ABSTRACT 
 
 
 
 
Title of Document: RFID SENSOR-DRIVEN STRUCTURAL 
CONDITION MONITORING IN 
INTEGRATED BUILDING INFORMATION 
MODELING ENVIRONMENT  
  
 Linjia Bai,  
Doctor of Philosophy, 2013 
  
Directed By: Yunfeng Zhang, Associate Professor 
Department of Civil and Environmental 
Engineering  
 
 
 
Steel structures are very popular structural forms for both buildings and bridges. 
Under extreme loading such as strong earthquakes or winds, structural members in 
steel structures could fail in buckling or yielding. In buildings, structural members are 
often hidden behind fire-proof coating and drywall, and thus buckling or yielding in 
steel members are very difficult to detect, often requiring removal of coverings and 
thus time consuming and costly. Fast and accurate assessment of damage conditions 
is important to the occupant safety and uninterrupted use. This research presents a 
method for integrating Building Information Modeling (BIM) and Radio Frequency 
Identification (RFID) based wireless strain sensor technologies to enable automated 
structural condition assessment. BIM can serve as an ideal 4-dimensional graphical 
computing environment for integrating structural health monitoring (SHM) with 
management practice, which often involves considerable amounts of sensor data of 
different modalities and changes of structural condition over time. To detect the 
 
  
exceedence of preset threshold strain levels, this research custom designed a novel 
breakage-triggered (BT) strain sensor for detection of threshold strain levels. This BT 
strain sensor uses RFID tag for wireless communication with RFID reader. With a 
special design, the sensor breaks at pre-set strain level and the connected RFID tag is 
activated. By altering its configuration to include multiple threshold strain levels, 
actual strain at the monitored location can be estimated after hazardous events. This 
system can rapidly identify and locate the spot where pre-set strain or crack has 
happened and the corresponding structure element of the BIM model is automatically 
highlighted to provide alert signal. Such damage spots require immediate action after 
hazardous events such as earthquake or extreme wind. The RFID tag can be scanned 
from a distance up to several meters by using an RFID reader. This system is highly 
automated, low power consumption, and low cost.  
A custom developed middleware was developed to enable automated sensor data 
entry into BIM environment and thus avoid possible human related errors. The BIM 
environment allows for capturing, organizing, processing (cleansing, fusion, mining, 
etc.), visualizing the sensor data from the monitoring system, updating relevant 
models and running finite element analysis for structural response and reliability 
computation as well as structural health prognosis. Process (inspection scheme) and 
work-flow automation can be realized. Therefore, the proposed method would lead to 
more user-friendly, highly automated and more economical way for structural 
condition assessment and SHM information management.   
 
  
 
 
RFID SENSOR-DRIVEN STRUCTURAL CONDITION 
MONITORING IN INTEGRATED BUILDING 
INFORMATION MODELING ENVIRONMENT  
 
 
 
By 
 
 
Linjia Bai 
 
 
 
 
 
Dissertation submitted to the Faculty of the Graduate School of the  
University of Maryland, College Park, in partial fulfillment 
of the requirements for the degree of 
Doctor of Philosophy 
2013 
 
 
 
 
 
 
 
 
 
 
 
Advisory Committee: 
Associate Professor  Yunfeng Zhang, Chair/Advisor 
Professor  Amde M. Amde   
Professor  M. Sherif Aggour 
Research Professor  Chung C. Fu 
Professor  Sung W. Lee 
 
  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
? Copyright by 
Linjia Bai 
2013 
 
 
 
 
 
 
 
 
 
 
 
 
 ii 
 
Acknowledgements 
The research presented in this dissertation was conducted at Department of Civil and 
Environmental Engineering, University of Maryland, College Park, Maryland. 
I would like to take this opportunity to express my sincerest appreciation to my 
advisor, Dr. Yunfeng Zhang, who has always provided guidance and direction to me. 
And I also appreciate the contributions of my other committee members, Dr. Amde 
M. Amde, Dr. M. Sherif Aggour, Dr. Chung C. Fu and Dr. Sung W. Lee.  
I would also like to express my gratitude to Department of Civil and Environmental 
Engineering for providing me with the great experience of teaching assistantship, and 
for the Structural Engineering Laboratory facilities in which the experimental tests of 
this study were conducted. 
Financial support from National Science Foundation and US Department of 
Transportation?s RITA Program is greatly appreciated.  
My friends, Zhen Li, Changjiang Zhou, Feng Shi and Ruipeng Li have been very 
helpful to me during my study. I would like to extend thanks to my friends at 
University of Maryland. 
Finally, I am deeply indebted to my family without whom I would never have 
accomplished my study. I cannot be grateful enough for their support. This 
dissertation is dedicated to my parents, Suqi Bai and Yuan Zhang, my dear wife, Hua 
Ge and my daughter, Emilly Bai. 
 
 iii 
 
Table of Contents 
 
 
Acknowledgements ....................................................................................................... ii 
Table of Contents ......................................................................................................... iii 
List of Tables ............................................................................................................... vi 
List of Figures ............................................................................................................ viii 
 
Chapter 1:  Introduction ................................................................................................ 1 
1.1. General ............................................................................................................... 1 
1.2. Objective ............................................................................................................ 2 
1.3. Dissertation organization ................................................................................... 4 
 
Chapter 2: BIM for Structural Health Monitoring (SHM)............................................ 7 
2.1.  Introduction on BIM ......................................................................................... 7 
2.2.  Current use of BIM in Architecture, Engineering, and Construction (AEC) 
industries ................................................................................................................... 8 
2.3.  Current research on RFID assisted BIM technology ........................................ 9 
2.4.  Potential of integrating BIM with SHM ......................................................... 10 
 
Chapter 3: Breakage-triggered (BT) Strain Sensor for Detection of Threshold Strains
 ..................................................................................................................................... 15 
3.1.  Introduction ..................................................................................................... 15 
3.2.  RFID-based strain sensor: overview ............................................................... 16 
3.3.  Basics of RFID antenna design ....................................................................... 20 
3.3.1.  Impedance match between RFID IC chip and antenna ............................ 21 
3.3.2.  RFID antenna tuning parameters ............................................................. 22 
3.4.  Breakage-triggered strain range sensor: device mechanics ............................ 24 
3.4.1.  Mechanics of BT strain sensor ................................................................. 24 
3.4.2.  Brittle bar design ...................................................................................... 25 
3.4.3.  Numerical simulation based design validation ........................................ 27 
3.4.4.  Experimental characterization test ........................................................... 29 
 
 iv 
 
3.5.  BT strain sensor design for detection of multiple threshold strains ................ 31 
3.6.  Pilot study of buckling detector for tube......................................................... 32 
3.7.  Summary ......................................................................................................... 33 
 
Chapter 4: Sensor-Driven Structural Health Monitoring Integrated with Building 
Information Modeling ................................................................................................. 52 
4.1.  Introduction ..................................................................................................... 52 
4.2.  BIM for SHM portal ....................................................................................... 56 
4.3.  Case study ....................................................................................................... 59 
4.4.  Summary ......................................................................................................... 65 
 
Chapter 5:  Resilient Structure Design with Fuse Members ....................................... 73 
5.1.  Introduction ..................................................................................................... 73 
5.1.1.  Background of steel structural framing .................................................... 73 
5.1.2.  Concept of force limiting device (FLD) .................................................. 74 
5.1.3.  Concept of concentrated damage zones ................................................... 76 
5.2.  Analysis procedure for dynamic collapse simulation ..................................... 77 
5.3.  Numerical analysis model ............................................................................... 80 
5.3.1.  Analytical model for dynamic member buckling .................................... 80 
5.3.2.  Analytical model for Force Limiting Device (FLD)................................ 82 
5.4.   Seismic analysis ............................................................................................. 83 
5.4.1.  Nonlinear static analysis .......................................................................... 86 
5.4.2.  Nonlinear time history analysis ............................................................... 87 
5.5.  Results and discussions ................................................................................... 88 
5.6.  Conclusions ..................................................................................................... 90 
 
Chapter 6: Case Study I: Demonstration of WSCA System on Truss Structure ...... 102 
6.1.  Introduction ................................................................................................... 102 
6.2.  Experimental test: hybrid simulation ............................................................ 103 
6.2.1. Software architecture for hybrid simulation ........................................... 104 
6.2.2. Hybrid simulation test setup ................................................................... 105 
6.3.  Component validation test on truss member ................................................. 108 
 
 v 
 
6.4.  System validation test on prototype truss ..................................................... 110 
6.5.  Summary and conclusions ............................................................................ 114 
 
Chapter 7: WSCA Case Study II: Eccentrically Braced Frame (EBF) with Shear Link
 ................................................................................................................................... 131 
7.1. Introduction .................................................................................................... 131 
7.2. BT strain sensor design .................................................................................. 132 
7.2.1. Prototype EBF modeling......................................................................... 132 
7.2.2. Shear link model in OpenSees ................................................................ 133 
7.2.3. Elements, materials and loads ................................................................. 136 
7.2.4. Eigen value analysis ................................................................................ 137 
7.2.5. Nonlinear static analysis ......................................................................... 137 
7.2.6. Nonlinear time history analysis .............................................................. 138 
7.3. FE analysis of shear link ................................................................................ 140 
7.3.1. FE model description .............................................................................. 140 
7.3.2. Orientation of BT strain sensor ............................................................... 141 
7.4. Experimental validation test .......................................................................... 145 
7.5. Summary and conclusions ............................................................................. 148 
 
Chapter 8: Summary, Conclusions and Recommendations ...................................... 180 
8.1. Research summary ......................................................................................... 180 
8.2. Conclusions .................................................................................................... 182 
8.3. Recommendations for future works ............................................................... 185 
 
Bibliography ............................................................................................................. 186 
 
  
 
 vi 
 
 
List of Tables 
 
Table 3- 1. Critical displacements for each brittle bar design derived from linear FE 
analysis (gage length 30 mm) ..................................................................................... 35 
Table 3- 2. Experimental characterization test results of the critical displacement of 
BT strain sensor for detection of single threshold strain level (unit: mm) ................. 35 
Table 3- 3. Experimental characterization test results of BT strain sensor capable of 
detecting two threshold strain levels (unit: mm)......................................................... 36 
Table 3- 4. Summary of the test results from epoxy paint patch tests ........................ 36 
 
Table 5- 1. Section sizes of the arch structure?s tubular member ............................... 92 
Table 5- 2. Natural frequency of the steel arch structure with and without fuse 
members (unit = Hz) ................................................................................................... 92 
Table 5- 3. Numbers of member failures under each earthquake records .................. 92 
Table 5- 4. Earthquake ground motion records used in this study (scaled to the 
maximum credible earthquake for Los Angeles, California) ...................................... 93 
 
Table 6- 1. Measured and design values of the trigger strain level for the two BT 
strain sensors on the PVC tube subjected to axial load ............................................ 115 
Table 6- 2. Measured and design values of the trigger strain level for the two BT 
strain sensors on the PVC tube in hybrid simulation test ......................................... 115 
 
Table 7- 1. EBF member sections and link properties (Richards and Uang 2006) .. 151 
Table 7- 2. Maximum link rotation angle (?a) in shear links of the prototype EBF 
building subjected to design basis earthquakes (DBE) ............................................. 151 
Table 7- 3. Maximum link rotation angle in shear links (?a) of the prototype EBF 
building (unit: radians) subjected to maximum credible earthquakes (MCE) .......... 152 
Table 7- 4. Maximum link rotation angle in shear links (?p) of the prototype EBF 
building (unit: radians) subjected to DBE earthquakes ............................................ 153 
 
 vii 
 
Table 7- 5. Maximum chord rotation in shear links (?p) of the prototype EBF 
building subjected to MCE earthquakes ................................................................... 154 
Table 7- 6. Distance and axial deformation between the BT strain sensor (inner) 
attachment points when shear link rotates by 0.08 rad (unit: cm) ............................ 154 
 
  
 
 viii 
 
List of Figures 
Figure 3- 1. RFID-based wireless strain or crack sensing devices developed by other 
researchers: (a) Morita and Noguchi (2008); (b) Chin et al.  (2009); (c) Yi et al. 
(2011) .......................................................................................................................... 37 
Figure 3- 2. Illustration of antenna tuning parameters ................................................ 37 
Figure 3- 3. Schematics of modifying RFID tag by adding by-pass to IC chip ......... 38 
Figure 3- 4. Configuration of RFID tag with different bypass locations .................... 38 
Figure 3- 5. Effect of adding bypass to IC chip on responding output power threshold 
level of RFID reader (see Figure 3-6 for explanation of different cases) ................... 39 
Figure 3- 6. Overall structure of the proposed threshold strain detection system with 
breakage-triggered (BT) strain sensors ....................................................................... 39 
Figure 3- 7. Schematic of BT strain sensor design ..................................................... 40 
Figure 3- 8. FEM model of BT strain sensor (unit: mm) ............................................ 40 
Figure 3- 9. Contour plot of first principal strain in the BT strain sensor (model 
H9W2, gage length = 30 mm) at breakage ................................................................. 41 
Figure 3- 10. Contour plot of first principal strain in the BT strain sensor (model 
H8W3, gage length = 30 mm) at breakage ................................................................. 41 
Figure 3- 11. Contour plot of first principal strain in the BT strain sensor (model 
H6W4, gage length = 30 mm) at breakage ................................................................. 42 
Figure 3- 12. Distribution of first principal strain in the model corresponding to 1% 
sensor strain in the BT strain sensor ........................................................................... 43 
Figure 3- 13. Test setup used for experimental characterization of BT strain sensors 
using a hand-driven precision slide............................................................................. 44 
Figure 3- 14. Brittle bar specimens made of acrylic sheet and epoxy paint ............... 44 
Figure 3- 15. Photos of cracked brittle bar (acrylic sheet) after test with vector plot of 
first principal stress superimposed .............................................................................. 46 
Figure 3- 16. Photos of cracked brittle bar (epoxy paint) after test ............................ 47 
Figure 3- 17. Histogram of critical displacement test results of BT strain sensors 
(acrylic sheet) with single threshold strain level ......................................................... 48 
 
 ix 
 
Figure 3- 18. Histogram of critical displacement test results of BT strain sensors 
(epoxy paint) with single threshold............................................................................. 48 
Figure 3- 19. Photo of BT strain sensors with two threshold strain levels (H9W2 and 
H6W4)......................................................................................................................... 49 
Figure 3- 20. Histogram of critical displacement test results of BT strain sensors 
(acrylic sheet) with two threshold strain levels........................................................... 49 
Figure 3- 21. Epoxy paint patch on paint bucket ........................................................ 50 
Figure 3- 22. Experiment setup for epoxy paint patch test ......................................... 50 
Figure 3- 23. Breakage of epoxy paint patch in test #1 .............................................. 51 
 
Figure 4- 1. Schematic of SHM framework integrated to BIM .................................. 66 
Figure 4- 2. Schematics of the workflow in the BIM based SHM system ................. 66 
Figure 4- 3. BIM database tables in MS Access ......................................................... 67 
Figure 4- 4. Pictures of the prototype steel highway bridge in case study ................. 68 
Figure 4- 5. BIM model of the prototype steel highway bridge ................................. 69 
Figure 4- 6. Time history and frequency spectrum of sensor data displayed in BIM 
environment ................................................................................................................ 69 
Figure 4- 7. ?Run Opensees? add-in for FE analysis called in BIM environment ..... 70 
Figure 4- 8. ?PlotBeamMemberMoment? add-in called in BIM environment ........... 70 
Figure 4- 9. ?PlotMidSpanDisp? add-in called in BIM environment ......................... 71 
Figure 4- 10. Modal analysis results from FE analysis program displayed in BIM 
environment ................................................................................................................ 71 
Figure 4- 11. Element-level condition state displayed for selected element in BIM 
environment ................................................................................................................ 72 
 
Figure 5- 1. Configuration of brace model considering mid-length mass .................. 94 
Figure 5- 2. Hysteresis model of buckled member, conventional fuse member (FM) 
and self-centering FM ................................................................................................. 94 
Figure 5- 3. Schematics of steel arch structure and location of fuse members ........... 95 
Figure 5- 4. First three mode shapes of the arch structure .......................................... 96 
 
 x 
 
Figure 5- 5. Locations of member failure under 20 earthquake records. (Numbers 
indicate the member of failure occurrences in total 20 earthquakes.) ........................ 96 
Figure 5- 6. Static pushover curve of the arch structure ............................................. 97 
Figure 5- 7. Pseudo-Acceleration response spectra of the seismic records in Table 5-4 
(viscous damping ratio=0.05) ..................................................................................... 97 
Figure 5- 8. Peak and residual arch top displacement of the arch structure ............... 98 
Figure 5- 9. Peak base shear and acceleration response of the arch structure ............ 99 
Figure 5- 10. Displacement response of the arch structure under LA11 record: 
horizontal displacement at arch top node U3 ............................................................ 100 
Figure 5- 11. Acceleration response of the arch structure under LA11 record: 
Horizontal acceleration at arch top node U3 ............................................................. 101 
 
Figure 6- 1. 3D model of the prototype truss structure for hybrid simulation test ... 116 
Figure 6- 2. Schematics of hybrid testing software framework (Schellenberg et al 
2006a) ....................................................................................................................... 116 
Figure 6- 3. Schematics of a hybrid testing facility for collapse simulation under wind 
loading....................................................................................................................... 117 
Figure 6- 4.  Schematic of test setup for component validation test on PVC tube ... 118 
Figure 6- 5. Dimensions of BT strain sensor (H6W4 as an example) ...................... 119 
Figure 6- 6. Schematic of BT strain sensor............................................................... 120 
Figure 6- 7. Comparison of measured strain of the PVC tube and triggering time point 
of the BT strain sensors............................................................................................. 122 
Figure 6- 8. Close-up view of the cracks in the BT strain sensors after triggering .. 122 
Figure 6- 9. Prototype truss model for hybrid simulation test .................................. 123 
Figure 6- 10. Hybrid simulation test setup for system validation of WSCA on truss 
structure..................................................................................................................... 124 
Figure 6- 11. Test results of the first hybrid simulation test on truss structure (HS Test 
# 1, Strain) ................................................................................................................. 125 
Figure 6- 12. Test results of the first hybrid simulation test on truss structure (HS Test 
# 1, H6W4)................................................................................................................ 126 
 
 xi 
 
Figure 6- 13. Test results of the first hybrid simulation test on truss structure (HS Test 
# 1, H9W2)................................................................................................................ 127 
Figure 6- 14. Test results of the first hybrid simulation test on truss structure (HS Test 
# 2, Strain) ................................................................................................................. 128 
Figure 6- 15. Test results of the first hybrid simulation test on truss structure (HS Test 
# 2, H6W4)................................................................................................................ 129 
Figure 6- 16. Test results of the first hybrid simulation test on truss structure (HS Test 
# 2, H9W2)................................................................................................................ 130 
 
Figure 7- 1. OpenSees model of the prototype EBF structure .................................. 155 
Figure 7- 2. Schematics of shear link model............................................................. 155 
Figure 7- 3. Combined behavior of parallel springs used to model shear hinge 
(Richards and Uang 2006) ........................................................................................ 156 
Figure 7- 4. Illustration of link rotation angle  ?a and ?p......................................... 156 
Figure 7- 5. First three mode shapes of the 3-story prototype EBF building ........... 157 
Figure 7- 6. Push-over curve of the prototype EBF building ................................... 158 
Figure 7- 7. Ground acceleration time history of the LA01 earthquake record (DBE 
level) ......................................................................................................................... 158 
Figure 7- 8. Roof displacement time history of the prototype EBF building subjected 
to LA01 earthquake record (DBE level) ................................................................... 159 
Figure 7- 9. Roof acceleration time history of the prototype EBF building subjected to 
LA01 earthquake record (DBE level) ....................................................................... 159 
Figure 7- 10. Typical shear link (L1L) rotation angle (?a and ?p) time history of the 
EBF building subjected to LA01 earthquake record (DBE level) ............................ 160 
Figure 7- 11. Axial force time history of the shear link L1L in the prototype EBF 
building subjected to LA01 earthquake record (DBE level) .................................... 161 
Figure 7- 12. Hysteresis (shear force vs. ?a) loop of L1L shear link subjected to 
LA01 earthquake record (DBE level) ....................................................................... 161 
Figure 7- 13. Schematics of the FEM model of 1/3-scaled shear link in ANSYS v.13 
(only part of the link is modeled; length is not 1/3 scaled to provide more space for 
sensor installation) .................................................................................................... 162 
 
 xii 
 
Figure 7- 14. Vector plot of first principal strain on the web at link rotation angle of 
0.08 rad. .................................................................................................................... 162 
Figure 7- 15. Vector plots of first principal strain from linear elastic analysis (by 
applying a 1-kips load).............................................................................................. 163 
Figure 7- 16. FEM models with integrated BT strain sensors oriented at different 
angles to the longitudinal axis (z axis) of shear link ................................................. 165 
Figure 7- 17. Schematics of the BT strain sensor attachment point movement when 
link rotation angle changes from zero to 0.08 rad .................................................... 166 
Figure 7- 18. Photo of the BT strain sensor oriented at 45 deg. on the shear link 
specimen web ............................................................................................................ 166 
Figure 7- 19. Evolution of maximum value of the first principal strain in the brittle 
bar of BT strain sensors with increasing link rotation angle (?a) ............................. 168 
Figure 7- 20. Contour plots of first principal strain in the BT strain sensors (with two 
brittle bars: H6W4 and H9W2) when link rotation angle reaches 0.03 rad. ............. 170 
Figure 7- 21. Schematic of test setup for validating the WSCA system on shear link: 
(a) shear link specimen; (b) test setup with actuator and reaction frame .................. 171 
Figure 7- 22. Close-up view of BT strain sensor installed on shear link web .......... 172 
Figure 7- 23. Test setup for validating the WSCA system on shear link .................. 173 
Figure 7- 24. Premature failure of shear link specimen due to local flange fracture 174 
Figure 7- 25. Test setup for compression test of stub column specimen .................. 175 
Figure 7- 26. Close-up view of BT strain sensors on stub column specimen ........... 176 
Figure 7- 27. Axial load applied to stub column specimen (vertical line indicates 
triggering of BT strain sensor) .................................................................................. 177 
Figure 7- 28. Axial stress (load divided by cross-section area vs. strain .................. 177 
Figure 7- 29. Measured strain from two metal foil strain gages in stub column 
specimen ................................................................................................................... 178 
Figure 7- 30. Screen shot of BIM (Revit structure) user interface ........................... 178 
Figure 7- 31. Photo of buckled stub column specimen ............................................. 179 
 
 1 
 
Chapter 1:  Introduction 
 
 
1.1. General 
There is a strong and growing interest in on-line structural health monitoring (SHM) 
systems among engineers, researchers, and decision makers in civil engineering 
applications. On-line SHM enables automatic collection of vital information about 
civil infrastructures? operating condition and in the aftermath of a catastrophic event 
such as strong earthquake, its ability to continue to carry the design load. Sensor data 
from SHM systems provides the base data for real-time updating of the structural 
condition and reliability. For example, probabilistic health prognosis method uses 
sensor data for system reliability updating, which can then be utilized for risk analysis 
and maintenance scheduling of engineering systems. However, the incompatibilities 
in user interface, data format, and computing environment in many current (mostly 
custom built) SHM software applications make it difficult if not impossible for most 
users to take advantage of the full set of resources available for SHM. With emerging 
monitoring system featuring continuous and online data collection, 3-dimensional 
graphical computing environment needs to be developed to enable effective 
communication of information between users at different levels in sensor-driven 
SHM practice.   
BIM has seen increasing popularity and implementation in the construction industry 
to produce data-rich models of buildings and structures (Wong et al. 2011; Eastman 
et al. 2011). In the USA, several government agencies including General Services 
 
 2 
 
Administration (GSA) have been actively promoting the BIM implementation for all 
major projects (GSA 2007). The U.S. GSA began using BIM in 2003. In July 2009, 
the state of Wisconsin issued guidelines to AE firms requiring the use of BIM on all 
projects totaling more than $5 million and new construction projects of more than 
$2.5 million (Beck 2012). BIM provides an integrative visual platform for 
information exchange between different modules of the system as well as a shared, 
coordinated repository for the structure?s lifecycle information. This addresses the 
needs for improvement in structural condition monitoring/assessment productivity, 
especially after hazardous events, when resources such as experts and timeline 
requirement are tight. The models are continually updated to reflect project-related 
adjustments. Accurate and comprehensive information, such as the digital 
information from a BIM model about a structure, its construction, its life-cycle 
maintenance and loading data, can be very beneficial to SHM and decision making in 
infrastructure operation. Therefore, BIM model can be leveraged as an integrative 
platform for other professional services such as sensor-driven SHM applications. 
This chapter presents the background information related to BIM with SHM system. 
It is followed by the motivation and objectives of the research and the organization of 
the dissertation. 
 
1.2. Objective 
This research aims to integrate the RFID and BIM technologies with SHM for an 
automated and graphical approach to structural condition monitoring, prognosis and 
health management to facilitate the implementation of resilient structure concept. 
 
 3 
 
Ability of rapid structural condition assessment after major hazardous events reduces 
the time to recovery and thus increases the resilience in disaster recovery. SHM is a 
cross-disciplinary field of emerging importance that involves different disciplines 
such as structural analysis, sensor technology, data fusion, and etc. In the past, ad hoc 
SHM systems with custom developed virtual user interface have been built for a 
variety of projects. Thus lack of universal virtual user interface and computing 
environment poses an impediment to wide adoption of SHM system and 
unnecessarily increases the cost to build SHM system for a project. BIM presents 
itself as a perfect solution to address this need. Many important infrastructure projects 
already have BIM model right from the start of the project design and construction 
phase; the availability of such BIM models can be taken advantage of by integrating 
the BIM with sensors and SHM technologies. BIM can integrate all types of condition 
assessment information collected in the life cycle of civil infrastructures. For 
example, information about sensors and monitoring equipments - model number, 
specification data, calibration history, can be readily accessed by selecting an object 
in the BIM model. The data displayed will reside and be maintained in a database.  In 
this graphical computing environment, SHM data can be shared in the BIM 
environment with other disciplinary applications in the infrastructure?s life cycle. 
There are many issues that come along with this new capability and the key modules 
for a functional BIM computing environment for SHM application are examined in 
this research. Specifically, the proposed BIM for SHM portal is linked to a database, 
finite element analysis software, providing the visual data displaying mechanisms that 
make the SHM information easily understandable and actionable. 
 
 4 
 
1.3. Dissertation organization 
The dissertation is organized into eight chapters including this introduction chapter. 
The other seven chapters are organized as follows. 
? Chapter Two briefly introduces the concept and characteristics of BIM, 
followed by a review of current adoption of BIM in Architecture, 
Engineering, and Construction (AEC) Industries and current research in RFID 
assisted BIM technology.  The potential of integrating BIM with RFID based 
wireless sensor and SHM as well as the challenges and important issues in 
their integration are discussed. A general framework for sensor-driven SHM 
integrated into the BIM environment is presented in this chapter. 
? Chapter Three presents a novel breakage-triggered (BT) strain sensor for 
detection of preset threshold strain levels. The BT strain sensor uses RFID tag 
for wireless communication with BIM-based SHM software program via a 
RFID reader device and a custom developed middleware. The principle of 
operation for the BT strain sensor is described and finite element (FE)  
analysis results is presented to gain insight on how to tune the design 
parameters to optimize its performance in strain sensing. The performance of 
the BT strain sensor is validated and characterized with experimental tests. 
Different threshold strain levels can be set to the BT strain sensor design by 
adjusting the geometry and size of the brittle bar in the sensor.  Two types of 
brittle materials, acrylic sheet and epoxy paint, were tested for use in the BT 
strain sensor. Additionally, use of epoxy paint patch to detect local buckling in 
steel structures was explored in this study and pilot test results are presented 
 
 5 
 
in this chapter. The BT strain sensor will be incorporated into the wireless 
structural condition assessment (WSCA) system to provide information on 
structural damage conditions such as yielding or buckling. In this chapter, the 
sensor technology components part of the BIM based WSCA system is 
established.  
? Chapter Four provides an overview of the sensor-driven SHM integrated into 
the BIM environment. Integration of various SHM components including 
RFID based sensors, FE analysis, database, condition assessment and 
interaction between these components are discussed in detail. Although many 
BIM software programs are available on the market for which the proposed 
framework can be equally well applied, Revit Structure is selected in this 
research to demonstrate the proposed technology. A case study of highway 
steel bridge located in Maryland is presented to illustrate the different aspects 
of the sensor-driven SHM system integrated into the BIM environment. This 
chapter focuses on the software part of the proposed WSCA system.   
? Chapter Five is related to the concept of resilient structure design by 
concentrating SHM on fuse members. New design trends for seismically 
resilient structure is to incorporate fuse members into structures so that only 
fuse members are to be damaged while other parts of the structure remain 
linear elastic during strong earthquakes. This resilient structure design is also 
appealing to SHM since condition assessment or monitoring work can now be 
focused on a limited number of fuse members instrumented with dense sensor 
networks. A steel arch truss with strategically selected fuse regions under 
 
 6 
 
earthquake is used as a prototype structure to demonstrate that damages could 
be concentrated within pre-selected damage zone to facilitate efficient 
condition assessment.  
? Chapter Six presents a hybrid test study on a truss structure to validate the 
WSCA system in a more realistic working condition. Test of BT strain sensor 
on a PVC tube before the hybrid test was first conducted to validate the sensor 
and debug any potential problem associated with implementing the BT strain 
sensor on a tubular member. In the hybrid test, the hardware and software 
components of the WSCA system were examined for their performance 
(individual and interaction) as a whole.  
? In Chapter Seven, the concept of resilient structure with focused SHM enabled 
by implementing the WSCA system is demonstrated using an eccentrically 
braced frame (EBF) building as a prototype structure. In EBF, shear links are 
considered as the fuse members that will develop large plastic deformation to 
provide the desired level of ductility and energy dissipation. To guide the 
design of BT strain sensors in the WSCA system implemented into the shear 
links, nonlinear time history analysis of EBF as well as FE analysis of shear 
link with BT strain sensors were performed. A reduced-scale shear link 
specimen was tested to validate and characterize the performance of the 
WSCA system in steel structures.  
? Lastly, Chapter Eight presents a summary of the research, conclusions and 
recommendations for future research work. 
 
