CIE 616

FALL 2004

ADVANCED DYNAMIC TESTING

TECHNIQUES
IN STRUCTURAL ENGINEERING
by
Andrei M Reinhorn
Xiaoyun Shao

Contents
Introduction of dynamic testing methods
Effective force testing
Pseudo dynamic testing
Real time hybrid dynamic testing

INTRODUCTION
Quasi-static loading test method (QST)
Shaking table testing method (STT)
Effective force method (EFT)
Pseudo-dynamic testing method (PDT)
Real time pseudo-dynamic testing method (RTPDT)
Real time dynamic hybrid testing method (RTDHT)

Quasi-static loading test method (QST)

a test specimen is subjected to slowly changing prescribed

forces or deformations by means of hydraulic actuators

inertial forces within the structures are not considered in

this method.

purpose is to observe the material behavior of structural

elements, components, or junctions when they are subjected
to cycles of loading and unloading.

dynamic nature of earthquakes are not captured

Shaking table testing method (STT)

test structures may be subjected to actual earthquake

acceleration records to investigate dynamic effects

inertial effects and structure assembly issues are well

represented

the size of the structures are limited or scaled by the size

and capacity of the shake table

Other testing methods (STT)

Effective Force Method
Pseudo-dynamic testing
Real Time Dynamic Hybrid Testing
(new developement)
:

m5x5a

m4x4a

m3x3a

m2x2a

m1x1a

F2

F1

Effective force testing method (EFT)

m4x4a
:

m3x3a
:

m2x2a
:

m1x1a
:

Real-Time PseudoDynamic Hybrid

Testing System
Real-Time Dynamic
Hybrid Testing
System

m5x5a
:

Effective Force
Technique
Hybrid Testing &
Computing

Applies the inertial ground motion generated forces through synchronized actuators - NEW

Effective force testing method (EFT)

applying dynamic forces to a test specimen that is anchored

rigidly to an immobile ground; perform real-time earthquake
simulation

these forces are proportional to the prescribed ground

acceleration and the local structural masses.
:

m5x5a

m4x4a

based on a force control algorithm

m3x3a
:

m2x2a

m1x1a

 Effective Force
Technique
Hybrid Testing &
Computing
Pseudo-Dynamic
Hybrid Testing
System
Real-Time Dynamic
Hybrid Testing
System

F2

F1

Applies forces in substructure through actuators only real time operation is a benefit but not a must

Pseudo-dynamic testing method (PSD)

applying slowly varying forces to a structural model

motions and deformations observed in the test specimens are

used to infer the inertial forces that the model would have
been exposed to during the actual earthquake

Substructure techniques

Real time pseudo-dynamic testing method

(RTPDT)

same as the PSD test except that it is conducted in the real

time

Introduce problem in control, such as delay caused by

numerical simulation and actuator

 Effective Force
Technique
Hybrid Testing &
Computing
Real-Time PseudoDynamic Hybrid Testing
System
Real-Time Dynamic
Hybrid Testing System

Applies forces in substructure through shake table and actuators real time operation is a must

Real-Time Seismic Hybrid Testing

INTERFACE FORCES
ACTIVE FEEDBACK FROM
SIMULATED STRUCTURE
APPLIED BY ACTUATORS
AGAINST REACTION WALL

REACTION
WALL
SIMULATED
STRUCTURE

SHAKING TABLES
(100 ton)

FULL OR NEAR
FULL SCALE TESTED
SUBSTRUCTURE

Fig.1. Real-Time Hybrid Seismic Testing System

(Substructure Dynamic Testing)

Real time dynamic hybrid testing method

(RTDHT)

based on shaking table test combined with substructure

techniques.

part of the structure (the physical model) is constructed and

tested on the shaking table

The rest part of the structure (the numerical model) is

numerically modeled in the compute

the earthquake effect on the superstructure was calculated

as a interface force and applied to the substructure by the
actuators (force control based)

Block Diagrams of Various Testing Methods

Open Loop Test

Open Loop Control (in concept)

Effective Force Test

Closed Loop Test

Closed Loop Test

Pseudo-dynamic Test with Substructure

Closed Loop Test

Summary of dynamic test methods

Advantages

Size of the specimen can be large or very large

Disadvantages

Inertial forces are not true forces and distorted by discrete

parameter model, actuators and computers
Rate effects are neglected because of quasi-static loading

PDT

Size of the specimen can be large or very large

Inertial forces are not true forces and distorted by discrete

parameter model, actuators and computers
Actuator time delay is introduced

RTPDT

True inertia forces in assembly

Size of the specimen is limited

True inertia forces on the specimen

Specimen can be large or very large

Part of the inertia forces are simulated with errors (same as PDT)
Actuator time delay is introduced

STT

RTDHT

Effective Force Testing

Equation of motion

&
&a Cx& Kx 0
Mx
Subscript refers to motion relative to a fixed reference
frame (absolute displacement)

x x Ix
a
g

&
x& &
x& I&
x&
a
g

&
& Cx& Kx M&
Mx
x& P t
g
eff

Open Loop Control (in concept)

