soil-structure-interaction simulation models

SOIL-STRUCTURE-INTERACTION SIMULATION MODELS Juan M. Pestana University of California, Berkeley

PEER 2001 Annual Meeting Oakland, California. January 26, 2001 RESEARCH SUPPORTED BY THE PACIFIC EARTHQUAKE ENGINEERING RESEARCH CENTER IN COLLABORATION WITH THE CALIFORNIA DEPARTMENT OF TRANSPORTATION AND THE NATIONAL SCIENCE FOUNDATION

SEISMIC SSI- CONCLUSIONS • Key components are: (foundation) soil, foundation system & structure. • Key issue of Performance Based Design is the evaluation (estimation) of performance -> Requirement of robust, efficient & realistic numerical- analytical description of material response & system. • Development & refinement of laboratory experiments which can accurately represent the problem are essential to validate “individual” constructs (numerical/conceptual blocks). It therefore provides an invaluable resource for calibrating numerical Codes-> Reduce inherent modeling uncertainty. • Successful development of coupled site response and Soil-PileSuperstructure Interaction analyses provides a Comprehensive framework for Performance Based Design for increasingly higher shaking levels

Structural Engineer ’s Engineer’s View of the World

Geotechnical Engineer ’s Engineer’s View of the World

Courtesy of Dr. Lehman- University of Washington

SOIL-STRUCTURE INTERACTION • Types of structures (buildings, bridges, bunkers) • Transfer Function: Empirically or analytical description of the behavior of a structure given response at ground level. • Estimation of performance – Shallow foundations- A lot of work still needed (i.e., mat foundation). A lot of interest - effect of structure embedment. – Deep Foundations- A lot of interest- Highway/ retrofit of Bridges and other structures in soft soilsSSI is extremely important.

SSI- DEEP FOUNDATIONS Soil-Pile-Superstructure Interaction Coupled Model

Modal Mass

Overall Response

Earthquake Motion input time history

Seismic Soil- Pile GroupStructure Interaction Test

Large Scale Shaking Table Tests

Modal Mass

Conceptual Soil - Pile Model Soil Model

Interface Near field Far field

Interface element: gapping = f(plastic soil deformation)

Free field gap

Pile

Near field element: partitioned dynamic nonlinear p-y/t-z spring = f(strain rate, #cycles, strain reversals) Far field element: radiation damping = f(soil nonlinearity in near and free field)

input time history

Free field element: time domain site response inputs motions at nodes

k1

k2

k3

c1

c2

c3

NONLINEAR SITE RESPONSE ANALYSES- MODEL Modulus Degradation for Clay ( Vucetic & Dobry) 1 G/Gmax

model

0.5 (Vucetic & Dobry)

0 -4 10 30

-2

10 Damping for Clay (Vucetic & Dobry)

Damping Ratio (%)

• KEY ELEMENTS: • ABILITY TO MODEL SMALL STRAIN NON-LINEARITY AND DAMPING CHARACTERISTICS • FLEXIBILITY TO REALISTICALLY DESCRIBE RESPONSE FROM SMALL TO LARGE STRAINS • CORRESPONDENCE WITH MEASURED SOIL BEHAVIOR - ESTABLISHED PARAMETER DETERMINATION • CHARACTERIZATION OF UNCERTAINTY

10

0

model

(Vucetic & Dobry)

20 10 0 -4 10

-2

10 Shear Strain (%)

10

0

TEST SERIES 1.1 CROSSSECTION

SPECTRAL ACCELERATION input time history

5% Damped Spectral Acceleration (g

Modal Mass

5% Damped Spectral Acceleration (g)

NONLINEAR SITE RESPONSE ANALYSIS CALIBRATED with VERTICAL ARRAY INFORMATION

2

Elev 0

0 2

Elev -8

0 2

Elev -18

0 2

Elev -30

0 2

Elev -48

0

0

0.01

0.1

1

Period (sec)

10

Nonlinear element Gap element

Pile

Far field

Friction element

gap element

nonlinear element

1.0

+

y gap

=

-1.0 -1.0

y/yc

p/pult

pgap /pult

p-y element

p/pult

1.0

1.0

-1.0 -1.0

1.0

y/yc

+ y/yc

1.0

friction element

1.0

pfriction /pult

-1.0 -1.0

-1.0 -1.0

y/yc

1.0

1.0

Formulation of nonlinear 1-D element

NUMERICAL CAPABILITIES -NONLINEAR py, t-z springs with Gapping Capabilities 1

p-y curve for clay p-y w/ gap (loop 1)

p/pu

p-y w/ gap (loop 2)

