UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Shake table testing and performance evaluation of a post-seismic repaired and retrofitted Oak Street… Chen, Ge 1998

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1998-0124.pdf [ 12.23MB ]
Metadata
JSON: 831-1.0050184.json
JSON-LD: 831-1.0050184-ld.json
RDF/XML (Pretty): 831-1.0050184-rdf.xml
RDF/JSON: 831-1.0050184-rdf.json
Turtle: 831-1.0050184-turtle.txt
N-Triples: 831-1.0050184-rdf-ntriples.txt
Original Record: 831-1.0050184-source.json
Full Text
831-1.0050184-fulltext.txt
Citation
831-1.0050184.ris

Full Text

SHAKE TABLE TESTING AND PERFORMANCE EVALUATION OF A POST-SEISMIC REPAIRED AND RETROFITTED OAK STREET BRIDGE BENT WITH FIBERGLASS WRAPPING SYSTEM by G E C H E N B . A S c , The University of British Columbia, 1995 A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF M A S T E R OF A P P L I E D SCIENCE in T H E F A C U L T Y OF G R A D U A T E STUDIES D E P A R T M E N T OF CIVIL E N G I N E E R I N G We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y OF B R I T I S H C O L U M B I A March, 1998 © G e C h e n , 1998 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. 1 further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of C ) \ l ^ The University of British Columbia Vancouver, Canada Date \4kXL. B O , \<ft?s DE-6 (2/88) Abstract ABSTRACT A 27% scaled as-built model of the Oak St. Bridge bent S28 (denoted OSB-O) had been severely damaged when tested on the shake table at the University of British Columbia Earthquake Lab (E. Davey, 1996). During testing the structure's overall behavior was brittle, and little plastic deformation was observed before sudden shear damage occurred in the cap beam at a very low seismic load level. An earlier slow cyclic (or Quasi-static) test program on 45% scaled models of the same bent had generated similar failure mode and mechanism (Anderson et al. 1995). It was decided to repair and re-strengthen the 27% scaled bent model, and to test this retrofitted model (denoted OSB-R) by subjecting it to the same seismic loading conditions. The main objectives of this research program were: to develop a repair and retrofit scheme for the damaged Oak Street Bridge bent as-built model, to re-test this re-strengthened model on the shake table and experimentally determine the dynamic response of the model under seismic loading condition, and to examine the effectiveness of the retrofit system by data interpretation and performance comparison between the as-built and re-strengthened models. To repair the damage to the OSB-O, concrete patching and epoxy injecting techniques were employed. A newly developed FRP (fiber reinforced plastic) jacketing system, QuakeWrap™ was chosen to retrofit the repaired bent model. The performance of the re-strengthened bent model was overall satisfactory. It behaved in a ductile manner instead of a brittle one, as plastic hinges were developed at the top of bent columns as ii Abstract anticipated. The retrofitted model suffered far less damage than the as-built one, and no major shear damage was observed anywhere on the structure, especially on the cap beam. A significantly higher strength and ductility level were achieved by this re-strengthened model. The structural integrity and stability of the specimen were well maintained at the end of testing, when further testing was prevented by the shake table limitation. The results also matched the testing results from the earlier slow cyclic tests on a 45% scaled model retrofitted using a similar technique. The testing results indicated that the 27% scaled repaired and retrofitted model had an ultimate strength of 160kN, and a yield strength of 120kN at a displacement of 6.6mm; while the values scaled from the slow cyclic test results showed that, an untested as-built Oak Street Bridge bent model retrofitted with the FRP jackets had an ultimate strength of 166kN, and a yield strength of 129kN at a displacement of 6.5mm. The research results have indicated that the FRP wrapping system could provide the structural members with sufficient and reliable shear strength so that brittle shear damage could be