1
1
1
2
3
1
1
2
3
1
2
1
2
3
Acknowledgment: This work was supported by a research grant from the National Science Foundation under Award Number 0406812

MISST
TABLE OF CONTENTS
LIST OF TABLES ..................................................................................................................................................... 3
LIST OF FIGURES .................................................................................................................................................... 3
1.
INTRODUCTION ............................................................................................................................................ 4
2.
RATIONALE AND PROBLEM DESCRIPTION ............................................................................................ 4
2.1
R ATIONALE ........................................................................................................................................................................................... 4
2.2
MISST S TRUCTURE .............................................................................................................................................................................. 6
2.2.1
Prototype ................................................................................................................................................................................................................. 6
2.2.2
Structural Idealization......................................................................................................................................................................................... 7
2.3
M ULTI -S ITE S IMULATION ..................................................................................................................................................................... 8
2.3.3
UI-SIMCOR ............................................................................................................................................................................................................ 9
2.3.4
Module description .............................................................................................................................................................................................. 9
3.
ACTIVITIES AND RESULTS ........................................................................................................................ 13
3.1
V ERIFICATION S TUDIES ...................................................................................................................................................................... 13
3.1.1
Distributed analytical simulation .................................................................................................................................................................. 13
3.1.2
Preliminary integrated simulation ................................................................................................................................................................ 13
3.2
D ISTRIBUTED H YBRID S IMULATIONS ............................................................................................................................................... 14
3.2.3
Small amplitude tests ...................................................................................................................................................................................... 15
3.2.4
Large amplitude tests ...................................................................................................................................................................................... 16
4.
DISCUSSION ................................................................................................................................................. 22
4.1
M AJOR F INDINGS ............................................................................................................................................................................... 22
4.1.1
Verification studies ............................................................................................................................................................................................ 22
4.1.2
Distributed Hybrid Simulations ..................................................................................................................................................................... 22
4.2
C LOSURE ............................................................................................................................................................................................. 23
5.
REFERENCES ................................................................................................................................................. 24
6.
APPENDICES ................................................................................................................................................. 25
6.1
P ARTICIPANTS ..................................................................................................................................................................................... 25
6.2
C ONSTRUCTION D RAWINGS ........................................................................................................................................................... 26
2 of 26

MISST
Table 3.1 Maximum responses for small amplitude test .....................................................................................16
Table 3.2 Maximum responses for large amplitude test with Santa Monica record .....................................17
Table 3.3 Maximum responses for simulation with Newhall Fire record .......................................................18
Table 3.4 Comparison between UIUC and Lehigh piers ....................................................................................18
Figure 2.1 The role of the multi-site simulation in seismic performance evaluation ...................................... 5
Figure 2.2 Layout of Santa Monica Freeway (unit: in) ............................................................................................ 6
Figure 2.3 Observed pier damage ............................................................................................................................. 7
Figure 2.4 Layout of model structure ........................................................................................................................ 7
Figure 2.5 Subdivision of model structure ................................................................................................................ 8
Figure 2.6 Allocation of modules and communication amongst sites ................................................................ 8
Figure 2.7 NEES@UIUC large scale laboratory ....................................................................................................10
Figure 2.8 UIUC experimental setup .......................................................................................................................10
Figure 2.9 Pier renforcement details (in) ................................................................................................................11
Figure 2.10 NEES@Lehigh large scale laboratory .................................................................................................11
Figure 2.11 Lehigh experimental setup ....................................................................................................................12
Figure 3.1 Distributed analytical simulation ...........................................................................................................13
Figure 3.2 Response from preliminary simulation ................................................................................................14
Figure 3.3 Utilized Northridge records ..................................................................................................................15
Figure 3.4 Comparison of target and measured displacement ..........................................................................15
Figure 3.5 Comparison between analysis and observed response ...................................................................16
Figure 3.6 Lateral force-displacement relationship (Santa Monica record) ....................................................17
Figure 3.7 Lateral displacement history for Piers 1 and 2 ...................................................................................18
Figure 3.8 Shear cracking at UIUC ...........................................................................................................................19
Figure 3.9 Initial shear failure at UIUC ....................................................................................................................19
Figure 3.10 Shear cracking at Lehigh........................................................................................................................20
Figure 3.11 Shear failure at Lehigh ...........................................................................................................................20
Figure 3.12 Secondary shear-axial failure at UIUC ...............................................................................................21
Figure 3.13 Lateral force-displacement relationship (Newhall Fire record) ...................................................21
Figure 4.1 Comparison of observed behavior .......................................................................................................23
Figure 6.1 Elevation view of piers .............................................................................................................................26
3 of 26