 
 7 
 
Chapter 2: BIM for Structural Health Monitoring (SHM) 
 
2.1.  Introduction on BIM  
The concept of Building Information Model (BIM) has existed since the 1970s 
(Eastman at al. 1974). The term BIM first appeared in a paper by van Nederveen et 
al. (1992). According to the National BIM Standards Committee (NBIMS 2007), 
BIM?s overall scope is broad and can be described within the relations of three 
categorizations: (1) BIM as ?a product? of digital representation of physical and 
functional characteristics of a facility; (2) BIM as ?a collaborative process? which 
covers business drivers, automated process capabilities, and open information 
standards use to information sustainability and fidelity, and (3) BIM as ?a facility 
lifecycle management tool? of well understood information exchanges, workflows 
and procedures. 
Traditional building design was largely reliant upon two-dimensional drawings 
(plans, elevations, sections, etc.). BIM extends this beyond 3-D, augmenting the three 
primary spatial dimensions (width, height and depth - X, Y and Z) with time as the 
fourth dimension and cost as the fifth. BIM therefore covers more than just geometry. 
It also covers spatial relationships, light analysis, geographic information, and 
quantities and properties of building components. 
A three-dimensional model of building geometry alone that is used only in 
visualization is usually not a BIM (Bazjanac 2004). BIM involves representing a 
design as combinations of ?objects? ? vague and undefined, generic or product-
 
 8 
 
specific, solid shapes or void-space oriented (like the shape of a room), that carry 
their geometry, relations and attributes. Building components are represented with 
digital ?objects? that carry computable graphic data attributes. BIM design tools 
allow extraction of different views from a building model for drawing production and 
other uses. Building models are consistent and non redundant such that changes to 
component data are represented in all views of component and the assemblies. 
Parametric object is an important concept in BIM. The objects in BIM are defined as 
parameters and could be modified and related to other objects parametrically. If one 
object is modified, other related objects will also update automatically according to 
their defined relations (Eastman et al. 2011). 
2.2.  Current use of BIM in Architecture, Engineering, and Construction (AEC) 
industries 
The Architecture, Engineering and Construction (AEC) industries have long sought 
techniques to decrease project cost, increase productivity and quality, and reduce 
project delivery time. BIM offers the potential to achieve these objectives (Azhar et 
al. 2008). 
Most complex projects in the AEC industries involve multidisciplinary collaboration 
and the exchange of large building data set. Traditionally, the collaboration efforts 
across the disciplines have been based on the frequent exchange of 2D drawings and 
documents. However, during the past decade, the widespread adoption of object-
 oriented Computer-aided Design (CAD) tools has generated more interests in BIM. A 
number of BIM-compliant applications such as analysis tools, model checkers and 
facility management applications are being developed (Singh et al. 2011). 
 
 9 
 
Typically, the inter-disciplinary collaboration in ACE industries has revolved around 
the exchange of 2D drawings and documents. Even though the separate design 
disciplines have been using 3D models and applications for visualization and design 
development, the collaboration practices have remained more or less 2D-based until 
recently (Singh et al. 2011). 
The current application of BIM in the AEC industries includes: 3d visualization; 
fabrication/shop drawings; code reviews; cost estimating; construction sequencing; 
conflict, interference, and collision detection; forensic analysis; and facility 
management (Azhar 2011). 
2.3.  Current research on RFID assisted BIM technology 
RFID based systems have been used in different applications in construction and 
maintenance, such as component tracking and locating, inventory management, 
equipment monitoring, progress management, facilities and maintenance 
management, tool tracking and quality control (Motamedi and Hammad 2009). 
The use of attached RFID tags for lifecycle ID and data management in the aerospace 
sector has been proposed by Harrison and Parlikad (2006). The information to be 
stored includes unique ID and important lifecycle information on tags attached to 
aircraft parts for enhancing inspection and repair processes. They demonstrated that 
the RFID enable more efficient and complete data capture in various maintenance 
processes, with the benefit of reducing human error, and also generating information 
in electronic format for record-keeping and further analysis (Harrison and Parlikad, 
2006). 
 
 10 
 
Ergen et al. (2007) proposed using RFID tags attached to engineered-to-order (ETO) 
components and explored the technical feasibility of such system by analyzing 
component-related information flow patterns in ETO supply chains. Their 
experiments demonstrated that it is technically feasible to have intelligent 
components in construction supply chains by using RFID technology; that status 
information can be collected automatically; and that maintenance information can be 
stored and retrieved during the service life of a facility. 
Motamedi and Hammad (2009) proposed permanently attaching tags to components 
in the manufacturing stage as an integrated part of the components. Having the tags 
permanently attached, where the information on the tags is gradually updated with 
accumulated lifecycle information, is beneficial for all the stakeholders throughout 
the stages of the lifecycle, from procurement and supply chain management to 
maintenance and disposal. They also proposed to include broader data types on the 
RFID tags that are attached to building components and are spread in a building. In 
the proposed approach, the information on the tags represents chunks of the BIM as a 
distributed database. This coupling between the BIM and the RFID information 
would allow reconstructing the database of the BIM (or part of it) based on the pieces 
of information distributed in all the attached tags. 
2.4.  Potential of integrating BIM with SHM 
The advantages of 3D graphical display and model digitization make BIM a 
promising computing environment for integrating different components of an SHM 
system. The potential issues in the integration process are discussed next.   
 
 11 
 
An interchangeable data format becomes an important prerequisite for BIM-SHM 
integration due to the intensive need of data exchange in an SHM system. However, 
poor software interoperability has long been regarded as an obstacle to industry 
efficiency in general and to BIM adoption in particular (Wikipedia 2013).  
Isikdag et al. (2007) summarized five storage, sharing and exchange methods for 
different BIM models: (1) Data exchange by using physical file; (2) Data sharing 
through application programming interfaces (APIs); (3) Data sharing through central 
project database; (4) Data sharing through federated project database, and (5) Data 
sharing by web services.  
The problems of interoperability between engineering software systems have existed 
since the introduction of computer-aided design (CAD) in the 1970s (Pratt 1993). 
File-based data exchange methods such as DXF, IGES, and SAT were developed to 
exchange geometric entities between one CAD system and another. However, as 
CAD systems became more sophisticated and evolved into parametric and object-
 based modeling, the limitations of existing file-based methods became apparent. 
Parametric modeling involves rules and constraints that define how shapes are to be 
generated or modified in various conditions; CAD systems became object-based and 
included attributes and relations between objects as well as geometry (Light and 
Gossard 1982). There was no standard way of representing these aspects with existing 
geometry exchange standards (Eastman et al. 2010). 
IFC format is one of the open standards, which was developed by BuildingSMART 
(the former International Alliance for Interoperability (IAI)). In September 2007 the 
 
 12 
 
Charles Pankow Foundation awarded the Applied Technology Council (ATC) a 
research grant for the development of Industry Foundation Classes (IFCs) for the 
structural domain (ATC-75 project). The goal of the ATC-75 project (ATC 2013) was 
to develop a basis for incorporating and integrating structural design parameters, 
codes, analysis tools and methods into the Industry Foundation Class (IFC) data 
model, an open-source object-oriented exchange language (file format) that is being 
developed by the IAI for Building Information Modeling (BIM) software (ATC 
2013). Its main purpose of IFC is to provide CAD users with the ability to exchange 
data between different CAD softwares. The IFC provides a set of definitions for all 
object element types and a text-based structure for storing those definitions in a data 
file. 
To implement the IFC standard, the individual CAD software has to provide a 
command to export/import IFC format files so that could enable data exchange 
between them. Unless each information exchange within the project workflows has its 
specific contents and level of detail defined, the breadth and flexibility of the IFC 
schema leaves room for errors (Eastman et al. 2010). The IFC is so far the most 
promising standard for inter-platforms data exchange in BIM industry. Part of the 
reasons for this is because it uses a plain text file (.txt file), the only truly universal 
computer data format (Goldberg 2003). 
Despite all this, there are still some barriers that make integration between BIM and 
SHM has not been fulfilled yet. IFC takes a lowest common dominator approach to 
BIM. So even though some BIM software has IFC certified importers and exporters, 
there is invariably a loss of data fidelity (Golabchi and Kamat 2012). No single IFC 
 
 13 
 
adapter could be developed to connect all the software in SHM system. For specific 
software, an IFC adapter has to be developed to fit its complex architecture. It?s rather 
inconvenient for SHM end users to develop IFC adapter and incorporate it to each 
software components in the SHM system.  
The results of the IFC, on the other hand, have been difficult to implement and utilize 
because the use cases in which exchanges are to be made have not yet been clearly 
defined. For a building model framework schema such as IFC, lack of definition of 
specific task-oriented exchange content leads to implementation of translators that 
may be technically correct, but generate incomplete and incompatible data exchanges 
for specific tasks, because of lack of coordination regarding what specific information 
is to be included in the IFC (Eastman et al. 2010). 
A proposed solution to these problems is to define the workflows used in practice, the 
specific information exchanges they require, and define appropriate model views for 
the exchanges. The national BIM standard (NBIMS 2007) approach for resolving the 
incompatible event of information exchange is based on ?use cases?, which precisely 
define the data required in each information exchange between disciplines in 
engineering workflows (Eastman et al. 2010). The details of use case approach could 
be found in NBIMS (2007), administered by BuildingSMART within the National 
Institute of Building Sciences (NIBS). According to this ?use case? approach, in the 
case of BIM-SHM integration, the SHM industry specialists who know the workflows 
and their information flows required for information technology exchange could write 
appropriate export and import translators for their softwares.  
 
 14 
 
Following the NBIMS ?use case? approach, a middleware program termed as BIM 
for SHM portal was developed in this study following a structure similar to that of 
Isikdag et al. (2007). The middleware program utilizes the Autodesk Revit API to 
access the BIM database and automatically update and store SHM data in a database 
such as Microsoft Access.  The details of this BIM for SHM portal can be found in 
Chapter 4. 
 
  
 
 15 
 
Chapter 3: Breakage-triggered (BT) Strain Sensor for 
Detection of Threshold Strains 
 
3.1.  Introduction 
Strain measurement is essential to the condition assessment of structures since 
structural members are usually designed to work under a certain strain level to 
prevent overloading. A structural health monitoring (SHM) system should generate 
an alarm signal whenever the strain exceeds this limit. Furthermore, ductile seismic 
design emphasizes utilizing the inelastic deformation capacity of some structural 
members (e.g., fuse members) to provide the desired ductility and energy dissipation. 
Therefore, after earthquake, knowing the actual strain sustained by these deformation-
 controlled structural members is critical to judge their remaining ductility capacity. 
Strain measurements over a pre-specified level usually suggest a limit state is likely 
to be reached locally and engineers needs to be immediately alerted for timely action.  
Several types of sensors are available for strain measurement including metal foil 
strain gage, vibrating wire strain gage, fiber optic sensors, and etc. However, these 
strain sensors require power supply and signal conditioning circuit to operate. 
Moreover, with the growing use of steel-concrete composite structures in high-rise 
buildings and bridges, strain monitoring of the steel member embedded in concrete or 
hidden behind fire-proof coating or drywall poses new challenges for conventional 
strain sensor technologies. Acquiring data from these strain sensors often requires 
time-consuming setup procedure, which might hinder the post-earthquake condition 
 
 16 
 
assessment process. In addition, power blackout is very likely to happen during 
earthquakes, limited power from accompanying battery poses another service 
challenges since they have to be checked frequently to ensure proper function in the 
next earthquake event.  
To address the above need for embeddable sensors that can operate without power 
supply, a passive wireless, low-cost breakage-triggered (BT) strain sensor for 
detection of threshold strains is proposed and its performance is characterized in this 
chapter. This BT strain sensor uses RFID (radio frequency identification) technology 
in wireless communication with BIM-based SHM software program and requires no 
power in detecting the threshold strain level experienced by the structural members. 
The principle of operation of the BT strain sensor and results of experimental 
demonstration are discussed in this chapter. 
3.2.  RFID-based strain sensor: overview 
Among many new SHM technologies, wireless sensing can help to significantly 
reduce instrumentation time and system cost (Lynch and Loh 2006). Such wireless 
sensors often require battery to supply the power.  
In recent years, Radio frequency identification (RFID) has received growing 
attentions for its fast development in many fields such as asset management, access 
control, transportation and logistics. Because of its passive wireless feature, 
researchers (e.g., Yi et al. 2013) are also looking into its potential to incorporate 
RFID technology into passive wireless strain sensors. Three types of such RFID-
 based strain sensors are shown in Figure 3-1. A passive wireless antenna sensor based 
 
 17 
 
on RFID technology was recently developed by Yi et al. (2013) for strain and crack 
sensing. In the device developed by Yi et al. (2013), when the antenna experiences 
deformation, the antenna shape changes, causing a shift in the electromagnetic 
resonance frequency of the antenna. A radio frequency identification (RFID) chip is 
adopted for antenna signal modulation, so that a wireless reader can easily distinguish 
the backscattered sensor signal from unwanted environmental reflections. The RFID 
chip captures its operating power from an interrogation electromagnetic wave emitted 
by the reader, which allows the antenna sensor to be passive (battery-free). Some 
other wireless passive strain sensors have also been proposed based on inductive 
coupling between two adjacent inductors (Butler et al. 2002; Loh et al. 2008; Jia et al. 
2006; Daliri et al. 2012). The interrogation distance achieved by inductive coupling is 
usually limited to several inches, which is inconvenient for practical applications. 
Because of the large scale in civil infrastructures, many locations are either 
inaccessible or difficult to reach.  
Morita and Noguchi (2008) developed a crack detection sensor using RFID tag and 
electrically conductive paint or printed sheet. When a crack occurs, the paint or 
printed sheet is broken, resulting in an increase in resistance. Crack width can be 
estimated by the ability of an RFID reader to communicate with the RFID tag. The 
crack detection sensor was experimentally studied by bending test of a concrete 
beam.  
Chin et al. (2009) developed a gage which is embedded into a passive RFID tag, so 
that its status can be queried by a nearby RFID reader. Their RFID tag consists of an 
antenna loop and a small microchip containing a small amount of information. The 
 
 18 
 
conductor bar is broken when significant cracks form across a gage. Communication 
with the tag attached to the gage is then lost. The loss of communication is reported to 
indicate the presence of the cracks. 
Yi et al. (2011, 2013) developed strain and crack sensors using an RFID-based folded 
patch antenna. They found that when strain stretches the patch antenna, the antenna?s 
electromagnetic resonance frequency changes accordingly. They did experimental 
study to correlate the change in the antenna?s electromagnetic resonance frequency to 
the crack width on the structure element so that the crack length could be obtained by 
the change of resonance frequency. The resonance frequency change can be 
interrogated and recorded by an RFID reader.   
RFID technology operates on the basis of radio-frequency electromagnetic field to 
wirelessly (up to several meters away) interrogate an RFID tag which contains 
electronically stored information. Each RFID tag is assigned a unique ID, which 
makes it popular for identification applications such as asset management. Most 
notable is that the RFID tag does not need to be within the line of sight of the RFID 
reader and can be read through many non-metallic materials (Shepard 2005, Glover 
2006). An RFID tag reading penetration test has been reported by Morita and 
Noguchi (2008), in which several fire-resistant materials were tested, such as pearlite 
board, rock wool board, and ceramic wool. The maximum readable distance varies 
depending on material thickness, although all materials allow tags to be read up to a 
distance of 100-110 cm (Morita and Noguchi 2008). 
 
 19 
 
In choosing the proper type of RFID hardware for strain sensing or crack detection, a 
few key parameters including the read range and data transfer rate should be 
evaluated. In this study, the read range of the prototype RFID tag is the first and 
foremost important parameter to consider. The maximum read range depends on a 
variety of factors including the output power of RFID reader, antenna type and size, 
working frequency range and environmental factors. Because of civil engineering 
structures? large scale, wireless interrogation of RFID tags installed on the structures 
is often desired, especially for locations difficult to access. Therefore, RFID tag that 
can be interrogated from a distance of several meters is preferred for this application. 
Additionally, the ease of modifying the tag antenna for impedance adjustment is also 
taken into account. UHF RFID has a greater read range and broader working 
frequency range. Therefore, UHF RFID tag suffers less effect when more than one 
tag needs to be read at a time. Considering all these issues in selecting the RFID tag, a 
UHF RFID tag (model #: Avery AD-224 RFID tag) has been selected in this study. 
This research aims to develop a wireless BT strain sensor that uses RFID for wireless 
communication with BIM-based SHM system. The strain sensor can be custom 
designed to detect whether the pre-specified strain threshold level has been surpassed 
or not. It is hoped that this strain information can assist engineers to infer whether the 
safety of the structure (e.g., due to yielding or fracture) is at issue. The proposed 
breakage-triggered wireless strain sensor is low cost, low power and requires very 
little maintenance in use. By using the RFID technology, the damage status of key 
structural components such as structural fuse elements can be rapidly assessed simply 
 
 20 
 
through scanning, which is highly desired for post-earthquake inspection. The salient 
features of this RFID-based passive wireless strain sensor are summarized as follows: 
? No power requirement: By using RFID tags, no external power supply is 
needed to power the RFID tags. In this way, the cost at installation and 
maintenance would be much lower compared with other types of strain sensor. 
More importantly, power outage is very likely during or after natural disasters 
like earthquake when sensors requiring power supply to function would 
become dysfunctional and no data would be collected.     
? On-demand strain data acquisition: Passive RFID tags do not emit any data 
signals unless the user scans the tag using an RFID reader. This feature is 
desirable for structural health monitoring to avoid radio frequency signal 
interference.  
? Programmable sensor with key SHM information stored: RFID tags can be 
programmed in advance to store key SHM data such as sensor 
location/specification and structural member parameters. Passive RFID tags 
can have up to 1 kilobyte of non-volatile memory available for such data 
storage. The stored data is allowed to be altered later to include updated data.  
? Do not require line-of-sight scanners for data reading from the RFID tags.  
3.3.  Basics of RFID antenna design 
This section gives a brief discussion of fundamental RFID antenna parameters and the 
key characteristics of an antenna. This introductory description provides a necessary 
theoretical background on UHF (ultra high frequency) RFID antenna design. 
 
 21 
 
3.3.1.  Impedance match between RFID IC chip and antenna 
Passive UHF RFID tag consists of an IC chip attached directly to an antenna. Proper 
impedance match between the antenna and the chip is crucial in RFID tag design. It 
directly influences RFID system performance characteristics such as the 
reading/writing range of a tag (Nikitin and Rao 2006). It is known that a RFID 
microchip is a nonlinear load whose complex impedance in each state varies with 
the frequency and the input power (Loo et al. 2008). The chip circuitry requires 
certain minimum voltage or power to function. The variation of the chip impedance 
with power and frequency can drastically affect the performance of the RFID tag. 
Usually, in order to maximize the tag range, the antenna impedance is matched to 
the chip impedance at the minimum power level required for the chip to work. The 
better the matching, the less power is reflected on the connection between IC bumps 
and antenna pads. This will result in a maximum power transfer to the IC. 
The complex impedance of antenna and IC can be expresses as: 
?? = ?? + ??? , ?? = ?? + ???    (3-1) 
Where Za is the antenna impedance and Zc is the IC impedance. The antenna 
impedance is the parameter describing the antenna input behavior as a circuit 
element. As usual in electronic circuit design, it is important to match this antenna 
impedance (Za) to a given source impedance, which is the IC chip impedance (Zc) in 
the case of RFID applications (Balanis 2004).  
 
 22 
 
The maximum power delivered from the source to the antenna is given if the 
antenna impedance is complex conjugate to the chip impedance (Dobkin 2008): 
?? = ???    (3-2) 
By adding a circuit to the antenna to bypass the IC, the impedance of the antenna 
will be altered and the impedance matching between the antenna and IC chip is 
voided. Due to this impedance mismatching, the reading range of the RFID tag 
would be altered. Therefore, the change of antenna status can be used to signify the 
status of breakage-triggered strain sensor. The reading range of the modified RFID 
tag is normally made shorter because the antenna of RFID tag is designed to reach 
the maximum reading range at impedance matching.  
3.3.2.  RFID antenna tuning parameters 
Considering the above described impedance matching and the impact of each part 
on the matching between antenna and IC chip, three parameters can be adjusted to 
tune the antenna parameters: Loop size (Area), Dipole length (L), Dipole distance 
(D), as illustrated in Figure 3-2. To adjust the impedance of the RFID tag antenna, 
some changes need to be made to the RFID tag antenna. Clearly, these two 
parameters, L and D, are not convenient to adjust and relate to the strain 
(deformation) sustained by the structure. However, the loop size could be adjusted 
by adding a by-pass over the two terminals of the RFID IC chip, as shown in Figure 
3-3. When the by-pass is closed, the loop size is reduced to almost zero, which 
 
 23 
 
deactivates the RFID tag. After the by-pass is broken, the impedance of the RFID 
tag is restored to its normal operating value. 
To test the effect of by-pass on the RFID tag, wireless interrogation was conducted 
on the RFID tag. In this wireless interrogation test, the RFID reader swept through a 
pre-defined interrogation frequency range (902 to 927 MHz, which is allocated for 
RFID UHF bandwidth in North America) with a frequency interval of 1 MHz. At 
each interrogation frequency, the output power of RFID reader is increased at 0.1 
dBm interval until the RFID tag is detected to identify the output power threshold, 
which is the minimum power level to activate the RFID tag. Through a USB 2.0 
port, a computer interface is used to operate and retrieve measurement data from the 
RFID reader. The configuration of the RFID tag and by-passes are shown in Figure 
3-4. The bypass was added to the antenna at difference distance respectively in 
order to form different loop area. It can be seen from the interrogation results that by 
adding the bypass, the responding power threshold of the RFID tags is increased to 
higher level, that is, without changing the reading distance and antenna orientation, 
higher power level is now required to communicate with the RFID tag after adding 
the bypass to the IC chip in the RFID tag. The closer the bypass to the chip, the 
higher power level is required for the RFID reader to talk to the tag. This feature can 
be utilized to identify the close or open status of the bypass which is part of the BT 
strain sensor proposed in this study. The results of the wireless interrogation test are 
shown in Figure 3-5. It?s noted that the RFID reader?s output power range is 10 to 
27 dBm. If the responding power threshold is higher than the maximum power level 
 
 24 
 
(i.e., 27 dBm), the RFID reader would not be able to detect the tag and the RFID tag 
is completely deactivated by the bypass.   
The Skyetek M9 UHF RFID reader is a multi-protocol ETSI 302 208 compliant UHF 
(862 - 955 MHz) RFID reader platform that supports a wide variety of UHF RFID 
tags. It can read and write based on the EPC Class1 Gen1, ISO 18000-6B and ISO 
18000-6C (EPC C1G2/Gen2) air interface and communications standards. The RF 
output power of the M9 is software-adjustable from 10 to 27 dBm (Skyetek 2012). 
3.4.  Breakage-triggered strain range sensor: device mechanics 
3.4.1.  Mechanics of BT strain sensor 
The working principle of the BT strain sensor for detection of threshold strains is to 
connect a crack-sensitive brittle bar with an RFID tag which changes the impedance 
of the tag by short-circuiting the IC chip. Because of this impedance mismatch, the 
RFID tag is no longer able to communicate with the RFID reader at previously set 
output power level. When the pre-set strain threshold is surpassed in the brittle bar 
due to increasing deformation in the monitored structure, the bypass to the IC chip 
is broken and the normal communication between the RFID reader and tag attached 
to the brittle bar is restored. Therefore, the breakage of the brittle bar triggers the 
restored capability of RFID tag to communicate with RFID reader about the updated 
status of threshold strain being surpassed in the BT strain sensor.  
The overall structure of the threshold strain detection system with BT strain sensors 
mounted on the monitored structure (e.g., an I-section beam) is shown in Figure 3-6. 
The C mark in Figure 3-6 indicates the proposed BT strain sensor which would break 
 
 25 
 
and open the bypass once a pre-set strain is surpassed. It thus acts as a threshold strain 
sensor which uses RFID tag for wireless communication. Before the breakage of the 
BT strain sensor, the voltage or power on the IC chip would fall below its minimum 
threshold power level for the RFID reader to communicate with the RFID tag even if 
it?s within its nominal reading range. If the structure member yields due to the 
extreme load and the strain at point C reaches the sensor?s breakage threshold strain 
level, then the RFID circuit resumes to its normal state and the voltage between point 
A and B returns to be the normal level for the RFID tag to communicate with the 
RFID reader. The RFID reader can detect the tag and send the tag ID to the 
corresponding reader software. Also, by retrieving the other information from the 
RFID tag, it provides important information about the damage locations that need 
attention and repairs. 
3.4.2.  Brittle bar design 
The following characteristics are desired for BT strain sensor design: 1) conductive 
trace is broken once the brittle bar is fractured. 2) low-cost strength material for the 
brittle bar; 3) ease of application. The most important requirement is the brittle bar 
breaks at the pre-set strain level. The crack propagation must be fast so that the 
conductive trace will break immediately after the breakage of the brittle bar.  
Conductive paint is applied on the brittle bar to form a conductive trace which is 
connected to the RFID tag at Point A and A? in Figure 3-7. When the brittle 
component is fractured due to strain on the structural member, the RFID tag is made 
 
 26 
 
active and its stored data can be interrogated by a RFID reader. The design for the 
brittle bar in the BT strain sensor is shown in Figure 3-7. 
The gage length between the two end fixtures of the BT strain sensor, Ls, 
determines how large the monitored area would be. The BT strain sensor is attached 
to the structural element at these two end fixtures. Also the threshold strain can be 
adjusted by changing the value of Ls. The threshold strain is defined as the relative 
displacement between the two ends fixtures divided by the gage length Ls at the 
time point of breakage. The width and height of the brittle bar (typically with a 
rectangular cross-section shape) is denoted as W and H respectively. The values of 
these two parameters should be determined in such a way that the brittle bar 
fractures when the overall deformation over the sensor gage length surpasses a pre-
 determined threshold strain value. Additionally, other parts in the sensor should be 
made much stronger and stiffer so that the deformation in the brittle bar contributes 
almost exclusively to the strain sustained by the sensor.  
Conductive paint is applied to cover the side area of the brittle bar and the two 
terminal pads (marked A and A? in Figure 3-7) of the conductive trace are connected 
to the RFID tag.  Therefore, the conductive trace completes the bypass to the IC chip 
but would open the bypass immediately after crack is formed in the underlying brittle 
bar. This way, by checking the status of the corresponding RFID tag, the BT strain 
sensor should yield the information as to whether threshold strain levels have been 
surpassed in the monitored structure. 
 