prevented or minimized during a real seismic event. Better confinement in the section comers also can reduce the severity of cover concrete spalling and minimize the steel buckling. Installed at designated locations of the structure, the system would help the structure to deform in a plastic ductile manner and to dissipate seismic energy in the process, and eventually to reach large displacements and ductility levels without significant reduction of the structure's load resisting ability and without losing its structural stability. iii Table of Contents TABLE OF CONTENTS A B S T R A C T ii T A B L E O F C O N T E N T S iv L I S T O F F I G U R E S viii L I S T O F T A B L E S xi A C K N O W L E D G M E N T S xii CHAPTER 1 INTRODUCTION 1 1.1 Background 1 1.2 Objectives 4 1.3 Scope 5 1.4 Literature Study And Other Related Research 6 1.4.1 Slow cyclic tests of two column bridge bents 6 1.4.2 Reversed cyclic tests on columns retrofitted with FRP wraps 8 1.4.2.1 Test carried out at the University of Arizona 8 1.4.2.2 Test carried out at the University of California at San Diego 9 1.4.3 Static lateral loading test on bridge columns retrofitted with CRFP 11 1.4.4 General remarks 13 CHAPTER2 PREVIOUS TESTS AND DAMAGED BENT MODELS 14 2.1 Introduction 14 2.2 Oak Street Bridge Slow Cyclic Tests 15 2.3 Shake Table Test of Oak Street Bridge Bent 27% Model 17 2.3.1 Test set-up 18 2.3.2 Instrumentation 21 2.3.3 Input motions and test program 21 2.3.4 Test results 22 iv Table of Contents 2.4 Damaged Specimen Investigation 22 CHAPTER 3 DEVELOPMENT AND DESIGN OF REPAIRING AND RETROFITTING SCI 1EME 26 3.1 Introduction 26 3.2 Design Objectives 27 3.3 Seismic Assessment of The Bent 28 3.3.1 Analytical assessments 29 3.3.2 Experimental assessments 29 3.4 Remedy of Deficiencies 31 3.5 Design of Retrofitting System 32 3.5.1 FRP composite wrapping material 32 3.5.2 Wrapping design 33 3.5.2.1 Section moment capacities 34 3.5.2.2 Demands in the cap beam 35 3.5.2.3 FRP wrap design for the cap beam 37 3.5.2.4 Column wrap design 38 3.5.2.5 Summary 39 3.5.3 Longitudinal post-tensioning 39 3.5.4 Design summary 41 CHAPTER 4 REPAIRING AND RE-STRENGTHENING TECHNIQUES 42 4.1 Introduction 42 4.2 Repairing of The Damaged Bent 43 4.2.1 Concrete patching 43 4.2.2 Epoxy Injection 45 4.3 Retrofitting of The Repaired Bent with FRP Wrapping 47 4.4 Post-Tensioning System Installation 49 CHAPTER5 RETROFTITED MODEL AND ITS PREDICTED BEHAVIOR 50 5.1 Physical Description 50 5.2 Predicted Model Behavior 52 5.3 Scaling Factors 53 v Table of Contents CHAPTER 6 TEST ARRANGEMENT 55 6.1 Introduction 55 6.2 Test Set-Up 55 6.2.1 Loading system 55 6.2.2 Supports 56 6.3 Input Motion And Test Program 57 6.3.1 Input motion 57 6.3.2 Test program 61 6.4 Instrumentation And Data Acquisition System 63 6.4.1 Strain gauges 63 6.4.2 Displacement transducers and accelerometers 64 6.4.3 Hammer test data 64 CHAPTER 7 TESTING AND OVERALL EXPERIMENTAL RESULTS 66 7.1 Introduction 66 7.2 Runs 67 7.3 Tests Description 68 7.4 Hammer Test Results 71 CHAPTER8 DATA INTERPRETATION 75 8.1 Introduction 75 8.2 Displacement And Acceleration Time History 75 8.2.1 Displacement time history 76 8.2.2 Acceleration time history and inertia force 82 8.3 Hysteresis Curves And Energy Dissipation Mechanism 90 8.4 Ductile Level 92 8.5 Frequency Responses 95 vi Table of Contents CHAPTER9 TEST RESULT COMPARISON OF RIvSmENGTHENED AND ORIGINAL MODELS. 98 9.1 Introduction 98 9.2 Damage to Specimen 98 9.3 Stiffness Degradation 101 9.4 Hysteresis Loops 103 9.5 Load-Deformation Characteristics 105 9.6 Displacement Ductility Level 109 CHAPTER 10 SUMMARY AND CONCLUSIONS I l l 10.1 Summary H I 10.2 Conclusions REFERENCES 114 APPENDIX A - DRAWINGS 116 APPENDIX B - DATA INFORMATION AND TECHNICAL SPECIFICATIONS 122 APPENDIX C - PHOTOGRAPHS 130 APPENDIX D - DATA SAMPLES 143 vii List of Figures LIST OF FIGURES Figure 1.1 Load-Deflection Curve for the column before repair 10 Figure 1.2 Load-Deflection Curve for the column after repair 10 Figure 1.3 Comparison of load-deflection envelopes 10 Figure 1.4 Comparison of Load-displacement Curves 11 Figure 1.5 Load-Displacement Curves 12 Figure 2.1 Oak