Effective Force Test

Effective Force Test Hardware

Components
Servo-Hydraulic Actuators
Servo-Hydraulic Control System
Elastic Spring
Measurement Instrumentation (DAQ)
Computer
Simulator
Controller

Effective Force Test Hardware

Configuration

Effective Force Test -Dynamic force

control
Series elasticity and displacement feedback
Target
Force

Measured
Force

1 / KLC

Command
Signal
Actuator with
Displacement Control

Compensator

Series
Load
Cell Spring, KLC

Structure

Effective Force Test -Dynamic force

control
Series elasticity and displacement feedback
Actuator
Displacement
Feedback
Desired
Force

1/K LC

G Actuator

G Actuator

Actuator
Displacement

Actuator in
Closed-loop
Displacement Control

Compensation

Structure
Displacement

K LC
Series Spring

1
ms 2 cs k
Structure

Achieved Force
Desired Force

CG

ms 2 cs k

ms 2 cs k K LC 1 CG

Ideal: C = 1/G

Achieved
Force

Effective Force Test -Dynamic force

control
The advantages of using the series spring
the actuator can be well tuned and operated in displacement
control
it provides for one more parameter than can be altered in the
control design (the oil stiffness cannot be)
the term KLC(1-CG) in the transfer function indicates that the
smaller the value of KLC the less sensitive is the transfer
function to deviations of C from 1/G

Effective Force Test Effect of Time

Delay
The dynamic characteristics of hydraulic actuators inevitably
include a response delay , which is equivalent to negative
damping
2.0

Experimental

1.8
1.6

Numerical

Magnitude

1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0.0

2.0

4.0

6.0

Frequency (Hz)

8.0

10.0

Effective Force Test Predictive

Control
1/K LC

Corrective
Displcement

T = e -s

Predictive
Displcement

T0

1
m0 s c0 s k 0 K LC 0

Delay
Model

Model of StructureSpring System

Actuator

K LC
Series Spring

1
ms 2 cs k
Structure

Smith Predictor

Effective Force Test Predictive

Control
2.0
1.8
1.6

Without compensation

Magnitude

1.4
1.2
1.0

With compensation

0.8
0.6
0.4
0.2
0.0
0.0

2.0

4.0

6.0

Frequency (Hz)

8.0

10.0

Effective Force Test Software

Simulink
Realtime Workshop5
XPC Target

Pseudo dynamic testing

Define a model of the structure system
Define the desired excitation usually base
acceleration
Calculate the expected response of structure
displacement
Use an actuator to apply the desired displacement in
the structure
Measure the resistance force in the structure (or
estimate it from measurements)
Repeat the above steps start from second

&
& Cd& R f
Md
i
i
i
i

Pseudo-dynamic testing

Pseudo dynamic testing

Pseudo-dynamic testing Hardware

Components

Servo-Hydraulic Actuators

Servo-Hydraulic Control Systems

Measurement Instrumentation

On-line computer

Pseudo-dynamic testing Hardware

Configuration (Local)

Pseudo dynamic testing

Discretized equation of motion of the structure at time
intervals

ti it

for

, i 1toN

&
& Cd& R f
Md
i
i
i
i
Equation solved in computer step by step, with R i as the reaction
force measured from the specimen under test. Result is the
displacement command of next step that will be applied to the
specimen at each node of mass by actuators.

Pseudo dynamic testintegration

algorithm
Both explicit and implicit time-stepping integration algorithm can be
applied for solving equation of motion in Pseudo-dynamic tests.
Explicit methods compute the response of the structure at the end
of current step based on the state of the structure at the
beginning of the step.

Central difference method (Takanashi et al. 1975),

Newmark- Beta method (1959),
Modified Newmarks method (1986),
The -function pseudodynamic algorithm (Chang et al. 1997)
Unconditionally stable explicit method(Chang, 2002)

(continued on next)

Pseudo dynamic testintegration

algorithm
(continued)
Implicit methods require knowledge of the structural response
at the target displacement in order to compute the response .

the displacement is dependent on other response parameters

at the end of the step

iteration is required in the algorithm to satisfy both the

imposed kinematic conditions and the equilibrium conditions at
the end of the time step
Newmark Alpha method (Hilber et al. 1977)
Hybrid implicit algorithm (Thewalt and Mahin, 1987)
Newton iteration (Shing, 1991)

Pseudo dynamic testintegration

algorithm
(continued)
implicit iteration algorithm provide improved stability
characteristics and permit the used of larger
integration time steps
iteration on experimental model is not practical since
structure materials are path dependent
explicit methods are easier to implement
Explicit integration methods are preferred for PSD
simulation when stability limits are satisfied for the
structural model under investigation

Pseudo dynamic testintegration

algorithm
(continued)
Example: Modified Newmarks Method

Md i 1 2 M (d i 1 d i ) (1 )R i 1 R i f i 1
t
2

d i 1 d i td i
d
i
2

t
d i 1 d i (d
i d i 1 )
2

Substitute into and solve

for

d
i 1

2
2
di
d i M 1 f i 1 R i (1 )R i 1
(2 )t
2

Pseudo-dynamic testing
substructuring
principle
may fabricate only part of the structure whose hysteretic
behavior is complex and apply the test to this part

remaining part treated in the computer

Pseudo-dynamic testing
substructuring
principle
M ee
M T
ea

C ee
M ea d
e

T

M aa d a C ea

C ea

C aa

d e R e
fe

d R a
fa
a

subscripts a and e denote the degrees of freedom within the

analytical and experimental substructures.