 27 
 
3.4.3.  Numerical simulation based design validation 
To obtain the deformation of the brittle bar right before it breaks, a finite element 
(FE) analysis was conducted to guide the design. A FE analysis software program ? 
ANSYS v13 was used for this purpose in this study. Acrylic was selected as the 
brittle bar material used in this study. For acrylic sheet, the tensile strain at break is 
assumed to be 3%.  In this study, threshold strain for the BT sensor is defined as 
sensor strain at the breakage point when the first principal strain in the brittle bar 
reaches 3%. The overall deformation of the sensor at breakage is called critical 
displacement. 
The overall dimensions of the sensor are, Hs=37 and Ls=30 mm. The width and 
height of the brittle bar is varied to achieve different values of the ratio between its 
local principal strain and the sensor strain. For convenience in notation, each design is 
denoted as HxWy, where x is the height value of the brittle bar and y is the width 
value with a unit of mm. The left side of the sensor is fixed in both the vertical and 
horizontal degree-of-freedoms while the right side of the sensor is allowed to move in 
the horizontal direction. The sensor strain is calculated as the ratio of the horizontal 
displacement underwent by the right side of the sensor over a gage length of Ls equal 
to 30 mm. Figure 3-8 shows the meshing scheme for the H9W2 model in ANSYS. To 
achieve better accuracy, elements were refined at the brittle bar and surrounding 
areas.  
According to the Maximum Normal Stress Theory (Rankine 1857), a brittle material 
will fail when the maximum principal stress exceeds some value, independent of 
 
 28 
 
other stress components present. Therefore, the failure criteria for the brittle bar 
adopted in this study is that first principal strain reaches the fracture strain in a simple 
tension test. For example, the acrylic brittle bar will fail by fracturing whenever the 
maximum first principal strain around the corner surpasses 3%. For brittle material, 
the first principal strain is defined as the first principal stress divided by the modulus 
of elasticity (E) of the material, that is, ?1 = ?1/?.  
Contour plots of the first principal strain for each of the three brittle bar designs from 
a linear elastic FE analysis are shown in Figures 3-9 to 3-11 respectively. The first 
principal strain for the three cases always occurs at the corners where cracks are 
observed to first appear. The strains in other parts of the brittle bar are found to be 
much smaller by several orders of magnitude. The FE simulation results suggest that 
for the proposed configuration of brittle bar, a significant amount of stress 
concentration is seen near the corner region which makes the failure mode (break 
point and deformation) predictable and repeatable for brittle materials. The critical 
displacements which correspond to the first principal strain of 3% in the brittle bar for 
each design are listed in Table 3-1. It is seen that the critical displacements at the 
breakage point are: 0.295 mm, 0.430 mm, and 0.600 mm for H6W4, H8W3 and 
H9W2 respectively. When connected with an RFID tag, BT strain sensor with these 
three designs can detect three different threshold strain levels so that the actual strain 
range in the structure can be estimated by scanning the BT strain sensors. To 
visualize the first principal strain distribution in different designs, a constant 
displacement of 0.3 mm (corresponding to 1% sensor strain with a gage length of 30 
mm) is imposed to the right-hand side of the BT strain sensor. The contours of the 
 
 29 
 
first principle strain in the corresponding brittle bar are shown in Figure 3-12. The 
first principal strain observed in the FE model of H9W2, H8W2, H6W4 are 1.5%, 
2.15%, and 3% respectively. 
3.4.4.  Experimental characterization test 
To experimentally characterize the BT strain sensor, a test setup shown in Figure 3-13 
was used. The test setup includes a hand-driven precision slide to impose the 
displacement on the test specimen and attaching fixtures. The linear motion slide has 
an accuracy of 0.001 inches (0.0254 mm) in measuring linear excursion distance. The 
BT strain sensor was attached to the linear motion slide with four screws. 
Test specimens are shown in Figure 3-13. In this study, two types of brittle materials 
were used for the brittle bar in the BT strain sensor: (i) The first material is acrylic 
sheet with a thickness of 2.54 mm (0.1 inches). The acrylic specimens were cut from 
a 6 x 12 inches acrylic sheet using a carbon dioxide laser cutter (CO2 cutter). (ii)  The 
other material is an epoxy paint that was custom made by mixing epoxy with fine 
calcium carbonate powder. The calcium carbonate powder is added to the epoxy resin 
mixture to increase the brittleness of cured epoxy paint. One advantage of using 
epoxy paint is that its fracture strain can be adjusted by changing the mix ratio of 
carbon carbonate powder in the paint. To degas the epoxy paint, vacuum pump was 
used to suck air bubbles out of the fresh paint mixtures. The fresh epoxy paint 
mixture was cast in a pre-fabricated mold and it generally took 72 hours for the epoxy 
paint to cure. As shown in Figure 3-14, the test specimens made of epoxy paint were 
comprised of two parts: the brittle bar is made of epoxy paint and the rest part is made 
 
 30 
 
from acrylic sheet. The two ends of the conductive trace were connected to the two 
terminals of the IC chip in the RFID tag. SPI flash-dry silver paint was used in this 
study to form the conductive trace. This silver paint is a highly concentrated 
suspension (43% silver solids by weight) of silver powders combined with an organic 
suspending and binder system.  
Figures 3-15 shows the cracks in the acrylic test specimen along with superimposed 
vector plots of first principal stress in the brittle bar. The vector plots are based on a 
linear elastic FE analysis of the brittle bar at breakage point. Clearly, the first 
principal stress orientation at the corner of the brittle bar agrees well with the crack 
direction at the same location. Figures 3-16 shows the cracks in the epoxy paint test 
specimen. 
The test results of the critical displacement corresponding to the first breakage of the 
brittle bar (thus the bypass to the IC chip is eliminated due to broken conductive 
trace) are list in Table 3-2. It is seen that for the acrylic material, the standard 
deviation of the test results on critical displacement are 8.15% and 7.27% for H9W2 
and H8W3, while this becomes 10.95% for H6W4. This might be partly due to the 
variation in the pre-tension of the test specimen induced by unevenly tightening the 
screws in different tests. Cracks in the brittle bar seen using a digital microscope (65x 
magnification factor) are shown in Figure 3-15. The histogram of the test results for 
acrylic BT strain sensor is shown in Figure 3-17. The brittle bar was found to fracture 
quickly after the initiation of the crack at the corner. No slow fracture process was 
observed in all the tests conducted in this study. As the conductive trace was 
deposited over one side of the brittle band, the conductive trace broke immediately 
 
 31 
 
after the crack was formed, which led to the disconnection between the bypass and 
the RFID tag and RFID tag is restored to normal working condition due to removal of 
the short-circuit. For each design, 15 tests were repeated to get statistics of the BT 
strain sensor performance. Even greater standard deviation values are observed in the 
epoxy paint specimens as listed in Table 3-2. This might be due to the random 
material defects in the epoxy paint material such as micron-size air voids trapped in 
the paint. The histogram of the test results for the epoxy paint based BT strain sensor 
is shown in Figure 3-18.  
3.5.  BT strain sensor design for detection of multiple threshold strains 
The test results described in the preceding section validate the BT strain sensors for 
detecting single threshold strain level although different design can be used for 
different threshold strains. In this section, a BT strain sensor with multiple brittle bars 
on a single device is proposed to detect multiple threshold strain levels. One of the 
advantages of such BT strain sensor with multiple brittle bars is to occupy smaller 
space at installation location in the monitored structure. The configuration of a BT 
strain sensor with two brittle bars is shown in Figure 3-19. It?s worth noting that due 
to the space limit and load capacity of the linear motion slide, only two threshold 
strains levels were considered in this device. The test results of the critical 
displacement of this sensor design are listed in Table 3-3. The histogram of the test 
results on the critical displacement of the BT strain sensor with two brittle bars is 
shown in Figure 3-20. It is seen that the mean value of critical displacements is 0.34 
mm for H6W4 and 0.69 mm for H9W2 respectively. By assuming a gage length of 2 
 
 32 
 
inches (50.8 mm), the two threshold strain levels of this BT strain sensor are 0.66% 
and 1.3%.   
3.6.  Pilot study of buckling detector for tube 
The BT strain sensor with brittle bar design was validated in the above described 
laboratory tests. It has been demonstrated that it?s capable of detecting pre-set 
threshold strain levels and this information can be acquired wirelessly with an RFID 
reader. The threshold strain recordings are repeatable with a small tolerable variance. 
However, one of the limitations of the BT strain sensor is its difficulty to detect the 
buckling or local buckling in structures members such as global buckling of tubular 
members. This is because the above presented brittle bar design is intended to work 
under longitudinal deformation of the sensor without out of plane bending. Out of 
plane bending will cause torsion in the brittle bar and thus complication in 
quantifying the threshold strain. 
To overcome the difficulty associated with buckling detection, particularly on 
structures with curved surface such as tubular members, another design for buckling 
detection is proposed here using epoxy paint patch. In this pilot study, this epoxy 
paint patch based buckling sensor can only tell whether buckling occurs or not. 
Further research works need to be done to quantifying the strain sustained by the 
epoxy paint patch during the buckling process.  
A schematic of the epoxy paint patches installed on a paint bucket are given in Figure 
3-21. The epoxy paint patch is cut into proper size and is attached to the tubular 
specimen when it?s still flexible before fully cured. The test specimen in this test is a 
 
 33 
 
1-qt. paint bucket with lid (diameter = 4.35 inches, height = 5 inches). To detect the 
breakage of the epoxy paint patch due to buckling of the bucket, silver paint was 
deposited to form conductive traces on the surface of the paint patch, as shown in 
Figure 3-21.  When the epoxy paint patch breaks, the conductive trace also breaks and 
thus the RFID tag connected to the conductive trace will be restored to normal 
operation condition. 
Four equally-spaced epoxy paint patches were bonded to each bucket specimen at 
their upper and lower sides using M-bond from Vishay. The experiment setup is 
shown in Figure 3-22. The RFID reader is placed 34 inches (86.4 cm) away from the 
center of the bucket specimen during the test. The servo-controlled SATEC universal 
test machine was used to applied axial load to the bucket specimen. The epoxy paint 
patches in test #1 after breakage are shown in Figure 3-23. A total of five tests were 
conducted. The epoxy paint patches are numbered as X-Y, number X denotes test 
number (1 to 5) and number Y denotes the epoxy paint patch number on each bucket 
(1 to 4). The test results of these five tests are listed in Table 3-4. It can be seen that 
for a bucket with obvious bucking, it?s likely that some epoxy paint patches on the 
bucket may not break. However, at least one epoxy paint patch was found to break in 
all tests.  
3.7.  Summary 
From the results presented in this chapter, it is shown that by modifying the RFID 
antenna, its responding power to the RFID reader would change. This feature can be 
utilized to detect whether a threshold strain is surpassed in the monitored structure by 
 
 34 
 
relating the strain to the change in the RFID antenna property. The two brittle 
materials used in this study each have their own advantages and limitations. For the 
acrylic sheet, its physical properties are more uniform compared with custom-
 formulated epoxy paint. This yields the test results that acrylic specimens have 
smaller variance. However, the drawback of acrylic sheet is that acrylic strain sensors 
are difficult to mount on curved structural surface. For epoxy paint, any arbitrary 
shape is possible to form and can be made to conform to curved surface. If air void in 
the paint can be completely removed by adopting a professional degassing process, its 
physical properties after full cure should be more uniform and repeatable. In this 
study, both materials have been demonstrated that they could break at a pre-
 determined threshold strain level and can be satisfactorily used for the BT strain 
sensor. In conjunction with the RFID technology, the BT strain sensor can fulfill the 
goal of rapidly scanning the strain condition at the monitoring location and wirelessly 
transfer the data to BIM database at request.  
  
  
 
 35 
 
 
Table 3- 1. Critical displacements for each brittle bar design derived from linear FE 
analysis (gage length 30 mm) 
Design 
Critical 
displacement 
(unit: mm) 
First principal 
strain in the 
model 
Sensor strain Force (N) 
H9W2 0.60 3.00% 2.00% 10.1 
H8W3 0.43 3.01% 1.40% 24.6 
H6W4 0.30 3.01% 0.98% 39.8 
 
 
 
 
Table 3- 2. Experimental characterization test results of the critical displacement of 
BT strain sensor for detection of single threshold strain level (unit: mm) 
 
Acrylic plate (thickness = 2.54 
mm) 
Epoxy paint (thickness 
= 2.03 mm) 
Test No. H6W4 H8W3 H9W2 H8W3 H9W2 
1 0.25  0.46  0.69  0.53  0.58  
2 0.30  0.53  0.61  0.43  0.76  
3 0.33  0.48  0.66  0.51  0.64  
4 0.36  0.56  0.64  0.46  0.86  
5 0.36  0.53  0.61  0.53  0.74  
6 0.30  0.48  0.69  0.53  0.76  
7 0.36  0.53  0.66  0.56  0.66  
8 0.36  0.43  0.76  0.41  0.64  
9 0.41  0.48  0.74  0.56  0.79  
10 0.38  0.51  0.71  0.53  0.76  
11 0.33  0.48  0.61  0.56  0.81  
12 0.36  0.58  0.66  0.48  0.84  
13 0.36  0.56  0.74  0.43  0.81  
14 0.38  0.51  0.64  0.51  0.74  
15 0.38  0.53  0.66  0.43  0.71  
Max. 0.41  0.58  0.76  0.56  0.86  
Min. 0.25  0.43  0.61  0.41  0.58  
Mean 0.35  0.51  0.67  0.50  0.74  
Std. Dev. 0.04  0.04  0.05  0.05  0.08  
 
 
 
 
 36 
 
Table 3- 3. Experimental characterization test results of BT strain sensor capable of 
detecting two threshold strain levels (unit: mm) 
 Acrylic plate (thickness = 2.54 mm) 
Test No. H6W4 H9W2 
1 0.33  0.66  
2 0.43  0.71  
3 0.33  0.69  
4 0.25  0.71  
5 0.38  0.66  
6 0.33  0.71  
7 0.30  0.74  
8 0.30  0.61  
9 0.30  0.69  
10 0.28  0.66  
11 0.36  0.69  
12 0.41  0.69  
13 0.28  0.71  
14 0.36  0.64  
15 0.38  0.74  
Max. 0.43  0.74  
Min. 0.25  0.61  
Mean 0.34  0.69  
Std. Dev. 0.05  0.04  
 
 
Table 3- 4. Summary of the test results from epoxy paint patch tests 
(?X?- Epoxy paint patch broke,  ?-?- No breaking) 
Test No. 1-1 1-2 1-3 1-4 
Results X - X X 
Test No. 2-1 2-2 2-3 2-4 
Results X X X X 
Test No. 3-1 3-2 3-3 3-4 
Results X X - X 
Test No. 4-1 4-2 4-3 4-4 
Results X X - - 
Test No. 5-1 5-2 5-3 5-4 
Results X X X X 
 
 
 
 37 
 
 
Figure 3- 1. RFID-based wireless strain or crack sensing devices developed by other 
researchers: (a) Morita and Noguchi (2008); (b) Chin et al.  (2009); (c) Yi et al. 
(2011) 
 
 
 
 
 
Figure 3- 2. Illustration of antenna tuning parameters 
 
 
 
 
 
 
(a)
 (c)(b)
 Area
 D
 L
 
 38 
 
 
 
 
 
Figure 3- 3. Schematics of modifying RFID tag by adding by-pass to IC chip 
 
 
 
 
 
 
Figure 3- 4. Configuration of RFID tag with different bypass locations 
 
By-pass
 IC
 8 mm
 20 mm
 15 mm
 8 mm
 IC
 Case A
 Case B
 Case C
 
 39 
 
 
Figure 3- 5. Effect of adding bypass to IC chip on responding output power threshold 
level of RFID reader (see Figure 3-6 for explanation of different cases) 
 
 
 
 
 
 
Figure 3- 6. Overall structure of the proposed threshold strain detection system with 
breakage-triggered (BT) strain sensors 
 
 
 
0
 4
 8
 12
 16
 20
 24
 28
 902 907 912 917 922 927
 Frequency (MHz)
 O
 ut
 p
 ut
  p
 ow
 er
  th
 re
 sh
 ol
 d
  (d
 B
 m
 )
 Case A
 Case B
 Case C
 Za
 Chip ZcAntenna
 A
 B
 C
 Electrically
 conductive 
material
 
 40 
 
 
Figure 3- 7. Schematic of BT strain sensor design 
 
Figure 3- 8. FEM model of BT strain sensor (unit: mm) 
 
Ls
 Fixtures
 Brittle bar
 H
 W
 Hs
 Conductive 
trace
 A B
 A? B?
 Ls
 Hs
 W
 H
 5
 4
 D
 is
 pl
 ac
 em
 en
 t
 
 41 
 
 
 
Figure 3- 9. Contour plot of first principal strain in the BT strain sensor (model 
H9W2, gage length = 30 mm) at breakage 
 
 
 
Figure 3- 10. Contour plot of first principal strain in the BT strain sensor (model 
H8W3, gage length = 30 mm) at breakage 
 
 
 
 
 42 
 
 
 
Figure 3- 11. Contour plot of first principal strain in the BT strain sensor (model 
H6W4, gage length = 30 mm) at breakage  
 
 
(a) H9W2 
 
 
 43 
 
 
(b) H8W3 
 
 
(c) H6W4 
 
Figure 3- 12. Distribution of first principal strain in the model corresponding to 1% 
sensor strain in the BT strain sensor  
 
 
 
 44 
 
 
Figure 3- 13. Test setup used for experimental characterization of BT strain sensors 
using a hand-driven precision slide 
 
 
 
 
 
 
 
Figure 3- 14. Brittle bar specimens made of acrylic sheet and epoxy paint 
 
 
 
Epoxy
 paint
 Acrylic
 
 45 
 
 
(a) H6W4 
 
 
(b) H8W3 
 
 
 46 
 
 
(c) H9W2 
Figure 3- 15. Photos of cracked brittle bar (acrylic sheet) after test with vector plot of 
first principal stress superimposed 
 
 
 
 
 
 
 
 
 
 
 47 
 
 
(a) H8W3 
 
(b) H9W2 
Figure 3- 16. Photos of cracked brittle bar (epoxy paint) after test  
 
 
 48 
 
 
Figure 3- 17. Histogram of critical displacement test results of BT strain sensors 
(acrylic sheet) with single threshold strain level 
 
Figure 3- 18. Histogram of critical displacement test results of BT strain sensors 
(epoxy paint) with single threshold 
 
 
0
 1
 2
 3
 4
 5
 6
 7
 0.25 0.30 0.36 0.41 0.46 0.51 0.56 0.61 0.66 0.71 0.76
 Critical displacement (mm)
 O
 cc
 ur
 re
 nc
 e 
fre
 q
 ue
 nc
 y
 H6W4
 H8W3
 H9W2
 0
 0.5
 1
 1.5
 2
 2.5
 3
 3.5
 4
 4.5
 0.38 0.43 0.48 0.53 0.58 0.64 0.69 0.74 0.79 0.84
 Critical displacement (mm)
 O
 cc
 ur
 re
 nc
 e 
fre
 q
 ue
 nc
 y
 H8W3
 H9W2
 
 49 
 
 
 
 
Figure 3- 19. Photo of BT strain sensors with two threshold strain levels (H9W2 and 
H6W4) 
 
 
Figure 3- 20. Histogram of critical displacement test results of BT strain sensors 
(acrylic sheet) with two threshold strain levels 
 
0
 0.5
 1
 1.5
 2
 2.5
 3
 3.5
 4
 4.5
 0.25 0.30 0.36 0.41 0.46 0.51 0.56 0.61 0.66 0.71 0.76
 Critical displacement (mm)
 O
 cc
 ur
 re
 nc
 e 
fre
 q
 ue
 nc
 y
 H6W4
 H9W2
 
 50 
 
 
Figure 3- 21. Epoxy paint patch on paint bucket 
 
 
Figure 3- 22. Experiment setup for epoxy paint patch test 
 
 51 
 
 
(a)                                                       (b) 
 
(c)                                                    (d) 
Figure 3- 23. Breakage of epoxy paint patch in test #1 
 
 52 
 
Chapter 4: Sensor-Driven Structural Health Monitoring 
Integrated with Building Information Modeling 
 
4.1.  Introduction 
Building information modeling (BIM) is a new technology which provides a platform 
for integrating data of different types and communication of information for asset 
management and preventive maintenance. BIM can also serve as an ideal 4-
 dimensional graphical computing environment for SHM and management practice 
which constantly involves considerable amounts of sensor data of different 
modalities. BIM software defines objects parametrically; that is, the objects are 
defined as parameters and relations to other objects, so that if a related object is 
amended, dependent ones will automatically also change (Eastman et al. 2011). Each 
model element can carry attributes for selecting and ordering them automatically, 
providing cost estimates and well as material tracking and ordering. BIM data can be 
stored as a digital file or in a database, and can be shared and exchanged between 
several applications (Isikdag et al. 2007).  
A framework of SHM integrated with BIM was developed using Revit Structures 
software to illustrate sensor data storage and display, finite element (FE) simulation, 
and structural condition assessment integrated in a BIM-enabled 4D setting, as shown 
in Figure 4-1. The framework includes the following four major modules: BIM model 
and a database, BIM for SHM portal, sensor clusters with RFID tag that automates 
 
 53 
 
the activation of the relevant area in the BIM model, and FE simulation software, 
which are briefly described below: 
? The BIM for SHM portal module, which provides an interface and 
middleware to facilitate the access to other modules in the framework. This 
was developed using C# programming language. The C# main program 
provides the interfaces to these components and sends commands to them, 
such as start RFID reader and select and highlight elements. It also 
coordinates the work order of the components and acquires and sends 
necessary date between components. Details of this module will be given in 
the next section. This modular architecture is also designed to support future 
expansion and adoption of other SHM utilities. 
? The BIM and database server module, which provides data access and storage 
in two software programs ? a BIM software such as Revit and a database 
server such as SQL server. In this study, MS Access is used for illustration. 
The BIM model built using Revit is linked to a MS Access database which 
includes all the information of the elements in the BIM model. The two files 
(i.e. Revit model file .rvt and MS Access file .mdb) created by the two 
software applications can update each other when either one is modified.  
? The sensor cluster module, which has a RFID tag for automated activation of 
the particular view area in the BIM model to display the sensor objects in the 
3D BIM model. Monitoring data from sensor networks will be sent to 
database servers either on a pre-arranged scheme or on request.  
 
 54 
 
? The FE simulation module, which consists of a set of basic functional units of 
a finite element structural analysis program. A FE analysis software program 
?OpenSees (Mckenna 1997) is used in this study. Other commercial general-
 purpose FE simulation software applications such as ANSYS, SAP2000 can 
be used for this purpose equally well as long as an adapter software 
application is available.  
There are several major software developers such as Revit, Tekla, Bentley and 
ArchiCAD that supply products with functionality in the BIM world (Eastman et al. 
2011). In this study, Revit - Autodesk?s main platform for BIM is selected. Revit has 
very similar functionality to the other major BIM software. Revit was designed from 
the ground up as a BIM platform to specifically address problem areas of the 
architecture, engineering, and construction (AEC) industry. It is a complete discipline 
specific building design and documentation system supporting all phases of design 
and construction documentation. Revit structure 2012 was used as BIM software to 
develop the BIM model. Revit Structure is a technological platform that currently 
supports architectural, structural, and mechanical disciplines. It allows the user to 
design with both parametric 3D modeling and 2D drafting elements. Building 
Information Modeling is a Computer Aided Design (CAD) program that employs 
intelligent 3D objects to represent real physical building components such as walls 
and doors. Revit uses .RVT files for storing BIM models. Typically, a building is 
made using 3D objects to create walls, floors, roofs, structure, windows, doors and 
other objects as needed. These parametric objects ? 3D building objects (such as 
windows or doors) or 2D drafting objects (such as surface patterns) ? are called 
 
 55 
 
"families" and are saved in .RFA files, and imported into the RVT database as 
needed.  
The SHM framework integrated with BIM can significantly reduce the amount of 
time that would be spent performing searching, accessing and validating the 
information. As the structure?s complexity increases, the BIM model becomes very 
large and identification and selection of the appropriate SHM locations for relevant 
sensor data viewing and information processing/display becomes very time-
 consuming if this is done in a manner of manual operation. This drawback 
undermines the advantage of BIM technology.  
In recent years, Radio frequency identification (RFID) has been given lots of 
attentions for its fast development in many fields such as asset management, access 
control, transportation and logistics. RFID uses radio frequency electromagnetic 
fields to transfer data from a tag attached to an object, for the purposes of automatic 
identification and tracking. Commonly adopted frequency bands for RFID 
communication are 120-150 kHz (LF), 13.56 MHz (HF), 902-928 MHz (UHF), and 
2.4 GHz (UHF). Passive RFID tags require no battery and are powered by the 
electromagnetic fields used to read them. The tag contains electronically stored 
information which can be read from up to several meters away. Also, each RFID tag 
was assigned a global unique ID, this make it ideal for identification application.  
The major advantages of using RFID are that the tag does not need to be within line 
of sight of the reader and may be embedded in the tracked object, tags can be read 
through many non-metallic materials, and many tags can be identified at the same 
 
 56 
 
time, each tag can contain more data inside and the radio signal transferring distance 
is longer. These key features enable the RFID system to fast access the BIM and 
automatically identify and select appropriate structural elements being monitored in 
BIM. The data stored on the tags could be read, updated and changed by RFID 
readers. In the SHM application in question, it is desirable that the RFID tag could be 
read from far distance. Hence, the system can detect the structure component even if 
the component is hidden or not visible. 
This chapter describes the initial efforts in developing a BIM for SHM portal targeted 
for urban infrastructure monitoring and management applications. A custom 
developed middleware enables automated sensor data entry into BIM environment to 
avoid possible human related errors.  A prototype bridge model was developed using 
Revit Structures software to illustrate sensor data storage and display, FE analysis, 
photorealistic visualization, dynamic simulation, and structural condition assessment 
integrated in 3D visual setting. 
4.2.  BIM for SHM portal 
A 3D BIM model for the monitored structure provides a visual clue about the location 
and condition of the key monitoring area and associated sensor clusters. This 
facilitates the visualization of the processed SHM information and its variation with 
time. In this way, a 4D model that integrates a 3D drawing with time is made 
available and allows the consideration of maintenance actions for deteriorating 
structural components through their life cycle. Changes that occur in the structural 
condition and SHM system over time can be highlighted in such a 4D BIM model. 
 
 57 
 
Furthermore, structural safety can be quantified in terms of structural reliability, 
taking into account the uncertainty involved. It is worth noting that structural safety is 
also time variant. This is because the loading condition and the properties of a 
structure may change over time. For example, many developed countries are currently 
experiencing a problem of aging and deteriorated bridge networks (Pablo 2009). 
Considering the fact that the ultimate goal of SHM is to evaluate the structural safety 
of the monitored structure, a 4D BIM model with time varying structural condition 
and system reliability information is very appealing for SHM applications.  
As shown in Figure 4-2, the main computer program developed in the C# 
environment ? the BIM for SHM portal, serves as the coordinator and interfaces with 
other components in the BIM based SHM system. The BIM for SHM portal 
coordinates the workflow in the system, such as activating the RFID reader and 
selecting the desired elements for condition assessment. Then it searches the BIM 
database for the ?RFID Tag? of each element. If a match is found, the proper BIM 
elements would be highlighted in a user-defined color. In this study, the BIM for 
SHM portal, was developed using the Revit API and RFID reader API and 
programmed with Visual C# language. These APIs (Application Program Interfaces) 
are usually Dynamic-link library (DLLs) that allow the user access and control 
particular applications? work environments. The Revit .NET API allows users to 
program with any .NET compliant language including Visual Basic.NET, C#, and 
C++/CLI. The user can use the Revit Platform API to gain access to model?s 
graphical and parameter data to perform tasks. The Skyetek API is built in C++ 
 
 58 
 
language, which allows user to manage basic and advanced RFID tag and reader 
manipulation tasks. 
To be able to fast locate and highlight the area of interest corresponding to the 
selected sensor cluster in the BIM model, a shared parameter ?RFID Tag? is added to 
the elements. Each BIM element is assigned an ID that uniquely identifies itself. This 
ID is not the RFID tag number; rather, this is the ID number assigned to each BIM 
element by Revit. A user can create custom parameters for any element or component 
category in the BIM software. In Revit, this is realized by creating a ?Shared 
Parameter?, which will be assigned to a specific structural member and are also 
available to be shared with other members in the family. They offer the power to add 
custom data or information that is not predefined in the family file or project 
template. For example, ?RFID Tag? and ?Damage Flag? are two shared parameters 
which are custom defined here for SHM application and can be assigned to any 
structure element. It should be noted that in Revit only ?Shared Parameters? can be 
shared by different projects and families and exported to external database in the 
ODBC (Open Database Connectivity) format. In Revit, export/import in ODBC 
format can be made compatible with a variety of database software such as SQL and 
Microsoft Access. This study uses Microsoft Access as the ODBC database and 
Figure 4-3 shows an ODBC database sheet created in BIM and tested in Microsoft 
Access for its compatibility. 
Additionally, an Autodesk Revit Add-in program has been developed to display the 
sensor data such as time series in the BIM environment. Visualization and imaging of 
SHM results are very helpful in communicating with practitioners and engineers. 3D 
 
 59 
 
models and data display enhance the vividness, clarity, pinpointing, and thereby 
improve the users? ability to visualize information and increase the users? 
effectiveness in communicating with others. The visualization capacity is especially 
appealing to SHM applications because this can expose users to view a space and its 
health condition that is not easily accessible in a real world environment. For 
example, it could be used by inspection engineers to walk through a bridge structure 
that otherwise has to be accessed through snooper truck and often requires road traffic 
control and even road closing.  
4.3.  Case study  
This section reports a demonstration project which showcases the functionality of the 
BIM based SHM system for infrastructure monitoring. A prototype system including 
the Revit Structures, a FE analysis software, sensors and custom-developed 
middleware code (i.e. BIM for SHM portal), are presented in a case study involving a 
multiple steel I-girders bridge in Maryland. The bridge is a single-span structure of a 
span length of 140 ft (42.67 m), as shown in Figure 4-4. The bridge was built in 
1980?s and fatigue cracks were identified in some of the diaphragm connection welds. 
Types of sensors used in the field test include piezoelectric film AE sensors, wireless 
accelerometers, laser distance sensors and strain transducers. 
Users can directly use the Revit GUI interface to create the structure model by adding 
architectural elements, structural elements, and other elements such as sensors or 
other SHM components. The BIM model of the steel bridge built in Revit is shown in 
Figure 4-5. 
 
