Professional Documents
Culture Documents
25.08.2009
Introduction
Hardening Soil (HS) and Hardening Soil-small (HS-small) models are designed to reproduce basic phenomena exhibited by soils:
densication stiness stress dependency plastic yielding dilatancy strong stiness variation with growing shear strain amplitude in the regime of small strains ( = 106 to = 103 ) this phenomenon plays a crucial role for modeling deep excavations and soil-structure interaction problems
Introduction
HS model was initially formulated by Schanz, Vermeer and Bonnier (1998, 1999) and then enhanced by Benz (2006) Current implementation is slightly modied with respect to the theory given by Benz:
simplied treatment of dilatancy for the small strain version (HS-small) modied hardening law for preconsolidation pressure modied form of the cap yield surface (2009)
This model seems to be one of the simplest in the class of models designed to handle small strain stiness It consists of the two plastic mechanisms, shear and volumetric Small strain stiness is incorporated by means of nonlinear elasticity which includes hysteretic eects
Eo
200 150 100 50 0 0 0.05 0.1
qun
q [kpa]
1 1 q50
0.25
Remark: All classical soil models require specication of Eur modulus (Cam-Clay, Cap etc..)
Andrzej Truty ZACE Services Hardening Soil model with small strain stiness
Atkinson 1991
Andrzej Truty ZACE Services Hardening Soil model with small strain stiness
0.7
07
Results for PI < 30 are conrmed by other researchers while these for PI > 30 should be used with a special care (Benz) Stokoe proposed linear interpolation for 0.7 0.7 = 104 for PI = 0 to 0.7 = 6 104 for PI = 100
Andrzej Truty ZACE Services Hardening Soil model with small strain stiness
Ed Es
10
co h es i ve so i ls
Ro ck
gran
u l ar
soil
1 1000
Es [kPa]
10000 100000 1000000
Go = A f (e)OCR Soil Clean sands Undisturbed clayey soils and crushed sand Undisturbed cohesive soils Loess emin 0.5 0.6
Hardin and Black (1978) A [kPa] 57 33 f (e) m 2 (2.17 e) 0.4 1+e 2 (2.97 e) 0.5 1+e Ref. Iwasaki Hardin and Black Kim Kim
0.6 1.4
1.5 4.0
16 1.4
Poisson ratio varies in the range = 0.1..0.3 in small strain domain Its value in further derivations will be kept constant (by default = 0.25)
Cap surface
qf 200
q [kPa] 150 100 50
M-C limit
E50 1
qf 1
0.01 0.02
Eur
0 0
0.03 eps-1
0.04
0.05
=0.1=const.
=0.01=const.
=0.001=const. =0.0001=const.
300 p [kPa] 400 500 600
sin psi_m
Domain of contractancy
10 20
Domain of dilatancy
30
phi_m [deg]
40
50
60
d PS = d1
g1 g1 g1 1 2 3
= d1
Cap mechanism
q2 2 + p 2 pc M 2 r 2 () r () is dened via van Ekelens formula (like in Cam-Clay model q2 Plastic potential: g2 = 2 + p 2 M pc + c cot m Hardening law: d pc = d2 2H p ref + c cot Yield condition: f2 = Remarks:
1
NC M and H parameters can be estimated for assumed Ko and tangent Eoed modulus set up at a given vertical stress
NB. Here same plastic ow rule is used as for the shear mechanism f1 Rankine yield condition (tensile cut-o) f3 = 1 ft = 0 where: ft is the assumed tensile strength (default is ft = 0) Rankine plastic ow rule(associated ow rule is used) g3 = f3
Andrzej Truty ZACE Services Hardening Soil model with small strain stiness
3 + c cot
Remarks
1
Strain dependency is added to the stress-strain relation, for stress paths penetrating the elastic domain The modied Hardin-Drnevich relationship is used to relate current secant shear modulus G and equivalent monotonic shear strain hist Reversal points are detected with aid of deviatoric strain history second order tensor Hij ; in addition the current equivalent shear strain hist is computed by using this tensor Hardening laws for PS and pc are modied by introducing hi factor; this factor for very small strains is much larger than 1.0 and decreases to 1.0 once the shear strain hist exceeds certains strain amplitude c Certain constractancy is allowed in the plastic ow rule for shear mechanism
Andrzej Truty ZACE Services Hardening Soil model with small strain stiness
plot from paper by Ishihara 1986 At step N : histN1 = 8 105 histN = 104 At step N + 1 : histN = 0 histN+1 = 2 105 max Primary loading: histN+1 > hist max Unloading/reloading: histN+1 hist Go Hardin-Drnevich law: G = hist (secant modulus) 1+a 0.7
Andrzej Truty ZACE Services Hardening Soil model with small strain stiness