MISST
The Multi-Site Soil-Structure-Foundation Interaction Test (MISST) was performed on May 16 and 17,
2006. MISST, a Pre-NEESR project, was conducted by collaborating partners at the University of Illinois at Urbana-Champaign (UIUC), Lehigh University, and Rensselaer Polytechnic Institute (RPI) and was established to provide a realistic test-bed application in which to verify and extend all components of
NEESgrid as well as all components of the sites taking part in the distributed simulation. Concurrently, it has extended and represents the current state-of-the-art in distributed and integrated experimental-analytical earthquake engineering simulation.
The rationale for the investigation and description of the simulation are initially provided. A discussion of the development activities leading to the culminating test on May 16 and 17 is then provided and followed by a complete description of the MISST test. Finally, findings and contributions stemming from the investigation are detailed.
MISST was established to extend and build upon the Multi-Site Online Simulation Test (MOST) which took place on July 30, 2003 (Spencer et al, 2004). MOST was the first geographically distributed experimental simulation to make use of the relatively young George E. Brown, Jr. Network for
Earthquake Engineering Simulation (NEES). The MOST simulation utilized NEESgrid, the NEES cyber infrastructure, to combine a numerical model run at the National Center for Supercomputing
Applications (NCSA) with experimental tests being conducted at UIUC and the University of Colorado at
Boulder. Similarly, MISST applied this distributed hybrid approach to simulate the response of a bridge structure which was partitioned into five separate modules distributed at three of NEES equipment sites: namely, UIUC, Lehigh, and RPI.
2.1
R ATIONALE
Earthquakes affect vulnerable communities in a broad-ranging manner. It is therefore imperative that all available tools should be deployed to mitigate earthquake effects. The tools of earthquake performance assessment are (i) testing, (ii) analysis, and (iii) collecting data from regions hit by destructive earthquakes.
Each of the three tools has advantages and drawbacks. Their relative merits and application potential are briefly assessed below.
Experimental testing is an essential and powerful investigation tool in earthquake engineering applications.
However, it is not without shortcomings. The three approaches (monotonic, cyclic, and dynamic loading) each have both positive and negative aspects. Dynamic testing (performed using shaking tables under computer control) is arguably closest to reality, but scale effects, full dynamic similitude, and difficulties with soil-structure interaction modeling pose serious hurdles. The development of on-line, computer-controlled (pseudo-dynamic or PSD) testing has by-and-large liberated dynamic testing from the constraints of scale effects, but strain-rate insensitivity is a pre-requisite for the acceptance of the obtained results.
4 of 26

MISST
Analytical investigations are fast-developing. However, similar to the situation with experimental testing, decisions are needed with regard to the type of model, method, and input. The model may vary from a single-degree-of-freedom idealization of the whole structural system to a detailed 2D or 3D representation of the structure, foundation and surrounding soil. The method may be static or dynamic, with various sub-classes of either of the two options. Great advancements have taken place in recent years in the aforementioned ingredients of seismic analysis. Indeed, developments in analytical capabilities have outpaced the advances in testing, due to the faster turnover in analysis and the relatively low costs involved, underpinned by the leaps in computer technology. Notwithstanding, there are modeling problems that continue to be unsolved, such as shear deformation and failure of RC structures, detailed three-phase (matrix, voids and moisture) soil response under variable amplitude dynamic loading, or even bolt-hole ovalization in steel connections.
Field observations are ultimately the most realistic since all aspects of source, path, site, foundation, structural, and non-structural components are taken into account. However, this result is for a specific set of the aforementioned parameters, all of which influence the damage inflicted on the exposed building stock. Even if the seismological quantities are those sought for a particular application, the soil substrata, structural material, resistance system, age, configuration and foundation type are unlikely to be those under investigation. Consequently, the data collected is generically regional in nature and cannot be readily applied in a predictive fashion. Finally, the frequency of earthquake occurrence is such that its applicability and usefulness is serendipitous.
Figure 2.1 The role of the multi-site simulation in seismic performance evaluation
Assessment of the relative merits and drawbacks of the three seismic performance evaluation methods leads naturally to the concept of multi-site, on-line, computer-controlled integrated testing-analysis of
complex systems (Figure 2.1). This approach, which is the underlying vision of NEES, combines the most
attractive features of each assessment approach, thus providing test results far superior to those obtained
5 of 26