 60 
 
Modeling for expediency alone may not work if a BIM for SHM model is desired. 
For example, consider the modeling of the steel I-girder of the bridge in BIM. 
Although it may be tempting to model the I-girder monolithically, the monolithic I-
 girder will need to be remodeled and segregated to match the FE analysis and hot spot 
areas required by the SHM application.  
Revit DB Link is an add-in program provided by Autodesk that allows user to 
maintain a relationship between a Revit project and a Microsoft Access, Microsoft 
Excel, or ODBC database. A user can use Autodesk Revit DB Link to export Revit 
project data to the database, make changes to the data, and import it back into the 
project. The database displays Revit project information in a table view that can be 
edited before import. By using Revit DB link, BIM database was developed in MS 
Access 2007 format. The BIM database includes database tables for each element 
types and instances.  
The workflow of the BIM based SHM system is shown in Figure 4-2. Eight Revit 
add-ins have been developed in the Microsoft C# environment. They can be called by 
clicking the pull-down menu shown in Figure 4-5. Prior to staring the main BIM for 
SHM portal program, certain parameters are allowed to be edited in the two files - 
parameter.txt and RFID parameter.txt. These files allow user to set the path for BIM 
database and parameters for Skyetek reader. Once these two parameter files are set 
up, the system is ready to start the main BIM for SHM portal program. The Skyetek 
RFID reader will be called and starts to scan tags sequentially. During each updating 
cycle, the Skyetek RFID reader will perform scanning every once in a while at a pre-
 defined interval, and record the IDs of all the RFID tags within its reading range. The 
 
 61 
 
BIM for SHM portal program compares the current tag ID pool with the one in 
previous cycle. If any change is found, the BIM for SHM portal program would send 
a request to the Revit to update its database and highlight the corresponding elements. 
By performing this check before sending out the request command, the program can 
avoid sending unnecessary update command to the Revit, and thus save the 
computing time. In the meanwhile, all the tag IDs would be saved to RFIDList.txt, 
which is located in the same directory. It can be used for verification purpose or other 
post-processing. If the loop time and interval for each scanning are set to small 
numbers, each scanning action would take less time at the price of increased 
likelihood for induced inaccuracy.  
The BIM based SHM system is highly automated to minimize human intervention. 
However, if for some reason manual update is necessary, the Revit GUI also allows 
the user to run RevitAccessDB process directly. If the user wants to highlight certain 
elements whose tag IDs are known, he can type in the tag IDs in the file RFIDList.txt, 
and then check the RevitAccessDB GUI button. The damage flags can also be reset 
by clicking the ResetDamageFlag GUI button. In this example, sensor data from 
accelerometer has been stored at a pre-specified location on the user?s computer. The 
user can display the acceleration data for the selected element using Revit GUI. 
Fourier transform was conducted on the sensor data. The sensor data and its 
frequency spectrum are visualized in the model, as shown in Figure 4-6.  
This BIM for SHM portal also facilitates the communication between the BIM model 
and FE analysis program. The FE model is a 140-ft (42.67 m) long simple beam built 
in the Opensees navigator (Mazzoni et al. 2009). Moving truck load was applied to 
 
 62 
 
the beam. The linear weight per foot length of the beam (including the bridge deck 
weight) is 487.1 lbf per foot (7.12 kN/m). The moment inertial I is 87,032 in4 (0.036 
m4) and the cross-sectional area A is 113.75 in2 (6.45 cm2). To run the FE analysis, 
simply check the ?RunOpenseesTCL? box in the Revit GUI drop-down menu, as 
shown in Figure 4-7. After the FE analysis is done, all the results files would have 
been saved to a pre-determined location. In the subsequent post-processing add-ins 
such as visualization the data, the program would search the numbers that match the 
ID of the selected elements in the BIM model, and then plot the desired response data 
such as internal forces. For illustration purpose, the bending moment of the 
highlighted beam location is plotted in Figure 4-8.  
As mentioned above, if one and only one element is selected, the plot window will 
show as Figure 4-8 after checking the ?PlotBeamMemberMoment? box. Within this 
window, the time series curve can be zoomed-in or out. Similarly, other response data 
such as middle span displacement of the bridge under specified loading or from 
monitoring can be visualized as well, which is shown in Figure 4-9. 
A modal analysis was also conducted in the FEM model through the Revit GUI, as 
shown in Figure 4-10. It?s seen that the fundamental frequency of the prototype 
bridge is 2.9 Hz (natural period displaced in the figure as 0.311 seconds). From the 
field test data using wireless accelerometers attached to the flange of the bridge 
girder, the measured fundamental frequency of the bridge is 3.2 Hz, which is close to 
the FEM result considering the simplicity of the FE model used in this study.   
 
 63 
 
As mentioned earlier, since the ultimate goal of SHM is to evaluate the structural 
safety of the monitored structure, a 4D BIM model with time varying structural 
condition and system reliability information is highly desired for decision making in 
the SHM applications. A time-dependent reliability analysis requires the knowledge 
of deterioration models which predict the future structural condition and form the 
basis for life-cycle monitoring and maintenance planning. However, the reliability 
analysis is only as good as the models which support it. The models must be updated 
over time to revise the optimum maintenance strategy based on how a structure 
actually behaves (Estes and Frangopol 2003). Sensor data from the SHM system 
installed on the monitored structure has been shown to be very useful for updating the 
deterioration models for future condition forecasting (Catbas et al. 2008; Li et al. 
2011).  
Condition rating data based on periodic inspection of bridges components provides an 
overall characterization of the general condition of a bridge that has been routinely 
used by inspection engineers. According to the National Bridge Inspection Standards 
(NBIS), condition ratings are one-digit numbers that characterize the general 
condition of the entire component being rated. For example, the condition of different 
parts of a bridge may be rated on a scale of 0 to 9 (with 9 being ?excellent? and zero 
being ?failed?). Condition ratings assigned to elements of a component must be 
combined to establish the overall component condition rating. Unlike the condition-
 rating inspection approach, the element-level inspection approach is quantity-based 
and allows the bridge owner to track deterioration of each bridge element over time 
(AASHTO 2011). This is advantageous, in that, by better understanding the rate of 
 
 64 
 
deterioration, the bridge owner can thus make more intelligent and cost-effective 
decisions related to preserving their bridge assets. Element-level inspection thus 
provides the means to help make better performance-based decisions for bridge 
management and many states in the US are beginning the process of collecting 
element level data (Rehm 2013).  
To accommodate the need for element-level condition data entry and display, the 
bridge structure can be divided into small, user-defined elements. The condition data 
for each element is recorded and stored in a database, which could be displayed in the 
BIM environment at user?s request for selected element as shown in Figure 4-11. To 
illustrate the concept, condition rating along with a Markovian-based modeling 
approach for estimating the deterioration of condition states of structural elements is 
adopted in this study. Markovian-based model is one of the most popular methods for 
bridge deterioration modeling and have been used by major Bridge Management 
System software including PONTIS. A Markov chain is a stochastic process in which 
the probability distribution in the next time point depends only on the current state 
(Cesare et al. 1992). The probability transition matrix key to the Markovian-based 
bridge deterioration model can be derived from the overall condition rating of bridges 
in a network (e.g., the National Bridge Inventory in the US). The Markovian 
transition matrix reflects the typical deterioration rates of the bridges in the dataset. 
Once updated with monitoring data, they can be used to forecast the future condition 
states of the monitored structural components. To implement this in the BIM 
environment, ?Element CS? add-in has been added to the Revit GUI and by clicking 
this button from drop-down menu, the selected element?s condition as well as an 
 
 65 
 
action button for forecasting the future condition state with the Markovian-based 
modeling approach is seen in Figure 4-11. With this visual display of the structural 
condition at element level, the bridge owner can make more intelligent and cost-
 effective decisions related to preserving their bridge assets with timely maintenance 
actions. 
4.4.  Summary 
This chapter presents a sensor-driven structural health monitoring (SHM) system that 
integrates building information modeling (BIM), sensors, FE analysis, and database 
technology. BIM can serve as an ideal 4-dimensional graphical computing 
environment for SHM and management practice which constantly involves 
considerable amounts of sensor data of different modalities. The ability to visualize 
the inspection and monitoring work done in the life cycle of a structure provides 
engineers and decision makers with a photorealistic view of the SHM process, 
simplifying their understanding of current condition of structural elements, structural 
reliability, maintenance work to schedule, response of structures under natural 
hazardous events such as earthquakes. In many instances, this improved 
understanding can facilitate the decision-making process often required when 
resources are scarce. To accommodate the need for element-level condition state data 
entry and display, the structure can be divided into small, user-defined elements. The 
condition data for each element is recorded and stored in a database, which could be 
displayed in the BIM environment at user?s request for selected element. 
 
 
 66 
 
 
Figure 4- 1. Schematic of SHM framework integrated to BIM 
 
 
Figure 4- 2. Schematics of the workflow in the BIM based SHM system 
BIM for SHM 
portal
 Database ServerBIM software
 Sensor clusters
 RFID tag
 TagID
 FEM 
software
 FE 
simulation
 BIM for SHM portal,
 user interface
 Database server
 BIM 
software
 Sensor clusters, RFID
 tag
 FEM software
 R
 FI
 D
  T
 ag
 ID
 FE simulation
 Highlight element, 
display sensor data and FEM 
results
 
 67 
 
 
Figure 4- 3. BIM database tables in MS Access 
 
 
 
 
 
 
 
 
 68 
 
 
 
Figure 4- 4. Pictures of the prototype steel highway bridge in case study 
 
 69 
 
 
Figure 4- 5. BIM model of the prototype steel highway bridge 
 
 
 
 
Figure 4- 6. Time history and frequency spectrum of sensor data displayed in BIM 
environment 
 
 
 70 
 
 
Figure 4- 7. ?Run Opensees? add-in for FE analysis called in BIM environment 
 
 
 
Figure 4- 8. ?PlotBeamMemberMoment? add-in called in BIM environment 
 
 
 71 
 
 
Figure 4- 9. ?PlotMidSpanDisp? add-in called in BIM environment 
 
 
 
 
 
Figure 4- 10. Modal analysis results from FE analysis program displayed in BIM 
environment 
 
 
 72 
 
 
Figure 4- 11. Element-level condition state displayed for selected element in BIM 
environment 
 
  
 
 73 
 
 
Chapter 5:  Resilient Structure Design with Fuse Members 
 
5.1.  Introduction 
5.1.1.  Background of steel structural framing 
Steel structural framing has been used for large span industrial buildings and facilities 
for public gathering. In the emergency of a strong earthquake, public spaces covered 
by large span arch structures, such as public halls or school gymnasiums, are 
expected to be converted to refuges for shelter provision for which a higher level of 
seismic performance (e.g., immediate occupancy) is preferred under strong 
earthquake events. However, under extreme earthquake, either part of the roof 
enclosure or the entire roof structure can be damaged. Collapse of roof could cause 
severe economic loss and poses safety risk to building residents. Furthermore, with 
recent emergence of building integrated renewable energy production system such as 
solar panels integrated with building facades and roofs, damage inflicted on the solar 
panels by excessively large deflection or collapse of roof would cause economic loss 
and power disruption to building residents.  
The collapse of a steel roof frame structure can be initiated by the buckling of a few 
members, as a result of load redistribution causing a subsequent progressive 
overstress condition in other members and thus its load carrying capacity is usually 
limited by the failure of first member or set of members to fail.  
 
 74 
 
5.1.2.  Concept of force limiting device (FLD) 
In an attempt to alter the brittle collapse behavior of steel roof frame structures under 
extreme earthquake loading, force limiting devices (FLD) can be used for robustness 
enhancement. These devices are fitted to critical compression members, and designed 
to provide a purely plasto-elastic behavior with a long plateau of member ductility 
(e.g., as exhibited in buckling-restrained braces). Since collapse is initiated by only a 
few critical members, the use of force limiting devices can be limited to a small 
selection of the most critical compressive chord or web diagonal members. The 
feasibility of using force limiting devices for controlling the behavior of space truss 
was validated in a research conducted by El-Sheikh (1999). The principle behind 
using these devices is to introduce artificial ductility in truss compression members, 
which would otherwise possess brittle post-buckling characteristics involving a loss 
of both stability and strength upon buckling. 
Also, to minimize the residual deformation of the roof structure associated with 
conventional FLD exhibiting bilinear hysteresis, self-centering (SC) members similar 
to the one developed by Zhu and Zhang (2008) provides a promising FLD alternative 
for use in steel roof frame structures. 
Large residual deformation associated with conventional ductile structural design can 
often make the structure appear unsafe to occupants, impair its ability to resist 
subsequent aftershock earthquakes and significantly increase the cost of post-
 earthquake retrofit (Ruiz-Garcia and Miranda 2006a, 2006b). Residual structural 
deformation thus starts to be recognized as a complementary parameter in the 
 
 75 
 
evaluation of structural (and non-structural) damage in performance based earthquake 
engineering (Pampanin et al. 2003; Christopoulos and Pampanin 2004). Recognizing 
the importance of controlling the residual deformation, self-centering seismic 
resisting system has recently been attracting considerable attention from the 
community (e.g., Kurama et al. 1999; Ricles et al. 2001; Christopoulos et al. 2002; 
Mahin et al. 2006). Such self-centering system is characterized by a flag-shaped 
hysteresis loop with certain energy dissipation capability and small or zero residual 
structural deformation after strong earthquakes.  
Special metals like shape memory alloys (SMAs) also exhibit a flag-shaped stress-
 strain curve (Casciati and Faravelli 2004; Desroches and Smith 2004; Song et al. 
2006). Due to its unique energy dissipation behavior and high fatigue life, SMA has 
been studied for use as a self-centering damping device in seismic resistant structures 
(Dolce et al. 2000; Zhu and Zhang 2008; Miller et al. 2011). Dolce et al. (2000) 
tested Nitinol-based devices with full re-centering and good energy dissipation 
capabilities. Zhu and Zhang (2008) studied a special type of bracing member termed 
self-centering friction damping brace (SFDB) which exhibits a flag-shaped hysteresis 
loop and has a potential to be used for force limiting devices with self-centering 
behavior. In the SFDB, the self-centering ability is realized using super elastic SMA 
wires while its enhanced energy dissipation capacity is achieved through friction. 
Miller et al. (2011) studied the seismic behavior and performance of self-centering 
buckling restrained braces (SC-BRBs). A SC-BRB specimen consisting of a typical 
BRB component, which provides energy dissipation, and pre-tensioned superelastic 
NiTi shape memory alloy (SMA) rods, which provide self-centering, was 
 
 76 
 
experimentally tested and its self-centering hysteresis is verified.  
Self-centering fuse members can be used for ductility enhancement and residual 
deformation reduction. These self-centering members would be fitted to strategically 
selected locations to create a self-centering hinge zone, and designed to provide a 
flag-shaped hysteretic behavior. To concentrate the deformation in such a self-
 centering hinge zone, the use of self-centering fuse members can be limited to a small 
selection of only a few critical members through properly weakening.  
5.1.3.  Concept of concentrated damage zones 
A structural weakening concept was employed in this study to improve the structural 
behaviour under dynamic loading. This concept is similar to selective weakening 
retrofit in earthquake engineering (Reinhorn et al. 2005).  
For a different application on seismic response mitigation, Reinhorn et al. (2005) and 
Viti et al. (2006) introduced the concept of weakening structures (e.g. reducing 
strength) to reduce simultaneously total accelerations and inter-story drifts. The main 
features of this procedure were to reduce maximum acceleration and associated forces 
in buildings subjected to seismic excitation by reducing their strength 
(weakening).This concept was employed in this study to improve the structural 
behaviour. Note that the yield force of FLD was designed to be equal to the buckling 
force (i.e., critical load) of a member with imperfection.  
The purpose of member weakening in the roof framing is to make the selected 
weakened member as a fuse element where favorable ductile damage of the overall 
 
 77 
 
structure is concentrated. This concentrated structural weakening strategy could also 
bring other benefits such as ease in damage inspection and rapid structural repair with 
member replacement post extreme events. Strengthening critical members only would 
not improve the structure collapse behavior because doing so will increase the forces 
on their neighboring members and cause them buckle instead. To strength all 
structural members is clearly not economical.  
Typically, a compromise has to be struck between the improvement to structural 
behavior associated with these devices and the cost addition caused by their use. 
Therefore, the use of FLDs is limited to a small selection of the most critical 
compression chord members in this study. This includes a strategy for selectively 
replacing key compressive members in steel roof structures with FLDs to prevent 
collapse or at least alter the collapse sequence to a favorable collapse mechanism. 
Simple linear finite element analysis of the original steel roof frame structure without 
FLDs was conducted to determine the internal forces in all members. The locations of 
the FLDs are chosen based on the axial force from this analysis. Several other 
plausible pressure distribution cases and found these selected places are also fine in 
order to improve the structural collapse resistance behavior compared with the 
original frame without FLDs. FLDs work as fuse members, where the deformation of 
the structure would concentrate; and because its strength is weakened compared to 
regular member, FLD would first be activated under many load cases.  
5.2.  Analysis procedure for dynamic collapse simulation 
To accurately simulate the dynamic collapse of steel roof framing structures subjected 
 
 78 
 
to combined gravity and earthquake load, non-linear time history analysis was 
performed in this study. The finite element model, that considers dynamic member 
buckling, yielding of tension members, and geometric nonlinearity due to large 
deflection of the roof structures undergoing collapse, is established using the 
OpenSees simulation platform (McKenna and Fenves 2005).  
A progressive collapse refers to a structural failure that is initiated by localized 
structural damage and subsequently develops, as a chain reaction, into a failure 
involving a major portion of the structural system (Ellingwood and Dusenberry 2005). 
Several mechanisms might contribute or lead to the progressive collapse of steel roof 
frame structures including: buckling of compression member, yielding of a tension 
member, facture of a tension member or connector, and nodal instability. The collapse 
of a steel frame structure can be initiated by the buckling of a few members (Blanford 
1997), and its load carrying capacity is usually limited by the failure of first member 
or set of members to fail. 
In this study, a 2-dimensional force-based element with corotational formulation and 
fiber discretization of the cross section was adopted for the simulation of buckling, 
post buckling, and hysteretic responses of members in steel roof frames. In this force-
 based corotational element, both material and geometric nonlinearities are considered. 
The formulation of force-based elements is based on interpolation functions for the 
internal force variation (Spacone et al. 1996). Neuenhofer and Filippou (1998) 
formulated a force-based element for geometric nonlinear analysis of plane frame 
structures, with linear constitutive relationship and small rotations. Following that, de 
 
 79 
 
Souza (2000) extended the material nonlinear force-based element proposed by 
Neuenhofer and Filippou (1997) to include geometric nonlinearity through deriving 
the transverse displacements from the curvatures using Lagrangian polynomial 
interpolation. The adopted kinematics is based on the assumption of moderately large 
deformations along the element, thus rigid body displacements and rotations can be 
arbitrarily large. However, at the price of further subdividing the structural member 
into smaller elements, large deformation problems can also be solved. 
In the analysis of frame elements, the material nonlinearity is considered by 
integrating the material stress-strain relations defined for each fiber over the section 
area, which is commonly referred to as ?fiber discretization?. Distributed plasticity is 
obtained by integration of the section force-deformation response over the member 
length in the corotating frame of reference. As shear effects are neglected, a uniaxial 
stress-strain relation is employed at the material point. The effect of spreading of 
plastic deformation along the member axis is thus considered in this study. 
The following equation of motion is used for nonlinear time history analysis, 
 ???(?) + ???(?) + ???(?)? = ??(?)    (5-1) 
Where the first term is the acceleration-dependent inertial force vector, R is the 
displacement-dependent restoring-force vector. ??(?) is the external applied-force 
vector.  
The standard approach for dynamic collapse analysis, as well as other nonlinear 
structural dynamic problems, is to time discretize the governing equations by 
 
 80 
 
Newmark time integration then solve them via the Newton-Raphson algorithm (Scott 
and Fenves 2010). In this study, the Newmark constant average acceleration method 
(?=0.25 and ?=0.50) is employed. This requires that iteration must be performed at 
each time step in order to satisfy equilibrium. Also, the incremental stiffness matrix 
must be formed and triangularized at each iteration or at selective points in time.  
  
5.3.  Numerical analysis model 
A finite element analysis software program ? OpenSees (McKenna and Fenves1997) 
was used in this study. This section describes the modeling technique for dynamic 
member buckling and self-centering fuse members in the numerical simulation of the 
nonlinear dynamic behavior of the steel arch structure under earthquake loading.  
5.3.1.  Analytical model for dynamic member buckling 
Modeling of the buckling behaviour of steel bracing members has been examined by 
a number researchers (e.g., Jin and El-Tawil 2003; Uriz et al. 2008; Krishnan 2010). 
An initial imperfection has to be assigned to the axially loaded member to simulate its 
buckling behaviour. However, few studies have considered the dynamic buckling 
effect of a compression member caused by the mid-point-inertia-mass (mid-mass)-
 induced lateral force at its mid-length (Tada and Suito1998).  
The member under examination is a steel tube (O.D= 6.26 in (159 mm), I.D= 5.78 in 
(147 mm)) with a material yield strength of 34 ksi (235 MPa). The member length is 
set to be 76.7 inches (1.95 m). To capture the inertia effect on member dynamic 
 
 81 
 
buckling, a lumped mass which represents half the mass of the member is assigned to 
the middle-length point of the member following a procedure described by Tada and 
Suito (1998). A parametric study has been conducted and the results revealed that a 
finer meshed model has very little effect on the member behavior. Therefore, a single 
mass lumped at mid-length was used in all cases. The end conditions are specified as 
pin-roller supported, as shown in Figure 5-1. Thus, the member was divided into two 
inelastic beam-column elements with an initial camber of 0.1% at the mid-node of the 
member in the finite element model.  
In this element, fiber discretization was used and the force components were obtained 
by integration over the cross-section. The tubular section of the member was 
discretized into two layers of fibers in the radial direction and 20 fibers in the 
circumferential direction respectively. As shear effects are neglected, a uniaxial 
stress-strain relation is employed at the material point. The material properties for the 
simulation model are based on the Menegotto-Pinto model for steel with extensions 
for kinematic and isotropic hardening (Filippou et al. 1983) with the following values: 
Young?s modulus = 29,000 ksi (200 GPa), yield stress = 34 ksi (235 MPa), and strain-
 hardening ratio = 0.005.  
Figure 5-2(a) shows the load-displacement (P-?) relations of the member under 
cyclic load, P. Due to the initial camber introduced at the mid-length point, the 
member is subjected to second-order bending effect. At a critical value of ?, once the 
second order bending moment at the mid-length point reaches the member?s plastic 
moment capacity, a plastic hinge with spread plasticity would form there. Further 
 
 82 
 
pushing the member in the longitudinal direction would result in the spreading of the 
plastic deformation near mid-length region and reduction in load. 
5.3.2.  Analytical model for Force Limiting Device (FLD) 
Fuse members with a long plateau of member ductility can be used for ductility 
enhancement in steel arch structures by fitting them to substitute critical compression 
members. The fundamental concept behind fuse members is to restrain low-order 
member buckling modes, creating full and stable hysteretic loops under tension-
 compression cyclic load. For example, with the use of such fuse members as buckling 
restrained braces (Uang and Nakashima 2003), member buckling can be held off till a 
fairly large compressive strain, without significant strength degradation.  
In numerical modeling of self-centering fuse members, a simple, one-dimensional, 
pin-ended truss element is used to have the appropriate uniaxial force-displacement 
properties. This is based on the Menegotto-Pinto model for steel with extensions for 
kinematic and isotropic hardening (Filippou et al. 1983). Flag-shaped hysteretic 
model, as shown in Figure 5-2(c), has been widely used to represent the cyclic 
loading behavior of self-centering system for sake of its simplicity (e.g. Christopoulos 
et al 2002; Seo and Sause 2005; Zhu and Zhang 2008). Two key parameters that 
define this flag-shaped hysteretic model are post-yield stiffness ratio ?  and energy 
dissipation factor ? that relates to the energy dissipation capacity of the system. For 
example, its lower bound (i.e., ?= 0) describes a piecewise nonlinear elastic system 
while the upper bound (i.e., ?=1.0) corresponds to a self-centering system with 
greatest possible energy dissipation capacity. For a given system with known initial 
 
 83 
 
stiffness and yield strength, the flag-shaped hysteretic model can be completely 
defined with these two parameters - ? and ?. In this study, ? value is set as 0.5 and ? 
value is set as 0.05.  
For comparison purpose, conventional fuse members exhibiting bilinear hysteresis are 
also considered in this study and its typical hysteresis loop is shown in Figure 5-2(b). 
The hysteresis loop of the conventional fuse member was calibrated with the 
experimental data reported by Clark et al. (1999).  
5.4.   Seismic analysis 
An arch truss is selected as the prototype structure, as shown in Figure 5-3. The 
member?s sizes of this arch structure are listed in Table 5-1. The dead loads applied to 
the arch truss nodes include its self-weight and the gravity load due to enclosure 
coverings. The self-weight of the arch truss is calculated to be 20 lbs/ft (0.29 kN/m) 
and the additional dead load is 11.23 psf (0.47 kPa). The total dead load is thus equal 
to 11.4 psf (0.55 kPa). The dead loads for the arch spaced 19.7-ft (6 m) apart is then 
converted into concentrated load applied to the top chord nodes. The lumped mass 
assigned to the top chord nodes are determined based on the corresponding dead 
loads. The natural frequencies of this arch structure are calculated from the finite 
element model and listed in Table 5-2. The viscous damping ratio of the arch structure 
is assumed to be 2%. Figure 5-4 shows the first three mode shapes for this arch 
structure. Members were modeled by subdividing into two beam-column elements, as 
illustrated in Figure 5-1. A 0.1% camber of the original member length was assumed 
for their initial imperfections.  
 
 84 
 
Four cases were considered in this analysis.  First case is the original structure 
without fuse members (labeled as ?Original?). Second case is an arch structure with 
conventional fuse members (FM) on locations shown in Figure 5-3(b) (labeled as 
?Conv. FM?). Third case is an arch structure with self-centering fuse members 
(energy dissipation factor ? = 0.5; cross-section of fuse members reduced to 70% of 
its original strength to create a weakened zone) which is labeled as ?SC 0.5-A?. The 
fourth case is an arch structure with self-centering fuse members (but the size of all 
members are increased by 1.33 times compared to case ?SC 0.5-A?) which is labeled 
as ?SC 0.5-B?. In practice, the case ?SC 0.5-A? could represent a retrofit for an 
existing structure by substituting critical members with self-centering fuse members 
while the case ?SC 0.5-A? might be a new design with weakened self-centering hinge 
zone. The location of fuse members are indicated in Figure 5-3(b). The supports at 
node L1 and L6 are pin supported. 
Some explanations are in order for the adopted strategy in selecting the locations of 
self-centering fuse members which collectively would create self-centering hinge 
zones at particular region. For example, three neighbouring self-centering fuse 
members (one upper chord, one lower chord, one web diagonal) could form a self-
 centering hinge zone in a truss.  
Note that the strength of the self-centering fuse members was reduced to 70 % of the 
critical load (i.e., buckling force) of a compression member with imperfection 
(around 40% of its original strength). Although the reduced member strength is lower 
than the critical load, it is worth noting that self-centering member still maintain a 
stable load carrying capacity and stiffness after ?yielding? while after member 
 
 85 
 
buckles, both strength and stiffness of regular members degrade rapidly. This is the 
essential distinction between using weakened self-centering members and regular 
buckled member even though the critical load for the buckled member is slightly 
greater. This distinction leads to the favorable behavior of steel arch structure with 
self-centering members.  
An analysis has been conducted to demonstrate the benefit of the fuse member 
selecting and weakening strategy. The nonlinear dynamic analysis on the original 
structure without fuse members under 20 earthquake records was conducted.  For 
each earthquake record, the member failures were counted and listed in Table 5-3. 
The failure occurrence varies from 1 to 11. And after check, it?s found that 31 
locations have ever failed under at least one earthquake records. The failure locations 
are indicated in Figure 5-5, where are marked in red. Due to the uncertainty of 
occurrence of earthquakes, it?s needed to consider these 31 members as possible 
failure locations after an earthquake disaster. The numbers of failure occurrences are 
also indicated on each failure location. It can be seen that failure occurrence on 
member L1L2 and L3L4 are 10 and 11 respectively. Thus these two locations are 
selected as fuse member locations. By incorporating weakened FLD member as fuse 
member, the damages can be concentrated only on the four pre-selected members and 
thus bring the ease to inspect and replace damaged member after earthquakes. 
 
 
 86 
 
5.4.1.  Nonlinear static analysis 
Nonlinear static analysis provides an insight into the expected seismic performance of 
the structure under specified lateral load profile. A pushover analysis of the arch truss 
was performed to identify the nonlinear static response of the arch structure with self-
 centering fuse members. The inverted triangular vertical load pattern was used to 
distribute the base shear at several nodes on outer chords along the height of the arch 
in pushover analysis according to its first mode shape. The loads were increased 
monotonically until the horizontal displacement at the top node of the arch structure 
reached to the drift ratio of 2%.   
The roof drift ratio (top node displacement divided by the overall height of the arch 
structure) vs. the base shear also known as pushover curve, recorded during the 
nonlinear static analysis is shown in Figure 5-6. It is seen that the slope of the post 
yield branch of the pushover curve changes as the chord members enter buckling 
phases. During the analysis, the base shear increases linearly until the roof drift ratio 
reaches about 0.5% and the base shear reaches 189 kN. Then the base shear decreases 
quickly as the drift ratio continues increase, suggesting the arch structure losing load 
carrying capacity and fail. On the contrary, the base shears for the three arch 
structures with fuse members did not degrade after the turning point. Case SC 0.5-B 
shows that by increasing all members 1.33 times than SC 0.5-A, the base shear 
turning point can be increased to a level comparable to the original structure. After 
checking member responses, members L4L5 was found to buckle in case Original 
case. This cause the base shear drops at around 0.5% drift ratio. However, after 
member L4L5 buckled, no other members yield or buckled. It?s seen from the Figure 
 
 87 
 
5-6 that no strength or stiffness degradation in the case SC 0.5-A and SC 0.5-B. This 
implies that no other members buckled after the base shear turning point in the two 
SC cases. The post yielding stiffness for Case Conv. FM is lower than two SC cases 
because the material for the fuse member has lower strain-hardening ratio. The 
stiffness of Case SC 0.5-B before yielding is slightly lower than the original case due 
to the reduced section for those fuse members compared to the substituted members 
in the original case. 
5.4.2.  Nonlinear time history analysis 
The performance of a structural system under strong earthquakes can be evaluated by 
examining the nonlinear time history analysis results of the key engineering demand 
parameters (EDPs). For this purpose, nonlinear time history analysis of the arch 
structure was performed under a suite of 20 strong earthquake records, which were 
originally developed by Somerville et al. (1997) for the Los Angeles, California 
region with a probability of exceedance of 2% in 50 years. These scaled earthquake 
records correspond to the maximum credible earthquake of the downtown Los 
Angeles area. Details of these earthquake records are listed in Table 5-4. These 
acceleration time histories have been derived from historical recordings or from 
physical simulations and have been altered so that their mean response spectrum 
matches the 1997 NEHRP design spectrum, modified from soil type of SB-SC to soil 
type SD and having a hazard specified by the 1997 USGS maps. The pseudo-
 acceleration response spectra for these earthquake records are shown in Figure 5-7. 
The time interval is set as 0.001 seconds for the time history analysis. 
 