C d K d f M d

M ee d
e
ee e
ee e
e
ea a C ea d a K ea d a

C d K d f M T d
C T d K T d
M aa d
a
aa a
aa a
a
ea
e
ea
e
ez

Dee de fe Dea da
Daa da fa Dea Tde

Tested part. Calculate displacement command for next step.

Interface force: Dea da
Analytical part. Calculate interface state used in interface
force.

Pseudo-dynamic testing Hardware

Configuration (Internet)
SIMULATION
COORDINATOR
`

Matlab Software
TCP/IP
Integration Algorithm

Analysis Site
LAN/WAN

SIMULATION
INTERFACE
NTCP Server
TCP/IP

NTCP Plugin
SCRAMNet
Simulation
Interface

CONTROLLER

DAQ

xPc Software
SCRAMNet
DSP Read/
Write

xPC Software
SCRAMNet
Analog I/O

Test Software
Control

DSP
Signal
Generation

Analog I/O
SCRAMNet to
Analog I/O
Bridge

ServoController
Servo-Hydraulic
Control

Flex Test

Remote
Substructure Site

SV

DAQ Hardware
Signal
Conditioners
`

SCRAMNet

Pseudo-dynamic testing Software

Response analysis Matlab Simulink

Controller implementation Matlab Stateflow

Dynamic hybrid testing - I

Combined use of earthquake simulators, actuators and
computational engines for simulation
Details later in the presentation
Response Feedback
Computational
Substructure

Physical
Substructure

Physical
Substructure
Shake Table

Computational
Substructure

Ground/Shake Table

Structural Actuator

Dynamic hybrid testing - II

Well understood

Structural
Actuator

Foundatio
n
Laminar
Soil Box
Shake Table

Focus of
interest

Real-time dynamic hybrid testing - II

Response
Feedback

Structural
Actuator
Has to operate in
Force Control
Distributed
mass

Foundatio
n
Laminar
Soil Box
Shake Table

Acceleration input:
Table introduces
inertia forces

Substructure Testing Unified

Approach
`

Response Feedback
Computation
al
Substructure
Physical
Substructur
e

Physical
Substructur
e

Structural
Actuator

Shake Table
Computation
al
Substructure
Ground/Shake
Table

&
&
Shake table acceleration, u
t

&
&
1 s u
1
1
42 4
3

First story contribution

to shake table acceleration

k3
3 s
x3 x2
m2
1 4 44 2 4 4 4
3
Third story contribution
to shake table acceleration

&
& 1 3 s k3 x3 x2
Actuator Force, Fa 1 1 s m2 u
1 4 4
42 4 4 4
31
1 4 4 4
42 4 4 4 4
3
First story contribution
to actuator force

Third story contribution

to actuator force

1 s 0 and 3 s 0

Unified approach to substructure

testing
If s 0 and s 0 , then the control requires a shake table and
an actuator to implement the substructure testing.
1

1 s 0 and 3 s 0

, then the controller require just an actuator

If
to implement the substructure testing as pseudo-dynamic
testing:
Note:
In pseudo-dynamic testing, inertia effects are computed.
s 0 or s 0
), the actuator
In dynamic hybrid testing (
should operate in force control.
1

Hybrid Controller Implementation (UB-NEES

Flexible architecture using parallel processing

Optional

Optional

Computationa
l Substructure
MTS Actuator
Controller (STS)

Physical
Substructur
e
Shake Table

Computationa
l Substructure

Structural
Actuator

MTS Hydraulic Power

Controller (HPC)

Network Simulator

General Purpose
Data Acquisition
System

Compensation Controller Real-time Simulator

xPC Target

Data Acquisition

SCRAMNET I

Physical
Substructur
e

SCRAMNET II

Ground/Shake Table
MTS Shake Table
Controller (469D)

Hybrid Testing

CONTROL OF
LOADING SYSTEM

HYBRID CONTROLLER
UB-NEES NODE

Design done jointly between MTS and UB

Implementation of RTDHT

Actuator

Structure

Shake
Table

Actuator

Substructure response
0.014

Second
(simulated
)
floor

0.012

0.010
0.008
0.006
0.004

0.002
0.000
0.0

2.0

4.0

6.0

8.0

10.0

0.10

First
(physical)
floor

0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0.00
0.0

2.0

4.0

Hybrid
test

6.0

8.0

10.0

Shake
table

Structure
Shake
Table