0=2.0
thick layer
-0.05
* 0.10
-0.10
OCR0=1.4 * 0.02
-0.15 * 0.005
ur 0.15
-0.20
c’ 0.0 kPa
-0.25
OCR0=1.1
’ 25°
0°
-0.30 OCR0=1.0
K0nc 0.677
-0.35
0 200 400 600 800 1000 (1-sin
Time ’)
Geo-engineering
129 of 607
Soft-Soil Creep Model
Geo-engineering
Summary
• Soft soil creep:
• Is a time dependent cam-clay based model
• Distinguishes between elastic and creep strains
where creep includes all plastic deformations
• Elastic deformations only depend on *
• Plastic deformations depend on *,μ* and
the initial OCR
• Parameter determination can be difficult,
simulation of a laboratory test is advised to
check model parameters.
130 of 607
Thank you
Questions?
Geo-engineering
131 of 607
COMPUTATIONAL GEOTECHNICS COURSE
CG7a CONSOLIDATION
CONSOLIDATION: OUTLINE
Introduction
Basic theory of groundwater flow
Permeability
Confined and unconfined problems
Finite element formulation for consolidation
Mechanical problem
Hydraulic problem
Global equations
Time step
Boundary conditions
Singapore 2011
132 of 607
TYPES OF ANALYSIS
Drained
Loading/Construction/ excavation: very slow (in relation to the soil
permeability)
Undrained
Loading/Construction/ excavation: very fast (in relation to the soil
permeability)
EXAMPLE
Excess pore
water pressure
Consolidation
Singapore 2011
133 of 607
OTHER EXAMPLES
h
q k k i
L
h
q k k
dy dy
p
h y
w
Singapore 2011
134 of 607
BASIC THEORY OF GROUNDWATER FLOW
Permeability often anisotropic
qx k x q y k y
x y
q x q y
0
x y
PERMEABILITY
Dependence on grain size
Soil k (cm/s)
Clean gravel >1
Clean sand 1 - 10-2
(coarse)
Sand mixture 10-2 - 5x10-3
Fine sand 5x10-2 -10-3
Silty sand 2x10-3 -10-4
Silt 5x10-3 -10-5
Clay 10-6 and less
Harr (1962)
Singapore 2011
135 of 607
PERMEABILITY
Dependence on void ratio
PERMEABILITY
k e
log
k 0 ck
Default value for c k is 1015
Singapore 2011
136 of 607
TYPES OF FLOW PROBLEMS
TRANSITION SATURATED/UNSATURATED
qx K r kx
x
qy K r ky
y
Kr 1 saturated zone
K r 10 4 unsaturated zone
4h
K r 10 4 h hk log( K r )
hk
hk 0.7m (PLAXIS)
Singapore 2011
137 of 607
FINITE ELEMENT FORMULATION FOR CONSOLIDATION (1)
Effective stresses
Constitutive law
Discretization
Stiffness matrix
Coupling matrix
Singapore 2011
138 of 607
FINITE ELEMENT FORMULATION FOR CONSOLIDATION (2)
Flow matrix
Coupling matrix
Singapore 2011
139 of 607
FINITE ELEMENT FORMULATION FOR CONSOLIDATION (4)
Time step
Automatic time stepping is required
Critical time step
Consolidation analysis
Prescribed time
Maximum excess pore pressure
BOUNDARY CONDITIONS
Singapore 2011
140 of 607
BOUNDARY CONDITIONS
BOUNDARY CONDITIONS
Prescribed flow
Prescribed head
Singapore 2011
141 of 607
INITIAL PORE PRESSURE AND EXCESS PORE PRESSURE
Sand
Clay Δpw
Sand pw0
k e
log
k 0 ck
Default value for c k is 1015
There may be large contrasts of permeability between different materials in the same
problem
• Too much permeability contrast may cause numerical difficulties
• The ratio between the highest and lowest permeability value should not exceed
105
• To simulate an almost impermeable material (e.g. concrete), a value lower by a
factor 1000 is sufficient
Singapore 2011
142 of 607
1D CONSOLIDATION - NUMERICAL SIMULATION
Parameters for Soft Soil model are chosen accordingly, stiffness for MC
model represents oedometer stiffness of HS model in middle of layer
ISOCHRONES
permeable
Tv
D
permeable
degree of consolidation Ut
Singapore 2011
143 of 607
1D CONSOLIDATION - NUMERICAL SIMULATION
Singapore 2011
144 of 607
1D CONSOLIDATION - NUMERICAL SIMULATION
MC Model
This image cannot currently be display ed.
0,1
0,2
Degree of consolidation
0,3
0,4
2D Axi WithOut PVD
0,5
2D Axi PVD
0,6
2D Axi PVD&Smear
0,7
Terzaghi WithOut PVD (Analytical)
0,8
Barron PVD (Analytical)
0,9
Barron PVD&Smear (Analytical)
1
time (day)
Singapore 2011
145 of 607
1D CONSOLIDATION - COMPARISON WITH ANALYTICAL SOLUTION
HS Model
This image cannot currently be display ed.
0,1
0,2
Degree of consolidation
0,3
0,4
2D Axi WithOut PVD
0,5
2D Axi PVD
0,6
2D Axi PVD&Smear
0,7
Terzaghi WithOut PVD (Analytical)
0,8
Barron PVD (Analytical)
0,9
Barron PVD&Smear (Analytical)
1
time (day)
Singapore 2011
146 of 607
01223ÿ5617289ÿ
2698
6
6ÿ238ÿ
9ÿ11
89ÿ62ÿ861
9ÿ2
ÿ!"ÿ#$%&$'()
147 of 607
148 of 607
149 of 607
150 of 607
151 of 607
152 of 607
153 of 607
154 of 607
155 of 607
156 of 607
157 of 607
158 of 607
159 of 607
160 of 607
161 of 607
162 of 607
163 of 607
164 of 607
165 of 607
166 of 607
167 of 607
168 of 607
169 of 607
170 of 607
CG8 - Drained and Undrained
Analysis
Prof Harry Tan Siew Ann
Centre for Soft Ground Engineering
National University of Singapore
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES Drained and Undrained Behavior
Contents
• Definition drained / undrained
• Drained / undrained soil behaviour
• Typical results from drained and undrained triaxial tests
• Skempton‘s parameters A and B
• Modelling undrained behaviour with Plaxis
• In terms of effective stresses with drained strength parameters
• In terms of effective stresses with undrained strength parameters
• In terms of total stresses
• Influence of constitutive model and parameters
• Influence of dilatancy
• Undrained behaviour with Mohr-Coulomb Model
• Undrained behaviour with Hardening Soil Model
• Excavation Example
• Summary
Singapore 2011
171 of 607
Drained / undrained
Drained / undrained
k = Permeability
k E oed
T t Eoed = Oedometer modulus
γ w D2 w = Unit weight of water
D = Drainage length
t = Construction time
T = Dimensionless time factor
U = Degree of consolidation
Singapore 2011
172 of 607
Undrained behaviour
UNDRAINED BEHAVIOUR
q
e
lin
advanced lu re
fai
models
advanced
models
elastic-perfectly
cu,3 plastic models
cu,1 cu,2
pc’ p’
Results from undrained triaxial tests using simple and advanced constitutive models
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES
Singapore 2011
173 of 607
UNDRAINED BEHAVIOUR
q [kN/m 2 drained
q [kN/m2] ]
350 Mohr-Coulomb / Soft Soil
300
undrained
250
Mohr-Coulomb
200
undrained
Soft Soil
150
100
50
0
0 -50 -100 -150 -200 -250 -300 -350
p' [kN/m2]
p' [kN/m2]
Undrained shear strengths predicted by Mohr-Coulomb and Soft Soil Model
Singapore 2011
174 of 607
Triaxial test (OC) – drained / undrained
Typical results from drained (left) and undrained (right) triaxial tests on overconsolidated soils
Singapore 2011
175 of 607
Skempton’s parameters A and B
Skempton 1954: p w B 3 A 1 3
- Fully saturated soil
- No inflow / outflow of pore water
- Bulk modulus of soil grains is considered to be very high
- Isotropic linear elastic material behaviour (Hooke´s law)
1 2 3 3 pw K w
p w
3K ' n
1 1
leading to pw
nK ' 3 3 1 3
1
Kw
1 1
with B A
nK ' 3
1
Kw
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES 12
Singapore 2011
176 of 607
Skempton’s parameters A and B
Singapore 2011
177 of 607
Undrained behaviour with PLAXIS
PLAXIS automatically adds stiffness of water when undrained material type is
chosen using the following approximation:
Kw Eu 2 G 1 u
K total K'
n 31 2 u 31 2 u
E' 1 u
K total assuming u = 0.495
3 1 2 u 1 '
Notes:
• This procedure gives reasonable B-values only for ´ < 0.35 !
• Real value of Kw/n ~ 1.106 kPa (for n = 0.5)
• In Version 8 B-value can be entered explicitely for undrained materials
Example 2:
Singapore 2011
178 of 607
Undrained behaviour with PLAXIS
Method A (analysis in terms of effective stresses):
type of material behaviour: undrained
effective strength parameters c´, ´, ´
effective stiffness parameters E50´, ´
• Method A:
• Recommended
• Soil behaviour is always governed by effective stresses
• Increase of shear strength during consolidation included
• Essential for exploiting features of advanced models such as the Hardening Soil
model, the Soft Soil model and the Soft Soil Creep model
• Method B:
• Only when no information on effective strength parameters is avilable
• Cannot be used with the Soft Soil model and the Soft Soil Creep model
• Method C:
• NOT recommended
• No information on excess pore pressure distribution (total stress analysis)
Singapore 2011
179 of 607
Undrained strength from Mohr circle
Consider fully undrained isotropic elastic behaviour
(Mohr Coulomb in elastic range)
pw = p > p´ = 0
cu
1 'o
'yo sin ' c' cos '
2 x
Parameters for Soft Soil model and Mohr Coulomb model accordingly
Singapore 2011
180 of 607
UNDRAINED STRENGTH: MOHR COULOMB vs ADVANCED MODELS
125
Advanced models - Method A
100
(HS / HSS / SS)
75
2cu
50
q [kN/m ]
2
-25
-50
-75
-100
0.00 -25.00 -50.00 -75.00 -100.00 -125.00 -150.00
p' [kN/m 2 ]
-75
-50
Hardening Soil model
-25
cu
t [kN/m ]
2
50
s = ½(x + y)
t = ½(x - y)
75
0.00 -25.00 -50.00 -75.00 -100.00 -125.00 -150.00
2
s' [kN/m ]
Plane strain compression:
cu/v' = 0.32 (MC)
cu/v' = 0.29 (HS)
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES
Singapore 2011
181 of 607
UNDRAINED STRENGTH: MOHR COULOMB - HARDENING SOIL MODEL
200
150
Hardening Soil model 2cu (NC)
Test 1
100
cu = 0 (OC)
50 Test 2
Test 3
q [kN/m ]
2
-50
-100
-150
Mohr Coulomb model
-200
0.00 -25.00 -50.00 -75.00 -100.00 -125.00 -150.00 -175.00 -200.00 -225.00
p' [kN/m 2 ]
Hardening Soil
Method B
-10
0 s = ½(x + y)
0.00 -25.00 -50.00 -75.00 -100.00
Singapore 2011
182 of 607
UNDRAINED STRENGTH: MOHR COULOMB vs HARDENING SOIL MODEL
Mohr Coulomb
2
Method A
-10 Mohr Coulomb
Method B
(cu based on HS-Method A) Hardening Soil
Method B
(cu based on HS-Method A)
0
0.00 -10.00 -20.00 -30.00 -40.00
s' [kN/m 2 ]
Consequence: consolidation analysis following undrained analysis with method B by means of "model change"
(introducing effective
PLAXIS FINITE ELEMENT strength
CODE FOR SOILparameters) starts with incorrect pore pressures
AND ROCK ANALYSES
300
275
250
225
200
175
q [kN/m ]
2
150
125
100
75
M C non dil
50 M C dil
H S_1 non dil
25 H S_1 dil
0
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00
1 [% ]
Simulation of undrained triaxial compression test – MC / HS model - q vs 1
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES
Singapore 2011
183 of 607
UNDRAINED STRENGTH: INFLUENCE OF DILATANCY
300
225
200
175
q [kN/m ]
2
150
125
100
75 MC non dil
MC dil
50 HS_1 non dil
HS_1 dil
25 total stress path
0
0.00 25.00 50.00 75.00 100.00 125.00 150.00 175.00 200.00 225.00 250.00
2
p' [kN/m ]
Simulation of undrained triaxial compression test – MC / HS model - q vs p´
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES
100
90 M C non dil
M C dil
80
H S_1 non dil
excess pore pressure [kN/m ]
2
70 H S_1 dil
60
50
40
30
20
10
-10
-20
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00
1 [% ]
Simulation of undrained triaxial compression test – MC / HS model - pw vs 1
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES
Singapore 2011
184 of 607
UNDRAINED STRENGTH: INFLUENCE OF DILATANCY
1.0
0.9 M C non dil
0.8 M C dil
HS _1 non dil
0.7 HS _1 dil
0.6
0.5
parameter A
0.4
0.3
0.2
0.1
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00
1 [% ]
Simulation of undrained triaxial compression test – MC / HS model - A vs 1
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES
125
75 K0nc decreased
50
q [kN/m ]
2
25
-25
-50
-75
-100
0.00 -25.00 -50.00 -75.00 -100.00 -125.00 -150.00
2
p' [kN/m ]
Singapore 2011
185 of 607
UNDRAINED STRENGTH: INFLUENCE OF Eoed and K0nc
125
100
75 HS-TC-Eoed >
HS-TC
nc
HS-TC-K0 <
50
HS-TC-Eoed + K0nc >
q [kN/m2]
25 HS-TE-Eoed >
HS-TE
nc
HS-TE-K0 <
0
HS-TE-Eoed + K0nc >
CSL-TC
-25
CSL-TE
-50
-75
-100
0.00 -25.00 -50.00 -75.00 -100.00 -125.00 -150.00
p' [kN/m2]
p
sin 'm
sin
ˆ m sin m
sin '
e.g. SOREIDE 2003
Singapore 2011
186 of 607
EXAMPLE - EMBANKMENT CONSTRUCTION
Singapore 2011
187 of 607
EXAMPLE - EMBANKMENT CONSTRUCTION
"fast": failure
-50
slow
fast fast
-40
-30
-20
slow
-10
0
0 4 8 12 16
Time [day]
time [days]
Singapore 2011
188 of 607
EXAMPLE - EMBANKMENT CONSTRUCTION
s = ½ (1 + 3)
s = ½ (1 + 3)
from Ortigao, 1995
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES
-10
breiiger mariner TON
UPPER MARINE CLAY
-15
schluffiger TON
-20
Depth [m]
schluffiger TON
-40
altes Schwemmland feinschluffiger grober
SAND schluffiger TON
OLD ALLUVIUM
-45
0 20 40 60 80 100
cu-Profile according to geotechnical design parameter
c_u [kN/m²]
table
PLAXIS FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSES
Singapore 2011
189 of 607
COMPARISON METHODS A AND B
Practical example: deep excavation in soft clay
100
Cu (Mohr-Coulomb)
70
65
60
0 20 40 60 80 100
E (U) E (U)
Singapore 2011
190 of 607
SUMMARY
cv * t
T
H2
Cv=66 m2/day
H = 15 m
Cv=125 m2/day
Cv=53 m2/day
Singapore 2011
191 of 607
What is West Coast Station Situation ???
• Cv= 50 m2/day
• H=15 m
• t = 100 days
• T=50*100/(15*15) = 22.2 >>> 0.4
• Situation on Passive Side is likely to be DRAINED
Condition
Singapore 2011
192 of 607
Cases of k=1e-7 to 1e-9 m/s
Displacements at Formation Level
Singapore 2011
193 of 607
Wall Deflection at B (15/83.85 – 1.65m above FL)
U x at B
Ux at B [m]
0.2
DRN
UND
0.15
k=1e-7
k=1e-8
0.1 k=1e-9
0.05
0
0 50 100 150 200 250
Time [day]
0.03
UND
0.025
k =1e-7
0.02
k =1e-8
0.015 k =1e-9
0.01
5e-3
-5e-3
0 30 60 90 120
Time [day ]
Singapore 2011
194 of 607
References
Atkinson, J.H., Bransby, P.L. (1978)
The Mechanics of Soils, An Introduction to Critical State Soil Mechanics. McGraw Hill
Ortigao, J.A.R. (1995)
Soil Mechanics in the Light of Critical State Theories – An Introduction. Balkema
Schweiger, H.F. (2002)
Some remarks on pore pressure parameters A and B in undrained analyses with the Hardening Soil
Model. Plaxis Bulletin No.12
Skempton, A.W. (1954)
The Pore-Pressure Coefficients A and B. Geotechnique, 4, 143-147
Vermeer, P.A., Meier, C.-P. (1998)
Proceedings Int. Conf. on Soil-Structure Interaction in Urban Civil Engineering, Darmstadt, 177-191
Singapore 2011
195 of 607
Structural Elements &
Modelling Excavations in Plaxis
Based on original course note by Ronald
Brinkgreve, Plaxis B.V.
Uç
RF Shen
24 November 2011
• Plate element
• Anchor element
• Geogrids element
• Interface element
196 of 607
Application of structural elements
1. Plate Element
Overview:
• 3 or 5 noded line elements
(for 6‐noded or 15‐noded element mesh)
• 3 degrees of freedom per node
• Plates have:
– Axial forces
– Shear forces
– Bending moments
– Hoop forces (axisymmetry)
• Elastic or elastoplastic behaviour
• For modelling walls, floors, tunnels
197 of 607
Plate Element
Plates – elastic parameters
h3 b
EI E (b = 1 m)
12
EA E h b (b = 1 m)
EI (Equivalent rectangular
d h 12
EA plate thickness)
h h
b
b = 1 m in plane strain
b = 1 meter in axisymmetry
b
Plate Element
Plates – elasto‐plastic behaviour
Np
M
Mp
198 of 607
Plate Element
(Illustration: Mp‐Np.P2D):
1200 ‐100‐90 ‐80 ‐70 ‐60 ‐50 ‐40 ‐30 ‐20 ‐10 0 10 20 30 40 50 60 70 80 90 100
Envelope 0
Elasto‐plastic plate
1000
Elastic plate
‐5
800
Elasto‐plastic plate
‐10 Elastic plate
600
N
400 ‐15
200 ‐20
0
‐25
‐200 0 200 400 600 800 1000 1200 M
M
Plate Element
Effect on Global FOS by c/phi Reduction
• Elastic wall excludes possibility of wall plastic hinge; and over-estimate FOS=1.75
• Allowing for wall plastic hinge (Elasto-plastic wall) gave lower FOS=1.40 and
smaller soil yielded zone behind the wall
8
199 of 607
Plate Element
Plates – weight, in soil
dreal
Plate Element
Plates – weight, excavation
Actual problem In the model
dreal
1
wreal = concrete d real wmodel = soil d real wplate
2 Below GT
soil sat
1
wmodel = wreal w plate = ( concrete soil ) d real Above GT
2
soil unsat
200 of 607
Plate Element
Plates – connections
Spring data:
• Stiffness
6 8
• Min/Max moment
Rotation
spring
5 7
Hinged connection
Rigid connection
(default)
Illustration: Connection.P2D
Plate Element
Walls – thin wall vs. thick wall
• Thin wall
– Wall thickness << wall length
– No much end‐bearing, only friction
→ Plate element suffices
• Thick wall
– Wall thickness significant
– End‐bearing capacity needed
→ Use soil elements with material set represen ng wall material
→ In order to obtain structural forces a plate with fictitious
properties may be inserted
201 of 607
Plate Element
(Illustration: Beam.P2D): d
2. Anchor Element
Anchors – fixed‐end
• To model supports, anchors and struts
– Elasto‐plastic spring element
– One end fixed to point in the geometry,
other end is fully fixed for displacement
– Positioning at any angle
– Pre‐stressing option
Anchors – node‐to‐node
• To model anchors, columns, struts and rods
– Elasto‐plastic spring element
– Connects two geometry points in the geometry
– No interaction with the mesh along the anchor
rod
– Pre‐stressing option
202 of 607
Anchor Element
Anchors – material properties
Axial stiffness, EA (for one anchor) [kN]
Spacing, Ls (out‐of‐plane distance between anchors) [m]
Maximum anchor force for compression and tension,
|Fmax,comp| and |Fmax,tens| [kN]
Ls
Anchor Element
Anchors – pre‐stressing
• Defined in Staged construction phase
• Both tension (grout anchor) or compression (strut)
possible
Tension = positive
203 of 607
3.Geogrid Element
Geogrids
• 3 or 5 noded line element
• Elastic or elasto‐plastic behaviour
• No flexural rigidity (EI), only axial stiffness (EA)
• Only allows for tension, not for compression
Ground anchors
• Combination of node‐to‐node anchor and geogrid
• Node‐to‐node anchor represents anchor rod (free length)
(no interaction with surrounding soil)
• Geogrid represents grouted part (full interaction with surrounding soil)
• No interface around grouted part; interface would create unrealistic failure
surface
• Working load conditions only – no pullout
• If pullout force is known this can be used by limiting anchor rod force
204 of 607
Ground anchors
Axial forces in ground anchors:
Input
geometry
4. Interface Element
Interfaces – material properties
• Soil‐structure interaction
– Wall friction
– Slip and gapping between soil and structure
• Soil material properties
– Taken from soil using reduction factor Rinter
Cinter = Rinter * Csoil
tan(φinter) = Rinter * tan(φsoil)
ψinter = 0 for Rinter < 1
= ψsoil Rinter = 1
σt,inter = Rinter * σt,soil
Ginter = (Rinter)2 * Gsoil
– Individual material set for interface possible
205 of 607
Interface Element
Interfaces – reduction factor
Suggestions for Rinter:
– Interaction sand/steel = Rinter ≈ 0.6 – 0.7
– Interaction clay/steel = Rinter ≈ 0.5
– Interaction sand/concrete = Rinter ≈ 1.0 – 0.8
– Interaction clay/concrete = Rinter ≈ 1.0 – 0.7
– Interaction soil/geogrid (grouted body) = Rinter≈ 1.0
(interface may not be required)
– Interaction soil/geotextile = Rinter≈ 0.9 – 0.5 (foil, textile)
Pile‐soil interaction ‐ NSF on pile
(Ref: PhD thesis by RF Shen (2008)
-9 curve
-10
-11
-12
-13
-14
-15
-16
Rigid base
-17
-18
-19
-20
-21
When using MC model and MCC model, the friction angle of the interface
element should be defined as follows: 9
4 M
2
(1 ' ) 'i nt arctan 4
'i nt arctan
9
' 9 2 M
2
(MC model) 4 (MCC model)
206 of 607
Illustration: Interface_slide.P2D
4m
Unit weight = 25kN/m^3 1m
Interface properties: Cint = 2.5kPa; inter=26.6
Expected sliding resistance = Block length (4m) * Cint + Block weight (100kN) * tan (inter) =
60 kN
Material Behaviour in Excavation
• Unloading due to excavation
– Vertical unloading at excavation bottom
– Horizontal unloading behind wall (may accompanied by
shear plasticity)
• Primary loading due to pre‐stressing
• HS‐small model is preferred
– Non‐linear elastic unloading/reloading behaviour
– Shear plasticity due to horizontal unloading
– High far‐field stiffness for better settlement trough
prediction
207 of 607
Material behaviour: Stress paths
Construction phases:
Kactive K0
I 1st excavation σv
II Pre-stressing anchor K=1
III Final excavation Point A
II III
I
Point A
Point B
Kpassive
Point B
σh
Eur , E50
Eoed
Eur , E50
Eur
208 of 607
Material behaviour
• Mohr‐Coulomb: unrealistic deformations
– Use of single E fails to cater for the complex material at various zones
– Overestimation over bottom heave
– Sometimes heave of soil behind the wall
– Soils below excavation behaves with Eur, even soils behind wall behaves
between Eur and E50. Use of E50 is too conservative.