 88 
 
5.5.  Results and discussions 
The results of important EDPs are presented in this section. In the results presented in 
this sections, the ?peak? result refers to the highest absolute value of the specified 
parameter (e.g. drift ratio, etc.) recorded during the time history analysis. The 
?ensemble average? refers to the mathematical average value of the results obtained 
from the twenty time history analyses which is shown in some figures.   
Figure 5-8 shows the peak and residual arch top displacement responses of the arch 
structures under the twenty seismic records. It?s seen that use of conventional fuse 
members decreases the average peak displacement response from 0.37 m for the 
original structure to 0.35 m for the Conv. FM case. However, the Conv. FM case is 
seen to have 0.027 m residual displacement on average, while the original structure 
case has 0.09 m residual displacement on average. It is observed that self-centering 
fuse members can significantly decrease the residual displacement response to zero. 
Case SC 0.5-A has slightly higher average peak displacement response than the 
original structure case due to reduced fuse member section sizes and thus global 
stiffness of the arch structure. Average peak displacement for Case SC 0.5-B is 0.38 
m, which is lower than that of Case SC 0.5-A. 
Figure 5-9 shows the peak base shear and arch top acceleration response of the arch 
structures under the 20 earthquake records. It?s seen that the all three cases with fuse 
members have lower base shear force than the original structure. Use of fuse 
members decreases the base shear force from 200 kN for the original structure to 
around 162 kN for the conventional FM case. The base shear forces for two self-
 
 89 
 
centering fuse member cases are quite close. Use of fuse members also decrease the 
peak arch top acceleration response from 10.4 m/s2 for the original structure case to 
8.96m/s2 for Case Conv. FM and 9.15 m/s2 for Case SC 0.5-A . The average peak 
arch top acceleration for Case SC 0.5-B is 9.9 m/s2, which is still lower than the 
original structure.  
Since it is not possible to present all detailed results of all time history analyses, 
LA11 is chosen as a representative case for the nonlinear time history analysis and 
some of its detailed time history results are presented here. Figure 5-10 shows the 
displacement time histories of node U3, which has the largest displacement response 
and thus is used as a reference point. For the original structure case, its peak 
horizontal displacement response reaches 0.52 m. It is seen that there is a sharp 
downward turn in the displacement response curve, indicating occurrence of member 
failure around 5 seconds. After checking member responses, members L4L5 was 
found to buckle at that moment and thus cause the sudden horizontal motion. This 
also led to the 0.23-m residual displacement observed at the end of seismic record 
duration in the original arch structure. It is worth noting that strength reduction has 
been intentionally made in the fuse member members in order to activate these 
members for energy dissipation and control of member forces. Consequently, this 
leads to reduced load demand on their neighbouring members and thus prevents the 
failure of these members. However, due to its bilinear hysteresis, a residual 
displacement as large as 0.03-m is observed for the conventional fuse member case in 
Figure 5-10. In order to minimize the residual displacement, the self-centering fuse 
members are used and the corresponding results are shown in Figure 5-10.  
 
 90 
 
Figure 5-11 show the horizontal acceleration of node U3 for all four cases. The use of 
both conventional fuse members and self-centering fuse members is able to decrease 
the acceleration response of the arch structure. Clearly, incorporating fuse members 
into the steel arch structure can effectively control its seismic responses under strong 
earthquake loading.  
5.6.  Conclusions 
This research presents the results of nonlinear dynamic analysis of steel arch 
structures subjected to earthquake load. Both material and geometric nonlinearities 
are considered in the simulation study. A 2-dimensional force-based element with 
corotational formulation and fiber discretization of the cross section is used to 
simulate member buckling and associated strength degradation and large. Four cases 
were considered for the arch structures, distinguished by the type of fuse members 
they used or no fuse members: Original Structure without any fuse members, 
Conventional fuse members, self-centering fuse members (two cases: SC 0.5-A and 
SC 0.5-B). The fuse members are installed at strategically selected locations to 
substitute critical members, and designed to provide a high level of member ductility. 
Through weakening of the steel arch structure by using fuse members of reduced 
strength (about 30% to 40% of their original member strength), deformation and 
energy dissipation can be concentrated to a pre-selected self-centering hinge zone 
formed by the fuse members while the entire arch structure?s integrity is retained 
during strong earthquake event. Findings from this study include:  
 (1) Installing fuse members at intentionally weakened locations in steel arch 
 
 91 
 
structures can effectively control the displacement response of the arch structure 
under earthquake loading. Unlike the original critical member susceptible to buckling, 
the fuse member could maintain its strength and exhibit ductile load behaviour. The 
strengths of fuse members usually have to be reduced (compared to the critical load 
of its original substituted member) to prevent the failure (e.g., buckling) of their 
neighboring members in the arch structure. The structural weakening strategy can 
concentrate the ductile damage of the overall structure on the selected favorable 
locations where are easy for damage inspection and structural repair post earthquake 
events. 
(2) The use of conventional fuse members exhibiting bilinear hysteresis results in 
residual displacement in the steel arch structure. Self-centering fuse members with 
flag-shaped hysteresis is very effective to eliminate the residual displacement.  
(3) As an alternative that could be used for new design, all the member size could be 
proportionally increased as demonstrated by Case SC 0.5-B. It?s seen that the global 
structure strength is increased to the same level as that of the original structure. By 
doing so, the displacement response of the arch structure can be further reduced 
without any residual displacement after strong earthquake events. 
  
 
 92 
 
Table 5- 1. Section sizes of the arch structure?s tubular member 
Member Outer diameter, OD (mm) Tube Wall Thickness (mm) 
F1 75.5 3.75 
F2 88.5 4 
F3 114 4 
F4 140 4.5 
X1 88.5 4 
X2 114 4 
X3 140 4.5 
X4 159 6 
X5 159 8 
S1 159 6 
 
 
Table 5- 2. Natural frequency of the steel arch structure with and without fuse 
members (unit = Hz) 
 1st mode 2nd mode 3rd mode 4th 
mode 
5th 
mode 
6th 
mode 
Original 0.706  1.724 3.424 4.975 6.667 7.092 
Conv. FM 0.626 1.618 3.425 4.831 6.329 6.711 
SC 0.5-A 0.625 1.617 3.424 4.829 6.325 6.701 
SC 0.5-B 0.688 1.726 3.817 5.405 7.246 7.634 
 
 
Table 5- 3. Numbers of member failures under each earthquake records 
Code No. of members failed Code No. of members failed 
LA01 5 LA11 7 
LA02 2 LA12 2 
LA03 7 LA13 8 
LA04 3 LA14 5 
LA05 7 LA15 2 
LA06 2 LA16 2 
LA07 1 LA17 6 
LA08 2 LA18 11 
LA09 10 LA19 1 
LA10 9 LA20 5 
 
 93 
 
 
Table 5- 4. Earthquake ground motion records used in this study (scaled to the 
maximum credible earthquake for Los Angeles, California) 
Code Original EQ. Record 
EQ. 
Magnitude 
Epicentre  
Distance  
    (km) 
Scale 
Factor 
PGA  
(cm/s2) 
Vertical 
PGA 
(cm/s2) 
LA01 
Imperial Valley, 1940, El 
Centro 6.9 10 3.02 678.05 494.25 
LA02 
Imperial Valley, 1940, El 
Centro 6.9 10 3.02 994.32 724.79 
LA03 
Imperial Valley, 1979, Array 
#05 6.5 4.1 1.52 579.06 422.09 
LA04 
Imperial Valley, 1979, Array 
#05 6.5 4.1 1.52 717.98 523.36 
LA05 
Imperial Valley, 1979, Array 
#06 6.5 1.2 1.26 443.54 323.31 
LA06 
Imperial Valley, 1979, Array 
#06 6.5 1.2 1.26 345.12 251.57 
LA07 Landers, 1992, Barstow 7.3 36 4.80 619.47 665.59 
LA08 Landers, 1992, Barstow 7.3 36 4.80 626.24 672.86 
LA09 Landers, 1992, Yermo 7.3 25 3.26 764.55 821.47 
LA10 Landers, 1992, Yermo 7.3 25 3.26 530.03 569.49 
LA11 Loma Prieta, 1989, Gilroy 7 12 2.69 978.74 713.43 
LA12 Loma Prieta, 1989, Gilroy 7 12 2.69 1426.40 1039.75 
LA13 Northridge, 1994, Newhall 6.7 6.7 1.55 997.40 730.59 
LA14 Northridge, 1994, Newhall 6.7 6.7 1.55 966.74 708.13 
LA15 Northridge, 1994, Rinaldi RS 6.7 7.5 1.19 784.95 574.97 
LA16 Northridge, 1994, Rinaldi RS 6.7 7.5 1.19 852.87 624.72 
LA17 Northridge, 1994, Sylmar 6.7 6.4 1.49 837.65 613.57 
LA18 Northridge, 1994, Sylmar 6.7 6.4 1.49 1202.16 880.57 
LA19 North Palm Springs, 1986 6 6.7 4.46 1499.15 1092.78 
LA20 North Palm Springs, 1986 6 6.7 4.46 1451.42 1057.99 
 
 
 
 94 
 
 
Figure 5- 1. Configuration of brace model considering mid-length mass 
 
 
 
 
 
 
 
 
 
       (a) Regular member              (b) Conventional FM           (c) Self-centering FM 
Figure 5- 2. Hysteresis model of buckled member, conventional fuse member (FM) 
and self-centering FM 
 
 
 
 
 
 
 
Beam-Column 
Element
 Mid-mass
 Load, P
 Initial camber
 80 in (2.03m)
 -0.15 -0.1 -0.05 0 0.05 0.1
 -400
 -200
 0
 200
 400
 600
 800
 Displacement (m)
 F
 or
 ce
  (k
 N
 )
 Regular Member
 -0.1 -0.05 0 0.05 0.1
 -300
 -200
 -100
 0
 100
 200
 300
 Displacement (m)
 F
 or
 ce
  (k
 N
 )
 Conv. FM
 -0.15 -0.1 -0.05 0 0.05 0.1
 -300
 -200
 -100
 0
 100
 200
 300
 Displacement (m)
 F
 or
 ce
  (k
 N
 )
 SC FM
 
 95 
 
 
(a) Elevation view (without fuse members) 
 
 
(b) Elevation view (with fuse members) 
Figure 5- 3. Schematics of steel arch structure and location of fuse members 
 
 
 
96 m
 38
  m
 1.95 m
 U1
 U2
 U3
 L2
 L3 L4
 L5
 U4
 U5
 S1
 S1
 S1
 S1
 S1
 S1
 S1
 S1
 S1
 S1
 S1
 S1
 S1
 S1
 S1 S1 S1
 X4
 X4
 X5
 X4
 X4
 X4 X4
 X5
 X5
 X5
 X5
 X5
 X3
 X3 X2 X2 X1
 F4
 F2
 F4
 F1
 F4
 F1
 F3
 F1
 F3
 F1
 F2
 F2
 F1 F3
 F1 F3
 F1 F3
 F1 F3
 F1 F2
 F1 F2
 F1 F2
 F1
 F2
 F1
 F2
 F1
 F2
 F2
 F2
 L1 L6
 96 m
 38
  m
 1.95 m
 U1
 U2
 U3
 L2
 L3 L4
 L5
 U4
 U5
 FM
 FM
 FM FM
 L1 L6
 
 96 
 
 
(a) 1st mode (0.706 Hz)         (b) 2nd mode (1.724 Hz)         (c) 3rd mode (3.424 Hz) 
Figure 5- 4. First three mode shapes of the arch structure 
 
 
 
 
 
Figure 5- 5. Locations of member failure under 20 earthquake records. (Numbers 
indicate the member of failure occurrences in total 20 earthquakes.) 
1
 11
 10
 5
 52
 111
 1
 1
 2
 1
 1
 5
 3
 11
 1
 4
 9
 101
 3
 3
 3
 2
 2
 U1
 U2
 L1
 L2
 L3
 L4
 U3
 U4
 U51
 1
 
 97 
 
 
Figure 5- 6. Static pushover curve of the arch structure 
 
 
Figure 5- 7. Pseudo-Acceleration response spectra of the seismic records in Table 5-4 
(viscous damping ratio=0.05) 
0 0.004 0.008 0.012 0.016 0.02
 0
 50
 100
 150
 200
 250
 300
 Arch top drift ratio
 B
 as
 e 
sh
 ea
 r (
 kN
 )
  
 
Original
 Conv. FM
 SC 0.5-A
 SC 0.5-B
 0 0.5 1 1.5 2 2.5 3 3.5 4
 0
 10
 20
 30
 40
 50
 60
 Natural period Tn (Sec.)
 P
 S
 A
  (m
 /s
 2 )
 
 98 
 
 
(a) Peak arch top displacement 
 
(b) Residual arch top displacement 
Figure 5- 8. Peak and residual arch top displacement of the arch structure 
0
 0.2
 0.4
 0.6
 0.8
 1
 LA
  01
 LA
  02
 LA
  03
 LA
  04
 LA
  05
 LA
  06
 LA
  07
 LA
  08
 LA
  09
 LA
  10
 LA
  11
 LA
  12
 LA
  13
 LA
  14
 LA
  15
 LA
  16
 LA
  17
 LA
  18
 LA
  19
 LA
  20
 Earthquake Ground Motion
 P
 ea
 k 
ar
 ch
  to
 p 
ho
 riz
 on
 ta
 l d
 is
 pl
 ac
 em
 en
 t
 (m
 )??
 Original
 Conv. FM
 SC 0.5-A
 SC 0.5-B
 Original
 Conv. FM
 SC 0.5-A
 SC 0.5-B
 0.000
 0.050
 0.100
 0.150
 0.200
 0.250
 0.300
 LA
  01
 LA
  02
 LA
  03
 LA
  04
 LA
  05
 LA
  06
 LA
  07
 LA
  08
 LA
  09
 LA
  10
 LA
  11
 LA
  12
 LA
  13
 LA
  14
 LA
  15
 LA
  16
 LA
  17
 LA
  18
 LA
  19
 LA
  20
 Earthquake Ground Motion
 R
 es
 id
 ua
 l a
 rc
 h 
to
 p 
ho
 riz
 on
 ta
 l
 di
 sp
 la
 ce
 m
 en
 t (
 m
 )
 Original
 Conv. FM
 SC 0.5-A
 SC 0.5-B
 SC 0.5-A
 SC 0.5-B
 Original
 Conv. FM
 
 99 
 
 
(a) Peak base shear 
 
(b) Peak arch top point acceleration 
Figure 5- 9. Peak base shear and acceleration response of the arch structure 
 
 
0
 50
 100
 150
 200
 250
 300
 350
 400
 LA
  01
 LA
  02
 LA
  03
 LA
  04
 LA
  05
 LA
  06
 LA
  07
 LA
  08
 LA
  09
 LA
  10
 LA
  11
 LA
  12
 LA
  13
 LA
  14
 LA
  15
 LA
  16
 LA
  17
 LA
  18
 LA
  19
 LA
  20
 Earthquake Ground Motion
 P
 ea
 k 
ba
 se
  s
 he
 ar
  (k
 N
 ) ?
 Original
 Conv. FM
 SC 0.5-A
 SC 0.5-B
 Conv. FM
 Original
 SC 0.5-A
 SC 0.5-B
 0
 5
 10
 15
 20
 25
 LA
  01
 LA
  02
 LA
  03
 LA
  04
 LA
  05
 LA
  06
 LA
  07
 LA
  08
 LA
  09
 LA
  10
 LA
  11
 LA
  12
 LA
  13
 LA
  14
 LA
  15
 LA
  16
 LA
  17
 LA
  18
 LA
  19
 LA
  20
 Earthquake Ground Motion
 P
 ea
 k 
ar
 ch
  to
 p 
ho
 riz
 on
 ta
 l a
 cc
 el
 er
 at
 io
 n
 (m
 /s
 2 )
 Original
 Conv. FM
 SC 0.5-A
 SC 0.5-B SC 0.5-B
 SC 0.5-A
 Original
 Conv. FM
 
 100 
 
 
 
Figure 5- 10. Displacement response of the arch structure under LA11 record: 
horizontal displacement at arch top node U3 
 
 
 
 
 
 
0 10 20 30 40 50 60 70 80
 -0.6
 -0.4
 -0.2
 0
 0.2
 0.4
 0.6
 (a) Original
 A
 rc
 h 
to
 p 
ho
 riz
 on
 ta
 l d
 is
 pl
 ac
 em
 en
 t (
 m
 )
 0 10 20 30 40 50 60 70 80
 -0.7
 -0.5
 -0.3
 -0.1
 0.1
 0.3
 0.5
 0.7
 (b) Conv. FM
 0 10 20 30 40 50 60 70 80
 -0.7
 -0.5
 -0.3
 -0.1
 0.1
 0.3
 0.5
 0.7
 (c) SC 0.5-A
 0 10 20 30 40 50 60 70 80
 -0.7
 -0.5
 -0.3
 -0.1
 0.1
 0.3
 0.5
 0.7
 Time (Sec.)
 (d) SC 0.5-B
 
 101 
 
 
 
 
 
Figure 5- 11. Acceleration response of the arch structure under LA11 record: 
Horizontal acceleration at arch top node U3 
  
0 10 20 30 40 50 60 70 80
 -10
 -5
 0
 5
 10
 15
 (a) Original
 A
 rc
 h 
to
 p 
ho
 riz
 on
 ta
 l a
 cc
 el
 er
 at
 io
 n 
(m
 /s
 2 )
 0 10 20 30 40 50 60 70 80
 -10
 -5
 0
 5
 10
 15
 (b) Conv. FM
 0 10 20 30 40 50 60 70 80
 -10
 -5
 0
 5
 10
 15
 (c) SC 0.5-A
 0 10 20 30 40 50 60 70 80
 -10
 -5
 0
 5
 10
 15
 (d) SC 0.5-B
 Time (Sec.)
 
 102 
 
Chapter 6: Case Study I: Demonstration of WSCA System 
on Truss Structure 
 
6.1.  Introduction 
In this chapter the implementation of the wireless structural condition assessment 
(WSCA) system will be demonstrated on a truss structure through experimental test. 
After the BT strain sensors were validated using the countertop test setup described in 
Chapter 3, it?s natural to further test the performance of the WSCA system on a more 
realistic structure in order to accomplish the following three goals: (i) validate the BT 
strain sensors installed on a large-scale structural member in a structure subjected to 
load; hybrid simulation test will be conducted for this purpose; (ii) investigate the 
various aspects of the WSCA system including the integration of the wireless sensors 
(e.g., BT strain sensors) and the BIM-based computing environment; (iii) validate and 
evaluate the performance characteristics of the BIM based WSCA system in a real 
structure setting such as demonstrating its capability of quickly locating and 
visualizing the strain in the selected members. 
To faithfully capture the essential structural behavior of monitored structure on which 
the sensor will be installed and its performance will be evaluated, it is desired that the 
test structure shall be fabricated at or near full scale. Conventional testing technique 
for experimentally evaluating structures usually involves fabrication of large-scale 
structural assembly and simplification of loading patterns on the test structure. While 
this is a proven method of testing, it is very costly and requires the use of testing 
 
 103 
 
equipments that are available at limited number of large-scale testing facilities across 
the country.  A hybrid simulation method provides a more efficient and cost effective 
alternative to conventional testing methods. 
The use of this hybrid simulation method for structural condition assessment will be 
demonstrated through assessing the truss structure with tubular members in this 
chapter. Specifically, Section 6.2 describes the hybrid testing methodology. The 
details of the test setup for the hybrid simulation test conducted on the truss structure 
are given in Section 6.3. The WSCA implemented on the truss structure under hybrid 
simulation test and its performance observed during the test is discussed in Section 
6.4. 
6.2.  Experimental test: hybrid simulation 
The concept behind hybrid testing of a large scale structure like a roof or bridge truss 
is that a large portion of the structure can be easily simulated using the fundamental 
behaviors of the materials and geometries of the structural elements.  However, 
certain parts of the truss may be difficult to numerically model and simulate (such as 
a truss member susceptible to buckling).  Hybrid testing solves this problem by 
combining experimental testing of the large scale truss member susceptible to 
structural damage with a computer model that simulates the remainder of the structure 
and external loading.  By experimentally testing the selected member of the truss 
structure, a more realistic and controlled testing environment can be applied in testing 
structural members at large scale and validating structural sensors. 
 
 104 
 
The prototype truss shown in Figure 6-1 has a span length of 14.7 ft (4.47m). A 3D 
model of the prototype truss was built in Revit. An overall view of the Revit model of 
the truss is shown in Figure 6-1.  It?s noted that only the experiment member 
indicated in the figure will be tested while other members will be simulated in the FE 
software. Concentrated load was applied at the mid-span joint of the lower chords and 
the load was monotonically increased until the upper chord failed or the BT strain 
sensor mounted on the physically tested truss member was triggered.  Strain gages 
were also installed on the test truss member. During the test, synchronized recording 
of the strain gage and BT strain sensor data was carried out until a threshold strain 
exceedence signal is received from the BT strain sensor.  
6.2.1. Software architecture for hybrid simulation 
To perform hybrid simulation test, a software framework needs to be first established 
to integrate the physical testing elements with the computer model. Figure 6-2 shows 
the architecture of such a software framework that depicts its basic components and 
their interrelationships. The requirements for this middleware software framework are: 
ease in operation, and compatibility with different types of experimental system. It 
shall allow researchers to carry out hybrid simulation without specialized knowledge 
about the underlying software. Also, it shall enable any interested user to custom 
develop and extend it to other field by adding new control system, application model, 
and integration modules. 
During a hybrid test, the first task for the software framework is to transform the 
degree-of-freedoms from the coordinate system of the finite element (FE) software to 
 
 105 
 
those of the test specimens being physically tested, by considering the geometry and 
kinematics of the system. Subsequently, another class is responsible for 
communication between the laboratory control and data acquisition systems. For 
geographically distributed hybrid testing, another class is needed to facilitate the 
communication between distributed experimental sites and the master site which runs 
the computational software and works as the coordinator in the test (Schellenberg et 
al. 2009).  
OpenFresco is one of the hybrid test software frameworks which have a modular 
architecture. Each module can be modified and new ones can be added without 
affecting other modules. In OpenFresco, a module is called a class, including 
ExperimentalElement, ExperimentalSite, ExperimentalSetup and Experimental-
 Control classes. The relationship between these classes is shown schematically in 
Figure 6-2 (Schellenberg et al. 2006a, 2008). 
OpenFresco is independent of the FE software used, meaning that any FE software 
that allows the addition of new elements can be used, such as Abaqus, LS-Dyna, 
OpenSees, Matlab, etc. For example, OpenSees can be readily used with OpenFresco 
because of its object oriented design methodology. Unlike a pure numerical analysis 
conducted in OpenSees, the difference in hybrid testing is to replace the numerical 
element with the ExperimentElement from OpenFresco and add the numerical 
integration operators (Schellenberg et al. 2006b). 
6.2.2. Hybrid simulation test setup  
Hybrid simulation seamlessly combines the analytical and physical parts of a 
 
 106 
 
structure and simulates their response with a number of advantages over traditional 
large scale structural testing method. In contrast to a conventional numerical 
simulation, in which the entire structure is modeled numerically, the hybrid testing 
method obtains some of the forces corresponding to either restoring force or damping 
from a concurrent physical test performed on the real structure or selected 
components. The hybrid testing method therefore gives the researches the flexibility 
to model the well understood parts of a structure using a FE program on a computer, 
leaving the construction and testing of a component that is the highly nonlinear and/or 
numerically difficult to model to physical parts of the structure in the laboratory. 
A hybrid simulation test setup to simulate the nonlinear response of large scale frame 
structures under severe transient loadings (e.g., extreme wind load from downburst) 
established at the University of Maryland (UMd) is shown in Figure 6-3. The goal of 
this hybrid simulation test setup is to investigate and validate the application of 
hybrid simulation method to faithfully reproduce the truss member damage behavior 
such as buckling and effect on the collapse of full-scale metal frame structures under 
transient loading. The UMd hybrid testing system is based on the sub-structuring 
pseudo-dynamic testing concept, in which the dynamic behaviour of the test structure 
is simulated on a computer by solving the equations of motion during the test using 
the structural restoring forces measured from the nonlinear substructure being 
physically tested. The hybrid testing system combines physical testing and model-
 based simulation in a flexible manner for an efficient and realistic evaluation of the 
nonlinear behaviour of a structural system subjected to extreme loading or dynamic 
excitation.  
 
 107 
 
This hybrid simulation test setup had been demonstrated the capability of hybrid 
testing technique in faithfully capturing the post-buckling behaviour of a long span 
arch truss structure subjected to transient wind loading. The details about these 
validations tests can be found in the work by Bai and Zhang (2010). This hybrid 
simulation test setup was also used in the present experimental study on the WSCA 
system installed for the prototype truss. However, monotonic loading was applied on 
the truss for simplicity.  
In this hybrid simulation system, the OpenFresco software framework for hybrid 
simulation is used along with an MTS FlexTest 60 controller and hybrid testing driver. 
The OpenSees software program is used to solve for the displacement vector 
satisfying the governing equation of motion of the structural model at each time step. 
The hardware part of the hybrid testing system include: three MTS actuators with 
servo-control (two 55-kips (244 kN) double-rod actuators with 30-GPM (114L/Min) 
servovalve, a 60-GPM (227 L/Min) hydraulic power supply, three 50-GPM (189 
L/Min) hydraulic service manifolds, a 4-channel servo-control system with FlexTest 
60 controller and MTS test software. 
The advantages of using hybrid simulation to assess the ultimate structural behavior 
(e.g., collapse) of structural systems are numerous: physical masses are removed from 
the experimental setup as they are modeled numerically, improving the safety of the 
tests, especially at imminent collapse stage; the specimens can be tested at large scale; 
only the vulnerable substructure (e.g., buckled member) is tested experimentally 
while the remaining structure is modeled numerically, thus improving the economy of 
the tests; etc. Those benefits make hybrid simulation test a safe, efficient and less 
 
 108 
 
expensive alternative for experimental evaluation of ultimate structural behavior 
(Wang et al. 2009). 
6.3.  Component validation test on truss member  
Before conducting a full scale hybrid simulation test, a validation test with a simple 
test setup is needed to validate and debug the BT strain sensor which is now installed 
on a large scale structural member. This simple test setup includes a truss member 
made of Poly (vinyl chloride) (PVC) tube under monotonic loading. This PVC tube 
has an outer diameter of 4.5 inches (11.43 cm) and an inner diameter of 4 inches 
(10.16 cm). The length of the PVC tube is 2 ft (0.61m). The modulus of elasticity of 
the PVC material was measured to be 160 ksi (1.1 GPa).  
The test setup of this component validation test is shown in Figure 6-4. The test setup 
includes the following components: 2 strain gages (Vishay model #: EA-06-250AE-
 350), BT strain sensor with RFID tag antenna, UHF RFID reader (Skyetek model #: 
SkyeModule M9). The universal test machine used for this test is a SATEC 400-kips 
(181.4 kN) servo-hydraulic test machine (serial #: I-series 400BTE/483280). As 
shown in Figure 6-4, the Skyetek RFID reader was placed at a distance of 75 inch 
(190.5 cm) away from the PVC tube. Two strain gages were bonded to the mid-length 
point of the PVC tube to measure the axial strain of the PVC tube during the test. The 
two strain gages are positioned at 180? from each other along the circumference of the 
tube to cancel out any eccentricity induced strain if there is any. The strain gages also 
provide a calibration tool for the BT strain sensor by comparing synchronized 
recordings of each other. A video camera was used to record the strain reading during 
 
 109 
 
the test in order to identify when the RFID tags were triggered. The RFID tags were 
connected to the BT strain sensors using AWG 26 ribbon wires, as shown in the inset 
of Figure 6-4.  
The BT strain sensors were directly cut from a 6 by 12 inches (15.24 by 30.5 cm), 
0.08 inches (2 mm) thick acrylic sheet using a carbon dioxide laser cutter (CO2 
cutter). The distance between the two end fixtures of the BT strain sensor is 2 inches 
(5.08 cm), which is considered as the gage length Ls of the BT strain sensors. The 
dimensions of the BT strain sensors are shown in Figure 6-5. As reported in Chapter 
3, FE simulation is conducted to find out the maximum axial deformation in the 
sensor that will break the brittle bar. As aforementioned, BT strain sensor failure is 
defined as the maximum first principal strain in the BT strain sensor first exceeds 
3%. The design values for H9W2 and H6W4 from the FE analysis are listed in 
Table 6-1. These design values are calculated by dividing the relative displacement 
between the two end fixtures from the FE simulation with the BT strain sensor gage 
length Ls (which is 2 in or 5.08 cm in this study). 
The two BT strain sensors (labeled as H6W4 and H9W2 respectively) were attached 
to the PVC tube using screws and glue bonding at their ends, as shown in Figure 6-6. 
The two BT strain sensors were stacked on each other, this ?stacking? configuration 
is intended to minimize the effect of cracking of H6W4 on H9W2 in the BT strain 
sensor, which was observed in the test of a monolithic design and caused 
complication in the BT strain sensor response. It was observed that after H6W4 
fractured, rotation about its pivot point pushed the acrylic sheet outward and made the 
H9W2 break before its design value. 
 