Gur
c
c = 0.7 a Go 1 Gur
Modications: Dilatancy
PSI_m [deg]
PSI_m [deg]
5 0 -5 0 -10 -15
Rowes dilatancy
PHI_m [deg]
Cap surface
SR
600 500 400 q [kPa] 300 200 100 0 0 100 200 p [kPa] 300 400 500
Shear mechanism
Cap surface
SR
Procedure:
PS For given SR state compute o from plastic condition f1 = 0
Remarks
1
SR NC Ko = Ko 1 sin() in the standard applications (approximate Jakys formula) SR Ko = 1 for case of isotropic consolidation (used in triaxial testing for instance)
For sands notion of preconsolidation pressure is not as meaningful as for cohesive soils hence one may assume OCR=1 and eect of density will be embedded in H and M parameters
Eoed
1
Material properties
Parameter ref Eur ref E50 ref m ur Rf c emax ft D M H OCR/q POP ref Eo 0.7 Unit [kPa] [kPa] [kPa] [] [] [] [kPa] [o ] [o ] [] [kPa] [] [] [kPa] [/kPa] [kPa] [] HS-standard yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes no no HS-small yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes
User interface
Advanced
Advanced checkbox is
ref Eur is the unloading/reloading Young modulus given at the reference stress ref
ur is the unloading/reloading Poisson coecient; it varies from 0.15 to 0.3, hence for sands it is recommended to assume ur = 0.2..0.25 and for clays ur = 0.25..0.3 m is the exponent in stress dependency power law; it varies from m = 0.4 to m = 0..6; it is smaller for dense sands and larger for clays L is the minimum allowed reference stress value used for evaluation of stiness moduli
The HS-small model requires two additional parameters: ref Young modulus at very low strains Eo at the reference stress ref and threshold shear strain 0.7
ref In case of lack of information on Eo one may try to estimate ref based on Alpans diagram assuming E = E Eo s ur
4 5
In the current implementation 0.7 is assumed to be constant In case of lack of information on 0.7 the diagram by Vucetic and Dobry can be used for cohesive soils and diagram by Wichtmann and Triantafyllidis for cohesionless ones
Cap surface parameter M and hardening parameter H are derived by using a simple calculator which simulates an oedometric test; for given tangent oedometric modulus Eoed ref NC at a given reference vertical stress oed and for assumed Ko parameter (here Jakys formula can be used) values of H and M are evaluated (press button Evaluate M,H ); one may m ref oed + c cot ref assume Eoed = E50 as a default value ref + c cot PS Setting the initial state variables o and pco can be carried out by means of assumed OCR or preoverburden pressure q POP
NC To compute Ko from Jaky formula press button Use Jakys formula for KoNC
Remarks
1
SR SR Pairs Ko and OCR (OCR 1.0) or Ko and q POP are needed to setup the initial position of the cap surface and the initial value of the hardening parameter PS min pco is the minimum allowed value for the initial preconsolidation stress
q = 0.5 qult with qult being the approximate ultimate limit load density The template data les for MC and HS model can be found in the CFG directory under names: template-foot-MC and template-foot-HS
10m
10m
insitu , , , , m, , , c , OCR, K SR , Given: dry , Ko ur ref L o ref ref E50 Eur = .... and ref = .... and Young modulus that user ref E50 Eoed would assume in the simulation with a standard MC model ref Find: Eur
We know parameters to be used in the simulation with aid of insitu , , , , c a standard MC model: E , dry , Ko ur Now we want to use HS/HS-small model but we do not know ref on how to estimate Eur , H and M parameters We select a plane-strain problem of a strip loading q applied to a uniform layer of soil as a template problem We assume the additional parameters for HS model: L , m, E ref SR OCR, Ko and the two coecients ur = .... (default is 3) ref E50 ref E and 50 = .... (default is 1.0) ref Eoed
ref We run the optimization procedure which yields the Eur , M and H such that the settlement at point A obtained from MC and standard (!!!) HS model are the same
Andrzej Truty ZACE Services Hardening Soil model with small strain stiness
HS-std HS-small
0.001
0.01
0.1
(a) 1 (1 ) (Z Soil) 3
4 3.5 SIG-1 / SIG-3 [kPa] 3 2.5 2 1.5 1 0 0.002 0.004 0.006 0.008 0.01 -EPS-Y [-] HS-std HS-small
(b) G () (Z Soil)
