calibration of modified cam clay model with use … · calibration of modified cam clay model with...

CALIBRATION OF MODIFIED CAM CLAY MODEL WITH USE OF LOADING PATH METHOD AND GENETIC ALGORITHMS

Magdalena KOWALSKA

Department of GeotechnicsFaculty of Civil Engineering

Silesian University of TechnologyGliwice, POLAND

2

PRESENTATION PLAN

1. Calibration

2. Loading Path Method

3. Case Study

4. Conclusions

3

1. CALIBRATION

CALIBRATION= evaluation of appropriate values for

parameters so that the simulatedstress-strain relationship of soil ispossibly consistent with thebehaviour observed in field orlaboratory test.

4

1. CALIBRATION

7E*, m, f, g, ν, c, φ2005

Coulomb-Mohr/Fahey-Carter

nonlin. elastic / perfectly plastic

5M, G, l, k, N1965ModifiedCam-Clay

elasto-plastic+ isotr. harden.

4E, ν, c, φXVIIICoulomb-Mohr

elasto-perfectly-plastic

2E, νXVIIHooke's lawelastic isotropic

No.parametersyearnametype

5

2. LOADING PATH METHOD

1. Select representative point2. Evaluation of in situ conditions3. Theoretical loading path (FEM)4. Undisturbed soil sample5. TX test: loading path →

experimental response path6. Theoretical response path7. Optimization → set of parameters

6

peat, dark brown

clay, gray

silt, grayfine/medium sand, gray

clayey aggradate mud, gray light brown

2.20

3.40

4.40

5.10

3. CASE STUDY3.1. Subsoil investigation

MAN trucks factory in Niepołomice

7M = 1.0, λ = 0.15, κ = 0.02, ν = 0.3, e0 = 0.9, γ = 18 kN/m3

600 kPaZ_SOIL.PC v.6.27Modified Cam Clay

siltclaysiltcl. sand

f. sand

5.0 m 5.0

m

164

3. CASE STUDY3.2. Model

8

3. CASE STUDY3.3. Stress path

0

40

80

120

160

0 40 80 120 160 200 240 280p' [kPa]

q [kPa]

point 164 CSL K0 line YS

9

3. CASE STUDY3.4. Triaxial test

Bishop & Wesley’s apparatusPolish Academy of Sciences

10

3. CASE STUDY3.4. Triaxial test

0

100

200

300

400

500

600

0 100 200 300 400 p' [kPa]

q' [kPa]

controlled pathsimple shearingCSL

11

3. CASE STUDY3.4. Triaxial test

controlled pathsimple shearing

-5

0

5

10

15

-5 0 5 10 15 20 25

εs [%]

εv [%]

12

3. CASE STUDY3.5. Theoretical response

Modified Cam Clay

MATLAB 6.1

Genetic Algorithms

13

3. CASE STUDY3.6. Genetic Algorithm

fitness function

minn

FF

n

i

2

ev

ei,v

ti,v

2

es

ei,s

ti,s

→

∑⎥⎥

⎦

⎤

⎢⎢

⎣

⎡

⎟⎟⎠

⎞⎜⎜⎝

⎛

εΔ

ε−ε+⎟

⎟⎠

⎞⎜⎜⎝

⎛

εΔ

ε−ε

=

where:Δεx = max(εx

e)-min(εxe)

xt, xe – theoretical & experimental valuesn – number of points

14

3. CASE STUDY3.6. Genetic Algorithm

0

100

200

300

400

500

600

0 100 200 300 400 p' [kPa]

q' [kPa]

segment I

segment II

segment III

15

3. CASE STUDY3.6. Genetic Algorithm

9 different cases were analysed:A. strain path εs-εv

B. compressibilty characteristic p’- εv

C. shearing characteristic q’-εs

a. controlled stress path with shearing(segm. I+II+III)