 110 
 
Three identical component validation tests were conducted on PVC tubes. The 
loading rate for the axial load applied to the PVC tube was 1 kips/min (4.45 kN/min). 
The axial strains recorded in the tests are plotted in Figure 6-7. The trigger strain 
values of the BT strain sensors measured from the three component validation tests, 
FEM simulation based design analysis, and countertop tests are listed in Table 6-1 for 
comparison purpose. It is seen that the measured data of the two strain gages are very 
close to each other, suggesting that there is very little eccentricity effect and bending 
moment in the PVC tube. It is also found that the trigger strain levels from all three 
tests are smaller than the countertop test results. Considering the different end fixture 
configurations for the BT strain sensors in the PVC tube test and countertop test (see 
Chapter 3), this discrepancy may be attributed to the fact that after H6W4 fractured, 
rotation about its pivot point pushed the acrylic part outward. Photos of the fractured 
BT strain sensors after triggering are shown in Figure 6-8. It is seen that their failure 
mode is identical to that observed in the countertop tests reported in Chapter 3. 
6.4.  System validation test on prototype truss  
This section presents an experimental study focused on the implementation and 
validation of the WSCA system through hybrid testing of the aforementioned truss 
structure under monotonic loading. The experimental findings provide valuable test 
data of the WSCA system performance, which can be used for identifying 
challenging issues associated with the full scale implementation of WSCA system.  
For the convenience of specimen fabrication and safety in trial test, a planar truss 
made of PVC tubes was used in the experimental study described below. The 
configuration and dimension of the planar truss and external loads are shown in 
 
 111 
 
Figure 6-9. The left-side top chord (highlighted in red dash line in the figure) of the 
truss was designated as the experiment element which was to be physically tested 
during the hybrid simulation test. The ExpBeamColumn2d experimental element in 
the OpenFresco was used to represent this physical member (i.e., the PVC tube to be 
physically tested). The PVC tube member has an outer diameter of 4.5 inches (11.43 
cm) and an inner diameter of 4 inches (10.16 cm). The length of the PVC tube is 44 
inches (1.12 m). The modulus of elasticity of the PVC material was measured to be 
160 ksi (1.1 GPa). The support, joints, and external force were all modeled 
numerically in the OpenSees software program. 
The experiment setup is shown in Figure 6-10. The test setup includes the following 
components: 2 strain gages (Vishay model #: EA-06-250AE-350), BT strain sensor 
with RFID tag antenna, UHF RFID reader (Skyetek model #: SkyeModule M9), 
Autodesk Revit Structure, and a custom developed middleware. The MTS hybrid 
simulation test setup described in Section 6.2 was used for this experimental study. 
The Skyetek RFID reader is placed at 30 inch (76.2 cm) away from the PVC tube. 
The inset of Figure 6-10 shows close-up view of the RFID tags and BT strain sensors 
installed on the test specimen.  A single-actuator setup was adopted in this hybrid 
simulation test to control the degree of freedom in the axial direction of the truss 
member because the original member in the OpenSees model is a truss bar. 
Displacement control was used in the hybrid test and the control point, which is the 
middle-span joint at the lower chord level of the truss, was pushed downward 
gradually from its original position. The total duration of the hybrid testing lasted 
about 10 minutes, which depends on the value of the ramping time set for the 
 
 112 
 
hydraulic system and controller. A shorter ramping time leads to faster loading rate.  
During the test, a displacement increment of 0.01-inch (0.254 mm) was imposed at 
each step to the test member. For all other elements in the numerical model, their 
cross-sectional area is assumed to be 3.34 in2 (21.5 cm2).  A video camera was placed 
in order to record the failure process of the physically tested member during the 
experiment. 
Two identical hybrid simulation tests were conducted for this experimental study on 
full-scale implementation and validation test of the WSCA system. Figures 6-11 to 6-
 16 show the test results from these two hybrid simulation tests. Figure 6-11-(a) shows 
the measured strain data as well as the trigger strains of the two BT strain sensors 
(labeled H6W4 and H9W2 respectively) during the first hybrid simulation test (HS 
Test #1). For independent verification purpose, the strain of the PVC tube member 
derived from the displacement data measured with the built-in LVDT transducer in 
the MTS actuator (calculated by dividing the measured displacement with the PVC 
tube length (44 in or 111.76 cm)) is also included in the figure. It?s seen that all of 
these strain data agree very well with one another. This suggests that the bending 
moment in the PVC tube due to eccentricity is very small, and thus the section where 
the BT strain sensor is bonded on the PVC tube is subject to uniform axial 
deformation. The square marks in Figure 6-11 identify the trigger point for the two 
BT strain sensors. As shown in Figure 6-10, two BT strain sensors were stacked on 
each other after installation on the PVC tube. The BT strain sensor H6W4 which has 
a lower trigger strain level was triggered at 6,486 ??, (corresponding to the average of 
the two strain gage readings at trigger).  
 
 113 
 
Figures 6-12 to 6-13 show the hybrid simulation test setup and screen shots of BIM 
user interface at the trigger of the BT strain sensors respectively, during the first 
hybrid simulation test HS Test #1. In the BIM user interface where the Revit structure 
model is shown, two colors were used to highlight the trigger strain level - blue or red. 
When the member concerned is marked with a blue color bar, the lower level BT 
strain sensors H6W4 is known to be triggered. If it?s marked with a red color bar, the 
higher level BT strain sensor H9W2 has also been triggered. Figure 6-12 (b) shows 
the properties of the selected member in the hybrid simulation test.  For example, a 
unique RFID tag info is shown there which can be used to rapidly identify the 
member and BT strain sensor. It also includes a ?DataTag? which stores the strain 
level and time when the BT strain sensor is triggered. These properties will be 
updated automatically during the test. It is worth noting that other customized 
properties can be included easily depending on the interest of the user. 
Similarly, the results from hybrid simulation test #2 are shown in Figures 6-14 to 6-16. 
In this test, the strain data from the two strain gages appear to be slightly different 
especially when the strain increased to 5,000 ??. This may be caused by the initial 
misalignment of the test specimen and introduced bending moment to the specimen 
during the test. However, the average value can be used to cancel out the bending 
moment effect on the axial strain.  The trigger strain for the two BT strain sensors in 
the two hybrid simulation tests are listed in Table 6-2. 
 
 114 
 
6.5.  Summary and conclusions 
This chapter presents an experimental test that demonstrates the implementation of 
the wireless structural condition assessment (WSCA) system. Through the component 
validation test on truss member made of PVC tube, the configuration and installation 
approach for the two BT strain sensors were determined: the sensors were attached to 
the PVC tube by stacking on top of each other and using screws and glue bonding at 
their ends. This ?stacking? configuration is intended to minimize the effect of 
cracking of H6W4 on H9W2 in the BT strain sensor. It is found that the BT strain 
sensor?s failure mode is identical to that observed in the countertop tests reported in 
Chapter 3. 
A hybrid simulation test of a truss structure was conducted to validate the BT strain 
sensors installed on a nearly-full-scaled tubular structural member and the BIM-based 
WSCA system. The various aspects of the BIM-based WSCA system including the 
integration of the RFID based BT strain sensors and BIM-based computing 
environment have been evaluated in a real structure setting. For example, its 
capability of quickly locating and visualizing the threshold strain exceedence in the 
selected members is demonstrated. In the BIM user interface where the Revit 
structure model is shown, the two triggering strain levels were highlighted in two 
colors (blue or red). The automatic updating of the condition states of selected 
members has also been demonstrated in the hybrid simulation test. 
 
 
 
 
 115 
 
 
Table 6- 1. Measured and design values of the trigger strain level for the two BT 
strain sensors on the PVC tube subjected to axial load 
 
BT strain sensor 
H6W4 H9W2 
Test #1 0.62% 1.02% 
Test #2 0.66% 1.05% 
Test #3 0.64% 1.03% 
Average of test values 0.64% 1.03% 
Design value 0.55% 1.18% 
Countertop Test 0.68% 1.32% 
 
 
 
Table 6- 2. Measured and design values of the trigger strain level for the two BT 
strain sensors on the PVC tube in hybrid simulation test  
(HS = Hybrid Simulation) 
 
BT strain sensor 
H6W4 H9W2 
HS Test #1 0.65% 1.10% 
HS Test #2 0.61% 1.02% 
Average of HS test values 0.63% 1.06% 
Design value 0.55% 1.18% 
Countertop Test 0.68% 1.32% 
 
 
 
 116 
 
 
Figure 6- 1. 3D model of the prototype truss structure for hybrid simulation test 
 
 
Figure 6- 2. Schematics of hybrid testing software framework (Schellenberg et al 
2006a) 
  
Experiment 
member 
44
 in
  (1
 .1
 m
 )
 Experiment system in 
lab
 Experimental Site
 FE software
 Experiment Element
 Experimental Setup
 Experimental Control Interface to control 
system
 Transform degree of 
freedom
 Store data and 
communicate
 Represent physical part 
and provide interface
 
 117 
 
 
 
 
 
  
F
 igu
 re
  6
 - 
3. 
S
 ch
 em
 at
 ic
 s 
of
  a
  hyb
 ri
 d 
te
 st
 ing 
fa
 ci
 li
 ty 
fo
 r 
co
 ll
 ap
 se
  s
 im
 ul
 at
 ion
  und
 er
  w
 ind 
lo
 ad
 in
 g 
 
 
 118 
 
 
 
 
 
  
F
 igu
 re
  6
 - 
4.  
S
 ch
 em
 at
 ic
  o
 f 
te
 st
  s
 et
 up 
fo
 r 
co
 m
 pon
 en
 t v
 al
 id
 at
 ion 
te
 st
  on 
P
 V
 C
  tub
 e 
(l
 eg
 en
 d:
  (
 1)
  d
 at
 a 
ac
 qu
 is
 it
 io
 n 
sy
 st
 em
  u
 se
 r 
in
 te
 rf
 ac
 e 
fo
 r 
st
 ra
 in
  g
 ag
 e 
m
 ea
 su
 re
 m
 en
 t;
  (
 2)
  B
 IM
  u
 se
 r 
in
 te
 rf
 ac
 e 
(R
 ev
 it
  S
 tr
 uc
 tu
 re
 );
  (
 3)
  
R
 F
 ID
  r
 ea
 de
 r;
  (
 4)
  P
 V
 C
  tub
 e 
w
 it
 h 
B
 T
  s
 tr
 ai
 n 
se
 ns
 or
 s)
  
  
(1
 ) 
(2
 ) 
(3
 ) 
(4
 ) 
 
 119 
 
 
Figure 6- 5. Dimensions of BT strain sensor (H6W4 as an example) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5.08 cm
 3.
 5 
cm
 H6W4
 9.08 cm
 
 120 
 
 
(a) Exploded view of BT strain sensor 
 
 
(b) Close-up view of the BT strain sensor on the tube 
Figure 6- 6. Schematic of BT strain sensor 
 
 
 
 121 
 
 
(a) Strain test data from Test #1 
 
(b) Strain test data from Test #2 
0 50 100 150 200 250 300 350 400 450 500
 -12000
 -10000
 -8000
 -6000
 -4000
 -2000
 0
 Time (sec.)
 A
 xi
 al
  S
 tra
 in
  ( 
??
 )
  
 
Gage 1
 Gage 2
 H6W4-FEM
 H9W2-FEM
 H6W4
 H9W2
 0 50 100 150 200 250 300 350 400 450 500
 -12000
 -10000
 -8000
 -6000
 -4000
 -2000
 0
 Time (sec.)
 A
 xi
 al
  S
 tra
 in
  ( 
??
 )
  
 
Gage 1
 Gage 2
 H9W2
 H6W4-FEM
 H9W2-FEM
 H6W4
 
 122 
 
 
(c) Strain test data from Test #3 
Figure 6- 7. Comparison of measured strain of the PVC tube and triggering time point 
of the BT strain sensors 
 
 
 
Figure 6- 8. Close-up view of the cracks in the BT strain sensors after triggering 
0 50 100 150 200 250 300 350 400 450 500
 -12000
 -10000
 -8000
 -6000
 -4000
 -2000
 0
 Time (sec.)
 A
 xi
 al
  S
 tra
 in
  ( 
??
 )
  
 
Gage 1
 Gage 2
 H9W2-FEM
 H6W4-FEM
 H9W2
 H6W4
 
 123 
 
 
Figure 6- 9. Prototype truss model for hybrid simulation test 
4@44in (1.1m)
 4
 4
  in
  (1
 .1
  m
 )
 P
 
 124 
 
 
Figure 6- 10. Hybrid simulation test setup for system validation of WSCA on truss 
structure 
 
RFID 
reader
 44
  in
  (1
 .1
  m
 )
 30 in 
(0.76 m)
 Strain 
gage 1
 Strain 
gage 2
 Strain 
gage 1
 
 125 
 
 
(a) 
 
(b) 
Figure 6- 11. Test results of the first hybrid simulation test on truss structure (HS Test 
# 1, Strain): (a) strain test data (from PVC tube); (b) load data (on PVC tube) 
0 50 100 150 200 250 300 350 400 450 500
 -12000
 -10000
 -8000
 -6000
 -4000
 -2000
 0
 Time (sec.)
 A
 xi
 al
  s
 tra
 in
  ( 
??
 )
  
 
Gage 1
 Gage 2
 Derived strain
 H6W4
 H9W2
 0 50 100 150 200 250 300 350 400 450 500
 -35
 -30
 -25
 -20
 -15
 -10
 -5
 0
 Time (sec.)
 A
 ct
 ua
 to
 r L
 oa
 d 
(k
 N
 )
 
 126 
 
 
(a) 
 
(b) 
Figure 6- 12. Test results of the first hybrid simulation test on truss structure (HS Test 
# 1, H6W4): (a) picture of test setup at triggering of H6W4; (b) screen shot of BIM 
user interface 
H6W4
 
 127 
 
 
(a) 
 
(b) 
Figure 6- 13. Test results of the first hybrid simulation test on truss structure (HS Test 
# 1, H9W2): (a) picture of test setup at triggering of H9W2; (b) screen shot of BIM 
user interface 
 
 
 128 
 
 
 
(a) 
 
(b) 
Figure 6- 14. Test results of the first hybrid simulation test on truss structure (HS Test 
# 2, Strain): (a) strain test data (from PVC tube); (b) load data (on PVC tube) 
0 50 100 150 200 250 300 350 400 450
 -12000
 -10000
 -8000
 -6000
 -4000
 -2000
 0
 Time (sec.)
 A
 xi
 al
  S
 tra
 in
  ( 
??
 )
  
 
Gage 1
 Gage 2
 Derived strain H6W4
 H9W2
 0 50 100 150 200 250 300 350 400 450
 -35
 -30
 -25
 -20
 -15
 -10
 -5
 0
 Time (sec.)
 A
 ct
 ua
 to
 r L
 oa
 d 
(k
 N
 )
 
 129 
 
 
(a) 
 
(b) 
Figure 6- 15. Test results of the first hybrid simulation test on truss structure (HS Test 
# 2, H6W4): (a) picture of test setup at triggering of H6W4; (b) screen shot of BIM 
user interface 
H6W4
 
 130 
 
 
(a) 
 
(b) 
Figure 6- 16. Test results of the first hybrid simulation test on truss structure (HS Test 
# 2, H9W2):  (a) picture of test setup at triggering of H9W2; (b) screen shot of BIM 
user interface 
 
 131 
 
 
Chapter 7: WSCA Case Study II: Eccentrically Braced 
Frame (EBF) with Shear Link 
 
7.1. Introduction 
Eccentrically braced frames (EBFs) are a lateral load-resisting system for steel 
buildings that can be considered a hybrid between conventional moment-resisting 
frames (MRFs) and concentrically braced frames (CBFs). EBFs were proposed in an 
attempt to combine the individual advantages of MRFs and CBFs, while minimizing 
their respective disadvantages (Popov and Engelhardt 1988). The most attractive 
feature of EBFs for seismic-resistant design is their high stiffness combined with 
excellent ductility and energy-dissipation capacity. The excellent ductility of EBFs 
can be attributed to inelastic shear deformation in the links, which are designed and 
detailed to sustain large inelastic deformations without loss of strength. The shear 
links are easy to replace and they won?t fail before experiencing large deformations 
(Mansour et al. 2008). The yielding of the links (yield mostly in shear, therefore 
called shear link) in EBFs serves to limit the maximum force transferred to the brace. 
Under severe earthquake loading, shear links in EBFs function as structural fuses, 
experiencing plastic deformation while other components of the EBF remain 
essentially elastic (Popov et al. 1988). Rotation capacity is greatest for short links that 
yield primarily in shear which are typically used in the EBF design (Richards and 
Uang 2006). The damage to an eccentrically braced frame due to a major earthquake 
is designed to be confined to a highly ductile link. The link acts as a safety valve for 
 
 132 
 
the structure (Ghobarah and Ramadan 1990). Therefore, monitoring the peak strain 
levels in shear links during earthquakes and thus its remaining ductility capacity for 
aftershocks is essential to the safety and resilience of EBF structures. 
In this chapter, a three-story EBF building was chosen as the prototype structure to 
demonstrate the resilient structure concept with the use of the proposed BIM-based 
WSCA system. First, a numerical simulation analysis was conducted on the prototype 
EBF subjected to an ensemble of 20 earthquake records in order to collect the 
information for the design of BT strain sensors.  An experimental test of the WSCA 
mounted on a 1/3-scaled shear link was performed to demonstrate the use of and 
evaluate the performance of the BIM-based WSCA system. 
7.2. BT strain sensor design 
7.2.1. Prototype EBF modeling 
The three-story prototype EBF buildings was designed by Richards and Uang (2006) 
for location in Los Angles based on the moment frame protocol work by Gupta and 
Krawinkler (1999), following the local code requirements for Los Angles. All 
prevailing code requirements for gravity, wind, and seismic design were considered. 
The building was designed as a standard office building situated on stiff soil (Soil 
type S2 as per UBC 1994). It was redesigned according to the IBC (ICC 2000) and 
AISC Seismic Provision (AISC 2002) using eccentrically braced frames (Richards 
and Uang 2006). 
A two-dimensional model of the case study building was developed in this study 
using the computer program OpenSees (McKenna 1997). The three-story EBF with 
 
 133 
 
the shear link model was modeled as a two-dimensional planar structure in the 
OpenSees software as shown in Figure 7-1. The member sections are listed in Table 
7-1. Accurate modeling of the EBF and specifically the shear links are critical to the 
seismic performance evaluation since the inelastic deformation of the EBF is 
confined to these shear yielding links while other members are supposedly to remain 
linear elastic during design level earthquakes. A special element developed by 
Moghaddasi-B. (2011) in the OpenSees that simulates the cyclic loading behavior of 
the shear links was used for this study. A detailed description of this shear link 
element is given in the next section.  
7.2.2. Shear link model in OpenSees 
Shear links were modeled using a technique proposed by Ramadan and Ghobarah 
(1995). Ramadan and Ghobarah?s approach is based on the theory developed by 
Ricles and Popov (1987) and can be incorporated into general finite element (FE) 
analysis programs. In this study, shear link is modeled using the approach adopted by 
(Moghaddasi-B. 2011), namely, the following three elements are used collectively in 
the model: a beam element in the middle which accounts for elastic axial, bending 
and shear deformations and yielding in bending, and two zero length elements at two 
ends of shear link which account for yielding in shear.  
The middle beam used in the link model is an OpenSees element called ?beam with 
hinges?. This element is comprised of three parts: two hinges at the ends, and a 
linear-elastic region in the middle (see schematics in Figure 7-2). The hinges at the 
ends in the middle beam is modeled using the rigid-plastic hinges which are built in 
 
 134 
 
to the beam element (Mazzoni et al. 2009). The hinges were defined to yield at the 
plastic moment capacity calculated using the expected yield stress of 345 MPa (50 
ksi). The post-yield flexural stiffness of the link was equal to 5% of the elastic 
flexural stiffness. The plastic moment capacity of a link, Mp, is determined from the 
following equations (ANSI/AISC 341-10): 
?? = ??? (7-1) 
where ?? is the yield strength of the shear link material and Z is the plastic modulus 
of the link cross section.  The moment developed in the shear links are supposed to be 
lower than their plastic moment capacity since the links are shear critical. 
The shear stiffness and yielding in the link are modeled by two zero length elements 
(termed shear hinge springs in Figure 7-2) at two ends of the shear link. In Figure 7-3 
link shear V is plotted against the elongation ? of the zero-length element in the 
direction perpendicular to the shear link longitudinal axis in the undeformed 
configuration. This V- ? curve is calibrated based on the experimental test results 
from A992 steel short links (Arce 2002). The shear link behavior observed in 
experimental test, can be divided into three distinct phases: the elastic behavior up to 
the shear force of 1.1Vp and plastic and degradation phase before ultimate failure of 
the link by fracture in web section (Okazaki et al. 2006).  The plastic shear capacity 
of a link, Vp, is determined using the equation below (from ANSI/AISC 341-10 
(AISC 2010)): 
?? = 0.6??  ? ???? (7-2) 
 
 135 
 
where ?? is the yield strength of the shear link material and Aweb is the cross sectional 
area of the web in the link. 
Four parallel springs with isotropic bilinear behavior were used to model the shear 
hinges at both ends of the link element as shown in Figure 7-2. Elastic-perfectly-
 plastic (ElasticPP) material in OpenSees was used to model the hysteresis behavior of 
four springs connected in parallel, with zero-length elements positioned at the ends of 
the link element to simulate the piecewise-linear shear behavior of the shear link 
shown in Figure 7-3.  
In this study, two types of link rotation angle,  ?? and ??, are defined and recorded, as 
shown in Figure 7-4. The first type of link rotation angle, denoted as ??, is defined 
using Eqn. 7-3a. ?? is computed as the relative vertical displacement between the two 
ends of the link divided by the link length. The other definition for link rotation angle 
is adopted from ANSI/AISC 341-10 seismic code (AISC 2010), as illustrated in 
Figure 7-4. In the figure, the link rotation angle can be related to the plastic story drift 
angle ?? . L is the bay width of the EBF. The link rotation angle, noted as ?? , is 
calculated with the  equation below:  
?? =
 ?
 ?
  
 ?? =
 ?
 ?
 ?? 
(7-3a) 
 
(7-3b) 
 
 136 
 
7.2.3. Elements, materials and loads  
Nonlinear beam-column elements with inelastic fiber sections were used to model the 
beams in the braced bays and all columns. The beams in the middle two bays are 
modeled using truss elements in the OpenSees because connections in these unbraced 
bays were considered simple in the original design (Richards and Uang 2006). The 
bracing members which are intended to carry the load only through axial action, are 
also modeled with truss elements. The braces in the first story of the EBF were 
assumed to be fixed at their lower ends. 
Material behavior for all elements, including braces, beams and columns was 
modeled using the Giuffre-Menegotto-Pinto model (Menegotto and Pinto 1973) with 
isotropic strain hardening of 2%, Young?s modulus of 200 GPa, and yield strength of 
345 MPa (50 ksi) (i.e., Steel02 material in OpenSees).  
A gravity load combination of 1.2DL+0.5LL based on design code was applied to the 
structure during seismic loading. The dead load DL is calculated to be 80 psf (3.83 
kPa) for the roof and 85 psf (4.07 kPa) for other floors. The live load LL for each 
floor is assumed to be 20 psf (0.96 kPa) for the roof, but 50 psf (2.4 kPa) for other 
floors. Floor masses (only dead load is considered for seismic mass) were lumped 
into the beam end nodes at each floor level and lean-on column joints based on 
tributary area. A lean-on column along the height of the structure and pinned at its 
base, consisting of 3 elastic rigid link bar elements was used in the model to simulate 
the gravity columns which can account for the P-Delta effect. The lean-on column 
was connected to the center interior nodes of each floor of the EBF frame by 
 
 137 
 
assigning multi-point constraint to the corresponding nodes to transfer the seismically 
induced inertia force.  
Damping effect in the time history analysis was considered through Rayleigh 
damping. As a typical value for steel buildings, a 2% damping ratio for the first and 
third modal frequencies of the EBF was used for the OpenSees model. The stiffness 
proportional damping was applied only to the frame elements, since for accurate 
analysis no stiffness damping should be considered for the lean-on column and highly 
rigid truss elements that link the frame and lean-on column (Zareian and Medina 
2010).  
7.2.4. Eigen value analysis 
The Eigen value analysis was performed using the model developed in OpenSees. 
The first three vibration periods of the 3-story EBF were 0.61, 0.29 and 0.20 seconds 
respectively. The first three mode shapes are shown in Figure 7-5.  
If it is assumed that the first mode dominates the seismic response of the structure, the 
inverted triangular vertical load pattern seems to be appropriate to distribute the base 
shear along the height of the building in pushover analysis according to the first mode 
shape of the building. 
7.2.5. Nonlinear static analysis 
Nonlinear static analysis provides a insight into the performance of the structure. A 
pushover analysis of the 3-story EBF was performed to identify the nonlinear static 
 
 138 
 
response of the EBF. The inverted triangular lateral load pattern was used to 
distribute the base shear along the height of the building in pushover analysis.  
The lateral loads were applied as lumped forces at the left most exterior nodes at each 
floor level. The loads were increased monotonically until the horizontal displacement 
at the middle node of the roof reached the drift ratio of 3%.  
The roof drift ratio (roof displacement divided by the height of the building) vs. the 
normalized base shear response (also known as pushover curve) from the nonlinear 
static analysis, is shown in Figure 7-6. The pushover curve also shows the sequence 
of the plastic deformation development in the shear links. In Figure 7-6, L1, L2, L3 
refers to shear links at the 1st, 2nd, and 3rd floor levels begin plastic deformation 
respectively. Definition for ?1, ?2, ?3 can be found in Figure 7-3.  
It is seen that the slope of the post yield branch of the pushover curve changes as the 
shear links enter different phases of their inelastic deformation. During the pushover 
analysis, the base shear increases linear proportionally until the roof drift ratio 
reached about 0.5% when the corresponding base shear ratio was about 0.11.  
7.2.6. Nonlinear time history analysis 
The performance of a structural system under strong earthquakes can be evaluated by 
comparing the nonlinear time history analysis results of the key engineering demand 
parameters (EDPs) with code specified values and what is generally accepted in 
practice. For the purpose of obtaining the anticipated strain levels of the shear links in 
the prototype EBF building at design basis earthquakes (DBE) and maximum credible 
 
 139 
 
earthquakes (MCE), nonlinear time history analysis of the 3-story EBF was 
performed under a suite of 20 strong earthquake records, which were originally 
developed by Somerville et al. (1997) for the Los Angeles, California region with a 
probability of exceedance of 10% in 50 years. The MCE records were obtained by 
scaling up the DBE records by 1.5 times. These scaled earthquake records correspond 
to the design basis earthquake in the downtown Los Angeles area. The Newark 
average-acceleration (?=1/2, ?=1/4) time-step integration method with Newton initial 
stiffness interaction was used to solve the dynamic differential equations. The time 
interval for all time history analyses was set to be 0.001 seconds. 
The results of important EDPs are presented in Figures 7-7 to 7-12. Since it is not 
possible to present all detailed results of all time history analyses, LA01 is chosen as 
a representative case for the nonlinear time history analysis and some of its detailed 
time history results are presented here. For simplicity, shear link will be denoted as 
LxL and LxR, where x indicated the floor where the shear link is located; L indicated 
the left-hand side link while R indicates right-hand side link on each floor, as shown 
in Figure 7-1. 
Tables 7-2 to 7-5 list the maximum link rotation angle of shear links in the EBFs 
subjected to either DBE or MCE earthquake records. The mean value refers to the 
ensemble average of the results over the 20 earthquake records. The mean value of 
the link rotation angles derived for DBE and MCE records will be used to guide the 
design of the BT strain sensors in the next section. 
 
 140 
 
7.3. FE analysis of shear link 
7.3.1. FE model description 
A FE analysis of the shear link was first conducted to determine the principal strain 
distribution on the web plate in order to guide the design of the BT strain sensors. To 
lower the demand on the testing equipment, a reduced-scale shear link specimen that 
is 1/3 of its original size W16x77 is adopted. W16x77 is the section size of the two 
shear links located at the EBF?s first floor. Although the model length (8 inches) is 
slightly longer than that (6 inches) of 1/3-scale, it should have negligible effect on the 
findings since the loading of the shear link is displacement-control loaded to a link 
rotation angle of 0.08 rad. The extra 2-inch in length is also favorable to the 
experimental test because it gives more space to sensor installation. Therefore, a 
reduced-scale FE model is adopted to provide information for test planning. The FE 
model of the 1/3-scaled shear link created in ANSYS v.13 is shown in Figure 7-13. 
This FE model was developed in three dimensions to provide the details of stiffeners 
and flange. 3-d solid elements (ANSYS 8-node element SOLID45 with 3 degrees of 
freedom (DOFs) per node) was used to model all components in the shear link. The 
right-hand side end of the shear link was restrained from movement by applying 
translational constraints to all DOFs of the nodes over the entire end section. 
Therefore, the link model has a fixed end on its right side.  Material behavior for steel 
was modeled using the Bilinear Isotropic Hardening (BISO) with strain hardening of 
2%, Young?s Modulus of 200 GPa, and yield strength of 345 MPa (50 ksi).  
 