Street Bridge Bent S28, Model Dimensions 15 Figure 2.2 Test Set-up, Slow Cyclic Test 16 Figure 2.3 Comparison of Load-deflection Response, Slow Cyclic Test 17 Figure 2.4 Vertical Loading Set-up 19 Figure 2.5 Steel Mass Block Connection 20 Figure 2.6 Lateral Support System 20 Figure 2.7 'Scaled Acceleration Time History Record 21 Figure 2.8 27% Model Dimensions 23 Figure 2.9 Detailed Drawing for The West Shear Opening, 27% scaled model 24 Figure 3.1 Analysis Points on Bent 34 Figure 3.2 Free Body Diagram of Joint 36 Figure 3.3 Critical Region for Shear Demand 36 Figure 3.4 Post-tensioning System 40 Figure 3.5 End Anchor Plate 40 Figure 4.1 Patching of Shear Opening 44 Figure 4.2 Epoxy Inj ection 47 v i n List of Figures Figure 5.1 Re-strengthened Model 50 Figure 5.2 Later Load Displacement Response for OSB5 54 Figure 5.3 Idealized And Actual Load-Deflection Response of OSB5 54 Figure 6.1 Pin Connection 57 Figure 6.2 Response Spectra of Different EQ Records vs. Design Spectrum, £ = 5% 58 Figure 6.3 Time History of Joshua Tree E-W Record 58 Figure 6.4 Original And Scaled Acceleration Time History 60 Figure 6.5 Displacement Response Spectra, \ = 5% 60 Figure 6.6 Acceleration Response Spectra, % = 5% 61 Figure 6.7 Strain Gauge Layout 63 Figure 6.8 External Instrumentation 64 Figure 6.9 Hammer Test Setup 65 Figure 7.1 Sample FRF Plot (120% run) .72 Figure 7.2 First Natural Frequency of the Specimen Through Runs 73 Figure 8.1 Table and Bent Relative Displacement for 10%c run and 150%b run 77 Figure 8.2 Table and Bent Relative displacement for 200%b run, 1.40g 78 Figure 8.3 Maximum Displacements of the Cap Beam and Mass Block through Runs 80 Figure 8.4 Bent Displacement and Mass Block Displacement 81 Figure 8.5 Acceleration Time History Plots for 80%b run and 150%b run 83 Figure 8.6 Max. Bent Absolute Acceleration vs. Max. Table Acceleration 85 Figure 8.7 Acceleration Magnification Changes through Runs 86 Figure 8.8 Maximum Inertia Force Changes through Runs 86 Figure 8.9 Inertia Force - Displacement Plot (westward motion) 87 Figure 8.10 Maximum Bent Relative Displacement Change (westward motion) 87 ix List of Figures Figure 8.11 Load-deflection Curve for The Retrofitted Bent., 89 Figure 8.12 Hysteresis Plot for Six Typical Runs 91 Figure 8.13 Displacement Ductility Changes through Runs 94 Figure 8.14 Natural Frequencies from Hammer Test Data And Shake Table Test Data 95 Figure 8.15 Sample FRF Plots for The 20%b Run And 200%a Run 96 Figure 9.1 Comparison of Displacement Time History, 150% run 99 Figure 9.2 Max. Mass Acceleration vs. Max Table Acceleration, OSB-0 100 Figure 9.3 Max. Mass Acceleration vs. Max Table Acceleration, OSB-R 101 Figure 9.4 Comparison of Frequency Degradation, Hammer Tests 102 Figure 9.5 Comparison of Frequency Degradation, Shake Table Tests 102 Figure 9.6 Comparison of Hysteresis Loops 104 Figure 9.7 Comparison of Maximum Inertia Force 106 Figure 9.8 Comparison of Maximum Bent Relative Displacement 107 Figure 9.9 Comparison of Load - Deformation Curves 108 Figure 9.10 Comparison of Displacement Ductility 109 x List of Tables LIST OF TABLES Table 2.1 Scaling Values 18 Table 3.1 Code Shear Capacities of The Cap Beam, in kN 29 Table 3.2 Mechanical Properties of FRP 33 Table 3.3 Moment Capacity of the cap Beam 35 Table 3.4 Flexural Capacity of Column 35 Table 3.5 FRP Wrapping System 41 Table 3.6 Post-tensioning System 41 Table 5.1 Retrofitting System Descriptions 51 Table-6.1 Characteristics of The Joshua Tree W-E record 59 Table 6.2 Planned Testing Runs 62 Table 7.1 Testing Runs 67 Table 7.2 Hammer Test Results 72 Table 8.1 Maximum Displacements 79 Table 8.2 Maximum Accelerations 84 Table 8.3 Max Bent Relative Displacements and Ductility Levels 93 Table B. 1 Shake Table Data Files 124 Table B.2 Data Column Keys and Calibration Factors 125 Table B.3 Hammer Test Data Storage Information 126 xi Acknowledgments ACKNOWLEDGMENTS The author is extremely grateful for the guidance and constant encouragement provided by the thesis supervisor, Dr. Robert Sexsmith, through this research. His suggestions, practical advice and encouragement are largely responsible for the successful