MISST from existing approaches. With reaction walls, geotechnical centrifuges, shaking tables, structural and geotechnical field facilities, instrumented sites and a wave tank, NEES furnishes the earthquake engineering community with a comprehensive toolkit for the application of multi-site simulations, integrated by the NEESgrid cyber infrastructure. MISST demonstrates the application of this toolkit and cyber infrastructure through an investigation into the behavior of a bridge structure which was heavily damaged during the January 17, 1994 Northridge earthquake. Moreover, the simulation serves as a test-bed in which to critically evaluate this distributed and hybrid methodology.
2.2
MISST S TRUCTURE
2.2.1
Prototype
The prototype structure for the investigation was selected as Collector-Distributor 36 of the Santa
Monica (I10) Freeway which was damaged in the Northridge earthquake of 17 January 1994 (Broderick and Elnashai, 1995). The Collector-Distributor 36 forms part of a pair of off-ramps from the eastbound carriageway on the I-10 freeway at La Cienega-Venice Under- crossing and was designed and constructed
between 1962 and 1965. The structural configuration is shown in Figure 2.2. The ramp was located about
25km to the south-east of the epicenter. From the bifurcation point just to the west of bent 5 of the under-crossing, the RC ramp was carried first over the multi-column bent 5, then over three single piers
(6, 7 and 8) and finally over the pier wall of bent 9 to the east abutment. The deck consisted of a 3-cell continuous box girder which was rigidly connected to the supporting structure 10.
5 6
Expansion Joint
7 8 9 10
240.0
258.9 247.6 234.0
1071.0
Circular Pier (D=48)
186.5 746.5
Circular Pier (D=48)
5032.5
1270.0
Circular Pier (D=48)
Rectangular Wall
(B=18, H=354)
1212.5
Figure 2.2 Layout of Santa Monica Freeway (unit: in)
546.0
There was no visible damage on either the ramp deck or the abutment. However, the piers experienced
varying levels of damage of which the most severe is illustrated in Figure 2.3.
In particular, pier 6 experienced spectacular failure and was the most damaged of all the columns supporting the ramp 10. As
shown in Figure 2.3, shear failure occurred in the lower half of the pier. The concrete cover completely
spalled over the height and the concrete core disintegrated. Moreover, all the reinforcement bars buckled symmetrically and the transverse hoops opened, leaving the pier with large permanent axial deformation.
In addition to the geometric configuration, this complex failure renders the structure ideal for showcasing the benefits of a distributed hybrid approach.
6 of 26

MISST
(a) Pier 6 (b) Pier 8
Figure 2.3 Observed pier damage
2.2.2
Structural Idealization
To utilize the participating NEES experimental facilities and for simplification, the bridge was assumed to
have three piers. The overall structural configuration is similar to the real structure as shown in Figure 2.4.
Also shown in Figure 2.4 are the initial loads and lumped mass values. The initial load was calculated with
deck self-weight and applied to the top of piers. Masses were placed on the deck as it was desired to test the piers under static conditions with the dynamic response being determined from an analytical model of the deck. Using the Rayleigh formulation, 2% damping was applied to the first and second modes of the structure whose elastic periods were found to be 0.3710 and 0.0768 sec respectively. The adapted soil spring model is discussed in subsequent sections and assumed material properties are as follows:
Concrete
 Compressive strength: 5000 psi
 Tensile strength: 281 psi
 Crushing strain: 0.0025
Reinforcement
 Yield strength: 60 ksi
 Ultimate strength: 97 ksi
Figure 2.4 Layout of model structure
7 of 26

MISST
2.3
M ULTI -S ITE S IMULATION
As previously discussed, the MISST structure was partitioned into five distinct modules whose response was simulated at one of three separate NEES equipments sites. The partitioning and resulting modules are
illustrated in Figure 2.5. The allocation of the modules and communication amongst the sites is illustrated in Figure 2.6.
All communication and interaction was conducted via NEESgrid and UI-SIMCOR, simulation coordinator software developed at UIUC. A description of UI-SIMCOR and each of the modules is provided below.
1
3
Deck, Pier 2, Soil 2
4 2
5
UIUC
Pier 1
Lehigh
Pier 3
7
6
RPI
Soil 1
Soil 3
Figure 2.5 Subdivision of model structure
Figure 2.6 Allocation of modules and communication amongst sites
8 of 26