b. controlled stress path (segm. I+II)c. controlled stress path inside the YS

(segm, I)Named: Aa, Ba, Ca, Ab,…, Cc

16

(*) – the concrete value independent on the FF

( ) – with a trend in FF minimizing

3. CASE STUDY3.7. Estimation of parameters

2.4800.0100.0790.2870.0310.0340.2560.0670.072FF

0.4930.2450.2940.1400.4020.1040.188*0.0330.486ν

0.005*0.0060.0020.007*0.0060.0300.005*0.0540.001κ

0.7170.0690.0671.1520.0120.0990.0280.0590.067λ

0.9810.9581.410*19.4370.6671.414*1.0900.6401.403*M

CcCbCaBcBbBaAcAbAa

Compressib. charact. p-εv

Shearing charact. q-εs

Strain pathεv-εs

a – controlled path + simple shearingb – controlled pathc – controlled path inside YS

17

3. CASE STUDY3.7. Estimation of parameters

2.4800.0100.0790.2870.0310.0340.2560.0670.072FF

0.4930.2450.2940.1400.4020.1040.188*0.0330.486ν

0.005*0.0060.0020.007*0.0060.0300.005*0.0540.001κ

0.7170.0690.0671.1520.0120.0990.0280.0590.067λ

0.9810.9581.410*19.4370.6671.414*1.0900.6401.403*M

CcCbCaBcBbBaAcAbAa

Compressib. charact. p-εv

Shearing charact. q-εs

Strain pathεv-εs

a – controlled path + simple shearingb – controlled pathc – controlled path inside YS

(*) – the concrete value independent on the FF

( ) – with a trend in FF minimizing

0.30

0.02

0.15

1.00

FEM

18

3. CASE STUDY3.7. Estimation of parameters

2.4800.0100.0790.2870.0310.0340.2560.0670.072FF

0.4930.2450.2940.1400.4020.1040.188*0.0330.486ν

0.005*0.0060.0020.007*0.0060.0300.005*0.0540.001κ

0.7170.0690.0671.1520.0120.0990.0280.0590.067λ

0.9810.9581.410*19.4370.6671.414*1.0900.6401.403*M

CcCbCaBcBbBaAcAbAa

Compressib. charact. p-εv

Shearing charact. q-εs

Strain pathεv-εs

a – controlled path + simple shearingb – controlled pathc – controlled path inside YS

(*) – the concrete value independent on the FF

( ) – with a trend in FF minimizing

19experimentaltheoretical

3. CASE STUDY3.7. Estimation of parameters

A. Strain path εv – εs

-5

0

5

10

15

-5 5 15 25 35εs [%]

εv [%]

-2

0

2

4

6

8

-1 0 1 2 3εs [%]

εv [%]

-0,1

-0,05

0

0,05

-0,06 -0,04 -0,02 0 0,02εs [%]

εv [%]

Ab

Aa

Ac

20

3. CASE STUDY3.7. Estimation of parameters

B. Shearing charact.q’–εs

Bb

Ba

Bc0

50

100

150

200

-1 0 1 2 3εs [%]

q' [kPa]

15

20

25

30

35

-0,06 -0,04 -0,02 0 0,02

εs [%]

q' [kPa]

0

200

400

600

-5 0 5 10 15 20 25εs [%]

q' [kPa]

experimentaltheoretical

21

3. CASE STUDY3.7. Estimation of parameters

C. Compressibilty char.p’–εv

Cb

Ca

Cc

-4

0

4

8

12

0 100 200 300 400p' [kPa]

εv [%]

-2

0

2

4

6

8

0 100 200 300

p' [kPa]

εv [%]

-0,1

-0,05

0

0,05

30 35 40 45 50

p' [kPa]

εv [%]

experimentaltheoretical

22

4. CONCLUSIONSIN THE ANALYSED CASE:

The parameters in the optimum set aredifferent than the results of simplelaboratory tests.

It is also different than the one assumedin FEM analysis.

23

4. CONCLUSIONS

GENERAL:

Shape of the loading path imposed on a sample and the estimation method havegreat influence on values of parameters.

Modified Cam Clay model is effective for normally consolidated soils. The quality ofmatching is much better in the plasticrange of stress than in the YS interior.

24

THANK YOU FOR YOUR ATTENTION!