completion of this research project, and are very much appreciated. Gratitude is expressed to Dr. Carlos Ventura for sharing his expertise in the field of structural dynamics, to Dr. Don Anderson for his many helpful suggestions, and to Dr. Ricardo Foschi for final reading of this thesis. This research project was sponsored by the Natural Sciences and Engineering Research Council of Canada, and their financial support is gratefully acknowledged. The author would also like to thank Polycrete Restorations Ltd. for the service on repairing and installation work, and the Structural Rehabilitation Corporation for supplying QuakeWrap™ and other materials as well as background information. In addition, Dywidag Systems International Canada Ltd. provided equipment and labor for post-tensioning installation, and their support is appreciated. Many thanks should go to Howard Nichols and Vincent Latendresse, who assisted the author on setting up the test specimen and running the shake table test. Their assistance and cooperation was greatly appreciated. Appreciation is extended to Dick Postgate, Douglas Smith and other staff in the machine shop for their help during this research project. Also, the author is thankful to Elizabeth Davey for sharing her research experiences, to Ernie Wong for his help during the testing, and to Anne Maloney for proofreading the thesis. The author's friends have played important roles through these two years, and their friendship has made this part of his life as a graduate student at the University of British Columbia a enjoyable and memorable one . Finally, the author would like to express his gratitude to his mother for her unconditional moral support, and to Helen, for her understanding and the strength she has brought to him through last years. xii Chapter 1—Introduction Chapter 1 INTRODUCTION 1.1 Background The seismic vulnerability of many existing structures has became the focus of a great deal of research over the last decade. Several recent earthquakes, such as the 1987 Whittier Narrows, the 1989 Loma Prieta, the 1994 Northridge and the 1995 Kobe earthquakes have indicated that the seismic performance of these structures was inadequate, and demonstrated the urgent need for seismic retrofitting and upgrading. Many bridges serve as vital links in infrastructure systems, and the consequences of collapses of these bridges would be severe. Therefore many analytical and experimental projects have been carried out to study the seismic performance and vulnerabilities of these essential public service bridges. Since the early 1990's, a series of research programs has been carried out at the University of British of Columbia to assess the vulnerabilities and study the seismic performance of some reinforced concrete bridge components. In particular, several analytic and experimental studies have focused on the Oak Street Bridge approach bents, which are two-column concrete reinforced bents commonly constructed for bridge approaches in the Lower Mainland of British Columbia. Following a preliminary study done by Crippen International (Crippen International Ltd., 1993), now Klohn-Crippen, which indicated deficiencies on these bents, a slow cyclic test program on models of 1 Chapter 1—Introduction 45% scale to the prototype was carried out (Anderson et al, 1995). Six models were tested during the research. In five models of the Oak Street Bridge bent S28, one was an as-built model and four were as-built models retrofitted using a variety of schemes. One as-built model of the nearby Queensborough Bridge bent S26 was first tested and then retested after a post-seismic repair was made. Recently, an as-built 27% scale Oak Street Bridge bent S28 model was tested by means of shake table tests to examine the structure's performance when subjected to a real earthquake time history (Davey, 1996). The results of these projects have verified the deficiencies existing in these bridge bents. Many of these shortcomings could cause the bridges to fail in a brittle manner at low seismic load levels. The most critical hazard noted is that specimens in both slow cyclic tests and shake table