0.5

0 -8

-6

-4

-2

0

2

4

gap -0.5

-1

y/yc

6

Pile

8

8

NEAR FIELD RESPONSE

depth = 5d

depth = 6d

depth = 7d

depth = 8d

depth = 9d

0

Soil Resistance P (lb/in)

Modal Mass

depth = 4d

-8 8

0

-8 8

0

Calculated Test Data

input time history

-8 -0.5

0

Deflection Y (in)

0.5

-0.5

0

Deflection Y (in)

0.5

D

FOUNDATION RESPONSE- BENDING MOMENTS

S2

S3 ft)

input time history

S4

0 1 2 3 4 5 6

Depth (ft)

0 1 2 3 4 5 6

S3

0 1 2 3 4 5 6

Depth (ft)

Depth (ft)

S1

Depth (ft)

Modal Mass

0 1 2 3 4 5 6

0

0.5 Bending Moment (k-in)

1

PSA (g)

PREDICTION OF STRUCTURAL RESPONSE 5

S1

4

PSA (g)

PSA (g)

4 3

3

2

2

1

1

0 5

0 S2

5

3

4

PSA (g)

4 2

PSA(g)

1 0

S4

PSA (g)

Modal Mass

S3

5

3 2 1

PSA (g)

PSA (g)

0

input time history

0.01

0.1 1 Period (sec)

10

Computed Max. Sa (g)

5

STRUCTURAL RESPONSE

4 3

1 0

1.6

0.8

S1

S2

S4

0.4 0 0

Computed Tp (sec)

Computed Max. Sa (g)

(a)

0.4 0.8 1.2 1.6 Recorded amax (g)

2

Computed Tp (sec)

Computed amax (g)

2

S3

S3 S4

2

0

(b)

1.2

S2 S1

1 2 3 4 Recorded Max. Sa (g)

5

0.3 S4

0.2

S3

S1 S2

0.1

5

0 (c)

0

0.1 0.2 Recorded Tp (sec)

0.3

Spectral Accelerat ion (g)

Performance of Pile Groups Equivalent Pier Approach

Response of Structure

3.0 Damping=5%

Recorded e=1.0 e=0.8

2.0

e=0.6 e=0.4 e=0.2

1.0 1

0.0

Response of Pile Cap Damping=5%

0.6

p-y efficiency

3.0

Spectral Accelerat ion (g)

2x2 pile gr 3x3 pile gr

0.8

Recorded

0.4

e=1.0 e=0.8

2.0

e=0.6 e=0.4

0.2

e=0.2

1.0 0 0

0.2

0.4

0.6 Input MHA

0.0 0.01

0.1

1

Period (sec.)

10

0.8

1

Multidirectional Shaking 2-D & 3-D Site Response Analyses r u 0.08 S

Pacoima Dam record a = 0.05g

l 0.06 e l l 0.04 a r a 0.02 P

max

50 m Soil Profile

t 0.00 l u a-0.02 F -0.04

-0.06 -0.06

-0.04

-0.02

0.00

0.02

0.04

Fault Normal Surface Motion (m)

0.06

0.08

) rL 10.0 o. F d .u d 5.0 e z i sl 0.0 (a /m P r -5.0 o ,N e -10.0 c) rL 10.0 o. F d

Test B1D3 L/D= 2 w (Hz) 3.0 Symbol A/d 5% 10 % 20 % 50 % API

-0.4

-0.2

0

0.2

0.4

Normalized Deflection, y/d Test B3D4 L/D= 8 w (Hz) 3.0

.u d 5.0 e z i l 0.0 a m r -5.0 o N -10.0

Symbol A/d 5% 10 % 20 % 50 % API

-0.4

-0.2

0

0.2

Normalized Deflection, y/d

0.4

MULTIDIRECTIONAL P-Y ELEMENTS

Numerical Tool * Nonlinear Site

OpenSees Platform

Response (2-D) * Near Field, p-y New Elements

Near Field, t-z and Q-z Capabilities

CONCLUSIONS- DEJA VU • Key components are: (foundation) soil, foundation system & structure. • Key issue of Performance Based Design is the evaluation (estimation) of performance -> Requirement of robust, efficient & realistic numerical- analytical description of material response & system. • Development & refinement of laboratory experiments which can accurately represent the problem are essential to validate “individual” constructs (numerical/conceptual blocks). It therefore provides an invaluable resource for calibrating numerical Codes-> Reduce inherent modeling uncertainty. • Successful development of coupled site response and Soil-PileSuperstructure Interaction analyses provides a Comprehensive framework for Performance Based Design for increasingly higher shaking levels