• Hardening Soil model: qualitative realistic deformations
– Soil stiffness for Isotropic loading, shearing and unloading‐reloading can
be catered for automatically in the model.
– More realistic bottom heave
– Improved settlement trough behind wall.
• HS‐small model: qualitative and quantitative realistic deformations
– Improved version over HS to take care of far field small strain behavior
– More realistic settlement trough behind the wall (narrower and deeper)
Hands-on session
E3 – Excavation Exercise
209 of 607
Modelling groundwater in PLAXIS
Ronald Brinkgreve
Plaxis BV / Delft University of Technology
Contents
• Pore pressures
– Generation
– Definition
– Steady-state
• Groundwater flow
– Flow in unsaturated soil
– Material data sets
– Boundary conditions
– Finite Element Modelling
Singapore 2011
210 of 607
Pore pressures - generation
• Hydrostatic
– Based on defined phreatic levels
– Takes only water weight into account
– Aquifers possible
– Simple, horizontal water levels – no flow
– Undisturbed subsoil – simple excavations
• Steady-state groundwater flow
– No flow field change in time
– Long term solution
– Constant water levels
– Most excavations, some embankments
Singapore 2011
211 of 607
Pore pressures - generation
• Hydrostatic pore pressures
– PLAXIS 2D, PLAXIS 3D, (PlaxFlow 3D Tunnel, 3D Foundation)
• Steady-state groundwater flow
– PLAXIS 2D, PLAXIS 3D*, (PlaxFlow, 3D Tunnel)
• Transient groundwater flow
– PLAXIS 2D, (PlaxFlow)
Singapore 2011
212 of 607
Pore pressures - modes
• Classical mode
– Semi-coupled analysis.
– Full coupling between deformation and excess pore pressures.
– Steady-state pore pressures are generated prior to calculation and are thus INPUT.
– Changes in steady-state pore pressures may change excess pore pressures and
deformations, but not v.v.
– Soil is either fully saturated or dry
• Advanced mode
– Fully coupled analysis.
– Full coupling between deformation, steady-state pore pressures and excess pore
pressures.
– Changes in steady-state pore pressures, excess pore pressures and deformations
influence each other.
– Soil can be partially saturated
Advanced mode was introduced in Plaxis 2D 2010. Earlier versions only have classical
mode.
• Pore pressure based on distance below phreatic level and water weight
• Should only be used for horizontal water levels
• Different soil layers can have different phreatic levels
p = h * w
Singapore 2011
213 of 607
Pore pressures – steady-state
Phreatic levels
Layer 1
General phreatic level
Interpolate Layer 2
Singapore 2011
214 of 607
Pore pressures – steady-state
Steady-state flow
4 8 11 5
General General
General
16 17
21 26
19 18
20 23 24 27
22 25
13 14 12
15 7 10
0 1
Singapore 2011
215 of 607
Pore pressures – steady-state
Transient flow
k k re l k sa t , k re l f h p , S
(h p )
S (h )
n
Singapore 2011
216 of 607
Groundwater flow – flow in unsaturated soil
Linear Model
krel
1 hp 0 Saturated
1 k rel 1 m h p 0 hp Partially saturated
hp Dry
m
0 hp
β
Singapore 2011
217 of 607
Groundwater flow – flow in unsaturated soil
van Genuchten model
1 g n
S ( h p ) S res ( Ssat S res ) 1 g a h p
gn ( g )
n
2
g n 1
g n g n
S Sres
krel S S e l 1 1 S e n
g g 1
with Se
Ssat Sres
1 if hp 0
4h p Log-linear in Permeability
h
krel h p 10 pk if h pk h p 0
4
10 if h p h pk
Singapore 2011
218 of 607
Groundwater flow - material data sets
• Parameters:
– Permeabilities (kx, ky)
– Void ratio (to calculate storage)
– Elastic storage coefficient
(The volume of water that a unit volume of saturated soil loses due to
a unit reduction in the applied water head)
– Maximum unsaturated zone height
• Soil classification
– Particle fractions
– Predefined series (Staring, Hypres, USDA) with Van Genuchten and
Approx. van Genuchten parameters.
– User-defined
Upper soils:
< 1m below soil surface
Lower soils:
all deeper soils
Singapore 2011
219 of 607
Groundwater flow - material data sets
Soil classification: Hypres
Hydraulic Properties of
European Soils
Particle distribution:
• < 2μm
• 2μm - 50μm
• 50μm – 2mm
Particle distribution:
• < 2μm
• 2μm - 50μm
• 50μm – 2mm
No difference between
upper and lower soils
Singapore 2011
220 of 607
Groundwater flow - material data sets
Soil classification and Van Genuchten parameters
Relative permeability
Degree of saturation
Closed -- q1,2 0
Head General phreatic level (hw) p1,2 ( y1,2 hw ) w
Head(user-defined) h1 and h2, or p1 and p2 p1,2 ( y1,2 h1,2 ) w
Inflow q1 and q2
Outflow q1 and q2
Singapore 2011
221 of 607
Groundwater flow - boundary conditions
2
Boundary condition on line between points 1 and 2
1
Infiltration
q1, q2, max h y max
Ψ1,min, Ψ2,min ,Ψ1,max, Ψ2,max
min max inflow/outflow
min h y min
Singapore 2011
222 of 607
Groundwater flow - Finite Element Modeling
• GWF calculation generally needs finer mesh than deformation analysis
• GWF calculation generally needs often more additional steps than
deformation analysis
• GWF calculation usually converges, but not always to the correct solution
from an engineering point of view. This can be due to:
• Mesh coarseness
• Distorted elements
• Large differences in permeabilities
Singapore 2011
223 of 607
Singapore 2011
224 of 607
CG11a - Slope Stability by FEM
CONTENTS
Part 1
Advantages of FEM cf LEM
Cases of Slope without GWT effects
Part 2
CUT Slope – Undrained (Method A or B)
CUT Slope – Drained
CUT Slope – Undrained and Consolidation
DH = 10 m
c ' f c ' / FS
tan ' f tan ' / FS
Total strains
2:1 Slope
H = 10 m
DH = 15 m
NO GWT
' 20 o
c' / H 0.05
FS = 1.343
H cu2
u 0 o
cu2/cu1 = 0.6
FS=1.35
cu2/cu1 = 0.2
FS=0.47
cu2/cu1 = 0.6
FS=1.35
cu2/cu1 = 0.2
FS=0.47
cu2/cu1 = 0.6
FS=1.35
cu2/cu1 = 0.2
FS=0.47
cu2/cu1 = 1.0
FS=1.45
cu2/cu1 = 0.6
FS=1.35
cu2/cu1 = 0.2
FS=0.47
2:1 Slope
cu1=50 kPa
cu2/cu1 = 0.6
FS=0.97
cu2/cu1 = 1.5
FS=2.02
cu2/cu1 = 2.0
FS=2.08
cu2/cu1 = 0.6
FS=0.97
cu2/cu1 = 1.5
FS=2.02
cu2/cu1 = 2.0
FS=2.08
cu2/cu1 = 2.0
FS=2.08
cu2/cu1 = 1.5
FS=2.02
cu2/cu1 = 0.6
FS=0.97
Method B:
only when no information on effective strength parameters is available
(consolidation analysis will not change shear strength)
cannot be used with the Soft Soil model and the Soft Soil Creep model
Method C:
NOT recommended (should give deformation close to Method B)
no information on excess pore pressure distribution (total stress
analysis)
cu
1
2
x'o y'o sin ' c' cos '
Fig.6 Mohr Circle for evaluating undrained shear strength (plane strain)
u u
cc c' cos φo 'm
c ' cos ,
' m'sin '
sin
At top of clay; c u 2 cos 24 0 1 .83 kPa
K 0 1 sin ' 1 - sin 24 0 0.59
,v
,
m 1 K 0 6/2 1 0.59 4.77 kPa/m
2
,
m sin ' 4.77 sin 24 0 1.94 kPa/m
1.403
26
24
22
Water Level
20
18
Elevation (m)
16
14 Description: Clay Water
12 Soil Model: Mohr-Coulomb
10 Unit Weight: 16
8 Cohesion: 2 1:2 Cut
6 Phi: 24
4
2
0
0 5 10 15 20 25 30 35 40 45 50 55 60
Distance (m)
2.085
26
24
22
Water Level
20
18
Elevation (m)
16
14
Water
12
Description: Clay
10
Soil Model: S=f(datum) 1:2 Cut
8 Unit Weight: 16
6 C - Datum: 1.83
4 Rate of Increase: 1.94
2 Datum (elevation): 20
0
0 5 10 15 20 25 30 35 40 45 50 55 60
Distance (m)
Method A (undrained)
Effective strength parameters c´=2 kPa, ´=24o, ´=0o
Effective stiffness parameters E50´=15000 kPa, ´=0.2
A (Undrained) =2.27
2
B (Ignore Undrained) =2.13
B (Undrained) =2.14
1.5
1
0 3e3 6e3 9e3 1.2e4
|U| [m]
Conclusions
244 of 607
245 of 607
246 of 607
247 of 607
248 of 607
249 of 607
250 of 607
251 of 607
252 of 607
253 of 607
254 of 607
255 of 607
256 of 607
257 of 607
258 of 607
259 of 607
260 of 607
261 of 607
262 of 607
263 of 607
264 of 607
265 of 607
266 of 607
GENERATION OF INITIAL STRESSES & PHI-C
REDUCTION ANALYSIS
CONTENTS
a. Ko-Procedure
b. Gravity Switch On
b. Some Examples
c. Pointers
d. Appendix
e. References
Singapore 2011
267 of 607
Ko and Gravity Switch-On Procedure
PART 1: INITIAL STRESSES
INITIAL STRESSES
1. Initial stresses represent the equilibrium state of the undisturbed soil and consist of:
a) Soil weight
b) Loading history
a) K0 procedure
b) Gravity loading
Singapore 2011
268 of 607
K0-PROCEDURE
GRAVITY LOADING
For 1D compression: 'n 'v so K0
1 1
Singapore 2011
269 of 607
GRAVITY LOADING
1. Procedure
a. Initial phase
a. Skip K0 procedure, ΣMweight remains zero
b. Phase 1
a. Choose Plastic calculation, Total multipliers
b. Set weight multiplier ΣMweight = 1
c. Phase 2
a. Select Reset displacements to zero to discard all displacements from
raising the gravity
GRAVITY LOADING
Points
1. Undrained material
Singapore 2011
270 of 607
GRAVITY LOADING
Singapore 2011
271 of 607
PHI-C’ REDUCTION IN PLAXIS (STRENGTH REDUCTION METHOD1,2,3,4,5&6)
Main advantages:
2. Critical surface is found automatically as slope failure occurs naturally through the zones due
to insufficient shear strength to resist shear stresses.
5. Information such as stresses, movements and pore-pressures and numerical tool as for
deformation analysis
Safety factor
• Many possible definitions
a v a ila b le s o il r e s is ta n c e
1 .8
m o b iliz e d s o il r e s is ta n c e
fa ilu r e lo a d
5 .9
w o r k in g lo a d
Singapore 2011
272 of 607
SAFETY FACTOR ANALYSIS
Phi/c reduction:
a. Reduction of strength parameters c and tan(f) until failure is reached.
b. The factor of safety :
tan reduced
Lowered incrementally
c tan
Msf
creduced tan reduced
Calculation procedure:
a. Create a phi/c reduction phase
b. Accept the default increment for Msf=0.1 from the multiplier tab-
sheet.
c. Calculate
d. Carefully examine ΣMsf vs. displacement
curve in Plaxis Curves
Singapore 2011
273 of 607
SAFETY FACTOR ANALYSIS
Notes:
a. Select control point within (expected)
failing body
b. Use sufficient number of load steps
c. Use a sufficiently fine mesh
d. Limit the maximum structural forces by
choosing elasto-plastic behaviour for walls,
anchors and geotextiles.
1.16 1.16
1.12 1.12
Sum-Msf
Sum-Msf
1.08 1.08
1.04 1.04
1.0 1.0
0.0 0.3 0.6 0.9 1.2 1.5 0.0 0.3 0.6 0.9 1.2
displacement displacement
Singapore 2011
274 of 607
SAFETY FACTOR ANALYSIS
2
1
H = 12m
D cu = 50 kPa
cu = 100 kPa Plaxis:
F = 1,35
cu 50 D (Taylor,1948)
Stability charts: F N0 6.6 1.38 , N0 f ( , )
Pd 12 20 H
Singapore 2011
275 of 607
EXAMPLE 2: HOMOGENOUS SLOPE WITH AND WITHOUT FOUNDATION
LAYER
Note:
1 Griffiths & Lane (1999)
Model Slope G2 (757 elements-15n) : 2 Bishop & Morgenstern (1960)
FOS= 1.323 (1.4001, 1.7522,1.2794,1.3795, 1,3756) 3.Taylor (1937)
4.Janbu
5.Bishop
6.Morgenstern-Price
Cu2
Cu2/Cu1 = 2
Singapore 2011
276 of 607
EXAMPLE 4: UNDRAINED CLAY SLOPE WITH A THIN LAYER
2.0
1.8
1.6
1.4
Factor of safety)
1.2
Model Slope G3A cu2/cu1=1 (3436 elements- Model Slope G3D cu2/cu1=0.8 (3436 elements-
1.0
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1
Cu2 /Cu1)
Note:
Model Slope G3B cu2/cu1=0.6 (3436 elements- Model Slope G3F cu2/cu1=0.5 (3436 elements- 1 Griffiths & Lane (1999)
15n) 15n) 2 Bishop & Morgenstern (1960)
FOS= 1.319(1.401, 1.404) FOS= 1.112(1.251) 3.Taylor (1937) (green line)
4.Janbu
Model Slope G3E cu2/cu1=0.4 (3436 elements- Model Slope G3c cu2/cu1=0.2 (3436 elements-
15n) 15n) :
FOS= 0.903(1.051) FOS= 0.470 (0.591, 1.304)
SOME POINTS
1. Always inspect the incremental displacements or strains as computed in the last load
step to make sure that failure is reached.
3. Mesh: Refine and redo the phi-c analysis until the factor of safety remains constant
upon further refinement of the mesh.
4. Always use the arc-length time stepping procedure within the Phi-C reduction
(default)
Singapore 2011
277 of 607
Appendix : 2D vs. 3D (Benz)
1. 3D analysis may yield substantially different results from 2D analysis
2. The advantage of FEM over classical design tools is obvious. I
3. n the example: stability of a bentonite slurry trench.
REFERENCES
1. Matsui, T. & San, K.C. (1992) Finite element slope stability analysis by shear strength reduction technique. Soils and
Foundation, Vol.32 (1),pp.59-70
2. Zienkiewicz, O.C., Humpheson, C. & Lewis, R.(1975) Associated and non-associated visco-plasticity and plasticity in soil
mechanics. Geotechnique 25(4).pp. 671-689.
3. Ugai, K.(1989). A method of calculation of total safety factor of slope by elasto-plastic FEM. Soils and Foundation 29(2).
pp.190-195.
4. Farias, M.M., Naylor, D.J.(1998). Safety analysis using finite elements. Computer and Geotechnics.Vol 22(2) pp 165-181.
5. Griffiths, D.V., Lane, P.A. (1999). Slope stability analysis by finite elements. Geotechnique 49 (3), pp.387-403.
Singapore 2011
278 of 607
Introduction to Plaxis 3D
Ronald Brinkgreve
Plaxis bv
Plaxis 3D Input
General toolbar
Mode switches
Selection explorer
Drawing area
Model explorer
Mode toolbar
Command line
Singapore 2011
279 of 607
Plaxis 3D Input : Modes
SOIL STRUCTURES
Singapore 2011
280 of 607
Command line
• All the actions carried out using either the mouse or the explorers are
translated into commands.
• Alternatively, PLAXIS 3D allows to carry out actions using keyboard
input by directly typing the corresponding commands in the command
line.
– The Session tab displays the commands executed in the active session
– The Model history tab displays all the commands executed in the project
Other functionality
• Selection by either clicking individual objects or at once by defining a selection
box in the draw area. Criteria can be applied to the type of items to be
selected.
• Group creation for fast model creation when the same operations have to be
undertaken over a large number of objects
Singapore 2011
281 of 607
Soil Mode
Soil mode
Borehole 4
• Definition of soil volumes and initial water levels Borehole 3
• Based on the concept of boreholes Borehole 1
Singapore 2011
282 of 607
Borehole definition
• Defining soil layer heights in the Soil
layers tabsheet
• Defining water conditions in the
Water tabsheet:
– Specific Head
– Hydrostatic distribution,
– Interpolate from adjacent layers
– Dry
– User-defined pore pressures
• Defining Initial Soil conditions in the
Initial conditions tabsheet
– Specify OCR, POP, K0x and K0y for
the K0 procedure
Material Sets
Singapore 2011
283 of 607
Structure Mode
Singapore 2011
284 of 607
Defining the geometry: Array
• Generate multiple copies of a selection, arranged in a rectangular pattern
Singapore 2011
285 of 607
Defining the geometry: Other functions
• Decompose into surfaces: creates outer surfaces of selected volumes
• Decompose into outlines: creates contour of selected surfaces
• Intersect: splits selected geometric objects along their intersection
• Combine: merges selected geometric objects of the same kind
• These functionalities are only accessible from the RMB context menu oin
the draw area
Loads
• Generated by clicking the Create load button or by right-clicking on
any geometric objects in the draw area
Point load
Line load
Surface load
Singapore 2011
286 of 607
Default Boundary Conditions
Structures
Singapore 2011
287 of 607
Importing Geometry
• Possibility to import from external sources in different formats like
– 3D Studio files (*.3DS)
– AutoCAD native (*.DWG)
– Interchange (*.DXF) file format
Singapore 2011
288 of 607
Mesh density in Plaxis 3D
• Global coarseness:
– Defines an average element size based on model dimensions and relative
element size factor Re (Very coarse / Coarse / Medium / Fine / Very Fine)
– Reference element size = 0.05 * Re * (Model diagonal length)
• Local refinement (Fineness factor):
– Element size can be locally refined or coarsened
– Element size = (Fineness factor) * (Global coarseness) * (Ref. elem. size)
• Color code
– Depending on their degree of local refinement, geometric objects are
displayed in different color in the Draw area (gray for Fineness factor of 1
and green otherwise with darker colour when getting more refined)
Singapore 2011
289 of 607
Generate Mesh
Singapore 2011
290 of 607
Water Levels
Singapore 2011
291 of 607
The Staged Construction Mode
Phase explorer
Singapore 2011
292 of 607
The Phase explorer
• For creating and editing the calculation phases
Singapore 2011
293 of 607
Singapore 2011
294 of 607
Working in the Geometry Modes
of Plaxis 3D
William WL Cheang
Plaxis AsiaPac
Introduction
• The Geometry modes of Plaxis 3D comprises the Soil mode and the
Structures mode
• They are meant to fully define the model geometry in terms of:
– Soil stratigraphy
– Structural elements
– Soil structure interfaces
– Loads
– Boundary conditions
• The Geometry modes are indicated using blue tabsheets and precede the
Calculation modes (green tabsheets) when building up a model from
scratch
295 of 607
The Soil Mode
296 of 607
Soil Mode Toolbar
Create Borehole
Import Soil …
Show Materials
297 of 607
Boreholes Borehole 4
Borehole 3
• Boreholes are locations in the draw
area at which the information on the Borehole 1
height of the constitutive soil layers
and location of the water table is Borehole 2
given
• If multiple boreholes are defined,
PLAXIS 3D will automatically
interpolate between boreholes and
derive the corresponding position
and height of the soil layers from the
available borehole information.