tests were severely damaged by a brittle shear failure in the cap beams, because the cap beams in these bents lack sufficient transverse shear reinforcement, and potential plastic hinge regions were inadequately detailed to carry shear. Significant cracks in the joint regions also indicate the lack of proper reinforcement in both vertical and horizontal directions. After the 27% scale model was tested and failed on the shake table, it was considered worthwhile to carry out a further program of repairing and re-strengthening the model, and test it again on the shake table subjecting it to the same earthquake motions as suffered by the original model. The original specimen had been so badly damaged that the structure would appear to require demolition in the case of an actual seismic event. Such a post-seismic repairing and re-strengthening test program was justified by several considerations: 2 Chapter 1—Introduction • First, as the data from the original shake table test was available, it was possible to compare the performance of the as-built model and repaired model and to examine the effectiveness of a repair and retrofit scheme. • Secondly, the performance of a repaired 27% scale model on the shake table could also be compared with the performance of the repaired larger specimen tested under the slow cyclic loading, and the effect of the scale and loading conditions upon the models could be examined. • Thirdly, it would enhance understanding of not only the behavior of the re-strengthened damaged specimen but also the response of normally retrofitted structures under seismic loading. • Finally, such a repairing and testing program could provide a proper analytic base for the actual practice during post-seismic restoration, when the damaged infrastructure system has to be brought back to service in the shortest time possible. A systematic study of a fast, effective and economical post-seismic repair scheme would be of great significance to a smooth and successful restoration operation. It was decided to repair the damaged Oak Street Bridge bent model employing concrete patching and epoxy injecting techniques, which are commonly used in practice to repair mainly shrinkage cracks and accidental cracks that occur during construction in reinforced concrete structures. To re-strengthen the model, several alternatives were considered from among various seismic retrofitting techniques currently used to upgrade existing bridge structures. It has been understood that the design of retrofits in general is governed by an improvement in structure's ductility level rather than an increase in strength so that the structure would deform in a ductile manner instead of a brittle one. Examples of bridge retrofit practices in California and British Columbia were outlined by Mitchell, 3 Chapter 1—Introduction Sexsmith and Tinawi (1995). In this case, the FRP ( fiberglass reinforced plastic) wrapping technique was chosen to re-strengthen the model, because of the nature of the material and the fact that in the earlier slow cyclic test, the specimen retrofitted using this similar technique performed relatively well compared with others. It was also considered a good opportunity to study the performances of two models retrofitted using similar technique but subjected to different loading conditions, namely quasi-static cyclic loading and dynamic loading (shake table). 1.2 Objectives The main objectives of this test project are to develop a repair and retrofit scheme for the damaged Oak Street Bridge bent S28 model, to test the re-strengthened model on the shake table and experimentally determine the dynamic responses of the model under the dynamic loading condition, to quantify its ultimate strength and ductility level, to examine the effectiveness of this repair and retrofit scheme through data interpretation and performance comparison between the re-strengthened and original models, and to provide structural analytical bases and support for future practice. More specifically, the objectives of this thesis include • to develop a repair and retrofit scheme which ensures 1. reestablishment of the damaged