MISST
2.3.3
UI-SIMCOR
Recognizing the need for a central control system for multi-site testing the University of Illinois Simulation
Coordinator, UI-SIMCOR, was developed for multi-site substructure pseudo dynamic testing and simulation (Kwon et al, 2005). During the development of the coordination system, the following key components were sought:
 Integration scheme for PSD test
 Communication amongst sub-structured components
 Sub-structuring (sub-division) of the complex system
One of the notable advantages of UI-SIMCOR is that is allows all sub-structured components to be analyzed or physically loaded statically. The dynamic components of structural test are contributed by
UI-SIMCOR through a PSD algorithm. Currently, the  -Operator Splitting method is used as the integration scheme (Combescure and Pegon, 1997); however, it should be noted that other methods could easily be incorporated into the software. Another significant advantage of the simulation coordinator is the ease with which it allows integrated response to be determined from numerous separate subdivisions of the overall system. As demonstrated by MISST, distant geographically distributed sub-structured components can be integrated and tested as a fully interacting system, allowing multiple laboratories to be used for large complex systems.
2.3.4
Module description
As noted, three of fifteen NEES equipment sites were utilized during the investigation. Experimental models were constructed and tested at both UIUC and Lehigh while computational modules were
analyzed at both RPI and UIUC (Figure 2.6). Provided herein is a description of each module.
2.3.4.1
Experimental modules
Pier 1 - UIUC
Facility: The Multi-Axial Full-Scale Sub-Structured Testing and Simulation Facility (MUST-SIM) housed at the University of Illinois was employed to experimentally assess the behavior of Pier 1. The key components of the MUST-SIM Facility include the following (Elnashai et al, 2004):
 Reaction wall: L-shaped post-tensioned concrete strong wall of 50 x 30 x 28 x 5 ft (Length x
Width x Height x Thickness, respectively) (Figure 2.7)
 Loading and Boundary Condition Boxes (LBCBs): Self-reacting assemblies of actuators, swivel joints, and control software cable of imposing any combination of six actions (forces and moments) and six deformations (displacements and rotations) to test specimens connected to its
loading platform (Figure 2.7)
 Non-contact instruments: Includes Krypton’s K-600 Dynamic Measurement Machine, Stress
Photonics Gray Field Polariscope, and a close range digital photogrammetry system
 1/5 Scale Laboratory and User Studio: Fully functional training laboratory and visualization studio
9 of 26

MISST
(a) Reaction wall (b) Load and Boundary Condition Box (LBCB)
Figure 2.7 NEES@UIUC large scale laboratory
Specimen: To satisfy capacity limitations, a ½ scale model of prototype Pier 1 was constructed and
tested at UIUC (Figure 2.8). The cross section of the prototype is contrasted with that of the model
in Figure 2.9. Reinforcement was selected to match, to the extent possible, the prototype’s
reinforcement ratios. The final longitudinal ratio was 3.11% while 0.176% was provided transversely.
Detailed construction drawings are provided in Appendix 6.2 and as-built material properties were
found to be 6868 psi for the 28 day concrete compressive strength and 79 ksi for the reinforcement
yield strength of reinforcement. As shown in Figure 2.8, a single LBCB was utilized to control the top
node of the pier while the base was fixed to the laboratory strong floor.
Figure 2.8 UIUC experimental setup
10 of 26

MISST
24-#11 for the outer bars
8-#11 for the inner bars
Total: 32 - #11
24-#6 for the outer bars
8-#6 for the inner bars
Total: 32 - #6
1" Cover
2" cover
#4 Stirrups @ 16" cs.
#3 Stirrups @ 12" cs.
(a) Prototype (b) ½ scale model
Figure 2.9 Pier renforcement details (in)
Pier 3 – Lehigh
Facility: The Real Time Multi Directional (RTMD) Test Facility in ATLSS Center, Lehigh University was employed to experimentally assess the behavior of Pier 3. The key components of the RTMD Facility include the following:

ATLSS Center Reaction Wall: Pre-existing reaction wall with a maximum height of 50.2 ft (Figure
2.10)
 Dynamic Actuators: Five 20 in stroke dynamic actuators with 382 kips (3 actuators) or 517 kips
(2 actuators) force capacity and accompanying hydraulic equipment
 Advanced control and sensor systems: Includes control, video, tele-operation, and data acquisition systems as well as advanced sensors
(a) ATLSS Center Reaction Wall (b) Dynamic Actuators
Figure 2.10 NEES@Lehigh large scale laboratory
Specimen: Similarly to Pier 1, a ½ scale model of prototype Pier 3 was constructed and tested at
Lehigh (Figure 2.11). The cross section of this pier was identical to that of Pier 1 shown in Figure 2.9.
As shown in Figure 2.11, in this instance two actuators where utilized to control the rotation and
lateral displacement of top node while the axial load was applied though vertically tensioned
DYWIDAG rods. Detailed construction drawings are provided in Appendix 6.2
11 of 26

MISST
Figure 2.11 Lehigh experimental setup
2.3.4.2
Analytical modules
The Mid-America Earthquake Center program Zeus-NL was utilized for all of the three computational modules. Zeus-NL is an inelastic fiber analysis package which was specifically developed for earthquake engineering applications (Elnashai et al, 2004). Provided below is a description of each of the analytical modules.
Deck, Pier 2 and Soil 2
The bridge deck in the prototype structure is a prismatic multi-celled RC box girder. This bridge deck was simplified as equivalent rectangular section in Zeus-NL. The material and section properties in the prototype structure were employed for the middle pier. To account for soil-structure interaction (SSI), equivalent base-spring elements have been used to simulate the flexibility of the foundation-soil system.
For the calculation of equivalent soil stiffness, the shear wave velocity, Poisson ratio, and shear modulus were assumed to be 250m/sec, 0.40, and 1.67 ksi (11.5 MPa) respectively. For simplification, only x-direction (longitudinal displacement) was considered with a equivalent soil stiffness of 11620 kip/in.
Soil 1 and Soil 3
Soil 1 and Soil 3 were identical to the soil model discussed above (Soil 2).
12 of 26