• Each defined soil layer is used
throughout the whole model contour
Creating Boreholes
298 of 607
Defining Water Conditions
• Water conditions can be specified
from the Modify soil layers dialog box
in the Water tabsheet
• Available options are
– Head
– Hydrostatic
– Interpolate
– Dry
– User-defined
299 of 607
Importing Top and Bottom Model Surfaces
• Definition of the top and
bottom soil layer surfaces can
be achieved as a result of
surface import operation
• Import formats include 3DS,
ITS, DWG, DWF and SLT
• Feature available for VIP
members only
Importing Soils
• The geometry of the soil
can be imported from
predefined files instead of
using the Borehole tool
• The same import formats
as for importing soil
surfaces are available
(3DS, ITS, DWG, DWF and
SLT)
• Feature also only
available for VIP members
300 of 607
Material Sets
• Definition of material model parameters
for constitutive soil layers and
structural elements
• Dialog box consistent among all Plaxis
products
• Available from many places in Plaxis
3D:
– Present in each model
– Many shortcuts from different dialog
boxes where material sets assignment
is required
• Entry point to the SoilTest facility
301 of 607
Introduction to Structure Mode
• Meant to define structural
elements and loading
302 of 607
Changing Movement Limitation Settings
• Selecting one the six sides view
of the default views gives access
to the Movement limitation dialog
boxes
303 of 607
Points
• Can be generated by cilcking the Create point button
• The following items can be assigned to a point
– Point load
– Point prescribed displacement
– Fixed-end anchor
Point load
Fixed-end anchor
Lines
• Can be generated by clicking the Create line button
• The following items can be assigned to a line
– Beam
– Line load
– Line prescribed displacement
– Node-to-node anchor
– Embedded pile
304 of 607
Surfaces
• Can be generated by clicking the Create surface buttom
• The first three created points define the surface plane by default
• Existing surfaces can be edited from the Surface points dialog box of
from the pop-up submenu of Create surface button
Mode points/lines
Insert points
Delete points
Rotate
• Rotate any selection around a rotation point compared to global axis
• Possibility to directly rotate the selected objects from the draw area
using Euler angles
305 of 607
Extrude
• Lines and surfaces can be extruded to create surfaces and volumes
correspondingly:
– From the Extrude dialog box
– By dragging and dropping the bottom surface to the top surface location
Array
• Generate multiple copies of a selection, arranged in a rectangular pattern
306 of 607
Further Geometrical Operations
• On top of the aforementioned geometrical operations (rotate, extrude and
array), Plaxis 3D also offers
– Decompose into surfaces: create outer surfaces of selected volumes
– Decompose into outlines: create outer lines (including points) of selected
surfaces
– Intersect: Split select geometric objects along their intersection
– Combine: Merge selected geometric objects of the same kind
• These functionalities are only accessible from the RMB context menu oin
the draw area
Loads
• Can be generated by clicking the Create load button or by invoking the
RMB context menu on any geometric objects in the draw area
Point load
Line load
Surface load
307 of 607
Prescribed Displacements
• Can be generated by clicking the Create prescribed displacements
button or by invoking the RMB context menu on any geometric objects
in the draw area is a very similar way as for Load definition
308 of 607
Structures
• Can be generated by clicking the Fixed-end anchor
Structure button
Beam
• Can also be created from the RMD
context menu after selecting: Node-to-node anchor
– points (fixed-end anchor) Embedded pile
– lines (beam, node-to-node anchor
or embedded pile) Plate
Positive interface
Negative interface
Importing Geometry
• Possible to import from external sources in different formats like 3D
Studio files (*.3DS), AutoCAD native (*.DWG) and interchange (*.DXF) file
formats:
– Click to import surface
309 of 607
Working in the Calculation Modes of Plaxis 3D
William Cheang
Plaxis AsiaPac
Introduction
• The Calculation modes of Plaxis 3D comprises the Mesh mode, the Water
Levels mode and the Staged Construction mode
• They are meant to fully define the model geometry in terms of:
– The finite element mesh
– Changes in water pressure distribution
– Construction phases
– Calculation settings
• The Calculation modes are indicated using green tabsheets and follow the
Geometry modes (blue tabsheets)
310 of 607
The Mesh Mode
311 of 607
Mesh Density Definition in Plaxis 3D
• Global coarseness:
– Define an average element size based on model dimensions and relative
element size factor Re (Very coarse / Coarse / Medium / Fine / Very Fine)
– Ref. Elem. Size = 0.05 * Re * Model Diagonal Length
• Fineness Factor:
– Element size could be locally refine or coarsen
– Element size = Fineness Factor * Global Coarseness * Ref. Elem. Size
• Color code
– Depending on their Fineness Factor, geometric objects are displayed in
different color in the Draw Area (gray for fineness factor of 1 and green
otherwise with darker color as getting more refined)
• Refine mesh
Coarsen mesh
Reset local coarseness
Generate mesh
View mesh
312 of 607
Changing Local Mesh Density
• By clicking the toolbar
buttons Refine mesh or
Coarsen mesh and
selecting the desired
items in the Draw Area
Generate Mesh
• Can be done by clicking the
Generate mesh button from the
Mesh Toolbar or the RMB
context menu
313 of 607
View Mesh and Select Points
• The View mesh and the
Select points for curves will
open the Output Program
where
– mesh could be evaluated
– nodes and/or stress points
could be selected at the
location of which Output
needs to be generated
314 of 607
Water Levels
• Generated water level is created by specifying a Head in the boreholes (in
the Soil mode) and is the default water level
– A single borehole can be used to create a horizontal water surface that
extends to the model boundaries.
– When multiple boreholes are used, a non-horizontal water surface can be
created by combining the heads in the various boreholes
– Non hydrostatic distribution in the soil may be specified in the Water
tabsheet of theModify soil layers dialog box
• User water levels is available in the Water levels mode and can be defined
as a alternative to Generated water level
• Enable preview of
generation of water
pressure
315 of 607
Assigning Water Conditions to Soil Volume
• Water levels can be specified
for each individual volume
• Available options are:
– Global level (default)
– Custom level
– Head
– User-defined
– Interpolate
– Dry
• Can be done from
– The WaterConditions feature
in the Selection explorer.
– The RMB context menu
316 of 607
Introduction to the Staged Construction
Mode
• To define relevant
construction stages and
Deactivate
Show materials
Calculate
317 of 607
Activate/Deactivate
• Indicates which model parts
should be active or inactive in
the each construction stage
• Can be defined through
– RMB context menu
– The selection or model
explorer
• Selection tool can be
advantageously used for fast
selection of large number of
objects
• Be careful with hidden objects
which might still be active
318 of 607
Defining Calculation Stages
• Calculation phase settings can be
edited:
– Phase type
– Phase parameters
• Phase types are:
– Initial stress definition
– Plastic calculation
– Safety factor analysis (phi-c
reduction)
– Consolidation analysis (with closed
or open flow model boundaries)
Phase Parameters
• Advanced phase parameters
can also be controlled from the
Phases dialog box
– Maximum number of
calculation steps
– Reset displacement
– Updated Mesh
– Ignore undrained behaviour
– Parameters for automatic step
size
319 of 607
Singapore 2011 -3D
320 of 607
CG15 Deep Foundation
2D and 3D FEM
Outline
• Analysis of Single Pile in Axi-symmetry
– Influence of discretization
– Influence of interface
– Influence of dilatancy
• Case Study – Instrumented Spun Pile at
Butterworth 8 Condo Project
• Pile Raft Analysis
• Cases
– St Thomas Tower – large pile group
– River Gate Tower – Pile-raft jack-in piles
2
10
12
14
15
Fill
Upper MC
16
17
-25
-30
-35
-40
-45
-50
18
19
1.00E-02
-2.00E-02
S e t t le m e n t ( m )
-3.00E-02
-4.00E-02
-5.00E-02
QT-PLAXIS
-6.00E-02
-7.00E-02
Qb-PLAXIS
-8.00E-02 QT- Load Test
-9.00E-02 Qb-Load Test
-1.00E-01
20
-0.05
DRAINED
-0.1
-0.15
-0.2
-0.25
-0.3
-0.35
0 1e4 2e4 3e4 4e4 5e4
TOTAL PILE HEAD LOADS (kPa)
21
-0.05
DRAINED
-0.1
-0.15
-0.2
-0.25
-0.3
-0.35
0 -10000 -20000 -30000 -40000 -50000
PILE TOE LOADS [kN/m2]
22
Hyperbolic
Extrapolation
2*WL
Actual Pile
Plunge at
Chin’s Plot 2.2*WL
for 2*WL
24
´Helaba ´Eurotower´
Hochhaus´ 1974-77
1975-77 RF, s = 9 cm
RF, ´Japan Center´
´EUROTHEUM´ s = 10 cm
1997-99 1994-96
´Citibank´
PRF, PRF, s = 3.2 cm
1985-86
RF, s = 3.2 cm
s = 11 cm ´Commerz
bank I
Hochhaus´
1972-74
RF, s = 9 cm
27
Resistance Q
Single pile
Pile group
Pile group Piled raft
Settlement s
28
Resistance Q
C Single pile
I
E
Settlement s
Corner pile
Edge pile
Inner pile
Conclusion:
Group behaviour and pile-raft-interaction
reduce on the one hand the stiffness of
the piles and increase on the other hand
their bearing capacity!
After El-Mossalamy 1996
29
PRF
PRF
Q ( s ) P
Qtot ( s )
sPRF/sRF
30
Q P + QR
after Hanisch, Katzenbach, König
Qtot = Stot
QP = Q b + Qs
Qtot Stot D
z
Interaction influences:
Pile-soil interaction
Pile-Pile interaction s,j
Pile-raft interaction
31
32
33
Model 1 Model 2
I E
34
D = 1.5m
Load Q [MN]
0 10 20 30 40
Q single pile
L = 30m 0
Q inner pile
64 piles
e = 3D 5
Q edge pile
Q corner pile
10
Settlement [cm]
15
20
25
30
35
D = 1.5m
Load Q [MN]
0 10 20 30 40
Q single pile
L = 30m 0
Q inner pile M1
64 piles
e = 3D 5
Q inner pile M2
15
20
25
30
36
Conclusions
• Bearing resistance failure will not occur until well above
WL
• Load sharing between pile and raft depends on
settlement. No linear relation between load sharing and
settlement
• Load in pile depend of pile position and spacing
• Pile-raft interaction leads to increasing vertical and
horizontal stresses in soils
• Increased load in piled-raft will mobilize larger skin
friction in upper part of piles
• Assumption of piles as an action on the raft of constant
value is unrealistic and lead to poor design
38
39
Concept of Piled-Raft
40
Piled-Raft
41
42
´Helaba ´Eurotower´
Hochhaus´ 1974-77
1975-77 RF, s = 9 cm
RF, ´Japan Center´
´EUROTHEUM´ s = 10 cm
1997-99 1994-96
´Citibank´
PRF, PRF, s = 3.2 cm
1985-86
RF, s = 3.2 cm
s = 11 cm ´Commerz
bank I
Hochhaus´
1972-74
RF, s = 9 cm
43
Frankfurt Cases
44
45
Case 1 Messe-Torhaus
• First case in Frankfurt of
piled-raft during 1982-84
• 400 MN,130m high, 30
storeys
• Two rafts 24.5x17.5m and
2.5m thick on 42 bored
piles each
• Piles 0.9m D and 20m long
• Without piles settlement
about 250 mm; with piles
reduce to 120 mm
46
47
Case 1 Messe-Torhaus
80%
20%
48
49
1982-1984
h = 130 m
30 floors
Inner pile
Depth [m] Depth [m] 50
Edge pile
Messeturm
• Built in 1991
• Height = 256 m; 60 storeys
• Weight = 1900 MN
• Uplift = 200 MN
• Raft = 41x60m
• Raft alone estimate settle 400 mm
• Piled-raft estimate settle 200 mm
• Raft thickness = 6m centre and 3m edge
• 64 bored piles in 3 rings
51
52
53
1991
h = 256 m
60 floors
55
BH5
BH1
BH4
BH2
BH6
Introduction
• Aim of 3D study
– to predict total and differential settlements of piled
raft foundation
– to obtain likely raft forces and bending moments
• Pile load test model to obtain strength and stiffness
correlations with available SPT N-values
• Features in 3D model:
– Non-uniform soil profile by input of multiple
boreholes
– Unloading-reloading soil behaviour simulated
realistically by Hardening Soil models
– Large number of piles modelled using embedded
pile elements in Plaxis 3D Foundation
58
BH1
BH4
BH2
BH6
-9 m
-21 m
-50 m
60
4
parameters based on SPT N-values
5 Field test data • Hence use of assumed parameters
6
Simulated
tend to be conservative in settlement
predictions
7
63
25
represents field soil properties
30
35
50
64
BH5
BH1
BH4
BH2
BH6
Foundation 3D Model
BH3
BH5
BH1
BH4
BH2
BH6
Foundation 3D Model
Plate elements for raft & pilecaps,
with loading shown
68
Embedded piles
with varying toe
depths based on
piling plan
69
70
71
Foundation 3D Model
72
ν 0.2
Simulation Steps
• Phase 0: Initial stress generation by K0 procedure
– K0 = 0.7 for top firm silty clay layer (over-
consolidated) & K0 = 0.531 for others (normally-
consolidated)
• Phase 1: Installation of bored (embedded) piles
– Undrained (short-term) simulation
• Reset displacements to zero
• Phase 2: Excavation to pile C.O.L (RL114.3m)
– Undrained excavation throughout model and addition
of surcharge on active side of retaining wall
(simplified)
• Reset displacements to zero
• Phase 3: Installation of rafts and foundation loads
– Drained (long-term) simulation 74
75
Simulation Results
• Long-term ground settlements under loading
76
77
Simulation Results
• Long-term ground effective stress under
loading
78
79
Simulation Results
• Raft shear force Q23
80
81
Simulation Results
• Raft bending moment M22
82
46
12 6,515 1.3 4,908 < 7.5 Mpa
48
13 7,100 1.4 4,612 < 7.5 Mpa
14 9,295 1.4 6,038 < 7.5 Mpa
49
15 9,588 1.4 6,228 < 7.5 Mpa
44
47
16 8,965 1.4 5,824 < 7.5 Mpa
52
41
42
50
19 8,296 1.4 5,389 < 7.5 Mpa
39
51
20 9,618 1.4 6,248 < 7.5 Mpa
21 9,441 1.4 6,133 < 7.5 Mpa
22 8,429 1.4 5,476 < 7.5 Mpa
33
23
38
24 8,442 1.4 5,484 < 7.5 Mpa
25 8,711 1.4 5,659 < 7.5 Mpa
32
27
22
37
26
30
20
35
29
24
34
10
85
Site Soil Profile Calibration of single pile load test
87
iv. Jack-in spun piles are generally safe even with short piles so long
they have been jacked to refusal with two times of working load of
88
piles
91
William WL Cheang
Principal Geotechnical Consultant
Plaxis 3D Plaxis AsiaPac
Lee Siew Wei
Contributions
Dr And Pickles
Mr Tom Henderson
Ms Carmen Choi
Mr Tse Sai Chung
Dr Ronald Brinkgreve
Mr Frank van Gool
Ms Wendy Swolls
Contents
Singapore Advanced Course 2011
367 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Modelling of Tunnels and Tunnelling in 3D
2. Construction stages
3. Modelling anchors
5. Conclusions
Geometric modelling issues
Singapore Advanced Course 2011
368 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Geometric modelling issues (CM Line)
Geometric modelling issues
Hint: Draw cross section surface and use Extrude command to create shafts
PLAXIS 3D will automatically create intersections
Singapore Advanced Course 2011
369 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Geometric modelling issues (CAD)
Geometric modelling issues (Import)
Singapore Advanced Course 2011
370 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Construction stages (For sequence of events)
Modelling Anchors
Singapore Advanced Course 2011
371 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Modelling Anchors
Modelling volume loss
Singapore Advanced Course 2011
372 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Conclusions
2D & 3D MODELLING OF
TUNNELLING
Part 2
Singapore Advanced Course 2011
373 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Part 2‐Outlines
A. 2D modelling of tunnelling
B. 3D modelling of tunnelling
Tunnelling Observations
Singapore Advanced Course 2011
374 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Concept of Modelling Tunnelling in 2D
3D 2D 2D
Moller (2006)
Methods of Modelling Tunnelling in 2D
• Plaxis 2D provides
1. Lining Contraction Method
2. Stress Reduction Method (-method)
3. Applied Pressure Method (APM)
(from Grout Pressure Method by Moller & Vermeer, 2008)
Singapore Advanced Course 2011
375 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Lining Contraction Method
Stress Reduction Method ()
1 Pk 1
Pk = initial ground radial
= pressure
Pk ΣMstage = 1 -
Singapore Advanced Course 2011
376 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Applied Pressure Method (APM)
FE Prediction of Greenfield Surface Settlement
Singapore Advanced Course 2011
377 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Modelling of Tunnelling in Hong Kong Soils
60m
0
Fill -3
Marine Deposits -6
20m
40m
Completely Decomposed Granite
6m Ø tunnel (CDG)
-
Rock 40
• Ground conditions: 3m Fill, 3m MD, 34m CDG & rock; GWT at surface
• Tunnel 6m diameter with axis at 20 mbgl; 2700 nos of 15-noded elements
Soils Modelled by Mohr Coulomb Model
E ν c' / cu '
Soil
(kN/m3) (MPa) [-] (kPa) (Deg)
Fill 19 20 0.3 0 30
MD
16 6 0.3 15 0
(Undrained)
CDG 20 39 0.3 5 35
Soils Modelled by HS & HS‐small Models
Fill
20 60 0.5 0 30 100 0.2 - -
(HS)
MD
(HS) 6 18 1 0 22 100 0.2 - -
(Und.)
CDG
5E
(HS- 39 117 0.5 5 35 200 0.2 200
-5
small)
Singapore Advanced Course 2011
378 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Pre‐failure Stress‐strain Behaviour
1: Mohr Coulomb
Initial Stress Equilibrium
• K0 = 1 – sin'
' = drained effective friction angle (Fill=30°; MD=22°)
Soil K0 CDG
Fill 0.5
MD 0.625
Schnaid et al.
(2000)
Singapore Advanced Course 2011
379 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Details for Analyses
-2
Settlement (mm)
-4
Gaussian (K=0.5, VL 1%)
-6
Lining contraction - LC 1%, VL 0.32%
Mohr Coulomb Soil with Lining Contraction
Singapore Advanced Course 2011
380 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Comparison of MC and HS & HS‐small Models
-4
Gaussian (K=0.5, VL 1%)
-6
Lining contraction - LC 1%, VL 0.32%
-2
Settlement (mm)
-4
Gaussian (K=0.5, VL 1%)
-6
Lining contraction - LC 1%, VL 0.77%
Singapore Advanced Course 2011
381 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Summary of 2D Modelling of Tunnelling
A. Good prediction of greenfield surface settlement curve (Gaussian) in 2D requires
1. advanced soil constitutive model for nonlinear stiffness from small strains
2. refined method of modelling tunnel excavation
B. Tunnelling example investigated herein:
1. effect of advanced constitutive model is more significant than method of
modelling tunnel excavation
2. Applied Pressure Method gives steeper surface settlement curve, followed
by Stress Reduction Method & Lining Contraction Method
C. On realistic prediction of surface settlement curve & pattern of ground deformation
around tunnel:
1. Mohr Coulomb model + Lining Contraction Method gives unrealistic results
2. HS & HS-small models + Applied Pressure Method gives better results
Part 3.1
Singapore Advanced Course 2011
382 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Presentation Outline
Three‐Stage Building Damage Assessment Due to Tunnelling
Burland (1995)
Singapore Advanced Course 2011
383 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Tunnel‐pile‐soil Interaction
• A three-dimensional problem due to
1. progressive advance of tunnel face towards piles
2. movement of piles in 3D
Zones of Influence
Pile settlement
C B A
Depth
Singapore Advanced Course 2011
384 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Analysis of Tunnel‐pile Interaction
• Typically use the combination of
1. empirical relationships/closed-form solutions to estimate
greenfield ground movements; and
2. boundary element methods to compute pile deformations and
stresses
• Suitable for preliminary assessment, with some limitations
• Alternatively, use 3D numerical analysis
Pros: model tunnelling, tunnel-pile-building interaction &
geotechnical entities in one single analysis
Cons: complicated, relatively long analysis time & require
advanced constitutive model for soil non-linear behaviour
Example of Tunnelling Below Piled Building
25m
25m
0 mbgl P4 P5 Rear P6
2m Pile cap
5 mbgl Fill 9m 10m
1m
4m
MD
10 mbgl P1 P2 Front P3
CDG 10m 6m Ø tunnel
1m
4m
Tunnel 2m Ø pile
6m Ø Pile design load 15MN (~5MPa)
30 mbgl
31.5 mbgl Rock 3m Ø bell-out
P1/P4 P2/P5 P3/P6
Singapore Advanced Course 2011
385 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Information for Tunnel, Piles & Ground
Soil Small Strain Non‐linear Stiffness
0.01% 0.1% 1%
Singapore Advanced Course 2011
386 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
CDG Small Strain Non‐linear Stiffness
Laboratory small strain stiffness
• Hardening Soil + Small Strain
results for CDG samples
Overlay (HSsmall) constitutive
model to consider CDG small strain
Ng et al. (1998) non-linear stiffness
800 HSsmall_Baseline
600
400
200
0
0.0001 0.001 0.01 0.1 1 10
Shear strain (%)
3D Finite Element Model (Plaxis‐GiD)
Rear Load 15 MN
Building “Plate”
40m Pile cap
Bored pile
Front
Fill Tunnel face
Tunnel MD 149m
CDG
120m Rock TBM
length
Bell-out
43,000 elements Linings
Singapore Advanced Course 2011
387 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Tunnel Confinement Pressure
A
PIII A. Face support pressure (PI to PII) =
PIV
hydrostatic pore pressure + overpressure
PI
Rear Front B. Along TBM shield, tunnel support pressures
vary to consider
6m Ø TBM shield 9m
1. Conical shape of TBM shield / over-
cutting
PII
2. Ground loss into tail void in rear
PVI C. Any combination of support pressure
PV profiles can be modelled
A
PIII
Pressure
PV increases
Section A-A with depth
Modelling of Tunnel Face Advance
TBM shield
deactivated
(elements nulled)
2. Apply tunnel support pressure profiles
3. For each face advance, shift tunnel
support pressures forward &
1.5 1.5m
correspondingly erect new lining behind
TBM
4. The process is repeated as tunnelling
Lining
Lining
TBM shield
(elements nulled) progresses
1.5 1.5m
Lining
Lining
TBM shield
(elements nulled)
1.5 1.5m
Singapore Advanced Course 2011
388 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Modelling of Structures
1. Piles & pile cap modelled by solid elements
2. Interface elements along pile shafts & on pile cap vertical faces
3. Consider flexural stiffness (EI) & axial stiffness (EA) of superstructure by
incorporating a “Plate” structural elements on top of pile cap.