specimen's structural integrity 2. improvement of its shear strength in the cap beam and joint regions to force plastic hinging in the columns 3. ultimate control over the failure mode and mechanism 4 Chapter 1—Introduction • to implement such a scheme by repairing the bent and installing the fiberglass wrapping system • to test the re-strengthened model by subjecting it to several test runs with increasing magnitudes of a chosen earthquake motion on the shake table, and to record experimental data by several means • to evaluate the model's overall performance and determine its ultimate strength and ductility level by interpreting data, studying the model's damage mechanism, and examining its dynamic characteristics • to further determine the effectiveness of the scheme by comparing the experimental results of shake table tests of both re-strengthened and original models. 1.3 Scope This thesis presents an experimental study of the dynamic response of a post-seismic repaired Oak Street Bridge bent model when tested on a shake table. The project is a part of a research program focusing on the seismic performance and retrofitting techniques of the existing reinforced concrete bridges in the Lower Mainland of British Columbia. The main stages and topics of the project covered in the thesis are the review of the tests and investigation of the damage of the original model, the development and implementation of a post-seismic repair and retrofit scheme, testing of the re-strengthened specimen, interpretation of data obtained, and finally the evaluation of the bent performance by analyzing the test results and comparing them with the results from the original bent model tests. 5 Chapter I—Introduction The bent was 27% scale, and the original undamaged one was modeled after the Oak Street Bridge bent S28. The bent was repaired and re-strengthened with fiberglass wrapping system at critical regions of the cap and both columns. In addition, a post-tensioning system was applied externally. It was tested dynamically on the digitally-controlled shake table in the Earthquake Engineering Research Laboratory at the University of British Columbia. 1.4 Literature Study and Other Related Researches Despite the recent popularity of seismic retrofit research, little work has been done in the area of repairing and re-strengthening damaged reinforced concrete frame structures using the FRP wrapping technique. Furthermore, literature related to the study of the dynamic response of structures re-strengthened using this technique when subjected to dynamic loading have not been seen. A number of preliminary research studies have been done on the push-over tests and slow cyclic tests of reinforced concrete columns and beams repaired and/or retrofitted with similar techniques. In addition, some experimental reports have been found, which studied the dynamic tests of reinforced concrete structures retrofitted using other techniques, such as the steel jacketing method. 1.4.1 Slow cyclic tests of two column bridge bents The Oak Street & Queensborough Bridges two column bent test was carried out as a part of a research series on bridge seismic performance and retrofitting at the University of British Columbia (Anderson et al, 1995). Six 45% scale models were tested, one of which was an Oak Street Bridge bent retrofit 6 Chapter 1—Introduction with FRP jackets (specimen 0SB5), and the other of which was a restored Queensborough Bridge bent after it had been damaged (specimen QB1R). OSB5 was retrofitted using fiberglass wrap around the shear critical regions of the cap beam and the upper 40% of the columns. The wrapping system was designed according to a report detailing design recommendation provided by the Ffexcel-Fyfe Co. in California. The cap beam was fitted with 3 wraps at a thickness of 0.5 inches per wrap, which was designed to improve the overall shear capacity to meet the shear demand, while two wraps on the columns were designed to carry all the shear. The specimen was then subjected to a cyclic lateral load simulation earthquake load, while a vertical load simulation dead load was applied constantly. The imposed