MISST
Due to the complex nature and originality of the proposed simulation, the problem was approached incrementally. Provided herein is a discussion of the activities which led to culminating simulation that took place May 16 and 17, 2006. Moreover, a detailed description of the final simulation and its results is given.
3.1
V ERIFICATION S TUDIES
3.1.1
Distributed analytical simulation
Prior to incorporating any experimental modules, communication and protocol between the various sites was established and validated through a fully analytical distributed simulation. The substructure scheme
that was utilized is illustrated in Figure 3.1. The employed modules were identical to those of the final
simulation with the exception of piers 1 and 2:
 Module 1: Analytical model of deck, pier 2, and soil 2 – UIUC
 Module 2: Analytical model of pier 1 – UIUC
 Module 3: Analytical model of pier 3 – Lehigh
 Module 4: Analytical model soil 1 – RPI
 Module 5: Analytical model soil 3 – UIUC
1
3
Deck, Pier 2, Soil 2
4 2
5
UIUC
Pier 1
Lehigh
Pier 3
7
6
RPI
Soil 1
Soil 3
Figure 3.1 Distributed analytical simulation
3.1.2
Preliminary integrated simulation
Continuing with the incremental approach, a second simulation was conducted utilizing linear experimental specimens for piers 1 and 3 (modules 2 and 3, respectively). At Lehigh, an existing steel specimen was employed to model pier 3; whereas, Mini-MOST (Nakata et al, 2004), the scaled and portable version of the Multi-Site Online Simulation Test (MOST), was utilized at UIUC to model pier 1.
The behavior of the distributed simulation is compared to that the fully single analytical simulation in
Figure 3.2. As shown, the recorded response matched quite well to that of the analytical model despite
a downward drift displayed during the later portion of the experiment. The drift was later attributed to
13 of 26

MISST mechanical slip occurring at the Lehigh site. Regardless, the simulation adequately served as a foundation for subsequent fully nonlinear simulations.
10
0
-10
-20
-30
-40
-50
50
40
30
20
Analytical Simulation
Experiment
0 3 6 9 12
Time (sec)
Figure 3.2 Response from preliminary simulation
15
3.2
D
ISTRIBUTED
H
YBRID
S
IMULATIONS
Building upon the preliminary verification studies, several simulations were performed that utilized the
substructure scheme illustrated in Figure 2.6. These simulations included both small and large amplitude
tests. The small amplitude test was intended to verify the functionality of all components and equipment while the large amplitude tests were intended to replicate the observed damage in the prototype structure.
Two earthquake records that were captured during the Northridge earthquake of 1994 were employed during these simulations. The first record was strong motion data collected at the Santa Monica City Hall which had a peak ground acceleration (PGA) of 0.37g (Figure). The second record was collected at the
Newhall Fire Station and had a PGA of 0.58g. In both cases, the acceleration record was applied along the longitudinal direction of to the bridge structure.
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
0
PGA: 0.37g
3 6 9
Time (sec)
(a) Santa Monica City Hall station
12 15
14 of 26

MISST
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
0
PGA: 0.58g
3 6 9
Time (sec)
(b) Newhall Fire station
Figure 3.3 Utilized Northridge records
12 15
3.2.3
Small amplitude tests
To ensure proper communication and interaction was occurring amongst the sites, a small amplitude test was performed. This test utilized the Santa Monica record which had been scaled to ¼ of its original PGA or 0.093g. Thus, the experimental specimens at both UIUC and Lehigh remained near or within the linear range of behavior. The time step utilized in the integration scheme was 0.005 sec and a total 998 steps
(4.99 sec) were performed which corresponded to a total test length of about 3 hrs or 11 sec per step in
real time. As shown in Figure 3.4, commanded target displacement closely coincided with measured
response at both the UIUC and Lehigh sites. Thus, it was assured that proper communication and load
application was occurring at the geographically distributed sites. Table 3.1 shows the maximum
responses for pier 1 and 3.
0.1
0.05
0
-0.05
-0.1
0 100 200 300 400 500 600
Step
(a) Pier 1 - UIUC
Target Displ
Measured Displ
700 800 900 1000
0.1
Target Displ
-0.1
0
0
100 200 300 400 500 600 700
Measured Displ
Step
100 200 300 400 500 600 700 800 900 1000
0.1
0.05
0.1
0
0.05
Step
(b) Pier 3 - Lehigh
Figure 3.4 Comparison of target and measured displacement
Target Displ
Measured Displ
-0.05
0
15 of 26
Target Displ
Measured Displ
-0.1
-0.05
0
-0.1
0
100 200 300 400 500 600
Step
700 800 900 1000
100 200 300 400 500 600 700 800 900 1000
Step
0.00020
Target Rotation
Measured Rotation
0.00000
-0.00020
0 100 200 300 400 500 600 700 800 900 1000
Step