Superstructure EI estimated by (Potts & Addenbrooke, 1997)
1. Parallel Axis Theorem (bending about building neutral axis); or
2. Summation of EI for individual building storeys
4. Tunnel linings modelled by “Plate” elements
Prediction on Ground Surface Settlement
Overpressure 20 kPa, G/F VL 1.6% Overpressure 20 kPa
Distance from tunnel centreline (m)
-8 VL 1.61%
-12
Mid-building
CDG
-16
Greenfield
Tunnel -20
Gaussian
-24
Singapore Advanced Course 2011
389 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Prediction on Pile Transverse Displacement
Overpressure 20 kPa
Transverse horizontal disp. (mm) +10D
-5 -4 -3 -2 -1 0
0
-2D
Front 5
Rear +2D
10
+2D
Depth (mbgl)
15
+10D Rear
20 1m P2
Front
25
30 -2D
35
Tunnel advance
Prediction on Pile Longitudinal Displacement
Overpressure 20 kPa
Longitudinal horizontal disp. (mm) +10D
-3 -2 -1 0 1 2
0
10 +2D
Depth (mbgl)
15 Rear
-2D 20 1m P2
Tunnel advance
Front Front
25
Rear
+2D
30 -2D
+10D 35
Tunnel advance
Singapore Advanced Course 2011
390 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Prediction on Pile Settlement & Axial Force
Overpressure 20 kPa
Settlement (mm) Increase in axial force (MN)
0 -1 -2 0 1 2 3 4
0 0
P2 -2D P2 -2D
5 Front 5 Front
Rear Rear
10 +2D 10 +2D
+10D
Depth (mbgl)
Depth (mbgl)
+10D
15 7
15
20 20
25 25
A B
30 30
C
35 35
Pile toe
Prediction on Pile Bending Moment
Overpressure 20 kPa
Transverse moment (kNm) Longitudinal moment (kNm)
1500 500 -500 -1500 1500 500 -500 -1500
0 0
P2 P2
5 5
10 -2D 10 -2D
Depth (mbgl)
Depth (mbgl)
Front Front
15 15
Rear Rear
+2D +2D
20 20
+10D Tunnel +10D
25 25advance
30 30
35 35
Singapore Advanced Course 2011
391 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Check on Potential Structure Damage
OP 40kPa
25 OP 40kPa
_
Bldg. settlement (mm)
-0.4 0.2
/L (%)
15
5 -0.8 Cat.
0.1
-5 =0.14 2 3
-1.2 mm 1
-15 0.0 0
-15 -10 -5 0 5 10 15 0 0.1 0.2 0.3
Moment, M (MNm) -1.6 h (%)
Comparison with Closed Form Solution
Greenfield subsurface settle. (mm) Greenfield subsurface horiz. disp. (mm)
-35 -25 -15 -5 5 15 -15 -10 -5 0
0 0 0
Fill Fill
5 5
MD MD
10 10
Depth (mbgl)
Depth (mbgl)
15 CDG 15
CDG
20 20
25 25
Loganathan Loganathan
et al. (2001) 30 et al. (2001) 30
Rock
3D analysis Rock 3D analysis
35 35
Singapore Advanced Course 2011
392 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
3D FEA vs. Analytical Solution
Issues 3D FEA Analytical Solution
3D FEA vs. Analytical Solution
Issues 3D FEA Analytical Solution
Ground loss, VL 1. Model confinement 1. Assume a certain VL
pressure & predict VL
Effect on 1. Model tunnel, piles, 1. Different boundary element
piles/building building & their interaction in programs for pile axial and
one single analysis lateral responses
2. Results from piles & 2. Specific analysis for pile
building used directly in group effect
structural check 3. Dedicated modification
factors account for building
rigidity
Singapore Advanced Course 2011
393 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Part 3.2
6m ø tunnel P8 2m
P7
P4
P6
P3 P5 P6 P7 P8
P5 23m
P2 P1 P2 P3 P4
Bldg. footprint
P1 33×11m
y x
3m y
z 2m ø piles
31 mbgl
Tunnel advance direction z x
6m ø tunnel
Singapore Advanced Course 2011
394 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
3D Model by Plaxis-GID
Pile cap + “Plate”
45m
Building 23m
Pile
Settlement & Axial A B
P5 Force Pile P5 C
Rear
cap
Front
cap Settlement, Uy (mm) Pile axial laod, N (MN)
0 -5 -10 -15 -20 -25 -30 0 2 4 6 8 10 12
0 0
-5 -5
Depth (mbgl)
Depth (mbgl)
-10 -10
-20 -20
-25 -25
Front Rear +2D +15D Front Rear +2D +15D
Singapore Advanced Course 2011
395 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Channel Tunnel Rail Link (UK) Measurement
Selementas (2005)
2m
Approaching
Final Initial
Longitudinal Horizontal Displacement &
P5
Rear
Bending Moment Pile P5
cap
Front
Longitudinal hor. disp., Uz (mm) Longitudinal moment, Mz (kNm)
cap -1 0 1 2 3 -400 -200 0 200 400 600
0 0
-5 -5
Depth (mbgl)
Depth (mbgl)
-10 -10
-15 -15
Tunnel
-20 -20
advance
-25 -25
Front Rear +2D +15D Front Rear +2D +15D
Singapore Advanced Course 2011
396 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Transverse Horizontal Displacement &
P6
Rear
Bending Moment Pile P6
cap
Front
cap Transverse hor. disp., Ux (mm) Transverse moment, Mx (kNm)
-6 -5 -4 -3 -2 -1 0 -2000 -1500 -1000 -500 0 500
0 0
-5 -5
Depth (mbgl)
Depth (mbgl)
-10 -10
-15 -15
-20 -20
-25 -25
Front Rear +2D +15D Front Rear +2D +15D
Tunnel
Greenfield Surface & Building Settlements
-5
-10
-15
Greenfield
-20
Building
-25 VL 2.8%
-30 Building
Singapore Advanced Course 2011
397 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Building Settlements > Greenfield Surface Settlements
Greenfield
surface
settlements
Part 3.3
Singapore Advanced Course 2011
398 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Tunnelling Near a Group of 48 Piles
0 mPD
Pile cap 1.5m
13.6m
10m -20mPD
1m
0.6m Ø Franki piles
3D Model by Plaxis 3D
Bldg. load
“Plate” modelling
superstructure EI & EA
Building
40m
Fill 1m 48 Franki piles
(Embedded Piles)
CDG Tunnel
advance
Tunnel 120m
140m
6m Ø tunnel
Singapore Advanced Course 2011
399 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Effect of 3m Thick Annulus Grout
Grouted
annulus
21m
3m
Tunnel
Effect of Fixed Pile Head Connections
Pile cap
“Plate” modelling
superstructure
Tunnel
Singapore Advanced Course 2011
400 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Output of Results
Tunnel Tunnel
advance advance
Greenfield Surface & Building Settlements
Ho rizo
Horizontal ntal distance
distance fro mcentreline
from tunnel centreline (m)
(m)
-60 -40 -20 0 20 40 60
0
-2
Settlement (mm)
-4
-6
-8
Greenfield
-10
B uilding
Bldg.
-12
• For pile toes above tunnel, building settlements are greater than
greenfield surface settlements due to undermining below pile toes
Singapore Advanced Course 2011
401 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Building Settlements > Greenfield Surface Settlements
Greenfield
surface
settlements
Comparison of Building Settlements
Ho distance
Horizontal rizo ntal distance fro centreline
from tunnel m centreline
(m)(m)
0 5 10 15 20 25
0
Building settlement (mm)
-2
Annulus grout
-4
-6
Fixed pile heads
-8 A 1- B aseline
A 2 - A nnulus gro ut
-10
A 3 - Fixed heads
-12 Baseline
(Pinned pile heads)
Singapore Advanced Course 2011
402 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Transverse Horizontal Displ. of Closest Pile
P ile transverse Ux (mm)
-6 -5 -4 -3 -2 -1 0 1
0
-4
A 1- B aseline
Level (mPD)
(Pinned pile heads)
-8
A 2 - A nnulus
Gro ut
A 3 - Fixed -12
Heads Annulus grout
-16
-20
Tunnel
Part 3.4
Singapore Advanced Course 2011
403 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Tunnelling Intersecting Piles
17.8m
Building footprint
50 nos 0.6m Ø Franki 8.8m Plan View
piles @ 3Ø spacing
Tunnel 6m Ø tunnel
advance
0 mPD
Pile cap 1.5m
Front View
3D Model by Plaxis 3D
Upper half
Full annulus grout
annulus grout
Bldg. load
“Plate” models
superstructure
Building
40m
Fill
CDG
Tunnel Grout
120m 3m
150m 12m
Singapore Advanced Course 2011
404 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Output of Results
Isometric view Front view
6m Ø tunnel
Exaggeration scale 100x
Greenfield Surface & Building Settlements
Horizontal distance from tunnel centreline (m)
-60 -40 -20 0 20 40 60
0
Settlement (mm)
-10
-20
-30 Greenfield
Building
-40
Bldg.
Singapore Advanced Course 2011
405 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Effect of 3m Thick Annulus Grout on Bldg. Settlements
Baseline No grout
-30
Half grout
Full grout
-40
Effect of Annulus Grout on Pile Settle. & Axial Force
Critical pile
Tunnel 6m Ø tunnel
advance
Pile settlement (mm) Pile axial force (MN)
-40 -30 -20 -10 0 -1 0 1 2 3
0
Tension
-4
Level (mPD)
-8
Baseline -12
(No grout)
-16 Initial
Half grout
Baseline (No grout)
-20 Half grout
Full grout
-24 Full grout
Singapore Advanced Course 2011
406 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
Effect of Annulus Grout w.r.t Pile N‐M Capacity
Critical pile
Tunnel 6m Ø tunnel
advance
2500 Pile capacity
Baseline (No grout)
2000 Half grout
Full grout
Axial force (kN)
1500
1000
500
-500
-200 -150 -100 -50 0 50 100 150 200
Moment (kNm)
Summary
1. Details & results for 3D modelling of tunnel advance near a piled building are given:
a. pressure-controlled boundary on tunnel face & along TBM
b. integrated response of piles & building to tunnelling in 3D
c. bldg. on end-bearing piles: bldg. settlement < greenfield surface settle.
d. bldg. on friction piles (toes above tunnel): bldg. settlement > greenfield surface settlement
b. considers layered ground, tunnel face advance, TBM support pressure, bldg. stiffness &
combined piles-cap-bldg. behaviour
3. 3D analysis adds value to tunnel design & construction process, e.g. assessment of
requirement for protective measures
Singapore Advanced Course 2011
407 of 607
Tunnels & Tunnelling in 3D‐ Cheang & Lee (2011)
References
1. Atkinson, J. H. & Sallfors G. (1991). Experimental determination of soil properties. Proc. 10th ECSMFE, Florence, Vol.3, 915-956
2. Burland, J. B. (1995). Assessment of risk of damage to buildings due to tunnelling and excavation. 1st Int. Conf. on Earthquake Geotech. Engrg., IS Tokyo.
3. Geotechnical Control Office (GCO) (1985). Technical Note T4/85 - MTR Island Line: Effects of Construction on Adjacent Property. Civil Engrg. Services
Dept., Hong Kong.
4. Hake, D. R. & Chau, I. P. W. (2008). Twin stacked tunnels - KDB200, Kowloon Southern Link, Hong Kong. Proc. 13rd Australian Tunnelling Conference, 445-
452.
5. Loganathan, N., Poulos, H. G. & Xu, K. J. (2001). Ground and pile-group responses due to tunnelling. Soils and Foundations, 41(1), 57-67.
6. Moller, S. (2006). Tunnel induced settlements and structural forces in linings. PhD thesis, University of Stuttgart.
7. Moller, S. & Vermeer, P. A. (2008). On numerical simulation of tunnel installation. Tunnelling & Underground Space Technology, 23, 461-475.
8. Ng, C. W. W., Sun, Y. F. & Lee, K. M. (1998). Laboratory measurements of small strain stiffness of granitic saprolites. Geotechnical Engineering, SEAGS,
29(2), 233-248.
9. Pang, C. H. (2006). The effects of tunnel construction on nearby pile foundation. PhD thesis, National University of Singapore.
10. Potts, D. M. & Addenbrooke, T. I. (1997). A structure’s influence on tunnelling-induced ground movements. Geotechnical Engineering, Proc. ICE, 125, 109-
125.
11. Schnaid, F., Ortigao, J. A. R., Mantaras, F. M., Cunha, R. P. & MacGregor, I. (2000). Analysis of self-boring pressuremeter (SBPM) and Marchetti
dilatometer (DMT) in granite saprolites. Canadian Geotechnical J., 37, 796-810.
12. Selementas, D. (2005). The response of full-scale piles and piled structures to tunnelling. PhD thesis, University of Cambridge.
13. Storry, R. B. & Stenning, A. S. (2001). Geotechnical design & contraction aspects of the Tsing Tsuen Tunnels – KCRC West Rail Phase; Contract DB320.
Proc. 14th SEAGC, Hong Kong, 443-448.
14. Storry, R. B., Stenning, A. S. & MacDonald, A. N. (2003). Geotechnical design and construction aspects of the Tsing Tsuen Tunnels – contract DB320 KCRC
West Rail Project. Proc. ITA World Tunnelling Congress, (Re)claiming the Underground Space, Saveur (ed.), 621-626.
15. Vermeer, P. A. & Brinkgreve, R. (1993). Plaxis Version 5 Manual. Rotterdam, a.a. Balkema edition.
Singapore Advanced Course 2011
408 of 607
Embedded Elements in Plaxis 3D
By
RF Shen
25 November 2011
Embedded pile in Plaxis 3D
Embedded pile in Plaxis
3D can be used to
simulate piles to obtain
pile movement profiles,
internal forces readily.
HOWEVER, a good
understanding of the
behavior of embedded
pile in Plaxis 3D is critical
for the proper use of this
very useful element.
Illustration of a pile behind
excavation 2
409 of 607
Embedded pile in Plaxis 3D
An embedded pile = Beam element + Interface
element around the beam element to interact
with the surround soil elements.
As such, the definition of an embedded pile
element consists of 2 parts: properties of the
beam & properties of the interface element
(skin resistance and foot resistance).
Embedded pile in Plaxis 3D
For the definition of the
beam element part, it is
much the same as that
defined for the beam
elements.
Be cautious of non‐isotropic
pile (like H‐pile) with the
possibility of different pile
orientation, as will be
illustrated in the next slide!
4
410 of 607
Embedded pile in Plaxis 3D
Embedded pile in Plaxis 3D
Illustration: Embeddedpile Orientation.P3D
For a higher resistance
to excavation‐induce
bending moment, the H‐
piles are oriented with
major axis to be bending
toward the excavation
side, has the embedded
pile orientation being
properly configured in
the right‐hand figure?
Illustration of a pile behind
excavation 6
411 of 607
Embedded pile in Plaxis 3D
An embedded pile can cross
a 10‐node tetrahedral soil
element at any place with
any arbitrary orientation,
introducing 3 extra nodes
inside the 10‐node
tetrahedral soil element.
An embedded pile crossing an
tetrahedral soil element
7
Embedded pile in Plaxis 3D
Interaction between
embedded pile and
surrounding soil element at
each node is based on:
embedded pile interacting with
surrounding soils
8
412 of 607
Embedded pile in Plaxis 3D
While interface element has been
provided along the embedded pile
shaft (good for correct simulation of
axial pile‐soil interaction), take note
that NO interface element provided
for the lateral soil sliding around the
embedded pile. As such, for laterally
loaded case, embedded pile only
works well in working condition
when there is not much pile‐soil
lateral sliding occurs, it can not be
used for simulation of ultimate embedded pile interacting with
lateral loading scenario. surrounding soils
9
Performance of embedded pile versus
solid pile in Plaxis 3D
10
413 of 607
3D FEM mesh for solid pile
The pile is simulated
by dia. 1m solid
cylindrical object with
surrounding interface
with Rinter=1
The dia. 1m solid
cylindrical object has
concrete elastic properties
with E=3.0E+7kPa 11
3D FEM mesh for embedded pile
Defining the beam
properties: The
embedded pile has the
structural properties
match exactly the dia.
1m bored pile
12
414 of 607
Calculated pile capacity
Defining embedded pile interface properties:
Skin resistance: Cu=100kPa, Rinter =1, Thus, Tmax =
3.14*1m*100kPa = 314kN/m.
End‐bearing resistance: qb = 9Cu = 900kPa, Fmax =
0.25*3.14*(1m)^2*900kPa = 706kN
So, Total shaft resistance Fshaft = 314kN/m * 20m = 6280kN
Total base resistance Fmax = 706 kN
13
Total pile resistance Ftotal = Fshaft + Fmax = 6990 kN
Axially loaded embedded pile
Fz =7000kN
=1000kN
=2000kN
=3000kN
=4000kN
=5000kN
=6000kN
14
415 of 607
Punching of embedded pile under
theoretical axial load
Fz =7000kN
15
Embedded pile toe resistance under axial loads
Fz =7000kN
16
416 of 607
Axial load transfer curves
Axial force (kN)
0 2000 4000 6000 8000
0
‐5
Elevation (m)
‐10
Embeddedpile Fz=1000kN
‐15
Embeddedpile Fz=3000kN
‐20 Embeddedpile Fz=5000kN
Embeddedpile Fz=7000kN
17
‐25
Shaft skin resistance of embedded pile
0 20 40 60 80 100 120
0
‐5
Elevation (m)
‐10
Embeddedpile Fz=1000kN
‐15 Embeddedpile Fz=3000kN
Embeddedpile Fz=5000kN
‐20
Embeddedpile Fz=7000kN
‐25
Shaft friction (kPa) 18
417 of 607
Axial load‐settlement behavior
10
20
Pile settlement (mm)
30
40
50
60
70
80 Embeddedpile
EmbeddedpileSolid pile
90
100 19
Axial load (kN)
Axial load transfer curves
Axial force (kN)
0 2000 4000 6000 8000
0
‐5
Elevation (m)
‐10
Embeddedpile Fz=1000kN
Embeddedpile Fz=1000kN
Embeddedpile Fz=3000kN
‐15 Embeddedpile Fz=5000kN
Embeddedpile Fz=3000kN
Embeddedpile Fz=7000kN
Solid pile (1000kN)
‐20 Solid pile (3000kN)
Embeddedpile Fz=5000kN
Solid pile (5000kN)
Solid pile (7000kN)
Embeddedpile Fz=7000kN
20
Solid pile (8000kN)
‐25
418 of 607
3D FEM mesh for embedded pile
0 20 40 60 80 100 120
0
‐5
Elevation (m)
‐10 Embeddedpile Fz=1000kN
Embeddedpile Fz=3000kN
Embeddedpile Fz=5000kN
‐15 Embeddedpile Fz=7000kN
Solid pile (1000kN)
Solid pile (3000kN)
‐20
Solid pile (5000kN)
Solid pile (7000kN)
Solid pile (8000kN)
‐25
Shaft friction (kPa) 21
Concluding remarks: Embedded pile performs
satisfactorily under axial loads and conform to
theoretically values, while solid pile exhibit too high end
bearing resistance and much stiffer pile response, and
thus needed to be used with cautions.
Can tension loading case eliminate the end‐bearing
difference?
Total shaft resistance Fshaft = 314kN/m * 20m = 6280kN
Self‐weight of bored pile = 0.25*3.14*(1m)^2 * 20m *
24kN/m^3 = 377 kN
So, expected total pull out resistance = 6280 + 377 = 6660kN
22
419 of 607
Can tension loading case eliminate
the end‐bearing difference?
‐50
‐40 Embededpile_use actual load
Embededpile_use actual load
Pile movement (mm)
Solid pile
‐30
‐20
‐10
0
0 1000 2000 3000 4000 5000 6000 7000 8000 23
Tension load (kN)
Load transfer curves under tension loadings
Tension force (kN)
‐8000 ‐7000 ‐6000 ‐5000 ‐4000 ‐3000 ‐2000 ‐1000 0
0
‐5
Elevation (m)
‐10
Embeddedpile Fz=1000kN
Embeddedpile Fz=1000kN
Embeddedpile Fz=3000kN
Embeddedpile Fz=5000kN ‐15
Embeddedpile Fz=3000kN
Embeddedpile Fz=6000kN
Solid pile (1000kN)
Solid pile (3000kN)
Embeddedpile Fz=5000kN
‐20
Solid pile (5000kN)
Solid pile (7000kN)
Embeddedpile Fz=6000kN
Solid pile (8000kN) 24
‐25
420 of 607
In general, embedded pile performs satisfactorily
under both axial compression loads and tension
loads and generally conform to theoretically
values, while solid pile exhibit too high end
bearing resistance and much stiffer pile response
under axial compression load and develop very
large suction at the toe of solid pile under tension
load which may not be so reliable, and thus
needed to be used with cautions.
How about the performances under lateral loads?
