Introduction

Hybrid Simulation Concept

Hybrid simulation is an experimental method, used to study the dynamic response of complex engineering systems. It involves the combination of two systems, a physical and a computational subsystem interacting with each other. The interaction between the two subsystems can be attained with the use of actuators and/or shake tables which apply the interface conditions between the physical and the virtual subsystems. It can be broadly divided into two categories: pseudo-dynamic and dynamic hybrid simulation. In the commonly used pseudo-dynamic form of hybrid simulation, the inertia effects of the physical subsystem are taken into account in the numerical subsystem, usually as lumped masses. On the other hand, in dynamic hybrid simulation, the inertia effects are part of the physical subsystem and hence, this method should be used when the physical subsystem subsystems of interest cannot be represented using a simplified approach. Clearly, dynamic hybrid simulation is the appropriate technique to study complex interaction phenomena, such as the soil-foundation-structure interaction described above. One of the most important benefits of the hybrid SFSI simulation approach, is the fact that the properties of the superstructure can be easily changed by simply altering the computer model. As a result different superstructures can be emulated and the response in each case can be evaluated using the same experimental setup.

This work is driven by need for sufficiently rich experimental data to evaluate numerical models of complete soil-foundation-structure systems. There are clear practical impediments to performing such system-level experiments at full-scale.

Role of hybrid simulation: While it may not be possible by any simple approach to test complete soil-foundation-structure systems at full scale, hybrid simulation offers potential for a very good approximation to such a test, by systematically coupling subsystems. It allows adaptability by allowing a physical soil-foundation system to be coupled with multiple complex virtual superstructure models. Hybrid simulation would add an extra dimension by integrating many of these approaches enabling a fully coupled simulation as shown in Figure 1.

In this experiment, the physical substructure of interest is the soil-foundation system, while the superstructure is represented by small shaker applying the interface conditions. The specific experiment is the first step towards hybrid simulation of soil-foundation-structure systems

A simplified model is used as a first step towards hybrid simulation of a full soil-structure-foundation system. Elastomeric bearings are used to represent the soil properties.

Laminar soil box

Shake table (mimics superstructure)

Earthquake

Virtual superstructure

Full system

Hybrid simulation Substructuring

Multiple superstructures represented by

computer model

Ph

ysical sub

system w

ith d

istribu

ted m

ass

Motivation • Sufficiently rich experimental data are needed to evaluate numerical models of

complete soil-foundation-structure systems • Practical impediments to performing such system-level experiments at full-scale

Real-Time Hybrid Simulation of Soil-Foundation-Structure Interaction

Aikaterini (Katerina) Stefanaki1 , M. V. Sivaselvan2 and Anthony Tessari2

Hybrid Simulation of Soil-Foundation-Structure Concept

• Hybrid simulation • is an experimental method, used to study the dynamic

response of complex engineering systems. It involves the combination of two systems, a physical and a computational subsystem interacting with each other. The interaction between the two subsystems can be attained with the use of actuators and/or shake tables which apply the interface conditions between the physical and the virtual subsystems.

In dynamic hybrid simulation, the inertia effects are part of the physical subsystem and hence, this method should be used when the physical subsystem subsystems of interest cannot be represented using a simplified approach. Clearly, dynamic hybrid simulation is the appropriate technique to study complex interaction phenomena, such as the soil-foundation-structure interaction.

Role of hybrid simulation: While it may not be possible by any simple approach to test complete soil-foundation-structure systems at full scale, hybrid simulation offers potential for a very good approximation to such a test, by systematically coupling subsystems. It allows adaptability by allowing a physical soil-foundation system to be coupled with multiple complex virtual superstructure models. Hybrid simulation would add an extra dimension by integrating many of these approaches enabling a fully coupled simulation as shown in Figure 1.

The physical subsystem is a pile foundation system built in saturated soil contained in a large-scale geotechnical laminar box. The shake table used here was designed and constructed for the purposes of dynamic hybrid simulation experiments. Figures 2a,b,c show the experimental setup in the Structural Engineering and Earthquake Simulation Laboratory (SEESL) at University at Buffalo.

The specific experiment is the first step towards hybrid simulation of soil-foundation-structure systems

Soil-Structure Interaction Experiment Setup

Shake Table

Laminar Box

Experimental Setup General View

Shake Table

• Physical substructure: Soil-foundation system in saturated soil in large-scale geotechnical laminar box • Superstructure represented by small shaker applying the interface conditions

The specific experiment is the first step towards hybrid simulation of soil-foundation-structure systems

The physical subsystem is a pile foundation system built in saturated soil contained in a large-scale geotechnical laminar box. The shake table used here was designed and constructed for the purposes of dynamic hybrid simulation experiments.