 141 
 
The BT strain sensors are installed on the web using four attachment screws, as 
shown in Figure 7-18. For the acrylic BT strain sensor, its Young?s modulus is 2.76 
GPa and tensile strength is 69 MPa. In the FE model, the attachment screws were 
simulated by coupling the nodes on the strain sensor with corresponding nodes on the 
web for all DOFs. The dimensions of the acrylic BT strain sensors are the same as 
those shown in Figure 6-5. The gage length of the BT strain sensor is 60.8 mm.  
The shear link is loaded by applying a concentrated load at the point on the top 
flange, as shown in Figure 7-13. To rotate the shear link by a link rotation angle of 
0.08 radians, the downward displacement at the load application point is 0.64 inch 
(1.63 cm).  
7.3.2. Orientation of BT strain sensor 
As mentioned earlier, a brittle material will fail when the maximum principal stress 
exceeds some value, independent of other stress components present. Therefore, the 
failure criteria for the brittle bar adopted in this study is that first principal strain 
reaches the fracture strain in a simple tension test. For example, the acrylic brittle 
bar will fail by fracturing whenever the maximum first principal strain around the 
corner surpasses 3%. The first principal strain is defined as the first principal stress 
divided by the modulus of elasticity of the material, that is, ?1 = ?1/E. When the 
relative displacement between the two parts of the BT strain sensor reaches the 
critical value, brittle bar in the strain sensor breaks and corresponding RFID tag is 
restored to normal operation to signify the breakage and threshold strain 
exceedance. The BT strain sensors would be considered triggered whenever the first 
 
 142 
 
principal strain first exceeds 3%. Triggering of the BT strain sensor can thus be 
related to the link rotation angle.  
The BT strain sensors were installed onto the web using four attachment screws (two 
on each end), as shown in Figure 7-18. An initial concern was whether the rotation 
restraint by the group of two screws on each end would cause additional strain in the 
brittle bar because of the link?s curved deflected shape. It is desirable to check the 
effect of sensor?s orientation angle would have on the deformation of the brittle bar in 
the BT strain sensor. For example, if the sensor axis is oriented vertically, the two 
groups of end screws should rotate by the same amount because they are on the same 
cross-section of the shear link and there would be no additional deformation induced. 
Figure 7-14 and 7-15 show the vector plot of first principal strain on the web. It can 
be seen that the first principal strain on the web are about 45? to the longitudinal axis 
of the shear link. Therefore, the relative displacement between the two parts of the BT 
strain sensor reaches larger value if the sensor orientated at 45? to the longitudinal 
axis of the shear link. 
To investigate an optimal orientation angle for the installation of the BT strain 
sensor on the web plate, four cases were considered in this FEM study. The FE 
models in these four cases are shown in Figure 7-16. Two BT strain sensors each 
with different brittle bars (H6W4 or H9W2) are installed on each side of the link?s 
web plate. The angles between the axis of the BT strain sensor and the longitudinal 
axis of the shear link in these four cases are: 45?, 0?, 90? and 35? respectively. The 
 
 143 
 
BT strain sensors are placed near the neutral axis in the vertical direction (Y-
 direction) and its right most edge is 1 inch (2.54 cm) away from the fixed end.  
To visualize the effect of different orientation angles of the BT strain sensors on their 
deformation, the initial and deformed locations of the four attachment screws in the 
sensors were presented in Figure 7-17. The initial locations of the four attachment 
screws were marked as open circles and deformed locations (corresponding to link 
rotation angle of 0.08 rad.) were marked as circles with dashed line. It is seen that for 
all three cases, the four attachment screws were still in a straight line after 
deformation. This suggests that there is negligible uneven rotation at the two end 
screw groups. The changes in the distance between the two inner attachment screws, 
??, and the ratio of ??/?  for each case are calculated and given in Table 7-6. The 
distance between two inner attachment screws after deformation is denoted as L? in 
the table.  
The relationship between the link rotation angle and maximum value of the first 
principal strain in the BT strain sensor are plotted in Figure 7-19. Using the 3% 
principal strain failure criteria, the link rotation angle could be related to the breakage 
and triggering of the BT strain sensor. Figure 7-19(a) shows that for Case 1 (45? 
orientation angle), the first principal strain of strain sensor H6W4 and H9W2 reaches 
3% when the link rotation angles are 0.037 and 0.078 radians respectively. The 
maximum value of first principal strain in the BT strain sensor in Cases 2 and 3 are 
below 0.2%, meaning that the BT strain sensor will not break even if the link rotation 
angle goes beyond 0.08 radians. The contour plots of the first principal strain for four 
cases are shown in Figure 7-20.  
 
 144 
 
The ensemble average of the maximum link rotation angles in the corresponding 
shear link (L1L) from the previous nonlinear time history analysis of the EBF 
building subjected to 20 earthquake ground motion records are also shown in Figure 
7-19(a). The ensemble average values of link rotation angles of the shear link are 0.07 
and 0.15 for the DBE and MCE levels respectively. Based on the FE analysis results, 
the BT strain sensors H6W4 and H9W2 could provide two threshold level indication 
of the link rotation angle, close to Frequent Earthquake (FE) level and DBE levels. 
This enables the WSCA system to provide multi-level structural condition indications 
after earthquake events. 
It?s worth noting that although this study uses a two threshold level BT strain sensor, 
adjustments could be made to include three threshold levels or more so that indication 
of exceeding link rotation angles corresponding to earthquakes in between the DBE 
and MCE levels could be provided. Other adjustments include rotating the BT strain 
sensors to an orientation angle that reduces the sensitivity of the BT strain sensors to 
link rotation. For example, the BT strain sensor could be rotated counter-clockwise by 
10 degrees so that the angle between the sensor axis and link longitudinal axis is 35 
degrees, which is Case 4 presented in Figure 7-19(d). It can be seen that the triggering 
link rotation angles for the BT strain sensors H6W4 and H9W2 are increased to 0.041 
and 0.091 radians respectively.  
Contour plots of the first principal strain of the BT strain sensors for all four cases are 
shown in Figure 7-20. In these figures, the shear link was deformed to a link rotation 
angle of 0.03 radians. It?s seen that in Case 1, the first principal strain in sensor 
H6W4 reached 0.025. However, the first principal strains in the BT strain sensors for 
 
 145 
 
Cases 2 and Case 3 are both smaller than 0.01. The first principal strain in the sensor 
for Case 4 is smaller than that in Case 1. This confirms that the BT strain sensor in 
Case 4 was triggered at a larger value of the link rotation angle compared with that in 
Case 1. 
7.4. Experimental validation test 
This section presents an experimental test to validate the WSCA system in a shear 
link specimen designed based on the prototype EBF building. As mentioned in the 
previous section, the shear link on the first floor of the EBF building was selected as 
the prototype link for this study. The prototype link has a section size of W16x77 and 
a length of 0.91 m (3 ft.). Due to limitation of loading equipment and availability of 
steel shapes, a 1/3-scaled steel WF section, W6x12, was used for the link test 
specimen in this experimental study. It is worth noting that the web thickness of 
W6x12 is 0.23 inches, which is greater than 0.15 inches (=1/3-scale web thickness of 
W16x77). The steel grade of the shear link specimen was ASTM A-992 steel. The 
W6x12 link beam was welded to a 1-inch (2.54 cm) thick end plate to anchor it on the 
reaction frame. 
The test setup configuration and test specimen dimensions are schematically shown in 
Figure 7-21. The test setup for validating the WSCA system on shear link includes 
the following components: 2 strain gages (Vishay model #: EA-06-250AE-350), two 
BT strain sensors with UHF RFID tag, UHF RFID reader (Skyetek model #: 
SkyeModule M9), BIM software (Autodesk Revit Structure), and a custom developed 
middleware. The RFID reader was placed at a distance of 32 inches (81.28 cm) away 
 
 146 
 
from the shear link. MTS servo-hydraulic testing system (see description in Section 
6.2) was used to apply load on the shear link in the experimental test. Displacement 
control was implemented in the test and the control point, which is 8 inches (20.32 
cm) away from the end plate on the upper flange surface, was pushed downward 
gradually from its initial position.  
Figure 7-22 shows a close-up view of the RFID tags and acrylic BT strain sensors 
installed on the link test specimen. The BT strain sensors on the web plate were 
inclined by a 45-degree angle to the longitudinal axis of shear link and 1 inch (2.54 
cm) away from the end plate. In this test, combined BT strains sensors H9W2 and 
H6W4 were stacked on top of each other to provide a two threshold strain level 
indication. When the first threshold strain level is reached, H6W4 would break, and 
its trigger would highlight the shear link model in the Revit Structure in blue color. 
Similarly, whenever the second threshold level is exceeded, H9W2 would break, and 
the shear link model in the Revit Structure would be highlighted in red color.  
Unfortunately, the test was halted due to an accident. As shown in Figure 7-24, top 
flange of the link specimen was fractured due to local bearing failure at a load level of 
nearly 30 kips (133.4 kN). The test was called off and an alternative test plan was 
proposed and implemented.  
To evaluate the BT strain sensor?s performance on steel structure component, 
compression test was conducted on a stub column (by turning the link specimen 90 
deg.). The test setup for this compression test is shown in Figure 7-25. The test setup 
includes the following components: 2 strain gages (Vishay model #: EA-06-250AE-
 
 147 
 
350), two BT strain sensors with UHF RFID tag, UHF RFID reader (Skyetek model 
#: SkyeModule M9), and BIM software (Autodesk Revit Structure). The universal 
test machine used for this test is a SATEC 400-kips servo-hydraulic test machine 
(serial #: 400BTE/483280). The stub column specimen was loaded axially till it failed 
at about 711 kN. Two strain gages were installed on the web plate at the BT strain 
sensor location, one on each side of the web plate. The strain gage was aligned with 
the longitudinal axis of the stub column in order to record its axial strain during the 
test. The Skyetek RFID reader was placed at a distance of 32 inch (81 cm) away from 
the shear link, as shown in Figure 7-25.  
The axial load was applied to the stub column by using displacement control. The 
loading rate is 0.2 inches (0.508 cm) per minute. The applied load on the stub column 
during the test is shown in Figure 7-27. Figure 7-28 shows the stress vs. the strain of 
the stub column (denoted as derived s train) derived from the measured crosshead 
displacement data (calculated by dividing the measured displacement with the stub 
column length of 11 inches (28 cm)). In this figures, the trigger time or strains of the 
two BT strain sensors (labeled H6W4 and H9W2 respectively) is also shown. It is 
seen that the axial force reached its peak value of 160 kips (711.7 kN) and begun to 
drop right after the peak. The readings of the two metal foil strain gages placed on the 
web plate were shown in Figure 7-29. The strain data measured (denoted as measured 
strain) from the two metal foil strain gages appear to be different. This may be caused 
by the initial misalignment of the fractured test specimen after the previous shear link 
test accident and introduced eccentricity and thus bending moment to the stub column 
specimen during the test. The two strain gages stopped working at 15 and 20 seconds 
 
 148 
 
respectively due to the web local buckling that was observed during the test. The two 
BT strain sensors were triggered at around 6,000 and 6,600 ?? (derived strain) 
respectively. Both BT strain sensors were triggered below design value due to the 
local buckling on the web plate during the test. This is because local buckling caused 
out-of-plane bending and thus introduced additional strain to the BT strain sensors.  
Figures 7-30 shows a screen shot of the BIM user interface (Revit Structure) after 
both BT strain sensors were triggered. In the Revit model, two colors (blue or red) 
were used to identify whether the two pre-set threshold strain levels were surpassed in 
the stub column. When the member concerned is highlighted with a blue color bar, 
the lower threshold strain level (for H6W4) is known to be exceeded. If it?s also 
marked with a red color bar, the higher threshold strain level (for H9W2) has also 
been surpassed. The buckled stub column specimen after the test was shown in Figure 
7-31. Both flange and web plates were buckled during the test. 
7.5. Summary and conclusions 
To validate and demonstrate the use of WSCA system in real steel structures, an 
experimental study of the BT strain sensors installed on shear links from a three-story 
prototype EBF building was conducted in this chapter. To guide the design of BT 
strain sensors for shear link strain condition monitoring, nonlinear time history 
analysis of EBF building subjected to an ensemble of 20 earthquake records was 
performed. The earthquake records were scaled according to the design basis 
earthquake (DBE) and maximum credible earthquakes (MCE) for a location in the 
Los Angles, California region. It?s found from nonlinear time history analysis that the 
 
 149 
 
mean value of maximum shear link rotation angles of the shear links in the EBF 
building is 0.07 radians under DBE and 0.15 radians under MCE. Based on this 
information, the design of BT strain sensors were studied using FE analysis of a shear 
link integrated with BT strain sensor. It?s found that by installing the BT strain 
sensors on the web plate of the shear link in different inclination angles, the WSCA 
system can be adjusted to monitor different levels of link rotation angle of the shear 
link and thus offers alert signal for exceeding multiple preset threshold strain levels 
that can be displayed in the BIM user interface. By installing the BT strain sensor at 
an inclination angle of 45 degrees, the WSCA system can give the first level 
exceedence alert at 0.037 radians and second level exceedence alert at 0.078 radians 
for the shear link specimen considered in this study.  
An experimental validation test of the WSCA system using a 1/3-scaled shear link 
specimen was performed to demonstrate the use of and evaluate the performance of 
the BIM-based WSCA system. The BT strain sensors were installed on the web plate 
based on the findings from the FE analysis. Unfortunately, an accident happened 
during the test of the shear link and the test had to be halted due to premature fracture 
of the link specimen. An alternate test was conducted in which the link specimen 
(thus called stub column) was loaded axially with the two-level BT strain sensors 
installed on the specimen. This test demonstrates that in the WSCA system, breakage 
of the two BT strain sensors was successfully detected and corresponding member in 
the Revit model was highlighted in colors.  
Based on the analysis and experimental results presented in this chapter, it is 
concluded that the WSCA system provides a promising rapid structural condition 
 
 150 
 
assessment technology, especially for post-earthquake structural condition 
assessment. The WSCA system can also be used for real-time structural health 
monitoring of fuse members in the structure, which facilitate the implementation of 
the concept of resilient structure by shortening the condition assessment and thus 
recovery time.  
  
 
 151 
 
Table 7- 1. EBF member sections and link properties (Richards and Uang 2006) 
Story 
Member sections 
Link length e 
and 
(???/??) Braces 
Columns 
with link 
connection 
Columns 
without 
link 
connection 
Beams/Links 
1 
HSS 
16x16x5/8 
W14x90 W14x61 W16x77 3.0 ft. (1.07) 
2 
HSS 
14x14x5/8 
W14x90 W14x61 W14x74 3.2 ft. (1.15) 
3 
HSS 
12x12x5/8 
W14x90 W14x61 W10x45 2.6 ft. (1.21) 
 
 
Table 7- 2. Maximum link rotation angle (??) in shear links of the prototype EBF 
building subjected to design basis earthquakes (DBE) 
 
EQ Record 
  
Shear link and corresponding maximum link rotation angle (??) 
(unit: radians) 
L1L L1R L2L L2R L3L L3R 
LA01 0.08 0.09 0.03 0.03 0.03 0.03 
LA02 0.07 0.06 0.03 0.03 0.03 0.03 
LA03 0.04 0.05 0.01 0.01 0.01 0.01 
LA04 0.01 0.01 0.00 0.01 0.02 0.02 
LA05 0.03 0.03 0.01 0.01 0.01 0.01 
LA06 0.00 0.00 0.00 0.00 0.02 0.02 
LA07 0.03 0.03 0.01 0.01 0.02 0.02 
LA08 0.02 0.02 0.01 0.00 0.01 0.02 
LA09 0.07 0.07 0.02 0.02 0.03 0.03 
LA10 0.02 0.02 0.01 0.01 0.03 0.03 
LA11 0.03 0.03 0.01 0.01 0.03 0.04 
LA12 0.02 0.02 0.03 0.03 0.07 0.06 
LA13 0.12 0.12 0.06 0.06 0.06 0.07 
LA14 0.12 0.12 0.05 0.05 0.06 0.05 
LA15 0.11 0.10 0.03 0.04 0.04 0.04 
LA16 0.20 0.19 0.04 0.04 0.03 0.03 
LA17 0.09 0.09 0.04 0.04 0.04 0.04 
LA18 0.09 0.08 0.05 0.05 0.08 0.08 
LA19 0.05 0.04 0.04 0.04 0.03 0.03 
LA20 0.13 0.12 0.04 0.04 0.06 0.06 
Mean 0.07 0.07 0.03 0.03 0.04 0.04 
Max. 0.20 0.19 0.06 0.06 0.08 0.08 
 
 
 
 152 
 
Table 7- 3. Maximum link rotation angle in shear links (??) of the prototype EBF 
building (unit: radians) subjected to maximum credible earthquakes 
(MCE) 
 
EQ record 
  
Shear link and corresponding maximum link rotation angle (??) 
(unit: radians) 
L1L L1R L2L L2R L3L L3R 
LA01 0.14 0.14 0.06 0.06 0.05 0.06 
LA02 0.12 0.12 0.05 0.05 0.06 0.06 
LA03 0.08 0.07 0.03 0.03 0.03 0.03 
LA04 0.02 0.02 0.01 0.02 0.04 0.05 
LA05 0.06 0.06 0.02 0.02 0.02 0.02 
LA06 0.02 0.02 0.01 0.00 0.02 0.02 
LA07 0.06 0.06 0.02 0.02 0.05 0.04 
LA08 0.05 0.05 0.02 0.02 0.05 0.04 
LA09 0.27 0.27 0.05 0.05 0.05 0.05 
LA10 0.03 0.03 0.02 0.02 0.06 0.06 
LA11 0.10 0.11 0.03 0.03 0.04 0.04 
LA12 0.06 0.06 0.04 0.04 0.10 0.10 
LA13 0.24 0.23 0.11 0.11 0.12 0.12 
LA14 0.27 0.28 0.13 0.13 0.09 0.08 
LA15 0.29 0.29 0.09 0.09 0.08 0.08 
LA16 0.39 0.39 0.07 0.07 0.04 0.04 
LA17 0.26 0.26 0.07 0.07 0.05 0.05 
LA18 0.17 0.17 0.06 0.06 0.14 0.15 
LA19 0.12 0.11 0.08 0.08 0.06 0.05 
LA20 0.29 0.28 0.07 0.07 0.07 0.08 
Mean 0.15 0.15 0.05 0.05 0.06 0.06 
Max. 0.39 0.39 0.13 0.13 0.14 0.15 
 
  
 
 153 
 
Table 7- 4. Maximum link rotation angle in shear links (??) of the prototype EBF 
building (unit: radians) subjected to DBE earthquakes 
 
 
EQ record 
  
Shear link and corresponding maximum link rotation angle (??) 
(unit: radians) 
L1L L1R L2L L2R L3L L3R 
LA01 0.11 0.11 0.08 0.08 0.07 0.07 
LA02 0.09 0.09 0.07 0.07 0.07 0.07 
LA03 0.07 0.07 0.05 0.05 0.04 0.04 
LA04 0.03 0.03 0.04 0.03 0.05 0.05 
LA05 0.06 0.06 0.05 0.05 0.04 0.04 
LA06 0.02 0.02 0.03 0.03 0.05 0.04 
LA07 0.06 0.06 0.05 0.05 0.05 0.06 
LA08 0.04 0.04 0.04 0.04 0.04 0.04 
LA09 0.11 0.11 0.06 0.06 0.06 0.06 
LA10 0.05 0.05 0.04 0.04 0.06 0.06 
LA11 0.05 0.05 0.04 0.04 0.07 0.07 
LA12 0.05 0.05 0.06 0.06 0.10 0.10 
LA13 0.16 0.16 0.11 0.11 0.11 0.11 
LA14 0.16 0.16 0.10 0.10 0.10 0.10 
LA15 0.14 0.14 0.08 0.08 0.08 0.08 
LA16 0.24 0.24 0.09 0.09 0.07 0.07 
LA17 0.13 0.13 0.09 0.09 0.08 0.08 
LA18 0.11 0.11 0.09 0.09 0.12 0.12 
LA19 0.07 0.07 0.08 0.08 0.07 0.07 
LA20 0.16 0.16 0.08 0.08 0.08 0.08 
Mean 0.10 0.10 0.07 0.07 0.07 0.07 
Max. 0.24 0.24 0.11 0.11 0.12 0.12 
 
 
 
  
 
 154 
 
Table 7- 5. Maximum chord rotation in shear links (??) of the prototype EBF building 
subjected to MCE earthquakes 
 
EQ record 
No. 
  
Shear link and corresponding maximum link rotation angle (??) 
(unit: radians) 
L1L L1R L2L L2R L3L L3R 
LA01 0.16 0.16 0.11 0.11 0.10 0.10 
LA02 0.15 0.15 0.10 0.09 0.10 0.10 
LA03 0.11 0.11 0.06 0.06 0.07 0.07 
LA04 0.04 0.04 0.05 0.05 0.07 0.07 
LA05 0.09 0.09 0.07 0.07 0.05 0.05 
LA06 0.04 0.04 0.04 0.04 0.05 0.05 
LA07 0.08 0.09 0.06 0.06 0.08 0.08 
LA08 0.09 0.09 0.06 0.06 0.08 0.08 
LA09 0.33 0.33 0.11 0.11 0.09 0.09 
LA10 0.06 0.06 0.05 0.05 0.09 0.09 
LA11 0.15 0.15 0.06 0.07 0.08 0.08 
LA12 0.09 0.09 0.08 0.08 0.14 0.14 
LA13 0.28 0.28 0.16 0.16 0.16 0.16 
LA14 0.32 0.32 0.19 0.19 0.13 0.13 
LA15 0.35 0.35 0.14 0.14 0.12 0.12 
LA16 0.44 0.44 0.13 0.13 0.08 0.08 
LA17 0.30 0.30 0.12 0.12 0.10 0.10 
LA18 0.22 0.22 0.10 0.10 0.19 0.19 
LA19 0.14 0.14 0.13 0.13 0.10 0.10 
LA20 0.35 0.35 0.13 0.13 0.11 0.11 
Mean 0.19 0.19 0.10 0.10 0.10 0.10 
Max. 0.44 0.44 0.19 0.19 0.19 0.19 
 
 
Table 7- 6. Distance and axial deformation between the BT strain sensor (inner) 
attachment points when shear link rotates by 0.08 rad (unit: cm) 
 
 ? + ?? ? ?? ??/? 
Case 1: 45 deg. 5.761  6.08 0.319 5.24% 
Case 2: 0 deg. 6.108  6.08 0.028 0.46% 
Case 3: 90 deg. 6.083  6.08 0.003 0.05% 
 
 
 155 
 
 
 
Figure 7- 1. OpenSees model of the prototype EBF structure 
 
 
 
 
 
 
 
 
 
 
Figure 7- 2. Schematics of shear link model 
 
 
4@30?(9.14m)
 3@
 13
 ?(
 3.
 96
 m
 )
 L1L
 L2L
 L3L
 L1R
 L2R
 L3R
 Elastic Beam
 Zero-length 
Plastic hinge
 Shear hinge springs
 Floor Beam
 ?
 Floor Beam
 
 156 
 
 
Figure 7- 3. Combined behavior of parallel springs used to model shear hinge 
(Richards and Uang 2006) 
 
 
(a)  ??  
 
(b)  ?? 
 
Figure 7- 4. Illustration of link rotation angle  ?? and ?? 
 
Parallel springs 
G is shear modulus of the link material.
 e 
L 
h ? p 
? p 
e 
L 
h 
? 
 
 157 
 
 
(a)  1st Mode 
 
(b)  2nd Mode 
 
(c) 3rd Mode 
 
Figure 7- 5. First three mode shapes of the 3-story prototype EBF building 
 
 
 158 
 
 
Figure 7- 6. Push-over curve of the prototype EBF building 
 
 
Figure 7- 7. Ground acceleration time history of the LA01 earthquake record (DBE 
level) 
 
 
 159 
 
 
Figure 7- 8. Roof displacement time history of the prototype EBF building subjected 
to LA01 earthquake record (DBE level) 
 
Figure 7- 9. Roof acceleration time history of the prototype EBF building subjected to 
LA01 earthquake record (DBE level) 
 
 
 160 
 
 
(a)  L1L link rotation angle ?? 
 
 
(b)  L1L link rotation angle ?? 
 
Figure 7- 10. Typical shear link (L1L) rotation angle (?? and ??) time history of the 
EBF building subjected to LA01 earthquake record (DBE level) 
 
 161 
 
 
Figure 7- 11. Axial force time history of the shear link L1L in the prototype EBF 
building subjected to LA01 earthquake record (DBE level) 
 
 
Figure 7- 12. Hysteresis (shear force vs. ??) loop of L1L shear link subjected to LA01 
earthquake record (DBE level) 
 
 
 162 
 
 
Figure 7- 13. Schematics of the FEM model of 1/3-scaled shear link in ANSYS v.13 
(only part of the link is modeled; length is not 1/3 scaled to provide more space for 
sensor installation) 
 
 
Figure 7- 14. Vector plot of first principal strain on the web at link rotation angle of 
0.08 rad. 
 
 
 163 
 
 
 
Figure 7- 15. Vector plots of first principal strain from linear elastic analysis (by 
applying a 1-kips load)  
 
 164 
 
 
(a)  Case 1: BT strain sensor oriented at 45 degree (to z axis) 
 
 
 
(b)  Case 2: BT strain sensor oriented at zero degree 
 
H6W4
 (H9W2)
 H6W4
 (H9W2)
 
 165 
 
 
(c)  Case 3: BT strain sensor oriented at 90 degree 
 
 
(d)  Case 4: BT strain sensor oriented at 35 degree 
Figure 7- 16. FEM models with integrated BT strain sensors oriented at different 
angles to the longitudinal axis (z axis) of shear link 
H6W4
 (H9W2)
 H6W4
 (H9W2)
 H6W4
 (H9W2)
 
 166 
 
 
Figure 7- 17. Schematics of the BT strain sensor attachment point movement when 
link rotation angle changes from zero to 0.08 rad  
(circles shown in figure indicates the attachment screws) 
 
 
 
Figure 7- 18. Photo of the BT strain sensor oriented at 45 deg. on the shear link 
specimen web   
Case 1: 45?
 Case 2: 0?
 Case 3: 90?
 1-1
 1-2
 1-3
 1-4
 2-1 2-2 2-3 2-4
 3-1
 3-2
 3-3
 3-4
 1 cm
 Y
 Z
 
 167 
 
 
 
 
(a)  Case 1: BT strain sensor oriented at 45 degree (to z axis) 
 
(b)  Case 2: BT strain sensor oriented at 0 degree (to z axis) 
 
 168 
 
 
 
(c)  Case 3: BT strain sensor oriented at 90 degree 
 
 
(d)  Case 4: BT strain sensor oriented at 35 degree 
Figure 7- 19. Evolution of maximum value of the first principal strain in the brittle 
bar of BT strain sensors with increasing link rotation angle (??) 
  
 
 169 
 
 
 
(a)  Case 1: BT strain sensor oriented at 45 degree (to z axis) 
 
 
 
 
 
(b)  Case 2: BT strain sensor oriented at 0 degree 
 
H6W4
 H9W2
 H6W4
 H9W2
 
 170 
 
 
 
(c)  Case 3: BT strain sensor oriented at 90 degree 
 
 
(d)  Case 4: BT strain sensor oriented at 35 degree 
 
Figure 7- 20. Contour plots of first principal strain in the BT strain sensors (with two 
brittle bars: H6W4 and H9W2) when link rotation angle reaches 0.03 rad. 
H6W4
 H9W2
 H6W4
 H9W2
 
 171 
 
 
(a) 
 
 
(b) 
Figure 7- 21. Schematic of test setup for validating the WSCA system on shear link: 
(a) shear link specimen; (b) test setup with actuator and reaction frame 
 
W6X12
 Stiffener
 
 172 
 
 
 
Figure 7- 22. Close-up view of BT strain sensor installed on shear link web 
. 
 