25
Estimation of lateral pile capacity assuming pile is rigid
enough and has sufficiently high strength, and failure occurs
in the clay
Brom's Theory:
Free head, L/d=20, e/d=0, Hu/cud^2=60
26
Hu=6000 kN
421 of 607
Lateral loading on pile in Plaxis 3D
27
Lateral load – movement curves
400
350
Embeddedpile
Solid pile
300 Solid pile
Lateral pile movement (mm)
250
200
150
100
50
0
0 1000 2000 3000 4000 5000 6000 7000
28
Lateral load (kN)
422 of 607
Embedded pile in Plaxis 3D
Take note that NO interface element
provided for the lateral soil sliding
around the embedded pile. As such,
for laterally loaded case, embedded
pile only works well in working
condition (FOS=2.0~3.0) when there
is not much pile‐soil lateral sliding
occurring, it can not be used for
simulation of ultimate lateral
loading scenario.
embedded pile interacting with
surrounding soils
29
BM under lateral loadings
14000
12000 Embeddedpile
Solid pile
Pile bending momnet (kNm)
10000
8000
6000
4000
2000
0
0 1000 2000 3000 4000 5000 6000 7000
30
Lateral load (kN)
423 of 607
Comparison of pile deflection profiles
‐20 0 20 40 60 80 100 120
0
‐2
‐4
‐6
Elevation (m)
‐8
Embeddedpile Fx=1000kN
Embeddedpile Fx=1000kN
‐10
Embeddedpile Fx=2000kN
Embeddedpile Fx=2000kN
‐12 Embeddedpile Fx=3000kN
Embeddedpile Fx=3000kN
Embeddedpile Fx=4000kN
‐14
Solid pile Fx=1000kN
Solid pile Fx=1000kN
‐16 Solid pile Fx=2000kN
Solid pile Fx=2000kN
Solid pile Fx=3000kN
‐18
Solid pile Fx=3000kN
Solid pile Fx=4000kN
‐20
Lateral pile movemetn (mm) 31
Comparison of BM profiles
‐2000 0 2000 4000 6000 8000
0
‐5
Elevation (m)
‐10
Embeddedpile Fx=1000kN
Embeddedpile Fx=1000kN
Embeddedpile Fx=2000kN
Embeddedpile Fx=2000kN
‐15 Embeddedpile Fx=3000kN
Embeddedpile Fx=3000kN
Embeddedpile Fx=4000kN
Solid pile Fx=1000kN
Solid pile Fx=1000kN
‐20 Solid pile Fx=2000kN
Solid pile Fx=2000kN
Solid pile Fx=3000kN
Solid pile Fx=3000kN
Solid pile Fx=4000kN
‐25
32
Lateral pile movemetn (mm)
424 of 607
Conclusions
• Embedded pile is a good model of single pile response
subjected to both vertical compression loads and
vertical pull out loads.
• Solid pile may give very large end bearing resistance
when subjected to vertical compression load, and
generate large suction force at the base when subjected
to pull out load, and thus must be treated with caution.
• Under working load condition with FOS=2~3 when there
is no much pile‐soil lateral slide occurring, embedded
pile give very reasonable pile deflection and pile bending
moment. However, embedded pile can not be used for
simulation of ultimate lateral loading scenario.
33
425 of 607
0123345
ÿ455667
ÿ789
ÿ
ÿ
8
ÿ4ÿ2
ÿ
ÿ
23ÿ!"ÿ22
426 of 607
427 of 607
428 of 607
429 of 607
430 of 607
431 of 607
432 of 607
433 of 607
434 of 607
435 of 607
436 of 607
437 of 607
438 of 607
439 of 607
440 of 607
441 of 607
442 of 607
443 of 607
444 of 607
445 of 607
446 of 607
447 of 607
448 of 607
449 of 607
450 of 607
451 of 607
452 of 607
453 of 607
454 of 607
455 of 607
456 of 607
457 of 607
458 of 607
459 of 607
460 of 607
461 of 607
462 of 607
463 of 607
464 of 607
465 of 607
466 of 607
467 of 607
Simulation of laboratory tests
Computational Geotechnics 1
468 of 607
Simulation of laboratory tests
2 Computational Geotechnics
469 of 607
Simulation of laboratory tests
INTRODUCTION
In daily engineering practice soil parameters are obtained from one or more laboratory tests. In order to perform
the best possible Plaxis calculation these soil parameters have to be translated into input parameters for the
constitutive model used, taking into account the possibilities and limitations of the constitutive model. Most
parameters for the constitutive models used in Plaxis can be determined directly from standard laboratory tests
as triaxial tests and oedometer tests. However, due to the complexity of the models it is recommended to
not simply accept the parameters determined from those tests, but to actually model the tests and see if the
parameters found actually give a proper representation of the real laboratory test results within the limits of the
constitutive models. For this purpose the SoilTest module is available in Plaxis with which in a simple manner
laboratory tests can be simulated without the need for making a finite element model.
In this exercise the SoilTest tool will be used for the simulation of both oedometer and triaxial tests on sand and
clay.
CONTENT
• Simulation of laboratory tests
2. Perform the laboratory tests using SoilTest with the parameters found
3. Match SoilTest results with the original laboratory results to find the best matching model parameters for
the Hardening Soil model.
Parameter determination
On a sample of dense sand both oedometer tests and triaxial tests have been performed. The results of those
tests are given in the figures below. Use these figures to determine the parameters for the Hardening Soil model
and collect the parameters in Table 1 (see below the figures). Note that it is possible that some parameters
cannot be determined with the given laboratory results, in which case these parameters have to be estimated.
Computational Geotechnics 3
470 of 607
Simulation of laboratory tests
4 Computational Geotechnics
471 of 607
Simulation of laboratory tests
Computational Geotechnics 5
472 of 607
Simulation of laboratory tests
6 Computational Geotechnics
473 of 607
Simulation of laboratory tests
Figure 6: Undrained triaxial (CU) tests at cell pressures of 100 kPa and 400 kPa
Computational Geotechnics 7
474 of 607
Simulation of laboratory tests
8 Computational Geotechnics
475 of 607
Simulation of laboratory tests
For a cell pressure σ30 = 100 kPa a maximum value of approximately |σ10 − σ30 | = 400 kPa is reached at failure.
The Mohr-Coulomb failure criterium is:
1 0
2 |σ1 − σ30 | + 12 (σ10 + σ30 ) · sinϕ − c · cosϕ = 0
Considering it is sand we assume that the cohesion is zero and so the Mohr-Coulomb failure criterium reduces
to:
|σ10 −σ30 |
(σ10 +σ30 ) = sinϕ
Filling in σ30 = 100 kPa and σ10 = 500 kPa as obtained from the test we find for the
friction angleϕ0 = 420
The triaxial test stiffness E50 is the secant stiffness over the first 50% of the failure value for | σ10 − σ30 |. This is
indicated in red in the triaxial test graph of figure 8.
σ 0 =100 kP a 400
E503 = 0.013 = 30800 kP a
Computational Geotechnics 9
476 of 607
Simulation of laboratory tests
The triaxial test stiffness ,E 50 , is within the Hardening Soil model defined as:
m m
c cosϕ−σ30 sinϕ σ30
ref ref
E50 = E50 c cosϕ+pref sinϕ , c = 0 ⇒ E50 = E50 − pref
The reference stress pref is chosen equal to the cell pressure of this triaxial test then
ref σ 0 =100 kP a
E50 = E503 ≈ 30000 kPa
Similar to the determination of the reference stiffness for triaxial testing the reference unloading-reloading stiffness
can be determined. In the triaxial test results an unloading-reloading cycle is done for this. The Hardening Soil
model does not have unloading-reloading behaviour with hysteresis but simple non-linear elastic unloading-
reloading behaviour. Therefore a secant value is taken for the unloading-reloading behaviour, as given with the
green line in the triaxial test results.
σ 0 =100 kP a 400
Eur3 = 0.026−0.021 = 80000 kPa
Under the same assumptions as for the stiffness in triaxial testing counts:
ref σ 0 =100 kP a
Eur = Eur3
But this is a bit low value for the unloading reloading stiffness and so
ref
Eur = 90000 kPa
is chosen
Dilatancy angle
From the plot of axial strain versus volume strain the dilatancy angle can be determined according to
∆εv
sinψ = −2∆ε1 +∆εv
10 Computational Geotechnics
477 of 607
Simulation of laboratory tests
From the oedeometer test results we determine the stiffness Eoed for vertical stresses σy0 = 100 kPa en σy0 =
200 kPa, see figure 10. Note that Eoed is a tangent stiffness. Make sure to use the primary loading part of the
oedometer test results.
σ 0 =100 kP a
y 320−0
Eoed = 1.4%−0.33% = 29900 kPa
σy0 =200 kP a 400−0
Eoed = 1.4%−0.47% = 43000 kPa
Within the Hardening Soil model the stress dependent oedometer stiffness is defined as:
m m
c cosϕ−σy0 sinϕ σy0
ref ref
Eoed = Eoed c cosϕ+pref sinϕ , c = 0 ⇒ Eoed = Eoed − pref
ref σ 0 =100 kP a
Eoed = Eoed
3
≈ 30000 kPa
The power m for stress dependent stiffness can now be determined as:
σ 0 =200 kP a m
y
σy0
Eoed 43000 200 m
ref
Eoed
= pref ⇒ 30000 = 100 ⇒ m = 0.5
Computational Geotechnics 11
478 of 607
Simulation of laboratory tests
The K0 value for normal consolidation (K0N C )can only be obtained if measurements for horizontal stresses have
been performed during the oedometer test. If so, results as given in figure 11 may be obtained. From the primary
loading line can be obtained that
0
∆σx ∆σ30 100
K0N C = ∆σy0 = ∆σ10 = 300 = 0.33
12 Computational Geotechnics
479 of 607
Simulation of laboratory tests
If no triaxial test with unloading-reloading is available the unloading-reloading stiffness can also be determined
from an oedometer test with unloading. However, the unloading-reloading stiffness required for the Hardening
Soil model is stress dependent on σ3 while the oedometer test results presented in figure 10 give the strain vs
the vertical stress σy (= σ1 voor oedometer testing).
ref σ 0 =100 kP a
Eur = Eur3 ≈ 110000 kPa
ref
This is a bit high and so a value of Eur = 90000 kPa is chosen.
Computational Geotechnics 13
480 of 607
Simulation of laboratory tests
CLAY
We start with the determination of the strength parameters based on the CU triaxial tests.
The black dotted lines is the Mohr-Coulomb failure criterium in the p’-q plane. In principal stresses the Mohr-
Coulomb failure criterium is defined as:
With p0 = (σ10 + 2σ30 )/3 and q = σ10 − σ30 under triaxial test conditions this can be rewritten as:
2p0 + 13 q
q 6sinϕ 0 6c cosϕ
2 = 2 sinϕ − c cosϕ = 0 ⇒ q = 3−sinϕ p + 3−sinϕ
Hence, the slope M of the Mohr-Coulomb line in p’-q plane is defined as:
6sinϕ 195
M= 3−sinϕ = 200 ⇒ ϕ = 250
From the intersection between Mohr-Coulomb line and the vertical axis where p=0 the cohesion can be determined:
6c0 cosϕ
q= 3−sinϕ = 0 ⇒ c = 0 kPa
14 Computational Geotechnics
481 of 607
Simulation of laboratory tests
From the results of the oedometer test the oedometer stiffness as well as the unloading-reloading stiffness can
be determined. As the graph is given on logarithmic scale one cannot simply draw a tangent line as was done
for the oedometer test on sand.
Considering that both primary loading and unloading/reloading paths are straight lines in the log(p)-εv graph,
hence they have a relation of the form:
εy = εv = A · log(σy0 )
ε2 −ε1 0.370−0.270
A= log(σ2 )−log(σ1 ) = log(120)−log(30) =0.166
In order to determine the stiffness we calculate the derivative of the strain over the stress and change to natural
logarithm:
ln(σy0 )
εy = εv = A · ln(10)
dεy dσy0 ln(10)
dσy0 =A· 1
ln(10) · 1
σy0 ⇒E= dεy = A · σy0
σy0
ln(10)
E = Eoed = A · pref − pref
In the Hardening Soil model the oedometer stiffness is defined as (assuming c = 0) :
m
σy0
ref
Eoed = Eoed pref
Hence:
ref ln(10)
Eoed = A · pref and m=1
If we choose pref = 100 kPa and with the previously determined A = 0.166 we get:
Computational Geotechnics 15
482 of 607
Simulation of laboratory tests
εy = εv = B · log(σy0 )
ε2 −ε1 0.427−0418
B= log(σ2 )−log(σ1 ) = log(120)−log(30) =0.0149
dσy0 ln(10)
Eur = dεy = B · σy0
However, the Eur in the Hardening Soil model is dependent on the smallest principal stress, which is σx0 in an
oedeometer test and not σy0 .
During the unloading process there is no linear relation between horizontal and vertical stress, as in the beginning
of unloading σy0 > σx0 where as after much unloading σy0 < σx0 . Therefore the assumption is made that during
unloading on average σx0 = σy0 .
0
ln(10) ln(10) ln(10) σx
Eur = B · σy0 = B · σx0 = B · pref pref
0
m
ref σx
Eur = Eur − pref
ref
Follows, in a similar way as for the Eoed , that
As only undrained triaxial test data is available it is only possible to determine an undrained E50 and not an
effective E50 . Therefore the only solution is to estimate the E50 with several runs of the SoilTest program using
different input values for the reference E50 until the best fit for the undrained triaxial test data is found. Typically
for normally consolidated clays the effective reference E50 is in the range of 2-5 times the effective reference
ref
Eoed , hence this can be used as a start value for the estimation procedure. By doing so a value E50 ≈ 3.5 MPa
of is found.
The K0-value for normal consolidation can only be obtained if measurements for horizontal stresses have been
performed during the oedometer test. As this is not the case here we can only use the estimation according to
Jaky’s rule:
Poisson’s ratio
The Poisson’s ratio for unloading and reloading is again estimated as νur = 0.2
16 Computational Geotechnics
483 of 607
Simulation of laboratory tests
Computational Geotechnics 17
484 of 607
Simulation of laboratory tests
18 Computational Geotechnics
485 of 607
Simulation of laboratory tests
In the following paragraphs a step-by-step description is given on how to model both an oedometer test and a
triaxial test with the help of many screen shots of the SoilTest tool. Please note that any parameters given on
those screen shots have no relation with the actual exercise and are solely for illustrating the possibilities of the
SoilTest tool.
Computational Geotechnics 19
486 of 607
Simulation of laboratory tests
In order to model an oedometer test first the material data set has to be created. After doing so, press the
<SoilTest> button to start the SoilTest tool. The window that opens is show in figure .
In the main window select the Oedometer tabsheet and set the parameters as indicated in Figure .
After the the oedometer test has been calculating graphs with results appear at the bottom of the SoilTest window.
The user can double-click these graphs to view them in separate windows. Furthermore, custom charts can be
added, see figure 4.
20 Computational Geotechnics
487 of 607
Simulation of laboratory tests
After the triaxial test has been calculated graphs with results appear at the bottom of the SoilTest window. As
described above for the oedometer test, the user can double-click this graphs to view them in separate windows
as well as add custom charts.
Computational Geotechnics 21
488 of 607
Simulation of laboratory tests
The standard functionality in SoilTest for simulation of a triaxial test does not allow for an intermediate unloading-
reloading path. However, the SoilTest functionality contains a General option with which soil test can be defined
in terms of boundary stresses or strains on all sides of a soil test cube. Hereafter it will be shown how this can
be used for the simulation of a triaxial test with unloading/reloading path.
After opening the SoilTest option from the material set definition window the tabsheet General should be chosen.
On this tabsheet a list of calculation phases can be defined where stress or strain increments can be applied.
Initial phase
First of all we have to specify whether stresses or strains will be applied on the boundaries during the test. For
this exercise stresses will be applied. Now the values of the initial stresses on the soil sample have to specified.
For a triaxial test the initial stresses are the cell pressures acting on the soil, hence for σxx , σyy and σzz the cell
pressure has to entered. The cell pressure is a water pressure and so there will be no shear stress acting on the
soil: τxy = 0. See figure for details.
Figure 19: General option for simulation of laboratory tests used for triaxial test
Phase 1
Apply a stress increment in vertical direction (∆σyy ) until the stress level where the unloading path should start.
Note that the horizontal stresses (∆σxx and ∆σzz ) remain the same as they represent the cell pressure. Hence,
the horizontal stress increments are zero in this phase.
Phase 2
Press the Add button to add another phase to the phase list. This phase represents the unloading phase. See
figure for details.
Phase 3
Press the Add button once more in order to add the 3rd phase. This phase represents the reloading of the soil
as well as the continuation of primary loading until either failure or a higher stress level from where for instance
22 Computational Geotechnics
489 of 607
Simulation of laboratory tests
Figure 20: Unloading/reloading cycle in a triaxial test using the General option
Computational Geotechnics 23
490 of 607
Derivation of Soil Parameters from
Lab Test Results & Verification in
Plaxis SoilTest
by
RF Shen
23 Nov 2011
Singapore 2011
491 of 607
In this exercise, we are going to fully utilize the test
data to derive soil parameters for Hardening Soil
Parameters from most common stress-strain data
provided in a typical SI report, and subsequently
use Plaxis SoilTest to verify the derived parameters
Part 1: Sand
Singapore 2011
492 of 607
For sand, one of the most common lab tests is
Triaxial Isotropically Consolidated Drained (CID)Test
Fa/A = q (deviatoric
stress)
Typical sample size 38 mm Ø x 76 mm a = q + r
Singapore 2011
493 of 607
450
400
350
Deviator stress (kPa)
300
250
100 Test data
50
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
Axial strain
0.06
0.05
0.04
Volumetric strain
0.03
0.02
0.01
Test data
0
0 0.02 0.04 0.06 0.08 0.1
‐0.01
Axial strain
Singapore 2011
494 of 607
Another common lab test is Oedometer Test
Settlement dial gauge
Oedometer Cell
Sample: dia. =75mm
Protruded
lever arm Height = 20mm
Heavy dead weights
Singapore 2011
495 of 607
Another common lab test is Oedometer Test
Boundary
conditions
0
0.1
0.2
0.3
0.4
Vertical strain (%)
0.5
0.6
0.7 Test data
0.8
0.9
1
1.1
1.2
1.3
1.4
0 100 200 300 400
Vertical pressure (kPa)
Singapore 2011
496 of 607
400
300
Vertical pressure (kPa)
200
100
Test data
0
0 50 100 150 200
Lateral stress (kPa)
Singapore 2011
497 of 607
Part 1: Strength parameters
450
400
Since c’ = 0 for sand, it can be
350
simplified to:
Deviator stress (kPa)
300
250
100 Test data
50
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
500 100
sin ' 0.67
Axial strain
500 100
' 42
Singapore 2011
498 of 607
Part 1: Strength parameters
0.06
0.05
So,
0.048
0.04
Volumetric strain
0.004
Test data sin 0.27
0
0 0.02 0.04 0.06 0.08 0.1
0.03 0.09
16
‐0.01
Axial strain
pc p’
Singapore 2011
499 of 607
Part 2: Stiffness parameters
450
400 400
350
Deviator stress (kPa)
300
250
3’ = 100 kPa
200
150
100 Test data
50
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
0.013
Axial strain
400
E50ref 30800 kPa 30000 kPa
0.013
Singapore 2011
500 of 607
Part 2: Stiffness parameters
450
400400
350
Deviator stress (kPa)
300
200
150
100 Test data
50
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
0.021 0.026
Axial strain
400
Eurref 80000 kPa
0.026 0.021
Singapore 2011
501 of 607
Part 2: Stiffness parameters
0
0.1
0.2
0.3
0.33
0.4
Vertical strain (%)
0.5
0.6
0.7 Test data
0.8
0.9
1
1.1
1.2
1.3
1.4
0 100 200 300 320 400
Vertical pressure (kPa)
320
ref
Eoed 29900kPa 30000kPa
1.4% 0.33%
m
ref c cot ' '1
Eoed Eoed ref
c cot ' p
Singapore 2011
502 of 607
Part 2: Stiffness parameters
0
0.1
320
29900kPa 30000kPa
0.2 ref
Eoed
0.3
0.4
1.4% 0.33%
0.47
Vertical strain (%)
0.5
400
0.6 200 kPa
Eoed 43000kPa
0.7
0.8
Test data
1.4% 0.47%
0.9
1
1.1
1.2
1.3
1.4
0 100 200 300 400
Vertical pressure (kPa) 400
m
c cot ' '1 200
200 kPa m
Eoed 43000
m = 0.5
c cot ' p
ref ref
Eoed 100 30000
Jaki’s formula:
300
Vertical pressure (kPa)
200
100
Test data
0
0 50 100 150 200
Lateral stress (kPa)
x ' 100
K 0NC 0.33
y ' 300
Singapore 2011
503 of 607
Summary of Hardening Soil Parameters
Singapore 2011
504 of 607
Part 2: Clay
Singapore 2011
505 of 607
For Clay, one of the most common lab tests is Triaxial
Isotropically Consolidated UnDrained (CIU)Test
Fa/A = q (deviatoric
stress)
Close the valve = Undrained test =
a = q + r
Excess will accumulate with shearing
350
Test data
300
250
q (kPa)
200
195
150
100
50
0
0 50 100 150 200 250 300 350 400 450
p' (kPa)
Singapore 2011
506 of 607
CIU stress path
Gradient:
350
Test data
300
200
195
150
100 ’ = 25
50
0 Intercept:
0 50 100 150 200 250 300 350 400 450
p' (kPa)
c’ = 0
Singapore 2011
507 of 607
Another common lab test is Oedometer Test
Test data
0.1
Vertical strain (%)
0.2
0.3
0.4
0.5
1 10 100 1000
Vertical pressure (kPa)
Singapore 2011
508 of 607
Oedometer test for clay
0
d y '
Test data
Eoed
0.1
d y
Vertical strain (%)
0.2
d (log y ' )
Gradient _ k
d y
0.3
0.5
1 10 100 1000
Vertical pressure (kPa)
ln y ' 1
d ( y ' )
d (log y ' ) d( )
2.3 1 y' 1 d ( y ' ) 1
Gradient _ k Eoed
d yy d yy 2.3 d yy 2.3 y ' d yy 2.3 y '
Test data
0.1
ref
Eoed 2.3 100 6.02 1350 kPa
Vertical strain (%)
0.2
Eoed y '
0.27 ref
0.3
Eoed pref
0.37 m
0.4
ref c cot ' '1
Eoed Eoed ref
0.5
c cot ' p
1 10 30 100 1000
120 m
Eoed '1
Vertical pressure (kPa)
ref
Eoed 2.3 y ' gradient _ k
ref
Eoed p
gradient _ k
log(120) log(30)
6.02
m=0
0.37 0.27
Singapore 2011
509 of 607
Oedometer test for clay
Eur refers to when 3’ =
0
100kPa
Test data
0.1
0.3 x’<100kPa;
When y’ loaded to about
0.418
0.4
300kPa and unload to
0.427
0.5 100kPa, x’ is expected to be
30
closer to 100kPa. As such, we
1 10 100 1000
Vertical pressure (kPa) 120
can approximately accept the
derived Eur.
log(120) log(30)
gradient _ k 66.9 Eur 2.3 100 66.9 15000 kPa
0.427 0.418
80
3’ = 100kPa for
consolidation,
During shearing, 3 = 0
70
60
Excess pore pressure
Deviator stress (kPa)
50 accumulates during
40 shearing 3’ 100kPa
3’ = 100 kPa
30
Typically for NC clay, E50ref
20
may be about
Test data 2~5 times
10 Eoedref or about 2800kPa~7000kPa.
Trial runs to fit the test data gives
0
0 0.01 0.02 0.03 0.04 E50ref =0.06
0.05 3500kPa
0.07
Axial strain
Singapore 2011
510 of 607
Summary of Hardening Soil Parameters
Singapore 2011
511 of 607
Let’s call it a day!