Experimental Setup for Algorithm Development

Shake Table

Elastomeric Bearings

Measured Accelerations and Settlements

1 Ph.D. Candidate, Department of Civil, Structural and Environmental Engineering, University at Buffalo 2 Assistant Professor, Department of Civil, Structural and Environmental Engineering, University at Buffalo

Hybrid simulation • Experimental method to study dynamic response of complex engineering systems • Combination of two systems: a physical and a computational subsystem

interacting with each other • The interaction utilized with the use of actuators and/or shake tables

Why Hybrid Simulation? • It allows adaptability by allowing a physical soil-foundation system to be coupled

with multiple complex virtual superstructure models and meaningful boundary conditions for the foundation

• Prof. Amjad Aref, Prof. Andrew Whittaker, Prof. Sabanayagam Thevanayagam • SEESL Personnel: Mark Pitman, Scot Weinreber, Robert Staniszewski, Jeffrey Cizdziel, Lou Moretta, Christopher Budden, Chris Zwierlein and Duane Kozlowski • National Science Foundation grant CMMI-0847053, University at Buffalo RENEW seed grant

Measured Pore pressures

Measured Strains

Foundation System Measured Frequency Response

Dynamic Hybrid Simulation • Inertia effects are part of the physical subsystem • Appropriate to study complex interaction phenomena, such as soil-foundation-

structure interaction

0 20 40 60 80 100 12010

-4

10-2

100

102

Amplitude Response

Frequency (Hz)

Absolu

te G

ain

0 20 40 60 80 100 120-800

-600

-400

-200

0Phase Response

Frequency (Hz)

Phase S

hift

0 2 4 6 8 10 12 14 16 18 20-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

Time (sec)

Pore

pre

ssure

(psi)

Rigid Superstructure

Superstructure 1

Superstructure 2

0 2 4 6 8 10 12 14 16 18 20-1.8

-1.6

-1.4

-1.2

-1

-0.8

-0.6

-0.4

-0.2

0

Time (sec)

Sett

lem

ent

(in)

Rigid Superstructure

Superstructure 1

Superstructure 2

0 2 4 6 8 10 12 14 16 18 20-1.5

-1

-0.5

0

0.5

1

1.5

2

Time (sec)

Shake t

able

accele

ration (

g)

Rigid Superstructure

Superstructure 1

Superstructure 2

0 2 4 6 8 10 12 14 16 18 20-0.3

-0.25

-0.2

-0.15

-0.1

-0.05

0

0.05

0.1

0.15

0.2

Time (sec)

Accele

ration -

Em

bedded a

ccele

rom

ete

r (g

)

Rigid Superstructure

Superstructure 1

Superstructure 2

0 2 4 6 8 10 12 14 16 18 20-150

-100

-50

0

50

100

150

Time (sec)

Str

ain

s f

rom

em

bedded s

ensor

Rigid Superstructure

Superstructure 1

Superstructure 2

0 2 4 6 8 10 12 14 16 18 20-0.3

-0.25

-0.2

-0.15

-0.1

-0.05

0

0.05

0.1

0.15

0.2

Time (sec)

Accele

ration -

Em

bedded a

ccele

rom

ete

r (g

)

Rigid Superstructure

Superstructure 1

Superstructure 2

0 2 4 6 8 10 12 14 16 18 20-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

Time (sec)

Pore

pre

ssure

(psi)

Rigid Superstructure

Superstructure 1

Superstructure 2

0 2 4 6 8 10 12 14 16 18 20-150

-100

-50

0

50

100

150

Time (sec)

Str

ain

s f

rom

em

bedded s

ensor

Rigid Superstructure

Superstructure 1

Superstructure 2

0 2 4 6 8 10 12 14 16 18 20-1.5

-1

-0.5

0

0.5

1

1.5

2

Time (sec)

Shake t

able

accele

ration (

g)

Rigid Superstructure

Superstructure 1

Superstructure 2

0 2 4 6 8 10 12 14 16 18 20-1.8

-1.6

-1.4

-1.2

-1

-0.8

-0.6

-0.4

-0.2

0

Time (sec)

Sett

lem

ent

(in)

Rigid Superstructure

Superstructure 1

Superstructure 2

Embedded Sensors

Acknowledgements