 
 173 
 
 
 
Figure 7- 23. Test setup for validating the WSCA system on shear link 
 
RFID 
reader
 32 in
 
 174 
 
 
 
Figure 7- 24. Premature failure of shear link specimen due to local flange fracture 
  
 
 175 
 
 
 
 
  
F
 igu
 re
  7
 - 
25
 . T
 es
 t s
 et
 up 
fo
 r 
co
 m
 pr
 es
 si
 on 
te
 st
  o
 f 
st
 ub 
co
 lu
 m
 n 
sp
 ec
 im
 en
  
(l
 eg
 en
 d:
  (
 1)
  d
 at
 a 
ac
 qu
 is
 it
 io
 n 
sy
 st
 em
  u
 se
 r 
in
 te
 rf
 ac
 e 
fo
 r 
st
 ra
 in
  g
 ag
 e 
m
 ea
 su
 re
 m
 en
 t;
  (
 2)
  B
 IM
  u
 se
 r 
in
 te
 rf
 ac
 e 
(R
 ev
 it
  S
 tr
 uc
 tu
 re
 );
  (
 3)
  R
 F
 ID
  
re
 ad
 er
 ; (
 4)
  s
 tub 
co
 lu
 m
 n 
sp
 ec
 im
 en 
w
 it
 h 
B
 T
  s
 tr
 ai
 n s
 en
 so
 rs
 ) 
(1
 ) 
(2
 ) 
(3
 ) 
(4
 ) 
 
 176 
 
 
 
 
Figure 7- 26. Close-up view of BT strain sensors on stub column specimen 
 
 
 
 
 177 
 
 
Figure 7- 27. Axial load applied to stub column specimen (vertical line indicates 
triggering of BT strain sensor) 
 
 
Figure 7- 28. Axial stress (load divided by cross-section area vs. strain 
(derived from crosshead displacement divided by stub column length) plot for stub 
column specimen (vertical line indicates triggering of BT strain sensor) 
 
 
 178 
 
 
Figure 7- 29. Measured strain from two metal foil strain gages in stub column 
specimen 
 
 
 
Figure 7- 30. Screen shot of BIM (Revit structure) user interface 
(color highlights signify the triggering of BT strain sensor) 
H6W4
 H9W2
 
 179 
 
 
 
Figure 7- 31. Photo of buckled stub column specimen 
  
 
 180 
 
Chapter 8: Summary, Conclusions and Recommendations 
 
8.1. Research summary  
The main goal of this research is to develop a wireless structure condition 
assessment (WSCA) system that builds on an integrated BIM environment for an 
automated and graphical approach to structural condition monitoring, prognosis and 
health management. This WSCA system also facilitates the implementation of 
resilient structure concept. To achieve this goal, two major efforts have been made in 
this research: (1) development of a low-cost radio frequency identification (RFID) 
based wireless breakage-triggered strain sensor that can be interrogated on an demand 
basis from the BIM environment. (2) development of essential component 
technologies to enable the integration of BIM and structural health monitoring (SHM) 
technologies. A BIM for SHM portal is at the heart of this integration. A summary of 
specific research works performed in this study are presented as follows, 
? A literature review on RFID assisted BIM technology and information 
exchange between BIM and SHM-related software programs was performed 
and the review is summarized in Chapter Two.  
? A passive wireless, low-cost breakage-triggered (BT) strain sensor for 
detection of threshold strain level exceedence is proposed and its performance 
is characterized in Chapter Three. The principle of operation of the BT strain 
sensor and results of experimental demonstration were discussed in this 
chapter. Two types of brittle materials, acrylic and epoxy paint, were tested 
for the BT strain sensor. For important application of buckling detection in 
 
 181 
 
tubular members, epoxy paint patch was investigated by compression test of 
paint buckets. BT strain sensors that can detect multiple threshold strain levels 
were also examined experimentally. 
? With RFID based wireless sensors being developed in this research, a sensor-
 driven SHM system that integrates BIM, sensors, finite element (FE) analysis, 
and database technology was presented in Chapter Four. A prototype bridge 
model was developed using the BIM software (Revit Structures) to 
demonstrate sensor data storage and display, FE analysis, photorealistic 
visualization, prognosis, and structural condition rating in an integrated 3D 
visual setting.  
? Chapter Five presents the concept of resilient structure design by 
concentrating SHM on fuse members. To demonstrate this concept of 
integrating SHM with resilient structure design, numerical analysis of a long-
 span steel arch structure subjected to earthquake excitation was conducted. 
This steel arch truss has strategically selected fuse zone in which structural 
deformation and damage could be concentrated. This strategy of concentrated 
damage zone will facilitate rapid structural condition assessment and this 
concept of focused SHM on fuse members will also be demonstrated later 
using the RFID based wireless strain sensor developed in this research. 
? The performance of the WSCA system was tested in Chapter 6 through a 
hybrid simulation test conducted on a truss structure.  The performance of the 
BT strain sensor installed on a large-scale structural tubular member was 
examined experimentally. The various aspects of the WSCA system including 
 
 182 
 
the integration of the wireless sensors (e.g., BT strain sensors) and the BIM-
 based computing environment were investigated.  
? In Chapter Seven, a three-story eccentrically braced frame (EBF) building was 
chosen as the prototype structure to demonstrate the resilient structure concept 
with the use of the WSCA system. An FE analysis was conducted on the 
prototype EBF subjected to an ensemble of 20 earthquake records in order to 
collect the information to guide the design of the BT strain sensor. An 
experimental test of the WSCA system mounted on a reduced-scaled shear 
link specimen was performed to demonstrate the use of and evaluate the 
performance of the WSCA system.  
8.2. Conclusions 
Based on the research findings, the following concluding remarks can be drawn,  
? From the experimental results presented in Chapter 3, it is shown that by 
adding a bypass to the IC chip on the RFID tag, its responding power to the 
RFID reader would change. This feature can be utilized to detect whether a 
threshold strain is surpassed in the monitored structure by relating the strain to 
the change in the RFID antenna property.  
? Two brittle materials, acrylic and epoxy paint, that were used for the BT strain 
sensor in this study, each have their own advantages and drawbacks. In this 
study, both materials have been demonstrated to break at a pre-determined 
threshold strain level and can be satisfactorily used for the BT strain sensor. In 
conjunction with the RFID technology, the BT strain sensor can fulfill the 
 
 183 
 
goal of rapidly scanning the strain condition at the monitoring location and 
wirelessly transfer the data to BIM database at request.  
? BIM can serve as an ideal 4-dimensional graphical computing environment 
for SHM and management practice which constantly involves considerable 
amounts of sensor data of different modalities. The ability to visualize the 
inspection and monitoring information collected in the life cycle of a structure 
provides engineers and decision makers with a photorealistic view of the 
SHM process, visualizing current condition of structural elements, structural 
reliability, maintenance work to schedule, response of structures under natural 
hazardous events such as earthquakes.  
? Installing fuse members, especially self-centering members, at strategically 
weakened zone of steel structures can effectively control the responses of the 
steel structures under earthquake. The use of conventional fuse members 
exhibiting bilinear hysteresis results in residual displacement in the steel arch 
structure. Self-centering fuse members with flag-shaped hysteresis is very 
effective to eliminate the residual displacement. The strengths of fuse 
members usually have to be reduced (compared to the critical load of its 
original substituted member) to prevent the failure (e.g., buckling) of their 
neighboring members in the arch structure. The structural weakening strategy 
can concentrate the ductile deformation of the overall structure at favorable 
locations where inspection and rapid structural repair with member 
replacement are convenient to perform.  
 
 184 
 
? RFID based BT strain sensor can be installed on the pre-defined fuse member 
for easy inspection. This study provides a strategy for selecting the fuse 
member location to form target fuse zones in steel structures.  
? Through the validation test of truss member made of PVC tube, the 
performance of BT strain sensor for detection of preset threshold strain level 
was validated on a large-scale structural member. It is found that the BT strain 
sensor?s failure mode is identical to that observed in the countertop tests 
reported in Chapter 3. The breakage of the brittle bar in the sensor can trigger 
the RFID tag connected. The WSCA system was validated through hybrid 
testing of a truss structure under monotonic loading. The trigger strain levels 
for each configuration were close to the design values. BIM user interface 
automatically highlighted the trigger strain level on the concerned member 
and displayed the updated properties of the selected member once the BT 
strain sensor was triggered. 
? The resilient structure concept was demonstrated with the use of the WSCA 
system implemented on the shear link of a 3-story EBF building. It was found 
that the orientation of the BT strain sensor installed on the web of shear link 
directly affects the deformation in the BT strain sensor and thus the trigger 
strain level of the sensor. The link rotation angle could be related to the 
triggering time point of the BT strain sensor. In the experimental test, the 
WSCA system successfully detected the breakage of the two BT strain sensors 
on the web plate and highlighted the element in the BIM model with different 
 
 185 
 
colors and updated the corresponding member properties of the BIM model 
automatically. 
? The WSCA system is validated in the two experimental case studies. The 
custom-developed middleware program in the WSCA system exchanged and 
updated SHM information automatically as expected.   
8.3. Recommendations for future works 
This study focuses on developing a wireless structural condition assessment 
system integrated in BIM environment. Further research work could be conducted in 
the following areas, 
?  The BT strain sensor developed for the WSCA system has been validated in 
this study through experimental tests. The BT strain sensor can provide 
reliable strain level indication through the breakage of its brittle bar. The 
epoxy paint patch approach was also demonstrated to be able to detect the 
local buckling of tubular structure. More experiments are needed to fine tune 
the BT strain sensor for buckling detection. 
? The current strain measurement range of the BT strain sensor can be tuned 
from 0.6% to 1.2%. This working range could be reduced to below 0.2% if a 
more brittle material can be found.    
? Full scale experimental test of BT strain sensors are desired for characterizing 
their performance on link elements. This study used a 1/3-scaled shear link 
specimen but premature fracture happened due to an accident during the test.  
Testing of the BT strain sensors on longer links that has larger flexural 
deformation is also interesting to do in future study. 
 
 186 
 
Bibliography  
 
AASHTO. (2011). AASHTO Guide Manual for Bridge Element Inspection, 1st 
Edition, American Association of State Highway and Transportation Officials 
(AASHTO), Washington, DC, USA, ISBN Number: 1-56051-497-8.  
AISC.  (2010). Seismic Provisions for Structural Steel Buildings. ANSI/AISC 341-
 10. Chicago, IL, USA.  
Applied Technology Council (ATC) (2013). Improvements to BIM Structural 
Software Interoperability. ATC-75 report.  Redwood City, California, USA. 
Arce, G. (2002). ?Impact of higher strength steels on local buckling and overstrength 
of links in eccentrically braced frames.? MS Thesis, Univ. of Texas at Austin, 
Austin, Texas, USA. 
ASCE. (2010). ?Minimum design loads for buildings and other structures.? 
ASCE/SEI 7-10/ANSI,  Reston, VA, USA. 
Autodesk. (2011). Revit 2012 API- Developer?s Guide, Autodesk, Inc., San Rafael, 
CA, USA. 
Azhar, S. (2011). ?Building information modeling (BIM): Tends, benefits, risks, and 
challenges for the AEC industry.? Leadership and Management in Engineering, 
11(3):241-252. 
Azhar, S., Nadeem, A., Mok, J., and Leung, B. (2008). ?Building information 
modeling (BIM): A new paradigm for visual interactive modeling and simulation 
for construction projects.? Proceeding of First International Conference on 
Construction in Developing Countries, Karachi, Pakistan, 435?446. 
Bai, L., and Zhang, Y. (2010). ?Progress collapse simulation of roof truss structures 
with hybrid testing technique.? Proceedings of the International Association for 
 
 187 
 
Shell and Spatial Structures (IASS) Symposium 2010, November 8-12 2010, 
Shanghai, China. 
Balanis, C. (2004). Antenna Theory: Analysis and Design, John Wiley & Sons, New 
Jersey, USA.  
Bazjanac, V. (2004). ?Virtual building environments (VBE)?Applying information 
modeling to buildings.? URL (last accessed on Nov. 7, 2013): http://repositories.cd
 lib.org/lbnl/LBNL-56072. 
Beck, K. (2012). ?The State of Wisconsin: BIM ? Digital FM handover pilot 
projects.? Journal of Building Information Modeling, pp. 22 - 23.  
Blandford, G. (1997). ?Review of progressive failure analyses for truss structures.? J. 
Struct. Engrg., ASCE, 123(2):122?129. 
Butler, J., Vigliotti, A., Verdi, F., and Walsh, S. (2002). ?Wireless, passive, resonant-
 circuit, inductively coupled, inductive strain sensor.? Sensors and Actuators A: 
Physical, 102(1-2): 61?66. 
Casciati, F., and Faravelli, L. (2004). ?Experimental characterisation of a Cu-based 
shape memory alloy toward its exploitation in passive control devices.? J. Phys. IV 
France, 115: 299?306.  
Catbas, F., Susoy, M., and Frangopolc, D. (2008). ?Structural health monitoring and 
reliability estimation: Long span truss bridge application with environmental 
monitoring data.?  Engineering Structures, 30(9): 2347?2359.  
Cesare, M., Santamarina, C., Turkstra, C., and Vanmarcke, E. (1992). ?Modeling 
bridge deterioration with Markov chains.? Journal of transportation engineering, 
118(6): 820-833. 
Chen, J., Pujol, S., Yau, D., and Miao, D. (2010) ?Integrated RFID and sensor 
networks for structural monitoring.? RFID and Sensor Networks, Chapter 21, 
Edited by Zhang Y. et al., CRC Press, Florida, USA. 
 
 188 
 
Chin, J., Rautenberg, J., Ma, C., Pujol, S., and Yau, D. (2009). ?An experimental low-
 cost, low-data-rate rapid structural assessment network,? IEEE Sensors Journal, 
9(11): 1361-1369. 
Christotopolulos C., Filiatrault A., and Folz B. (2002). ?Seismic response of self-
 centering hysteretic SDOF system.? Earthquake Engineering and structural 
dynamics, 31(5): 1131-1150. 
Christopoulos, C., and Pampanin, S. (2004). ?Towards performance based seismic 
design of MDOF structures with explicit consideration of residual deformations.? 
ISET J. Earthquake Technology, 41(1): 53-73. 
Clark, P., Aiken, I., Kasai, K., Ko, E., and Kimura, I. (1999). ?Design procedures for 
buildings incorporating hysteretic damping devices.? Proceedings of the 68th 
Annual Convention, Structural Engineers Association of California, Sacramento, 
CA, USA. 
Crisfield, M. (1990). ?A consistent co-rotational formulation for non-linear three-
 dimensional beam elements.? Comp. Method in Appl. Mech. & Engrg., 81(2):131-
 150. 
Crisfield, M., and Moita, G. (1996). ?A unified co-rotational framework for solids, 
shells and beams.? Int. J. Solids Structures, 33(20-22): 2969-2992. 
Daliri A., Galehdar A., John S., Wang C., Towe W., and Ghorbani K. (2012) 
?Wireless strain measurement using circular microstrip patch antennas.? Sensors 
and Actuators A: Physical, 184: 86?92. 
De-Souza, R. (2000). ?Force-based finite element for large displacement inelastic 
analysis of frames.? Ph. D. Thesis, University of California, Berkeley, CA, USA. 
Deavours, D. (2009). ?Analysis and design of wideband passive UHF RFID tags 
using a circuit model.? 2009 IEEE International Conference on RFID, 283-290. 
 
 189 
 
Desroches R., and Smith B. (2004). ?Shape memory alloys in seismic resistant design 
and retrofit: a critical review of their potential and limitations.? Journal of 
Earthquake Engineering, 8(3): 415-429. 
Dobkin, D. (2008). The RF in RFID: Passive UHF RFID in Practice, Newnes, 
Burlington, MA, USA. 
Dolce, M., Cardone, D., and Marnetto, R. (2000). ?Implementation and testing of 
passive control devices based on shape memory alloys.? Earthquake engineering 
and structural dynamics, 29(7):945-968. 
Eastman, C., Fisher, D., Lafue, G., Lividini, J., Stoker, D., and Yessios, C. (1974). An 
Outline of the Building Description System, Research Report 50, Institute of 
Physical Planning, Carnegie Mellon Univ., Pittsburgh, USA. 
Eastman, C., Jeong, Y., Sacks, R., and Kaner, I. (2010). ?Exchange model and 
exchange object concepts for implementation of national BIM standards.? J. 
Comput. Civ. Eng., 24(1): 25?34. 
Eastman, C., Teicholz, P., Sacks, R., and Liston, K. (2011). BIM Handbook: A Guide 
to Building Information Modeling for Owners, Managers, Engineers and 
Contractors, 2nd Edition, John Wiley & Sons, Hoboken, New Jersey, USA.  
Ellingwood, B., and Dusenberry, D. (2005). ?Building design for abnormal loads and 
progressive collapse.? Computer-Aided Civil and Infrastructure Engineering, 
20(3): 194?205.  
El-Sheikh, A. (1999). ?Effect of force limiting devices on behaviour of space 
trusses.? Engineering Structures, 21: 34-44. 
Ergen, E., Akinci, B., and Sacks, R. (2007). ?Life-cycle data management of 
engineered-to-order components using radio frequency identification.? Advanced 
Engineering Informatics, 21(4):356-366. 
 
 190 
 
Estes, A., and Frangopol, D. (2003). ?Updating bridge reliability based on bridge 
management systems visual inspection results.? J. Bridge Eng., 8(6): 374-382. 
Filippou, F., Popov, E., and Bertero, V. (1983). ?Effects of bond deterioration on 
hysteretic behavior of reinforced concrete joints.? Report EERC 83-19, Earthquake 
Engineering Research Center, University of California, Berkeley. 
Filippou, F., and Fenves, G. (2004), ?Methods of analysis for earthquake-resistant 
structures.? Earthquake Engineering: From Engineering Seismology to 
Performance-Based Engineering, (Editors: Bozorgnia, Y. and Bertero, V.), 
Chapter 6, CRC Press, Boca Raton, FL., USA. 
General Service Administration (GSA) (2007). ?GSA's national 3D-4D BIM 
program.? US General Services. URL (last accessed on Nov. 18, 2012): 
http://www.gsa.gov/bim. 
Ghobarah, A., and Ramadan, T. (1990). ?Effect of axial forces on the performance of 
links in eccentrically braced frames.? Engineering Structure, 12(2): 106-113. 
Glover, B., and Bhatt, H. (2006). RFID Essentials, O?Reilly Media, Sebastopol, CA, 
USA. 
Golabchi, A., and Kamat, R. (2012). ?Interoperability issues between different BIM 
software products.? University of Michigan, UMCEE Report No 2012-02. 
Goldberg, E. (2003), ?New tools deliver BIM data-Building model yields valuable 
information.? Cadalyst Magazine, Dec. 2003. 
Gupta, A., and Krawinkler, H. (1999). ?Prediction of seismic demands for SMRFs 
with ductile connections and elements.? Report SAC/BD-99/06, SAC Joint 
Venture, Sacramento, Calif. 
 
 191 
 
Harrison, M., and Parlikad, A. (2006). ?Lifecycle ID and lifecycle data management.?
  AUTO-ID Labs, AEROIDCAM-005. URL (last accessed on Nov. 1, 2013): http://
 www.aero-id.org/research_reports/AEROID-CAM-005-Lifecycle.pdf. 
Isikdag, U., Aouad, G., Underwood, J., and Wu, S. (2007). ?Building information 
models: A review on storage and exchange mechanisms.? Bringing ITC 
Knowledge to Work, 24th W78 Conference, Maribor. 
Jia, Y., Sun, K., Agosto, F., and Quinones, M. (2006). ?Design and characterization 
of a passive wireless strain sensor.? Meas. Sci. Technol., 17: 2869?2876. 
Jin, J., and El-Tawil, S. (2003). ?Inelastic cyclic model for steel braces.? ASCE J. 
Engrg. Mech., 129(5): 548-557. 
Krishnan, S. (2010). ?Modified elastofiber element for steel slender column and brace 
modeling.? ASCE J. Struct. Engrg.,136(11): 1350-1366.  
Kurama, Y., Sause, R., Pessiki, S., and Lu, L. (1999). ?Lateral load behavior and 
seismic design of unbonded post-tensioned precast concrete walls.? ACI Struct. J., 
96(4): 622?632. 
Li, Z., Zhang, Y., and Wang, C. (2011). ?A sensor-driven structural health prognosis 
procedure considering sensor performance degradation.? Structures and 
Infrastructure Engineering, 9(8):764-776. 
Light, R., and Gossard, D. (1982). ?Modification of geometric models through 
variational geometry.? Computer Aided Design, 14(4): 209?214. 
Loh, K., Lynch, J., and Kotov, N. (2008). ?Inductively coupled nanocomposite 
wireless strain and pH sensors.? Smart Struct. Syst., 4(5): 531?548.  
Loo, C., Klmahgoub, K., Yang, F., Elsherbeni, A., Kajfez, A., and Elsherbeni, T. 
(2008). ?Chip impedance matching for UHF RFID tag antenna design.? Progress 
In Electromagnetics Research, PIER 81, 359?370. 
 
 192 
 
Lynch, J., and Loh, K., (2006). ?A summary review of wireless sensors and sensor 
networks for structural health monitoring.? Shock Vib. Dig., 38(2): 91?128. 
Mahin, S., Sakai, J., and Jeong, H. (2006). ?Use of partially prestressed reinforced 
concrete columns to reduce post-earthquake residual displacements of bridges.? 
Proc., 5th National Seismic Conf. on Bridges & Highways, San Francisco, CA, 
USA. 
Mansour, N., Shen, Y., Christopoulos, C., and Tremblay, R. (2008). ?Seismic design 
of EBF steel frames using replaceable nonlinear links.? Proc. 14th World 
Conference on Earthquake Engineering, Beijing, China.  
Mazzoni, S., McKenna, F., Scott, M., and Fenves, G. (2009). "The Open System for 
Earthquake Engineering Simulation (OpenSees) user command-language manual." 
Pacific Earthquake Eng. Research Center, Univ. Calif., Berkeley, CA. URL (last 
accessed on Nov. 7, 2013): http://opensees.berkeley.edu. 
McKenna, F. (1997). ?Object-oriented finite element programming: frameworks for 
analysis, algorithms and parallel computing.? Ph.D. Thesis, University of 
California, Berkeley, CA, United States. 
McKenna, F., and Fenves, G. (2005). ?Open system for earthquake engineering 
simulation.? Pacific earthquake engineering research center (PEER), Berkeley, 
California.  
Menegotto, M., and Pinto, P. (1973). ?Method of analysis of cyclically loaded RC 
plane frames including changes in geometry and non-elastic behavior of elements 
under normal force and bending, Preliminary Report.? IABSE, 13: 15-22. 
Miller, D., Fahnestock, L., and Eatherton, M. (2011). ?Self-centering buckling-
 restrained braces for advanced seismic performance.? ASCE Structures Congress 
2011, Las Vegas, Nevada, USA.  
 
 193 
 
Moghaddasi-B., N.  (2011). ?Performance evaluation of two new seismic resistant 
diagrid framing systems.? Ph.D. Thesis, University of Maryland, College Park, 
MD, United States. 
Morita, K., and Noguchi, K. (2006). ?Crack detection methods using radio frequency 
identification and electrically conductive material.? Technical Memorandum of 
Public Works Research Institute, No. 4022, Page 88. 
Mosalam, K., Talaat, M., and Park, S. (2008). ?Modeling progressive collapse & 
reinforced concrete framed structures.? Proc. 14th World Conference on 
Earthquake Engineering, Beijing, China. 
Motamedi, A., and Hammad, A. (2009). ?Lifecycle management of facilities 
components using radio frequency identification and building information model.? 
Journal of Information Technology in Construction (ITcon), 14: 238-262. 
NBIMS.  (2007). United States national building information modeling standard 
version 1?Part 1: Overview, principles, and methodologies, US National Institute 
of Building Science Facilities Information Council, BIM committee. URL (last 
accessed on Nov. 7, 2013): (http://www.wbdg.org/pdfs/NBIMSv1_p1.pdf. 
Neuenhofer, A., and Filippou, F. (1997). ?Evaluation of nonlinear frame finite 
element models.? Journal of  Structural Engineering, ASCE, 123(7): 958-966. 
Neuenhofer, A., and Filippou, F. (1998). ?A geometrically nonlinear flexibility-based 
frame finite element model.? Journal of Structural Engineering, ASCE, 124(6): 
704-711. 
Nikitin, P., and Rao, K. (2006). ?Theory and measurement of backscattering from 
RFID tags.? IEEE antennas and Propagation Magazine, 48(6): 212-218. 
NXP Semiconductors. (2008). ?An 1629 UHF RFID label antenna design-UHF anten
 na design.? Application note, Rev. 1.0, NXP B.V., 2008. URL (last accessed on   
Nov. 7, 2013): http://www.nxp.com/documents/application_note/AN162910.pdf. 
 
 194 
 
Okazaki, T., Engelhardt, M., Nakashima, M., and Suita, K. (2006). ?Experimental 
performance of link-to-column connections in eccentrically braced frames.? J. 
Struct. Eng., ASCE, 132(8): 1201-1211. 
Pablo, R. (2009).  ?Risk assessment of highway bridges: a reliability-based 
approach.? The Technology Interface Journal, 10(2). 
Pampanin, S., Christopoulos, C., and Priestley, M. (2003). ?Performance-based 
seismic response of frame structures including residual deformations. Part II: 
Multidegree-of-freedom systems.? J. Earthquake Eng., 7(1): 119?147. 
Popov, E., and Engelhardt, M. (1988). ?Seismic Eccentrically Braced Frames.? J. of 
Construction Steel Research, 10: 321-354. 
Pratt, M. (1993). ?Geometric methods for computer-aided design.? Fundamental 
Developments of Computer-aided Geometric Modeling, L. Piegl, ed., Academic 
Press, London, 271?320. 
Ramadan, T., and Ghobarah, A. (1995). ?Analytical model for shear-link behavior.? 
J. Struct. Eng., ASCE, 121(11): 1574-1580. 
Rankine, W. (1857). ?On the stability of loose earth.? Philosophical Transactions of 
the Royal Society of London, Vol. 147. 
Rao, K., Nikitin, P., and Lam, S. (2005). ?Antenna design for UHF RFID tags: a 
review and a practical application.? IEEE Transactions on Antennas and 
Propagation, 53(12): 3870-3876. 
Rehm, K. (2013). ?Bridge inspection: primary element.? Roads & Bridges, May 6, 20
 13. URL (last accessed on October 24, 2013): http://www.roadsbridges.com/bridge
 -inspection-primary-element.    
Reinhorn, A., Viti, S., and Cimellaro, G. (2005), ?Retrofit of structures: strength 
reduction with damping enhancement.? Proceedings of the 37th UJNR Panel 
Meeting on Wind and Seismic Effects, 16-21 May, Tsukuba, Japan. 
 
 195 
 
Richards, P., and Uang, C. (2006). ?Testing protocol for short links in eccentrically 
braced frames.? J. Struct. Eng. ASCE, 132(8): 1183-1191. 
Ricles, J., and Popov, E. (1987). ?Dynamic analysis of seismically resistant 
eccentrically braced frames.? Report No. VCBIEERC-87107, Earthquake Engrg. 
Res. Ctr., Univ. of California, Berkeley, California. 
Ricles, J., Sause, R., Garlock, M., and Zhao, C. (2001). ?Post-Tensioned seismic 
resistant connections for steel frames.? Journal of Structural Eng., ASCE, 127(2): 
113-121.  
Ruiz-Garcia, J., and Miranda, E. (2006a) ?Evaluation of residual drift demands in 
regular multi-storey frames for performance-based seismic assessment.? 
Earthquake Engineering and Structural Dynamics, 35(13): 1609-1629. 
Ruiz-Garcia, J., and Miranda, E. (2006b) ?Residual displacement ratios for 
assessment of existing structures.? Earthquake Engineering and Structural 
Dynamics, 35(3): 315-336. 
Schellenberg, A., and Mahin, S. (2006a) ?Application of an experimental software 
framework to hybrid simulation of structures through collapse.? First European 
Conference on Earthquake Engineering and Seismology, Geneva, Switzerland. 
Schellenberg, A., Mahin, S., and Fenves, G. (2006b) ?Application of an experimental 
software framework for international hybrid simulation.? 4Th international 
conference on earthquake engineering, Taipei. 
Schellenberg, A., Kim, H., Mahin, S., and Fenves, G. (2008) ?Environment 
independent implementation of a software framework for fast local and 
geographically distributed hybrid simulations.? The 14th World Conference on 
Earthquake Engineering, Beijing, China. 
Schellenberg, A., Mahin, S., and Fenves, G. (2009) ?Advanced implementation of 
hybrid simulation.? PEER report 2009/104, University of California at Berkeley, 
CA, USA. 
 
 196 
 
Scott, M., and Fenves, G. (2010). ?Krylov subspace accelerated Newton algorithm: 
applicationto dynamic progressive collapse simulation of frames.? ASCE Journal 
of Structural Engineering, 136(5): 473-480.  
Seo, C., and Sause, R. (2005) ?Ductility demands on self-centering systems under 
earthquake loading.? ACI Structural Journal, 102(2): 275-285. 
Shepard, S. (2005). RFID: Radio Frequency Identification. McGraw-Hill, New York, 
USA. 
Singh, V., Gu, N., and Wang, X. (2011). ?A theoretical framework of a BIM-based 
multi-disciplinary collaboration platform.? Autom. Constr., 20(2): 134?144.  
Skyetek (2012). SkyeModule M9 datasheet, Version 081313, Skyetek Inc., Denver, 
Colorado, USA.  
Somerville, P., Smith, N., Punyamurthula, S., and Sun, J. (1997). ?Development of 
ground motion time histories for phase 2 of the FEMA/SAC steel project.? 
SAC/BD-97/04, SAC Joint Venture, Sacramento, CA, USA. 
Song, G., Ma, N., and Li, H. (2006). ?Applications of shape memory alloys in civil 
structures.? Engineering Structures, 28(9):1266-1274. 
Spacone, E., Ciampi, V., and Filippou, F. (1996). ?Mixed formulation of nonlinear 
beam finite element.? Comput. Struct., 58(1):71?83. 
Susoy, M., Catbas, F., and Frangopol, D. (2008). ?Evaluation of time-variant bridge 
reliability using structural health monitoring.? Proc. IMAC XXVI Conference & 
Exposition on Structural Dynamics - Technologies for Civil Structures, Orlando, 
FL, USA.  
Tada, M., and Suito, A. (1998). ?Static and dynamic post-buckling behavior of truss 
structures.? Engineering Structures, 20(4-6): 384-389.  
 
 197 
 
Uang, C., and Nakashima, M. (2003). ?Steel buckling-restrained braced frames.? 
Chapter 16 in Earthquake Engineering: Recent Advances and Applications, CRC 
Press, Boca Raton, FL. 
Uriz, P., Filippou, F., and Mahin, S. (2008). ?Model for cyclic inelastic buckling of 
steel braces.? ASCE J. Struct. Engrg., 134(4): 619-628. 
Van-Nederveen, G., and Tolman, F. (1992). "Modelling multiple views on buildings." 
Automation in Construction, 1(3): 215?224.  
Viti, S., Cimellaro, G., and Reinhofrm A. (2006). ?Retrofit of a hospital through 
strength reduction and enhanced damping.? Smart Structures and Systems, 2(4): 
339-355. 
Wang, T., McCormick, J., Yoshitake, N., Pan, P., Murata, Y., and Nakashima, M. 
(2009). ?Collapse simulation of a four-story steel moment frame by a distributed 
online hybrid test.? Earthquake Engrg. Struct. Dyn., 37(6):955?974. 
Wikipedia. (2013). Building Information Modeling, web link.  URL (last accessed on 
October 31, 2013): http://en.wikipedia.org/wiki/Building_information_modeling. 
Wong, A., Wong, F., and Nadeem, A. (2011) ?Government roles in implementing 
building information modeling systems: Comparison between Hong Kong and the 
United States.? Construction Innovation: Information, Process, Management, 
11(1): 61 ? 76. 
Yi, X., Wu, T., Wang, Y., Leon, R., Tentzeris, M., and Lantz, G. (2011). ?Passive 
wireless smart-skin sensor using RFID-based folded patch antennas.? International 
Journal of Smart and Nano Materials, 2(1): 22-38. 
Yi, X., Cho, C., Cooper, J., Wang, Y., Tentzeris, M., and Leon, R. (2013). ?Passive 
wireless antenna sensor for strain and crack sensing - electromagnetic modeling, 
simulation, and testing.? Smart Materials and Structures, 22(8). 
 
 198 
 
Zareian, F., and Medina, R. (2010). ?A practical method for proper modeling of 
structural damping in inelastic plane structural systems.? Computers & Structures, 
88(1-2): 45-53. 
Zhu, S. and Zhang, Y. (2008). "Seismic analysis of concentrically braced frame 
systems with self-centering friction damping braces." ASCE Journal of Structural 
Engineering, 134(1): 121-131.