Thank you!
Singapore 2011
512 of 607
21/11/2011
Presented By
RF Shen
24 November 2011
Briefing of the Project
513 of 607
21/11/2011
Dia. = 1 m with 200mm overlapping
Secant wall
d = 1 m
I = 0.049 m^4/m
E= 2.70E+07 kPa
A = 0.79 m^2/m
c/c spacing = 0.80 m
so,
EA/m = 2.65E+07 kN/m Take c/c spacing of 0.8m
EI/m = 8.28E+05 kNm^2/m Take c/c spacing of 1.6m
3
weight = 19 kN/m/m
The ground anchors are made of 32mm dia. Steel
bars at c/c spacing of 1m. The steel bar have a
stiffness of Es=2.1*10^8 kPa and with ultimate
strength of 605kN per anchor.
The anchors will be pre‐stressed to 60% of the
ultimatee strength, namely 363kN/anchor.
The properties of the grout body can be ignored.
514 of 607
21/11/2011
Anchor grout body (geogrid element with Elastoplasticity):
d = 0.032 m
E= 2.10E+08 kPa
A = 8.04E‐04 m^2/m
c/c spacing = 1.00 m
so,
EA/m = 1.7E+05 kN/m
Max axial force = 605 kN/m 5
The upper 40m of the subsoil consists of a more
or less homogeneous layer of medium dense
sand. Typical soil parameters based on triaxial
tests are presented in the next slide.
Underneath this layer there is a very stiff layer of
gravel which can be acted as the bottom
boundary of the 2D FEM mesh.
The ground water table is very deep and does not
play a role in this analysis.
6
515 of 607
21/11/2011
Soil parameters
Simulation in Plaxis 2D
version 2010
516 of 607
Tied-back excavation using HSsmall
TIED-BACK EXCAVATION
Using the HSsmall model
Computational Geotechnics 1
517 of 607
Tied-back excavation using HSsmall
2 Computational Geotechnics
518 of 607
Tied-back excavation using HSsmall
INTRODUCTION
A building pit was constructed in the south of the Netherlands. The pit is 15 m deep and 30 m
wide. A diaphragm wall is constructed using 100 cm diameter bored piles; the wall is anchored
by two rows of pre-stressed ground anchors. In this exercise the construction of this building
pit is simulated and the deformation and bending moments of the wall are evaluated.
The upper 40 m of the subsoil consists of a more or less homogeneous layer of medium dense
fine sand with a unit weight of 18 kN/m3 . Triaxial test data of a representative soil sample is
given in figure 2. Underneath this layer there is very stiff layer of gravel, which is not to be
included in the model. The groundwater table is very deep and does not play a role in this
analysis.
AIMS
• Using interface elements
• Pre-stressing of anchors
0 x 4 1
Stage 1
Secant wall
11 12
Stage 2
13 14 Anchor rods
Stage 3
7 8 15
Grout bodies
16 17
9 5 18 10
6
3 2
Computational Geotechnics 3
519 of 607
Tied-back excavation using HSsmall
4 Computational Geotechnics
520 of 607
Tied-back excavation using HSsmall
MATERIAL PARAMETERS
In this exercise the HSsmall model is used and the model parameters for the sand layer
have been extracted from the triaxial test data (see figure 2). The HSsmall model takes into
account the stress-dependency of soil stiffness, elasto-plastic behaviour under both compres-
sion loading and shear loading and increased stiffness in areas with very low strain levels.
The soil parameters can be found in table 1, while the determination of the soil parameters
can be found in appendix A.
Secant wall
The secant wall consists of 100cm diameter bored piles with an intermediate distance of 80cm,
hence there is a 20cm overlap of the piles. This configuration is taken this into account for the
determination of the cross sectional area (A) and moment of inertia (I) per meter out-of-plane
(see Appendix B). The concrete stiffness is Ec =2.7•107 kN/m2 with a specific weight γ=16
kN/m3, which leads to the material parameters as given in Table 2. The determination of the
stiffness parameters can be found in Appendix A.
Computational Geotechnics 5
521 of 607
Tied-back excavation using HSsmall
Ground anchors
The anchors are made of 32mm diameter steel bars at an intermediate distance of 1m. The
steel bars have a stiffness of Es =2.1*108 kN/m2 . The anchors have an ultimate strength
of 605 kN per anchor. In combination with a secant wall the anchors may be prestressed
to a maximum level of 60% of the ultimate strength, hence up to 363 kN per anchor. The
maximum compression force of the anchor is not important as the anchors will not be loaded
under compression. The grout body that forms the bonded length of the anchor behaves
relatively weak under tension compared to the steel bar inside. Therefore it is assumed that
both stiffness and strength of the bonded part of the anchor are fully determined by the steel
bar. This leads to the material properties for both the anchor rod (free length) and grout body
(bonded length) as given in tables 3 and 4.
6 Computational Geotechnics
522 of 607
Tied-back excavation using HSsmall
Computational Geotechnics 7
523 of 607
Tied-back excavation using HSsmall
8 Computational Geotechnics
524 of 607
Tied-back excavation using HSsmall
GEOMETRY INPUT
• Start a new project
Project properties
• Accept the default values in the Project tab sheet of the Project properties (15-node
elements). For the dimensions see figure 3.
Geometry
(15,0)
(0,0) (70,0)
0 x 4 1
(0,-5) 11 12
(0,-10) 13 14
(30,-15)
(0,-15) 7 8 15
(37.5,-20)
(30,-20) 16 17
(0,-25) (70,-25)
9 5 18 10
6 (37.5,-25)
(15,-27)
(0,-60) (70,-60)
3 2
• Click the Geometry line button and draw the geometry contour and soil layers as
specified in figure 4.
Computational Geotechnics 9
525 of 607
Tied-back excavation using HSsmall
• Click the Plate button and draw the secant wall from (15, 0) to (15, -25).
• Click the Interface button and draw the interface from (15,0) to (15, -27) and back to
(15,0). This creates an interface on both sides of the secant wall.
• Click the Node-to-node anchor button and insert both anchor rods. These anchors
connect the beginning of the grout bodies to the wall.
• Finally, click the Geometry line button again to introduce the two levels of excavation.
Fixities
Material properties
• Enter the material properties for the four soil data sets, as determined in table 1of this
exercise.
• After entering all properties for the three soil types, drag and drop the properties to the
appropriate clusters.
• Enter material properties for the plates, anchors and ’geogrids’ as indicated in tables 2,
3and 4.
10 Computational Geotechnics
526 of 607
Tied-back excavation using HSsmall
Mesh generation
• From the Mesh menu, set the Global coarseness to Medium and press the Generate
button. This will result in a mesh as shown in figure 5.
• Select the geogrid and plate elements and press Refine line from the Mesh menu. This
will result in a refinement around the selected lines as shown in Figure 6.
Computational Geotechnics 11
527 of 607
Tied-back excavation using HSsmall
12 Computational Geotechnics
528 of 607
Tied-back excavation using HSsmall
CALCULATION
• When starting the calculation program choose the Classical mode.
The entire construction process consists of five phases. Define the phases, as shown graph-
ically below. For each phase, use the Plastic calculation, Staged construction.
Initial phase
For the initial phase choose the K0 procedure for calculating the initial stresses. As the phreatic
line is located below the geometry the generation of initial pore pressures can be skipped and
since it’s not necessary to switch off any soil for the initial situation it is not needed to define
the initial phase.
Phase 1
• In the first phase, the diaphragm wall is activated and the first excavation takes place.
• Note that though the the interfaces along the wall are activated automatically with the
activation of the wall, the extensions below the diaphragm wall have to be activated
manually.
Phase 2
In the second phase, a new option is used, namely the prestressing of anchors.
• First the grout-body (the geogrid) is switched on by clicking on the ’geogrid’ element.
The element will appear in yellow as soon as it is switched on. The light grey colour
indicates non-active elements.
Computational Geotechnics 13
529 of 607
Tied-back excavation using HSsmall
• Now that the grout-body is active, the anchor element needs to be prestressed. By
double clicking on a node-to-node anchor a window will appear as shown in figure 9.
• Select the option Adjust prestress, fill in a prestress force of 300 kN/m (tension) and
press OK.
• In the geometry a black node-to-node anchor indicates that the anchor is activated. The
letter P indicates that a prestress force will be active in the anchor.
Phase 3, 4 and 5
Now define the remaining phases according to figures 10, 11 and 12.
14 Computational Geotechnics
530 of 607
Tied-back excavation using HSsmall
Figure 10: Phase 3: Second excavation Figure 11: Phase 4: Activation and
prestressing of 2nd anchor
Computational Geotechnics 15
531 of 607
Tied-back excavation using HSsmall
16 Computational Geotechnics
532 of 607
Tied-back excavation using HSsmall
INSPECT OUTPUT
The results of fase 5 is presented in Figure 13. After this final stage the excavation bottom
heave calculated is about 5 cm.
• By double clicking on the node-to-node anchors, Plaxis will present a table, in which the
stress in all anchors may be inspected. Anchor forces are approximately 340 kN where
the lower anchor has a slightly higher anchor force than the upper anchor.
When double-clicking on one of the geogrids the change of axial forces within the grout body
can be investigated. What is immediately noticeable is that the axial force at the connection
with the anchor rod is significantly lower than the force in the anchor rod itself. This is due the
fact that the end of the anchor rod is not only connected to the grout body, but also to several
soil elements surrounding the end of the anchor rod. Therefore part of the anchor force is
transferred directly to those soil elements while part of the anchor force is transferred to the
geotextile representing the grout body. The amount of force transferred to the soil depends on
the stiffness of the soil; in this exercise it is 25-35% of the anchor force. However, this effect
has very little influence on other calculation results. That is, it is not so important for other
calculation results how the anchor rod transfers its force; directly to the soil or by means of the
grout body.
• By double-clicking on the wall the structural forces in the wall can be inspected. The
maximum bending moment should be in the order of 350 kNm/m (figure 14)
• When double-clicking on an interface only the results of part of the interface can be seen.
In order to see the results for the whole interface chain, keep Ctrl + Shift pressed on the
keyboard while double-clicking on the interface. In figure 15the left side are the passive
earth pressures and the right side are the active earth pressures. It can be seen that
only a small part of the maximum passive earth pressures has been mobilized at this
stage.
Computational Geotechnics 17
533 of 607
Tied-back excavation using HSsmall
Figure 14: Bending moments in the sec- Figure 15: Effective normal stresses in the
ant wall interface
Geometry size
For any project the geometry has to be made sufficiently large so that the boudary conditions
have no influence on the calculation results. This means in practice that close to the boundar-
ies (with exception of a axis of symmetry) displacements should be small and stresses should
be undisturbed. When using the HSsmall model there is an interesting plot that can be used
to check this.
• From the Stresses menu choose the option State parameters and then G/Gur .
This plot shows the actual shear stiffness divided by the unloading/reloading shear stiffnes
at engineering strain level. For areas with very small deformations the stiffness will be high
(small strain stiffness) and so the value of G/Gur > 1. Hence, the geometry is sufficiently large
if next to the boundaries, with exception of the axis of symmetry, G/Gur > 1, which indeed is
the case.
Hint: State parameters are additional quantities that relate to the state of the
material in the current calculation step, taking into account the stress
history. Examples of state parameters are the isotropic overconsolidation
pressure (pp ) and the hardening parameter γp that specifies the maximum
shear strain level reach in the stress history.
18 Computational Geotechnics
534 of 607
Tied-back excavation using HSsmall
Surface settlements
In Plaxis Output it is possible to see calculation results in a user-defined cross section. This
feature will be used to check the surface settlements behind the secant wall.
• Click the Cross section button . The Cross section points window appears, see figure
16.
It is possible to draw a cross section by hand and check in the Cross section points window
what the coordinates are of the start and end point of the cross section. However, it is also
possible to position the cross section at a specific location by defining the coordinates of the
start and end point manually.
• Move the mouse to the Cross section points window and fill in the coordinates (15, -0.1)
for the first point and (70, -0.1) for the second point and press OK. This will create a
cross section from the secant wall until the right boundary of the model just below the
soil surface. The cross section will open in a new window.
• From the Deformations menu select Total displacements and then u y to see the vertical
displacements of the soil surface. The maximum settlement is 12-13 mm, see figure 17.
Computational Geotechnics 19
535 of 607
Tied-back excavation using HSsmall
20 Computational Geotechnics
536 of 607
Tied-back excavation using HSsmall
Strength parameters
σ1 −σ3
σ1 +σ3
= sin ϕ
370−100
370+100
= sin ϕ
ϕ = 35o
ψ =ϕ − 30 = 5o
Computational Geotechnics 21
537 of 607
Tied-back excavation using HSsmall
Stiffness parameters
Since excavation is considered in this exercise, the input of Young’s modulus E should be
based on unloading, rather than on primary loading. For the same reason, Poisson’s ratio
should also be based on unloading, which results in a somewhat lower value.
The triaxial test has a cell pressure σ 3 = 100 kPa. This corresponds with reference pressure,
so E50 = Eref
50 .
22 Computational Geotechnics
538 of 607
Tied-back excavation using HSsmall
Compared to the original bored piles the repetitive sections have a reduced cross sectional
area. Though it can be analytically derived how much the reduction is, the fastest way to
determine this is to draw the repetitive section on paper with a fine grid based on the original
bored piles with a diameter of 1000mm and an overlap of 200mm and count squares. Using
this method the cross sectional area of the repetitive section is determined as As = 0.74 m2 .
Since the sections are at a distance D apart where D is given as 800mm, the cross sectional
area of the wall per meter is given as:
Awall = ADs = 0.74
0.8
= 0.93 m2 /m
For the moment of inertia is assumed that the influence of the reduced cross sectional area
is negligble as the reduction is close to the axis of bending and symmetric. Therefore the
moment of inertia per meter wall is determined as:
Ipile 4 π·(0.5)4
Iwall = D
= πr
4D
= 4·0.8
= 61.3 · 10−3 m4 /m
Computational Geotechnics 23
539 of 607
Geotextile reinforced embankment
GEOTEXTILE REINFORCED
EMBANKMENT WITH CONSOLIDATION
Computational Geotechnics 1
540 of 607
Geotextile reinforced embankment
2 Computational Geotecnics
541 of 607
Geotextile reinforced embankment
INTRODUCTION
In 1979 a test embankment was constructed in the Netherlands near the town of Almere. The
objective of this test was to measure the influence of geotextile reinforcement on the short
term stability of an embankment on soft soil. Two test embankments were constructed on top
of a layer, one with and one without geotextile. The construction procedure was such that a
ditch was excavated in the clay layer while at the same time a retaining bank was made with
the excavated clay. A cross-section of the reinforced test embankment is given in figure 1.
line of symmetry
geotextile
retaining bank
1
sand fill 2
2
soft clay
1,5
strong sand layer
1 3.5 3.5 1 3 14 7
Cone penetration tests gave an average cone resistance of qc = 150 kPa for the clay. The clay
is considered to be normally consolidated. The behaviour is assumed to be undrained (the
retaining bank should be drained, however). The saturated weight of the clay is 13.5 kN/m3.
A plasticity index of Ip = 50% is assumed. Due to the limited soil data, parameters should
be selected using engineering judgement and by using the correlations given in the lecture
"Evaluation of soil stiffness parameters". To obtain an undrained shear strength for the clay
layer it is suggested to use the correlation su ≈ qc /15. Having no data for the effective cohesion
and the effective friction angle, they have to be estimated from the undrained shear strength in
order to do a consolidation analysis. For the determination of a stiffness parameter for the clay
layer it is suggested to use the correlation Eu ≈ 15000 · su /Ip (%). The shear modulus G is one
third of the undrained Young’s modulus Eu . The effective Poisson’s ratio should be chosen
such that a realistic K0nc is obtained in one-dimensional compression (K0nc = ν 0 /(1 − ν 0 ) ≈ 0.5).
The effective Young’s modulus is calculated from the shear modulus E 0 = 2G(1 + ν 0 ). The
fill was reported to be fully saturated loose sand with a saturated weight of 18 kN/m3 . The
behaviour is considered to be drained. The effective strength properties are estimated at ϕ0 =
30° and c’ = 3 kPa. K0nc is assumed at 0.5. For the stiffness one should take E’ = 4000 kPa
and ν 0 =0.33.
Computational Geotechnics 3
542 of 607
Geotextile reinforced embankment
AIMS
• Calculation of two alternatives within one project.
SCHEME OF OPERATIONS
1. Determination of stiffness & strength properties (clay)
2. Geometry input
3. Calculation
(a) Initial conditions (Pore pressure generation, Initial geometry configuration, Genera-
tion of initial stresses)
(b) Switch on geotextile, excavate ditch + raise retaining embankment
(c) Apply first hydraulic fill
(d) Apply second hydraulic fill
(e) Determine factor of safety
(f) Repeat this using consolidation phases instead of plastic phases.
4. Inspect output
4 Computational Geotecnics
543 of 607
Geotextile reinforced embankment
Note: The main purpose of the exercise is to assess the failure mechanism and
the factor of safety, which has the following consequences for the model:
• The geometry size is chosen such that the failure mechanism fits
within the model boundaries. This means the geometry can be fairly
small.
Computational Geotechnics 5
544 of 607
Geotextile reinforced embankment
GEOMETRY INPUT
General settings
Start a new project and select appropriate General settings. Use 15-node elements as basic
element type since in this exercise we will deal with failure behaviour.
(9.5,7.5) (12,8.5)
10 11
(33,8.5)
(8,7.5)
7 9 8 (33,7.5)
(4.5,5.5) (12,5.5) (26,5.5)
(0,5.5) 1 6 12 13 2
(33,5.5)
(0,3.5) 4 5 (1,3.5)
(0,2) 14 15
(33,2)
y
(0,0) 0 x 3
(33,0)
• Enter the geometry as indicated in the previous graph. The order in which geometry
points are created is arbitrary.
• Click the Geogrid button to introduce the geotextile (from (4.5, 5.5) to (26.0, 5.5)).
• Click the Standard fixities button for the standard boundary conditions.
• After entering all properties for the three soil types, drag and drop the properties to the
appropriate clusters, as indicated in figure 3.
6 Computational Geotecnics
545 of 607
Geotextile reinforced embankment
3
2 3
1
1
Geotextile
• In the project database select the data type Geogrids and create a new material set. In
this material set, enter 2500 kN/m as stiffness. Note that this is the stiffness in extension.
In compression no stiffness is used.
• Drag the geogrid data set to the geotextile in the geometry and drop it there. The
geotextile should flash red once, indicating the properties have been set.
Mesh generation
• From the Mesh menu select the option Global coarseness. In the window that appears,
set the mesh coarseness to Medium and click on the Generate button, which will present
the following FE mesh composed of 15-node elements.
Computational Geotechnics 7
546 of 607
Geotextile reinforced embankment
• Select the clay layer (this consists of two clusters, see also hint) and press Refine cluster
from the Mesh menu. This will result in a refinement in the clay layer that will be needed
for the consolidation analysis. See figure 5.
Close the window showing the generated mesh and continue to the Calculations program.
8 Computational Geotecnics
547 of 607
Geotextile reinforced embankment
CALCULATION
The calculation consists of two alternatives for the construction of the embankment: without
and with consolidation taken into account. After both alternatives the factor of safety is
determined. In the calculations list 8 phases are needed, 4 phases for each alternative. First
start with the fully undrained construction, that is without taking consolidation into account.
When starting Plaxis Calculations, choose Classical mode.
Initial conditions
• Select the initial phase in the phase list and then press the Define button on the Para-
meters tabsheet in order to define the initial phase. The input window now opens in
Staged Construction mode.
• Deselect all material clusters and geotextile elements that are not present at the start of
the analysis. As we want to model the entire construction sequence from the beginning,
switch off:
– Geotextile elements
– Material clusters for the fill
– Material cluster for retaining bank
• Now continue to the Water conditions mode by clicking the equally named button.
• Enter a phreatic level at ground level by two coordinates (0, 5.5) and (33, 5.5). Click on
the Water pressures button to generate the pore pressures.
Computational Geotechnics 9
548 of 607
Geotextile reinforced embankment
After this, we will construct the embankment taking into account consolidation:
10 Computational Geotecnics
549 of 607
Geotextile reinforced embankment
Computational Geotechnics 11
550 of 607
Geotextile reinforced embankment
INSPECT OUTPUT
In order to get a good idea of the displacement mechanism, one can view the contours
of incremental displacements. Figure 6 shows this plot of the final calculation step for the
undrained construction. It clearly shows the effect of the geotextile reinforcement. Figure 7
shows the incremental displacement for the consolidated construction. Here the embankment
has a more gradual settlement without showing an upcoming failure mechanism.