MISST
Table 3.1 Maximum responses for small amplitude test
Site Lateral
(in)
Displacement
Vertical
(in)
Rotation
(rad)
Lateral
(kip)
UIUC (Pier 1) 0.103 0.056 7.736E-05 14.66
Lehigh (Pier 3) 0.104 0.017 3.151E-04 12.46
Force
Vertical
(kip)
Moment
(kip-in)
139.36 1276.80
159.30 808.08
3.2.4
Large amplitude tests
To replicate the pier damage observed in the Northridge earthquake, large amplitude tests were
conducted with the strong motion records shown in Figure 3.3.
3.2.4.1
Simulation with Santa Monica record
The Santa Monica record was selected for use in the first large amplitude test. The record was amplified with a scale factor of 1.5 thus producing a PGA of 0.55g. And similarly to the small amplitude test, a
0.005sec integration time step was utilized which corresponded to about 6.7 hrs and 10.3 sec per step in real time. However, for this simulation a total of 2354 steps (11.77 sec) were performed.
Following the simulation, minor flexural cracks were observed near the base of both experimental piers.
Additionally minor shear cracks had developed near the mid-height of pier 1 at UIUC. As shown in Figure
3.5, the displacement amplitude is markedly larger than that of analytical prediction. This is partially due to
the softening and period elongation caused by the observed flexural and shear cracking. Figure 3.6 illustrates the hysteretic response of both piers. As detailed in Table 3.2, the maximum lateral
displacement for both piers was approximately 0.96 in and the corresponding maximum lateral forces were 78.46 and 101 kips for UIUC and Lehigh, respectively.
1
Analytical Prediction
Test Result
0.5
0
-0.5
-1
0 200 400 600 800 1000
Step
1200 1400 1600 1800
Figure 3.5 Comparison between analysis and observed response
2000
16 of 26

MISST
100
80
60
100
80
40
20
0
-20
-40
60
40
20
0
-20
-60
-80
-100
-40
-60
-80
-100
-120 -120
-1 -0.5
0 0.5
1 -1 -0.5
0 0.5
1
Displacement X (in)
(a) Pier 1 - UIUC
Displacement X (in)
(b) Pier 3 - Lehigh
Figure 3.6 Lateral force-displacement relationship (Santa Monica record)
Table 3.2 Maximum responses for large amplitude test with Santa Monica record
Site Lateral
(in)
Displacement
Vertical
(in)
Rotation
(rad)
Lateral
(kip)
Forces
Vertical
(kip)
Moment
(kip-in)
UIUC (Pier 1) 0.962 0.083 2.492E-04 78.46 193.34 3582.00
Lehigh (Pier 3) 0.969 0.017 2.783E-04 101.70 162.47 4614.40
3.2.4.2
Culminating simulation with Newhall Fire record
As significant damage was not observed during the simulation with the Santa Monica record, the more demanding Newhall Fire record was selected for further investigation. This record is characterized by a
PGA of 0.58g and a time step of 0.005 sec was again utilized. And in this instance, a total of 1210 steps were taken which corresponded to about 4.6 hrs or 13.6 sec per step.
Table 3.3 summarizes the maximum response measured from the UIUC and Lehigh piers. Figure 3.7 and
Table 3.4 illustrate the longitudinal displacement history and highlights significant events which occurred
during this final simulation. The UIUC pier experienced significant shear cracking near step 727 which
corresponded to 1.2in displacement and 111kips force (Figure 3.8). The simulation continued and the
first shear failure was observed in the UIUC pier at step 941. This was marked by the brittle formation
of a large diagonal crack near the mid-height of Pier 1 (Figure 3.9). Continuing onward, significant cracks
began to develop on the Lehigh pier (Figure 3.10) which was shortly followed (step 954) by a similar shear
failure of Pier 3 (Figure 3.11). As shown in Table 3.4, shear failure occurred at 124kips and 1.6in at UIUC
while at Lehigh failure occurred at 125 kips and 2.1in. Despite the brittle nature of the two shear failures and the associated loss of stiffness, the simulation remained stable and was continued. At step 1208 a final
secondary failure was observed at UIUC (Figure 3.12). This mixed axial-shear failure was marked by a
dramatic widening of the crack on Pier 1. The simulation continued for two additional steps; however, was halted at step 1210 owing to the loss of axial stiffness which had accompanied the secondary failure of the Pier 1. Due to this loss of stiffness, the algorithm correctly began to command large axial displacements which were considered to be unsafe in the laboratory environment.
17 of 26