0 0.08 0.07 0.06 0.05 -EPS-V [-] 0.04 0.03 0.02 0.01 0 -0.01 -0.02 -EPS-Y [-] HS-std HS-small 0.02 0.04 0.06 0.08 0.1
(d) v (1 ) (Z Soil)
200000 180000 160000 140000 G [kPa] 120000 100000 80000 60000 40000 20000
0 0.02 0.04 0.06 0.08 0.1
3.5 3
HS-std HS-small
0 0.00001 0.0001
0.001
0.01
0.1
(a) 1 (1 ) (Z Soil) 3
4 3.5 SIG-1 / SIG-3 [kPa] 3 2.5 2 1.5 1 0 0.002 0.004 0.006 0.008 0.01 EPS-1 [-] HS-std HS-small
(b) G () (Z Soil)
0 0.08 0.07 0.06 0.05 EPS-V [-] 0.04 0.03 0.02 0.01 0 -0.01 -0.02 EPS-1 [-] HS-std HS-small 0.02 0.04 0.06 0.08 0.1
(d) v (1 ) (Z Soil)
300000 250000 200000 G [kPa] 150000 100000 50000 0 0.00001 0.0001 HS-std HS-small
4 3.5 3 2.5 2 1.5 1 0 0.02 0.04 0.06 0.08 0.1 EPS-1 [-]
0.001
0.01
0.1
EPS-1-EPS-3 [-]
(a) 1 (1 ) (Z Soil) 3
4 3.5 SIG-1 / SIG-3 [kPa] 3
EPS-V [-]
(b) G () (Z Soil)
0 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 -0.01 HS-std HS-small 0.02 0.04 0.06 0.08 0.1
HS-std HS-small
(d) v (1 ) (Z Soil)
Given 3 drained triaxial test results for 3 conning pressures: 3 = 100 kPa 3 = 300 kPa 3 = 600 kPa
Shear characteristics q 1 Dilatancy characteristics v 1 Stress paths in p q plane Measurements of small strain stiness moduli Eo (3 ) for the assumed conning pressures (through direct measurement of shear wave velocity in the sample)
M* =
1
6 sin 3 sin
c* =
6c cos 3 sin
p
If we know M and c then we can compute and c: 3 M 3 sin = arcsin c = c 6+M 6 cos
Andrzej Truty ZACE Services Hardening Soil model with small strain stiness
1500
1800
Here: = arcsin
c=0
1
EPS-V [-]
0.03 0.02 0.01 0 -0.01 -0.02 0 0.02 0.04 0.06 0.08 0.1
EPS-1 = - EPS-3 [-]
Dilatancy cut-off
= arcsin
d 2+d
1 d=0.75
V 0.03
0.02 0.01 0 -0.01 -0.02 0 0.01 0.02 0.03 0.04
0.05
0.06
0.07
0.08
0.09
0.1
= arcsin
0.75 2 + 0.75
16o
Analytical formula: Eo =
ref + c cot Measured: shear wave velocity vs at 1 = 106 and at given conning stress 3 Compute : shear modulus Go = vs2 Compute : Young modulus Eo = 2 (1 + ) Go 3 [kPa] 100 300 600 Eo [kPa] 250000 460000 675000
ref Eo
3 + c cot
Analytical formula: Eo =
ref + c cot Measured: shear wave velocity vs at 1 = 106 and at given conning stress 3 Compute : shear modulus Go = vs2 Compute : Young modulus Eo = 2 (1 + ) Go 3 [kPa] 100 300 600 Eo [kPa] 250000 460000 675000
ref Eo
3 + c cot
Plot Eo vs 3
800000 700000 600000 500000 400000 300000 200000 100000 0 0 100 200 300 400 500 600 700
Eo [kPa]
[kPa]
Reanalyze Eo vs 3 in logarithmic scales 13.1 12.55 Averaged slope yields m; here m = = 0.55 1.0 Find intersection of the line with axis ln Eo at 3 + c cot ln =0 ref + c cot Here the intersection is at 12.43 hence ref Eo = e 12.43 2.71812.43 = 250000 kPa
13.6 13.4 13.2
ln Eo 13
12.8 12.6 12.4 12.2 0 0.2 0.4 0.6 0.8 1
m 1
12.43
To estimate small strain modulus Go at a certain depth one may use empirical formula by Mayne and Rix: q 0.695 Go = 49.4 t1.13 [MPa] e qt is a corrected tip resistance expressed in MPa e is the void ratio Note: this is very rough estimation Best solution: Perform triaxial testing and project on CPT prole to adjust empirical coecient (49.4) for a given site
2000
1500
f q50 ( 3 = 100)
1000
f q50 ( 3 = 100) 500 f q50 ( 3 = 100)
0 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
Reanalyze E50 vs 3 in logarithmic scales Here we can x m to the one obtained for small strain moduli Find intersection of the line with axis ln E50 at 3 + c cot ln =0 ref + c cot Here the intersection is at 10.30 hence ref E50 e 10.30 2.71810.30 30000 kPa
11.4 11.2 11
ln E50
10.8 10.6 10.4 10.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
10.30
The unloading reloading modulus as well as oedometric moduli are usually not accessible
ref We can use Alpans diagram to deduce Eur once we know ref E ref Eo (default is ur = 3); for cohesive soils like tertiary clays ref Eo this value is larger
For oedometric modulus at the reference stress ref = 100 ref ref kPa we can assume Eoed = E50 0.7 = 0.0001...0.0002 for sands and 0.7 = 0.00005...0.0001 for clays Smaller 0.7 values yield softer soil behavior
Conclusions
Model properly reproduces strong stiness variation with shear strain It can be used in simulations of soil-structure interaction problems Implementation is rather heavy It should properly predict deformations near the excavations Model reduces excessive heavings at the bottom of the excavation