The axial forces of the geotextile can be visualised by double clicking on the geotextile. This
will first present the displacement of the geotextile. On using the menu item Forces, one can
select Axial forces N.
At the ends of the geotextile the axial force must be zero, but due to the discretisation and
some numerical inaccuracy this is not completely achieved. The maximum axial forces is
approx. 8 kN/m. figure 9 shows the axial forces for the consolidated construction. The
maximimum axial force here is only 5-6 kN/m.
Finally, the factors of safety are checked. In order to do so follow these steps:
• Start the curves manager by selecting the Curves manager option from the Tools menu.
12 Computational Geotecnics
551 of 607
Geotextile reinforced embankment
• In the curves manager (see figure 10) select New in the Charts tabsheet. This presents
the Curve Generation window as shown in figure 11.
• On the x-axis we want to show the displacements of the point at the toe of the embank-
ment, hence choose Point A and Deformations → Total displacements → |u|.
• On the y-axis we want to show the strength reduction factor, hence select Project and
Multiplier → ΣM sf on the y-axis.
The created curve indicates a safety factor around 1.4 for the undrained construction and a a
safety factor of 2.1 for the consolidated construction of the embankment, as can be seen in
figure 12.
From the graph above, the factor of safety can be determined. Always look for a steady state
solution (slight variations in the load multipliers, increasing displacements). In most case, the
phi/c reduction calculation shows some variation at the beginning of the calculation. Note
that the displacements resulting from a Safety analysis are non-physical. Hence the total
displacements are not relevant. An incremental displacement plot of the last step, however,
shows the failure mechanism that corresponds the calculated value for ΣM sf .
Addicionally, figures 13 and 14 show the failure mechanisms with the lowest factor or safety
for both the undrained and consolidated case.
Computational Geotechnics 13
552 of 607
Geotextile reinforced embankment
Consolidated: ΣMsf=2.1
Undrained: ΣMsf=1.4
14 Computational Geotecnics
553 of 607
Geotextile reinforced embankment
Computational Geotechnics 15
554 of 607
Geotextile reinforced embankment
FACTORS OF SAFETY
The factors of safety are checked with the Curves program, see figure 19.
16 Computational Geotecnics
555 of 607
Geotextile reinforced embankment
Consolidated: ΣMsf=1.4
Undrained: ΣMsf=1.1
Computational Geotechnics 17
556 of 607
Geotextile reinforced embankment
18 Computational Geotecnics
557 of 607
Geotextile reinforced embankment
Computational Geotechnics 19
558 of 607
E5: Exercise on Pile Analysis
Presented By
RF Shen
25 November 2011
Briefing of the Project
Exercise SG-E5-Shen
559 of 607
Briefing of the Project
y
X = ‐35m
to 35m
X
(‐4 ‐25 0)
(‐.4 ‐25 ‐.8) z y = ‐25m
(.4 ‐25 ‐.8) X to 25m
(4 ‐25 0)
z = 0 to
‐30m
3
Briefing of the Project
The subsoil consists mainly of tertiary
formations of highly plastic clay with lenses of
lignite coal (clay with brown coal). In this
analysis, a uniform clay layer was idealized with
OCR = 1.3
4
Exercise SG-E5-Shen
560 of 607
Soil parameters
Simulation in Plaxis 3D
Step 1: General setting
Step 2: Add in a borehole
Step 3: Define soil properties
Step 4: Create 6 piles
Step 5: Create 1 pile cap
Step 6: Clone another pile group
Step 7: Create the trench
Step 8: Assign vertical loads
Step 9: Generate mesh with refinement
Step 10: Define stages and view results 6
Exercise SG-E5-Shen
561 of 607
Flieden bridge piled-raft foundation
FLIEDEN BRIDGE
PILED-RAFT FOUNDATION
Computational geotechnics 1
562 of 607
Flieden bridge piled-raft foundation
INTRODUCTION
The foundation of the 4-span railway bridge of Flieden in Germany (figure 1) was the first
railway bridge in Germany founded on piled rafts.
The subsoil consists mainly of tertiary formations of highly plastic clay with lenses of lignite
coal (clay with brown coal). To ascertain the adequacy of the piles and determine appropri-
ate design values, pile load tests were first conducted on large diameter bored piles with and
without post shaft grouting (El-Mossallamy et al. 2003). These results conform to the me-
chanical sensitivity of the organic silty clay and lignite coal lenses. It was decided to install all
foundation piles applying post shaft grouting.
INPUT
The bridge piers are consisted of two pillars, each founded on a separate group of 6 piles
underneath a raft. The pile arrangements are shown in Figure 2. The rafts are 1.5 meters
thick and are embedded in the soil with the raft base at a depth of 2.3 meters below the soil
surface. The piles where designed with a diameter of 1.2 m and a length of 18 m. The pillars
transfer two working loads of 20 MN and 22 MN respectively from the superstructure to the
piled raft foundation.
Work flow
In this excercise the model is created in a specific order that has proven to be a rather efficient
way to create the model. Please note that many parts of the model can be created in any
other order as well and the work flow presented here is not the only correct method to create
the project. The work flow to create the project presented here is:
1. Enter dimensions of the project and some general visualisation options
2. Define the underground model using 1 borehole and the appropriate soil material sets
Computational geotechnics 2
563 of 607
Flieden bridge piled-raft foundation
4. Copy this 1 pile 5 times to create the 6 piles needed for 1 piled raft
5. Insert the raft, the lower column and the top load
6. Copy the complete piled-raft 1 time to create the second piled raft
8. Generate mesh
Computational geotechnics 3
564 of 607
Flieden bridge piled-raft foundation
Geometry
General settings
Start the PLAXIS 3D input program. A Quick select project dialog box will appear in which
you can select an existing project or create a new one; choose Start a new project so that the
Project properties window appears.
1. In the Project properties window on the Model tabsheet the size of the model contour
has to be set. In the Contour box fill in xmin = −35, xmax = 35, y min = −25 and y max = 25.
2. Close the Project properties window, the drawing area will now appear.
3. From the Options menu choose Visualization settings. A new window will open, contain-
ing 2 tabsheets: View and Visibility.
4. On the View tabsheet the grid point distance (Spacing) and number of snap intervals per
grid distance can be set. By default the Spacing is set to 1 m with only 1 snap interval
per grid distance. As can be seen from figure 2 many dimensions of this project have an
accuracy of 0.1 m and therefore just 1 snap interval per 1 m is not sufficient. Therefore,
set the Intervals to 10, this will results in having a snap distance of 0.1m (Spacing /
Intervals).
Subsoil
The first step in creating a model in PLAXIS 3D is the definition of the subsoil, which is done
using boreholes.
1. Select the Create borehole button ( ) and move the mouse to the origin of the system
of axis. Click at (x,y,z) = (0 0 0), this will open the Modify soil layers window.
2. In the Modify soil layers window click the Add button in order to define a new soil layer
in this borehole. Set the top of the borehole to 0.0 m and the bottom to -30.0 m.
3. In order to assign a material set to the newly defined model it is necessary to first define
a material set. To do so, press the Materials button ( ) to open the material
sets database.
4. Though the model only has one soil layer (clay) we will have to define two material sets:
the second material set will be used to represent the concrete needed for both raft and
piles. Therefore, create two material sets according to the material parameters specified
in table 1.
5. After defining the two material sets close the window by clicking OK in order to return to
the Material sets window.
Computational geotechnics 4
565 of 607
Flieden bridge piled-raft foundation
Dilatancy angle ψ 0 - o
Permeability kx , k y , k z 0 - m/day
Interface strength Rinter 0.6 (Manual) Rigid -
Coefficient for initial lateral stress K0 Automatic Automatic -
Overconsolidation ratio OCR 1.3 - -
6. Drag and drop the clay material set from the Material sets window onto the borehole. The
mouse cursor changes shape when the material set can be dropped. After dropping the
borehole should get the colour of the material set. Now close the Material sets window
in order to return to the Modify soil layer window.
7. In the Modify soil layer window directly above the graphical representation of the bore-
hole it is possible to specifty a general phreatic level for this borehole by changing the
Head value. In this project the water level is 0.5 meters below ground level, therefore
change the Head to -0.5 m.
8. Press OK to close the Modify soil layers window and return to the drawing area. In the
drawing area there is now a block of soil with the horizontal dimensions specified in the
Project properties window and a depth according to the borehole.
We have now finished defining the subsoil and we will continue defining the foundation. Press
the Structures option ( ) on the mode toolbar to move to Structures mode.
The two bridge foundations are equal with exception of the load from the bridge acting on
the foundation. Therefore it’s sufficient to define 1 foundation and then make a copy of the
Computational geotechnics 5
566 of 607
Flieden bridge piled-raft foundation
foundation to get the second. Similarly, each foundation is supported by six equal piles, hence
it is sufficient to define 1 pile and make 5 copies to model all piles to model 1 foundation.
In the current version of PLAXIS 3D the only possibility to insert a pile is by inserting a cylin-
derical volume using the command line. The syntax for inserting a cylinder is:
cylinder <R> <L> <num planes> (<start_x> <start_y> <start_z>) (<dir_x> <dir_y> <dir_z>)
In short, one specifies the radius (R) and length (L) of the cylinder, a set of 3 coordinates to
indicate the starting point of the cylinder and a vector to indicate the direction of the cylinder.
Special attention should be given to <num planes>. In PLAXIS 3D a cylinder is modelled with
a polygon cross section, hence <num planes> gives the number of sides of the polygon. The
higher the number the more accurate the polygon will represent the circular cross section.
9. Insert the first pile at (x,y) = (-8.4, -1.8). Note that the piles have a 1.2m diameter (hence
a radius of 0.6m), are 18 meters long, start at z = -2.3m and go down vertically, that is
in the negative z-direction. The number of planes is set to 15 to accurately model the
cylinderical shape. This results in the following cylinder command:
Type this command on the command line and press <Enter>. The cylinder is now in-
serted in the model as a volume.
10. In order to assign interfaces around the pile, the pile has to be split into its separate
surfaces. To do so, right click on the pile and from the popup menu choose Decompose
into surfaces.
11. Now select the outer surfaces of the pile, right-click and select Create negative interface.
This will create a negative interface along the outside of the pile.
12. In order to create an interface below the foot of the pile, select the bottom circular surface
of the pile. It is probably necessary to rotate the model in order to see the foot of the pile
from below. Right-click again and select Create negative interface to create the interface
below the foot as well.
Hint: Interfaces are drawn as planes at a certain distance from the surface they
belong to. Therefore, if a project requires a lot of interfaces it may become
difficult to see the underlying structure as the interfaces are surrounding it.
This can be solved by either reducing the distance between interface and
structure or by making the interfaces invisible.
The distance between interface and surface can be reduced in the
Visualization settings that can be found under the Options menu. On the
View tabsheet the field Interface size controls the distance. By default this
value is set to 1. Reducing this value will reduce the distance between
interface and surface.
Alternatively, in the Object explorer it is possible to make the interfaces
invisible by clicking on the small eye in front of the branch Interfaces (to
make them all invisible) or in front of individual interfaces (to make only a
selection of interfaces invisible).
Computational geotechnics 6
567 of 607
Flieden bridge piled-raft foundation
We have now finished creating the first pile. The next step is to make 5 copies of the pile to
create the group of 6 piles of the first foundation slab.
13. Click the button Select rectangle ( ), ignore the suboptions that become available.
Now draw a rectangle that fits the whole pile so that all parts of the pile are selected.
14. Now click the Create array button ( ) to specify the locations of the copies of the pile.
The Create array window appears, see figure 3.
In x-direction we need 3 piles with an intermediate distance of 3.4 meters and in the y-direction
we only need 2 piles with a distance of 3.6 meters in between.
15. Set the Shape of the array to 2D, in xy plane as we want to copy the piles in both x and
y direction, keeping the z coordinate constant
16. Fill in 3 columns with a distance of x = 3.4m in between and 2 rows with a distance of y
= 3.6m in between.
We have now created the 6 piles for one of the bridge foundations.
After creating the 6 piles now the raft has to be modelled on top of the piles, including the
lower part of the column supporting the bridge:
Computational geotechnics 7
568 of 607
Flieden bridge piled-raft foundation
1. From the horizontal button bar with general options, click the Top view button ( ). This
will show the model seen along the z-axis.
2. In the Movement limitation window that appears, fix the z-coordinate to z = -0.8m by
filling in -0.8 in the z-value field and clicking the Set button.
3. Select the Create surface button ( ) and draw the surface representing the top side of
the raft from (x y) = (-9.6 -3.0) to (-9.6 3.0), (-0.4 3.0) and (-0.4 -3.0).
4. Select the surface that has just been created and click the Extrude button ( ). In the
window that opens fill in an extrusion vector of (x,y,z) = (0 0 -1.5) in order to create the
1.5m thick raft and click OK.
Now the raft has been created as volume, in order to assign interfaces to all sides of the raft,
the raft volume has to be decomposed into its surfaces.
5. From the button bar with general options, click the Perspective view button ( ). .
6. Right-click on one of the vertical sides of the raft and select the option Decompose into
surfaces. This will created surfaces for all sides of the volume.
7. For all 6 sides, right-click on the side and add an interface. Note that all sides need
a negative interface with exception of the vertical side at y = 3.0m; this side needs a
positive interface. Check if all created interfaces are on the outside of the raft!
8. In order to make the lower part of the supporting column, click again the Top view button
and fix the z-coordinate to ground level.
9. Create a surface from (x y) = (-6.0 -1.0) to (-6.0 1.0), (-4.0 1.0) and (-4.0 -1.0).
10. Extrude the surface 0.8 meters downwards, hence in the negative z-direction. This cre-
ates the lower part of the column from groundlevel down to the raft.
12. For all 4 vertical surfaces created, create an interface on the outside. That is, negative
interfaces for all vertical sides but the vertical side at y = 1.0m. The latter side needs a
positive interface.
The only part missing now is the load representing both the weight of the bridge and a passing
train
13. Right-click on the top plane of the column, that is the plane at ground level.
14. From the popup menu that opens, select the option Create surface load to add the load.
Computational geotechnics 8
569 of 607
Flieden bridge piled-raft foundation
The second raft is equal to the first raft, hence creating the second raft is simply making a
copy of the first raft:
1. Click the button Select rectangle ( ), ignore the suboptions that become available.
Now draw a rectangle that fits the whole structure of piles, raft and column so that all
parts are selected.
2. Now click the Create array button ( ) to specify the location of the copy of the founda-
tion structure in the Create array window.
3. Set the Shape of the array to 1D, in x direction as we want to copy the foundation just
one time in x direction, keeping the y and z coordinates constant
4. Fill in 2 columns with a distance of x = 10m in between and press OK. Now the second
raft is created as copy of the first raft.
In order to be able to refine the mesh in the area around the rafts it is needed to define a
volume of soil around the rafts where a mesh refinement can be applied. To do so, follow
these steps:
Computational geotechnics 9
570 of 607
Flieden bridge piled-raft foundation
2. Draw a rectangular surface from (x y) = (-10.0 -4.0) to (-10.0 4.0), (10.0 4.0) and (10.0
-4.0).
3. Select the Perspective view, select the newly created surface and extrude it 25m up,
hence in the positive z-direction.
We have now created a volume around the foundation structure that we can use for local mesh
refinement.
Mesh generation
In the Mesh mode we will specify global and local refinements and generate the mesh. In order
to generate more accurate results a refinement of the mesh around the foundation structures
will be applied.
1. In the geometry click somewhere close to the origin. This will select the body of soil that
encloses the foundation structures.
2. In the Selection explorer on the left the selected soil body appears, showing a mesh
refinement factor of 1.0. Change this mesh refinement factor to 0.30.
3. Select the Generate mesh button ( ) in order to generate the mesh. The Mesh options
window appears.
4. In the Mesh options window choose a Very coarse element distribution and click OK to
start the mesh generator.
5. After mesh generation has finished one can already see an indication of the amount of
elements and nodes generated in the command line box below the draw area. For this
project about 22,000 elements should be generated.
After inspecting the mesh the output window can be closed. Mesh generation has now been
finished and so creating all necessary input for defining the calculation phases has been
finished.
Computational geotechnics 10
571 of 607
Flieden bridge piled-raft foundation
Computational geotechnics 11
572 of 607
Flieden bridge piled-raft foundation
CALCULATION
The calculation consists of the initial phase and three additional phases. Since water levels
will remain constant the Water levels mode can be skipped. Therefore, click on the Staged
construction mode button to move to the defintion of the calculation phases.
Initial phase
By default the Initial phase is set to the K0 procedure, which is fine for this example. No further
changes have to be made.
1. Click on the Add phase button ( ) to add the first calculation phase.
2. As the foundations are surrounded by soil they cannot be accessed directly. In order to
change their properties the surrounding soil has to be made invisible. To do so, right-
click on the soil somewhere far away from the origin and from the menu that pops up
choose Hide to hide the outer soil. Now only the foundations and the refinement zone is
left. Make sure the soil is hidden, not deactivated!
3. Right-click on the refinement zone volume and again choose the Hide option from the
popup menu. With the refinement zone hidden, only the foundations structures remain
visible.
4. Open the material sets database by clicking the Show materials button ( ). Drag and
drop the material set representing the concrete on all piles, the rafts and the two parts
of the column. When assigning the material set, the colour changes from the colour of
the material set representing the clay to the colour of the material set representing the
concrete.
5. In the Model explorer, activate all interfaces by clicking on the checkbox in front of the
interfaces branch so that a checkmark appears.
1. Click on the Add phase button ( ) to add the second calculation phase.
2. In the Model explorer open the Surface loads branch and change the value for the two
surface loads. Set the first surface load to a vertical stress of σz = −5000 kN/m2 (20 MN
dived by 4 m2 cross sectional area of the column) and set the second surface load to a
vertical stress of σz = −5500 kN/m2 .
3. Make sure the surface loads are activated, that is that they have a checkmark in the
Model explorer.
Computational geotechnics 12
573 of 607
Flieden bridge piled-raft foundation
1. Click on the Add phase button ( ) to add the third calculation phase.
2. In the Model explorer change the values of the Surface loads to σz = −10000 kN/m2 for
the first surface load and σz = −11000 kN/m2 for the second surface load.
Press the Calculate button ( ) to start the calculation. Ignore the message "No nodes or
stress points selected for curves" as we will not draw any load-displacement curves in this
example, and continue the calculation.
Computational geotechnics 13
574 of 607
Flieden bridge piled-raft foundation
OUTPUT RESULTS
Figure 6 demonstrates the calculated load settlement behaviour of the piled raft applying the
GAPR (Geotechnical Analysis of Piled Raft, El-Mossallamy 1996). Due to the non-linear re-
sponse of the foundation system the loads have been incrementally applied till the ultimate
limit state. Another aim of the analysis under working loads was to determine the pile/soil
stiffness and subgrade reaction distribution beneath the raft, which are necessary for the de-
sign of the foundation. However, within the framework of this exercise the subgrade reaction
distribution will not be checked. Figure 7 shows the measured settlements in comparison to
the calculated values
Figure 6: Load-settlement behaviour of the piled raft foundation (calculated by program GAPR,
El-Mossallamy)
Computational geotechnics 14
575 of 607
3D Excavation Supported By Struts
Three dimensional finite element modelling using
Plaxis 3D
William Cheang, Shen Rui Fu & Tan Siew Ann
11/20/2011
Three dimensional finite element modelling of a 50 x 10 excavation supported by two levels of struts.
Excavation depth is 8.5m and the toe of the retaining structure is located at 12 m below ground level. Struts
spaced at 10m centre‐to‐centre spacing.
576 of 607
INTRO
ODUCT
TION
A. Ge
eometry
y
1. The
e geometryy of the exe
ercise is sh
hown in Ta
able 1.
2. The
e excavatio
on is 50m in 10m in width and 8.5m deep.
i length, 1
3. The
e toe of the
e diaphragm
m wall is lo
ocated at 12m
1 below
w the grounnd.
4. The
e excavatio
on is done in three sttages and supported
d by two leevels of struts spaced
d
at 1
10m centre
e-to-centre distance a
along the le
ength.
each level there are 5 numberss of struts.
5. At e
6. The
e ground co
ondition is simple an
nd it is a sin
ngle layere
ed problem
m.
577 of 607
B. Dime
ensions
s
1. The
e size of th
he models and
a dimen
nsions of th
he excavation is indiccated in Ta
able 2.
2. The
e size of th
he finite ele
ement mod
del in Plax
xis 3D is se
et at 100 x 60 x 22m
m in length,
width and dep
pth respecttively.
Figure
F 2: M
Model size in
i Plaxis 3D
578 of 607
C. Flow
w of worrk
The
e flow of wo
ork consistt the follow
wing steps. In genera
al the stepss will be:
1. Settting up of the
t projectt informatio
on and model size in “Project S
Settings”.
2. In M
Mode-Soil the depth of the mod
del is set and
a the initial positionn of the wa
ater table iss
set..
3. In M
Mode-Structures the dimension
ns of the excavation
e n and excaavation lev
vels will be
e
con
nstructed. This
T will be
b followed
d by the creation of the retain ing wall and supporrt
sysstem.
4. In M
Mode-Mesh
h the mode
el is discre
etised. Refinement will
w be madee on imporrtant zoness
to e
enhance th
he analysis
s and resul ts.
Mode-Watter Levels the wate r condition
5. In M n at each stages oof excavatiion will be
e
seq
quentially lo
owered using the ‘Clu
uster Dry’ option.
6. In Mode-Stag
ged Cons
struction th
he excava
ation proc
cess and constructiion of the
e
reta
aining syste
em will be made. Fin
nally the ca
alculation will
w be madde.
7. Insp
pect the ca
alculated re
esults in O
Output and construct displacemeent chart versus
v time
e
usin
ng the Cha
art option.