MISST
Table 3.3 Maximum responses for simulation with Newhall Fire record
Site Lateral
(in)
Displacement
Vertical
(in)
Rotation
(rad)
Lateral
(kip)
Forces
Vertical
(kip)
Moment
(kip-in)
UIUC (Pier 1) 2.481 0.187 6.076E-04 124.19 396.32 7730.75
Lehigh (Pier 3) 2.540 0.018 5.968E-04 125.35 170.04 5920.86
0 0
Figure 3.7 Lateral displacement history for Piers 1 and 2
Table 3.4 Comparison between UIUC and Lehigh piers
Significant shear cracking
UIUC Lehigh
Shear failure
UIUC Lehigh
Secondary failure
UIUC
Step
Displ. X (in)
727
1.222
Force X (kips) 111.47
948
1.942
124.48
941
1.605
124.19
954
2.142
125.35
1208
2.468
98.83
18 of 26

MISST
(a) Near view (b) Distant view
Figure 3.8 Shear cracking at UIUC
(a) Near view (b) Distant view
Figure 3.9 Initial shear failure at UIUC
19 of 26

MISST
(a) Near view (b) Distant view
Figure 3.10 Shear cracking at Lehigh
(a) Near view (b) Distant view
Figure 3.11 Shear failure at Lehigh
20 of 26

MISST
(a) Near view (b) Distant view
Figure 3.12 Secondary shear-axial failure at UIUC
Figure 3.13 shows the hysteretic response of both piers. Inspection of the hysteretic response (Figure
3.13) and of the lateral displacement history (Figure 3.7) clearly demonstrates the interaction amongst the
geographically distributed modules and sites. Following the initial shear failure at UIUC and its associated change in stiffness, redistribution of forces within the structure drove the second shear failure which occurred at the Lehigh site. Thus, it can be assured full interaction amongst the geographically distributed modules was occurring.
150 150
100
50
0
-50
100
50
0
-50
-100
-1 0 1 2 3
-100
-1 0 1 2
Displacement X (in)
(a) Pier 1 - UIUC
Displacement X (in)
(b) Pier 3 - Lehigh
Figure 3.13 Lateral force-displacement relationship (Newhall Fire record)
3
21 of 26

MISST
Provided herein is a discussion of the major findings stemming from not only the culminating MISST simulation but also from the verification studies carried out as part of the investigation. Also included is discussion of the contributions and impacts MISST has made upon the NEES and earthquake engineering communities.
4.1
M
AJOR
F
INDINGS
4.1.1
Verification studies
Due to the complexity of the problem under question, an incremental approach was utilized throughout the investigation. This approach resulted in two distinct preliminary verification studies from which useful conclusions can be drawn.
The first verification study utilized identical partitioning to that used in the final simulation, however substituted analytical representations for the two experimental piers. This study demonstrates the successful completion of a distributed, although not hybrid, simulation. Thus, it also demonstrates successful communication and study amongst geographically distributed researchers. Furthermore, the transition from this fully analytical simulation to following hybrid simulation went flawlessly. Therefore it clearly demonstrates the ease with which UI-SIMCOR is able handle transitions amongst the module’s type and the conversion from analysis to experiment.
The second verification study utilized linear, but experimental, representations of the two piers. Not only did this simulation ensure and harden the communication and loading protocol at the three sites, but it also demonstrated an additional and successful application of the Mini-MOST system.
4.1.2
Distributed Hybrid Simulations
Three distinct distributed hybrid simulations were carried out as part of this investigation: namely, a small amplitude simulation with the Santa Monica record, a large amplitude simulation with the Santa Monica record, and a large amplitude simulation with the Newhall Fire record. The first two of these simulations continued the incremental approach utilized throughout the investigation. Although these two were vital to the development and progress of the study, derived findings are similar to those of the verification studies above. Thus, no further attention will be given to the simulations which used the Santa Monica record. Conversely, the final large amplitude test with the Newhall Fire record was the culmination of the investigation and thus several important observations and conclusions are discussed below.
The coordination and communication of the three sites, UIUC, Lehigh, and RPI, for the five component hybrid and geographically distributed NEES simulation worked seamlessly. Despite their brittle nature, the simulation was able to continue on well past the initial shear failures observed at both the UIUC and
Lehigh sites. Furthermore, the redistribution of forces between the two sites with the bridge piers as either of the two suffered partial failure shows that full interaction was taking place between the distant sites. Thus the simulation system which includes all NEESgrid components, UI-SIMCOR, the analytical modules, and all experimental equipment and components at both UIUC and Lehigh proved to be quite effective and robust.
22 of 26