ANAL
LYSIS
A. PR
ROJECT
T SETTINGS
Fiigure 3: Qu
uick Selectt option (Se
elect “Starrt a new project” for thhis exercis
se)
579 of 607
Figure 4: S
Setting the size
s of a Pla
axis 3D mo
odel via “Mo
odel Tab”. The
T point off origin is automaticallyy
set at the ccentre of the
e model (x-o
origin = 0, yy-origin = 0) when using
g the abovee stated valu
ues.
1. The
e size in x-y space is set using the “Conto
our” input box.
b
2. The
e dimensio
ons in x-y space
s are 100 x 60 m in length
h and widtth. The min
nimum and
d
maxximum vallues are chosen as such that the point of origin w
will be loca
ated at the
e
cen
ntre in x-p space.
s
3. The
e units and
d general quantities fo
or the anallysis are given in the above figu
ure.
4. The
e dimensio
ons in ‘z-direction is sset using the
t ‘Borehole’ input ooption in ‘M
Modify Soiil
Layyers’ box.
5. Oncce the inpu
ut is complete click ‘O
OK’.
580 of 607
B.MODE: SOIL
L
Plaxis 3D has 5 main mo
odes to deffine the prroject, that is Soil, Sttructures, Mesh,
M
Wate
er Levels and
a Stage
ed Constru
uction. Eve
ery mode has a deedicated ve
ertical
toolbar located between the explorrer boxes and
a drawin
ng area. Inn Soil mod
de we
define the soil layers for projects. In
n this curre
ent case th
here is onlly a single layer
of so
oil. Informa
ation of this soil laye
er and loc
cation of th
he water ttable is en
ntered
using
g a single borehole. To define
e use the borehole
b option
o to deefine the single
s
soil la
ayer, follow
w these ste
eps:
1. Selecct the Bore
ehole tool ffrom the ge
eometry toolbar.
at represents the soil.. It is suggested
2. Clickk at a location in the ccluster tha
to cliick on (x; y) = (-50; -30) but any
a other position
p woould be fin
ne as
well. This place
es a boreh
hole at loc
cation (x; y)
y = (-50; --30) and opens
o
the Modify
M soil layers win
ndow (Figure 5). The Modify soiil layers wiindow
provides a grap
phical reprresentation
n of the cu
urrent boreehole on th
he left
and a spreads
sheet for d
data inputt. However when crreating the
e first
hole in an
boreh a analyssis both the
t graph
hical repreesentation and
sprea
adsheet is empty.
Figure
e 5: Modify soil
s layers window
w
581 of 607
3. Presss Add button to defin
ne a new la
ayer in the borehole. By default both
the top
t and bo
ottom bou
undaries off the bore
ehole are set to z = 0.0.
Chan
nge the bo
ottom boun
ndary to -2
22m. This action in effect is setting
s
the depth
d of the
e finite elem
ment mode
el (see Figure 6)
4. To define
d the material properties
s for this layer of soil, press the
Mate
erial button. The Mate
erial sets window
w will appear
Fig
gure 6: Defin
ning a single
e layered soil using a single
s borehhole
Table
e 1: Material Properties
s for soil
Param
meter Symbol Soil nit
Un
Soil M
Model Name Mohr-Coulomb -
Drain
nage type Undrained (B) -
Unsaturated soill weight unsat 18.0 kN//m3
Saturrated soil we
eight sat 19.0 kN//m3
Young’s moduluss E-ref 15000 kN//m3
Poissson’s ratio Nu 0.25 -
Cohe
esion Su 30 kN//m2
Frictio
on angle 0 -
Interfa
faces Rigid -
Coeffficient of earth pressure
e at rest Ko Man
nual = 0.55 -
582 of 607
6. Creatte a new material set ffor the soil material ac
ccording to the data given in
Table
e 1. Any parrameters no
ot mentioned should be
e left at thei r default va
alue.
7. After entering the material p
properties for
f the soil close
c the w
window by clicking
OK in
n order to re
eturn to the Material se
ets window.
8. The material properties a d into the model byy ‘drag-and
are entered d-drop’
oach onto the boreho
appro ole layer (s
see Figure 7). Once this is don
ne the
desig
gnated soil layer in th
he borehole
e should be
e indicatedd with the colour
chose
en for this material
m set..
Figure
re 7: Boreho
ole with matterial properrties properrly in place.
9. By de
efault the water-table
w is set at 0 m. In this example thhe water ta
able is
ed at ground level and therefore no
locate n alteration
n is neededd. Therefore
e keep
the ground level and there
efore no alteration is needed.
n Th erefore kee
ep the
d at 0 m.
Head
We havve now fin
nished de
efining the subsoil and
a we w
will continu
ue to
Structure
es Mode to
o define the
e geometrry of the ex
xcavation.
C. MO
ODE: ST
TRUCTU
URES
Press tthe Structu
ures option
n on the m
mode toolb
bar to mov
ve to Strucctures mod
de. We willl
create the excavvation size by constru
ucting a vo
olume. This volume w
will be loca
ated at the
e
point o
of origin of the
t model..
583 of 607
To construct the diaphragm wall we will decompose this volume into surfaces. We
will change the surfaces located at the four sides of the volume to plate elements. Also
interface elements will be introduced by using the same approach. Finally we will
construct 3 excavation levels by creating 3 horizontal surfaces at 3 locations along the
z-axis.
The 5 individual struts located at a given level are constructed using Beam elements.
One single strut is drawn and set with the corresponding material properties. This will be
copied and replicated using the Array tool along the x-direction at 10 m centre-to-centre
distance. The waling system located along the perimeter of the retaining wall will be
constructed using Beam elements. The struts and walers located at the first level will be
grouped together to form ‘support system level 1’. This is to allow easy identification and
activation during the staged construction process.
Again by using the Array tool all the struts and waler system (support system level 1)
located at the first level will be copied to the second level and this therefore forms
‘support system level 2’. Create the following material sets and properties listed in Table
2, 3 and 4.
Table 3: Strut
Element type Beam Unit
Identification Name Strut -
Area A 1.225E-1 m2
Weight 78.5 kN/m3
Behaviour Linear -
Modulus of Elasticity E 210.0E6 kN/m2
584 of 607
Second moment
m are
ea I1=
=I2 1.800E-2
1 m4
Table 4: Waling
Element type Beam Unit
Identifica
ation Na
ame Waling -
Area A 8.682E-3
8 m2
Weight 78.5 kN/m33
Behaviou
ur Linear -
Modulus of Elasticity
y E 210.0E6
2 kN/m22
Second moment
m are
ea I1=
=I2 1.045E-2
1 m4
eate the ex
C1.Cre xcavation volume
Figure
e 8: Switch
h to Top Viiew (Plan vview) and movement
m limitation w
window ap
ppears
To expedite the
t constru
uction of th
he foot prin
nt of the excavation we switch
h to the top
p
view
w mode byy clicking the
t tool loccated on the menu. The viewppoint will be switched
d
to p
plan view. A ‘Movem
ment limitatiion’ window
w appears
s (see Figuure 8). Usin
ng this too
ol
it iss possible to
t fix the drawing
d pla
ane at a sp
pecific elev
vation. In tthis case th
he drawing
g
plan
ne is fixed at z = 0 m).
m This is to
o facilitate drawing of
o the excavvation foottprint
585 of 607
e volume is created by firstly d
The drawing a surface in x-y planee at z= 0 m and then
n
extrruding dow
wnwards by 12 m ussing the ex
xtrusion too
ol. From thhe vertical tool bar in
n
Stru
uctures Mo
ode select the Create
e surface option.
o
1. Construct a surface of 50 (L) a
and 10 (W)) m in size by locatinng the four points of a
rectangula
ar at the fo
ollowing c oordinates
s. A clockw
wise or annti-clockwis
se scheme
e
can be cho
osen (see Figure 9 a
and 10)
Table 5: Location
L of the surfa ce points at
a z=0m
Poiint 1 Poiint 2 Poin
nt 3 Poin
nt 4
x y x y x y x y
-25 -5 -25 5 25 5 25 -5
Figure 9
9: Construcct surface sttarting from
m point 1 to 3 ( and finallly 4) in clocckwise direc
ction
586 of 607
Figure 10: Constru
uction of surrface at z = 0 m (comp
pleted). The position off the points can
c still be
fine tun
ne using the
e table inputt seen here at the top right
r corner..
After inserrting the surface rep
presenting the cross section off the excavvation in plan
p we willl
extrude this surface to form the
e size (dep
pth) of the retaining structure
s (D
Diaphragm wall).
Figure
re 11: Extrussion along z-axis
z using
g a vector le
ength of 12 (vector z = -1)
587 of 607
5. Create a volume
v by filling an e
extrusion le
ength of 12
2m. To extrrude down
nwards that
is beneath
h the groun
nd surface the extrusion vector is z = -12 (see Figurre 11).
6. Click Applly to confirm
m the extru
usion process.
C2.Cre
eate the re
etaining sttructure w
with interfa
aces
1. Ensure tha
at the view
w is set as P
Perspectiv
ve
2. Select Vollume 1 by clicking th
he volume
e in the dra
awing areaa. Once se
elected the
e
volume is marked in red and in
n the Selec
ction explorer Volumee 1 is indic
cated.
3. Using the ‘Right Mo
ouse Butto
on (RMB)’’ gesture turn the vvolume into surfacess
using ‘Deccompose in
nto surface
es’ tool (se
ee Figure 12).
4. Six surface
es will be formed
f aro
ound the original volu
ume (in thiss case Volume 1) but
we will ne
eed on the
e 4 vertica
al surfaces
s to form the
t diaphrragm wall by turning
g
these 4 surfaces
s in
nto plates. One may
y choose to delete the top and
a bottom
m
surfaces to
o get a cle
ean model or mainta
ain these knowing thaat we can modify the
e
elevation of these 2 horizonttal surface
es into the
e excavatioon levels. The latter
option is straight
s forw
ward for exxperienced
d user.
5. In this exe
ercise we will
w use the
e first apprroach that is to deletee the top and
a bottom
m
surface. Select
S the to
op and botttom surfac
ce and delete them.
588 of 607
6. Select the
e 4 vertical surfaces located along the perimeter
p oof the volume. To do
o
this, click each of the
t vertica
al surfaces
s (while prressing thee ‘ctrl’ buttton on the
e
keyboard).
7. Perform the RMB gesture a
and select ‘Create plate’.
p Youu should obtain the
e
following schematic
c as show
wn in Fig
gure 13 showing
s thhe selecte
ed vertica
al
surfaces.
589 of 607
Figure 14: The plate elements
e rep
presenting the
t dimensiions of the D
D-wall.
590 of 607
Figure 15a Figure
e 15b Figure 15c
Figure
re 15: Locall axis on a plate
p (z axiss which is in
ndicated in blue, Fig.1 5a). A positive value iss
the ooutward direection. A neegative valuue is the innward direction. By seelecting ‘cre
eate positive e
interfa
face’ interface elementts will be crreated on thhe positive side (Fig.155b). Similarrly to create
e
an inwward interfa
ace the ‘crea
ate negative e interface’ option is ch
hosen (see Fig.15.c)
Figure
re 16: Com mplete interfface elemen nts along the
t perimetter of the pplate using the ‘create
e
positiive interfacces’ and ‘c create neg gative interffaces’ apprroach. Thee diagram shows the e
directtion of ‘local blue axis’ along the p
plate.
eate the ex
C3.Cre xcavation levels
To constru
uct the stag
ged excava
ation proce
ess 3 horiz
zontal surfa
aces locateed at (z = -3m),
- (z = -
6m) and (zz = -8.5m)) will be co
onstructed.. The first surface is drawn butt the subse
equent two
o
surfaces a t ‘Array’ tool.
are formed by using the
1. Construct a surface
e using the
e same po
oints as in
ndicated inn Table 5 but with z
et to -3 m.
position se
2. Change to
o ‘Top view
w’ and the ‘Movemen
nt limitation tool’ apppear again. Set the z
value to -3
3.0 m.
591 of 607
ate surface
3. Use ‘Crea e’ tool loca
ated in the
e vertical toolbar annd draw a horizonta
al
surface ussing the co
oordinates given in Ta
able 5. This is the firsst excavation level to
o
3 m below
w ground le
evel (see F igure 17)
4. To constru
uct the exc
cavation sta
age 2 the position off the seconnd surface will be at z
= -6m. The
e firs surfa
ace will be ccopied and
d place at z = -6m. T
The verticall difference
e
between th
he first and
d second ssurface is 3m.
3
5. Select the first surfac
ce (after se
election it will
w be marrked red).
6. Select the
e ‘Array tool’ from th
he vertical toolbar. The
T ‘Creatte array window’
w willl
appear as shown in Figure 18. By default the array
y pattern is ‘rectangullar’.
7. In the ‘Co
onfiguration’ section
n choose ‘1D,
‘ in z direction to copy th
he chosen
n
surface an
nd place th
he second horizontal surface at
a z = -6m
m. To perfo
orm this we
e
set the ‘Diistance between colu
umns’ as -3.
- The ‘ne
egative’ vallue is need
ded to shifft
the copied
d surface in the dirrection opp
posite to the
t global z-directio
on (upward
d
positive). A value off 3 is the d
difference in distance
e between the first and
a second
d
surface
8. The create
ed second surface re
epresenting
g the second excavaation level is shown in
n
Figure 19.
592 of 607
Figure 18: The ‘Create
e array’ tooll.
Figure 19
9: Horizonta
al surface 1 and 2 for sstaged exca
avation
593 of 607
Figure 20: The excava
ation levels represented by 3 horiz
zontal surfaaces located
d at
(z1= -3m, z2
z = -6m annd z3 = 8.5mm)
eate struts
C4.Cre s and wale
ers for sup
pport systtem at level 1(z = -1 .5 m)
The suppo m for the firrst excavattion stage is made of
ort system o a line oof ‘waling’ transecting
t g
along the perimeter of the wall at z1 = -1 .5m and 5 numbers of struts loocated at a centre-to-
centre spa
acing of 10
0m along x--direction.
1. The
e walers will be mode
elled using
g ‘beam’ ele
ements.
2. . Tw
wo approacches can be
b used to
o construct the walers
s using thee ‘beam’ ellements. In
n
the first, the beams
b can
n be drawn
n using the ‘beam’ too
ol found inn the vertic
cal tool barr.
The
e coordinattes in x-y plane
p is ag
gain the same in Table 5 but thee elevation
n (z) will be
e
-1.5
5m below ground.
g Th
his approa ch is named ‘Cad in
nput’ approoach. We will
w use the
e
seccond appro
oach whe
ere the ‘b
beam’ elem
ments rep
presenting the wale
ers will be
e
con
nstructed using the ‘C
Command line’ appro
oach.
3. The
e ‘Comman
nd line’ box n in Figure 21 and the
x is shown e coordinaates for the
e beam line
e
is sshown. In
n this exa
ample bra
ackets are
e used to
o indicate the five blocks of
coo
ordinates. The
T bracke
ets are nott necessary
y but used here for cclarity reaso
ons.
Figu
ure 21: A be
eam line is drawn using
g 5 points. The last point is similarr to the firstt point
4. The
e constructed waling
g system iss shown in
n Figure 22.
2 The maaterial for the waling
g
can
n be inserrted using the same
e ‘drag-and
d-drop’ ap
pproach orr by desig
gnating the
e
matterial set th
hrough the
e ‘Model exxplorer’ box
x (see Figu
ure 22).
594 of 607
Figu
ure 22: Inpu
ut of materia
al set ‘Walin
ng’ to ‘Beam
ms’ represen
nting first leevel of walin
ng.
Figure
e 23: First le
evel waling system
s with
h material set placed in
n.
5. The
e 5 nos. of strut will be
b construccted. The first
f strut will
w be placeed using tw
wo points.
6. The
e centre-to
o-centre sp
pacing is 1
10m and elevation
e of
o the struut is z= -1.5m below
w
ground level. The first point is lo
ocated at (x1,
( y1, z1) = (-20, --5, -1.5) and
a second
d
poin
nt located at (x2, y2, z3) = (-20 , 5, -1.5).
7. The
ere are two
o possible options. In
n the first option
o the strut can bbe constru
ucted using
g
the Cad inputt approach using the coordinate
es given above. Thiss is done by choosing
g
595 of 607
the ‘beam’ ele
ement option from th
he vertical toolbar and drawing the two points using
g
the coordinate
e set.
8. The second option
o gain using the ‘Comm
is ag mand line’ approach..
9. Forr the secon
nd approac
ch type “be
eam -20 -5 -1.5 -20 5 -1.5”.
10. Pla
aced the material
m set for the strut eith
her using the
t ‘Drag--and-drop’ or ‘Mode
el
exp
plorer inputt’ approach
h.
11. The
e complete
ed model at
a this with one single
e strut with
h the walingg system is shown in
n
Figu
ure 24
Figu
ure 24: Loca
ation of firstt level ‘walin
ng’ with one
e strut. Firstt excavationn level and interface
i are
e
swittched off.
12. The
e remaining 4 nos. of
o struts a
are replicatted using the ‘Array tool’. The
e important
parrameters when
w using
g this too l are settiing the sh
hape = 1D
D, x directtion, no of
o
umns = 5 and
colu a the dis
stance bettween the columns = 10 (note ‘positive’). The strutss
will be replica
ated along x direction
n. The outtcome of this processs is shown
n in Figure
e
25.
596 of 607
Figu
ure 25: The
e complete support
s sys tem for leve
el 1 (z=-1.5m
m)
C5.Cre
eat struts and walerrs for supp
port syste
em at level 2 (z = -4. 5 m)
597 of 607
ments loccated here
elem ein are grrouped tog
gether. We
W can renname this
s group to
o
‘Gro
oup_1_Sup
pport_Leve
el_1’ unde
er the ‘Mod
del explorerr’ box (seee Figure 28
8)
6. The
e second level su
upport sysstem is also
a ‘Grou
uped’ andd then re
ename ass
‘Gro
oup_2_Sup
pport_Leve
el_2’ unde
er ‘Model ex
xplorer’ bo
ox (see Figgure 28)
598 of 607
C. MO
ODE: MESH
Figure 29: M
Mesh option
n using the default ‘Ele
ement Distri
ribution’ optiion of ‘Veryy Course’ mesh
Mesh
h> Mesh Quality
Q
599 of 607
Figure 31:: Mesh quallity check
Click the ‘U
Update’ bu
utton locate
ed at the to
op of the window
w and
d return to ‘Mode: Me
esh’.
4. We
e will now proceed
p to ‘Mode: Wa
ater Levels
s” to design the channge of the water-table
w e
with
hin the exccavated zone.
D. MO
ODE: WATER
W LEVELS
L S
1. In ‘Mode: Wa
ater Levels
s’ the conffiguration of
o the initial groundw
water cond
dition is ass
succh that the ‘General Phreatic
P Su
urface’ is set
s right at ground levvel as seen
n in Fig.32.
Thiss is the inittial groundwater leve
el that was set at ‘Borrehole Inpuut via Head
d’.
Figure 32: M
Mode: Wate
er Levels. Visual
V in deffault Fig
gure 33: Distribution of hydrostatic
c porewater
settings pressuure
600 of 607
2. Clicck ‘Preview
w’ button lo
ocated at th
he ‘Vertica
al Toolbar’.. The initiaal porewate
er pressure
e
disttribution fo
or this sta
age (Initia l phase) is shown in Fig 333. This is the input
porrewater pre
essure thatt we want tto adopt fo
or this phas
se.
In summa
ary the su
urrounding soils and
d soil bene
eath zone 2 is refe rred to the ‘Genera
al
Phreatic S
Surface’. Zone 1 is set
s to ‘Dry’ and Zone
e 2 is set to
t ‘Interpollate’ betwe
een zone 1
eral Phreattic Surface’.
and ‘Gene
601 of 607
Phase 2: S
Support System
S 1 (Mode
( Wa ter Levels
s)
Phase 3: E
Excavatio
on Stage 2 (Mode W
Water Levels)
1. Thiss phase re
epresent ex
xcavation o
of zone 2 to
t 6m below
w ground llevel.
2. The
e ground water
w is low
wered by selecting zone
z 2 and setting tthe condition to ‘Dryy.
The
e procedurre is similarr to the me
ethod state
ed in Phase
e 1.
3. The
e porewate
er pressure
e within zon
ne3 that is beneath zone
z 2 is s et to ‘Interpolate’
Phase 4: S
Support System
S 2 (Mode
( Wa ter Levels
s)
1. Add
d phase 4 and name
ed this sta
age as ‘Support System 2’ Thiss phase will
w be used
d
late
er.
2. The
e porewate
er pressure
e is kept the same as
s in Phase 3.
Phase 5: E
Excavatio
on to Finall Level (Mo
odel Wate
er Levels)
602 of 607
Figure 35b:: Changing the water condition
c to ‘Dry’ for ex
xcvation zon
ne 1
Figure 36
6b: Select ‘In
nterpolate’ option
o for zzone 2 and a preview of the input h ydrostatic porewater
cond
dition for Phase 1
603 of 607
E. MODE: STAGED CONSTRUCTION
Initial Phase
By default ‘Initial Phase’ is already included. All the structural elements are switched off.
In this stage:
In this stage:
In this stage:
604 of 607
Proceed to
o ‘Calculatte’. Prior to
o calculatio
on we nee
ed to selec
ct two poinnts. These two pointss
are used to plot the ‘Displacem
‘ ment vs. tim
me’ curves.
Table 6: Co
oordinates
s for ‘A’ and
d ‘B’
Point x y z
A 0 5 0
B -24.5
5 5 0
Figure 37: S
Selection of point ‘A’ and ‘B’ for pllotting
605 of 607
F. OU
UTPUT OF
O RES
SULTS
Figure 38: P
Plot: Displa
acement usiing iso-surfa
aces
Figure 39: A
Axial forcess for struts at
a final exca
avation leve
el.
Figure 40: D
Deformation
n of retainin
ng structure
e
606 of 607
Figure 41: Wall displaccements at point A and
d B with tim
me.
607 of 607