MISST
Moreover, the failure modes obtained are similar to those in the prototype observed following the 1994
Northridge earthquake (Figure 4.1). Thus, the observed and complex field behavior of a complicated
structural system was successfully reproduced. Not only does this create an opportunity to address or propose new design approaches for bridge structures, but also clearly demonstrates how NEES can be applied to address problems which have previously been unapproachable to the earthquake engineering community. Thus, MISST serves as both proof test and a benchmark for the NEES community.
(a) Pier 1 - UIUC (b) Santa Monica Freeway (c) Pier 3 - Lehigh
Figure 4.1 Comparison of observed behavior
4.2
C LOSURE
MISST was intended to serve as a realistic test bed application in which to verify and extend all components of NEESgrid and as well as all components of the sites taking place in the distributed simulation. Concurrently, it was intended to extend the state-of-the-art in earthquake engineering simulation. The successful completion of distributed hybrid simulations on May 16 and 17, 2006 not only demonstrate that these objectives were achieved, but also showcase the power and advantages offered by the George E. Brown, Jr. Network for Earthquake Engineering Simulation. Distributed hybrid simulation enables the integration of testing capabilities at multiple NEES sites, the strength of existing computational models, and the expertise of individual research groups to explore previously unapproachable problems.
23 of 26

MISST
Broderick, B.M., and Elnashai, A.S., (1995), “Analysis of the failure of Interstate 10 freeway ramp during the Northridge earthquake of 17 January 1994”, Earthquake Engineering and Structural Dynamics,
Vol. 24, 189-208.
Combescure, D., and Pegon, P., (1997), “
α
-operator splitting time integration technique for pseudodynamic testing. Error propagation analysis”, Soil Dynamics and Earthquake Engineering, Vol.
16, 427–443.
Elnashai, A.S., Papanikolaou, V., and Lee, D., (2004), “Zeus-NL – A System for Inelastic Analysis of
Structures”, Mid-America Earthquake Center, University of Illinois at Urbana-Champaign,
CD-Release 04-01.
Elnashai A.S., Spencer, B .F., Kuchma, D, Ghaboussi, J., Hashash, Y., and Gan, Q., (2004), “Multi-axial full-scale sub-structured testing and simulation (MUST-SIM) facility at the University of Illinois at
Urbana-Champaign”, Proceedings of 13th World Conference on Earthquake Engineering, Vancouver,
Canada, Paper No. 1756.
Kwon, O.S., Nakata, N., Elnashai, A.S., and Spencer, B.F., (2005), “A framework for multi-site distributed simulation and application to complex structural systems”, Journal of Earthquake Engineering, Vol. 9,
No. 5, 741–753.
Nakata, N., Yang, G., and Spencer, B. F., (2004), “System Requirements for Mini-MOST
Experiment”,University of Illinois at Urbana-Champaign, Department of Civil and Environmental
Engineering, MUST-SIM Facility, http://nees.uiuc.edu/ .
Spencer, B. F., Elnashai, A.S., Nakata, N., Seliem, H., Yang, G., Futrelle, J., Glick, W., Marcusiu, D., Ricker,
K., Finholt, T., Horn, D., Hubbard, P., Keahey, K., Liming, L., Zaluzec, N., Pearlman, L. and Stauffer,
E., (2004), "The MOST experiment: Earthquake engineering on the grid," Technical Report
NEESgrid-2004-41, http://it.nees.org/ .
24 of 26

MISST
6.1
P ARTICIPANTS
The success of the MISST was due to the sustained efforts of many individuals. Provided here is a list of all those involved. Their efforts are greatly appreciated.
 The University of Illinois at Urbana-Champaign
Bill Spencer
Amr Elnashai
Dan Kuchma
Sung Jig Kim
Curtis Holub
Narutoshi Nakata
Guangqiang Yang
Quan Gan
Jianfeng Tang
Oh-Sung Kwon
Nick Burdette
Tom Nagle
 Lehigh University
James Ricles
Sougata Roy
Tommy Marullo
Chad Kusko
Peter Bryan
 Rensselaer Polytechnic Institute
Tarek Abdoun
Hassan Radwan
 Shore Western Manufacturing
Matthew Schroeder
Bill Franks
25 of 26

MISST
6.2
C ONSTRUCTION D RAWINGS
#9@3" cs. for the Top bar
Pipe with 2.5" dia
#3 stirrup ties in both sides
#4@3" cs. for the Bottom bar
#3 Stirrups.
9" lap splices
77" Circumference with lap splices
#4@3" cs. for the top bar
#9@3" cs. for the bottom bar
(a) Pier 1 (UIUC)
Figure 6.1 Elevation view of piers
(b) Pier 3 (Lehigh)
26 of 26