deflection was increased in successive sequences of cycles until the specimen failed. The specimen reached a maximum seismic base shear around 100 kips at a displacement of 2.8 inches, and at a displacement of 4 inches subsequent cycles showed more rapid degradation though no sudden failure was observed. Although the seismic base shear and the maximum ductility for OSB5 were approximately the same as the other specimens retrofitted with steel jacketing and post-tensioning, it performed significantly better in that the degree of damage, spalling, and large cracking was less severe. The as-built 45% model of the Queensborough Bridge bent performed better than the Oak Street as-built model due to the increased amount of transverse reinforcement in the cap beam but suffered more severe damage in the joint regions. The restoration of the damage model was carried out by Polycrete Restorations Ltd. The epoxy injection technique was employed to repair cracks in the cap beam and joint regions. This QB1R specimen was then tested under the same loading conditions as other models. Overall, QB1R performed better than QB1. The testing results indicated that when the displacements were taken to the same maximum values used in the QB1 test, the base shear for QB1R was 24% higher than the original specimen QB1 (105 kips vs. 85 kips). 7 Chapter 1—Introduction 1.4.2 Reversed cyclic tests on columns retrofitting with FRP wraps Some preliminary work in the area of FRP materials and re-strengthening techniques using FRP was done at the University of Arizona by Saadatmanesh, Ehsani and others (Saadatmanesh et al, 1995). Around the same time, similar research was carried out at the University of California at San Diego by Priestly, Seible and others in conjunction with the Hexcel-Fyfe Co. (Priestly et al, 1993). Similar results were obtained from these studies, although retrofitting materials and techniques were slightly different between research conducted by Saadatmanesh et al. and by Priestly et al. 1.4.2.1 Test carried out at the University of Arizona The project carried out at the University of Arizona was cyclic testing of repaired earthquake damaged columns with prefabricated FRP wraps. Four 1/5-scale prototype bridge column specimens, two circular and two rectangular, were tested under reversed inelastic cyclic loading until failure. At the end of the tests, spalling of the concrete cover, opening of the 90-deg hooks of transverse reinforcement, yielding of stirrups, buckling of longitudinal bars, and dropping of the lateral load were marked and recorded. These specimens were repaired by chipping out loose concrete and filling the. gaps with fresh concrete, and then re-strengthened by applying FRP wraps. The FRP materials used were made of E-glass in the form of a unidirectional fabric. In general, all repaired columns performed extremely well under the same loading conditions, and the following conclusions were drawn from the test results: • The FRP composite wraps were effective in restoring the flexural strength and ductility capacity of earthquake damaged concrete columns. • After repair with FRP wraps, columns with lapped starter bars developed stable hysteresis loops up to the displacement ductility of fx = 4 (JJ = total imposed displacement / yield 8 Chapter 1—Introduction displacement). In columns with continuous reinforcement, the hysteresis loops were stable up to /i=6 without showing any sign of structural degradation. • In all repaired specimens, the rate of stiffness deterioration under large reversed cyclic loading was lower than that of the corresponding original columns. However, the initial stiffness of repaired columns was lower than that of the original columns. (Saadatmanesh et al., 1995) 1.4.2.2 Test carried out at the University of California at San Diego Another type of fiber composite material, Tyfo-s, was involved in the tests conducted by Priestly et al., and similar results were obtained (Priestly et al, 1993). In this investigation, a 0.4 scale circular shear column (as-built) was tested. It experienced brittle shear failure at column mid-height and lost 75% of its maximum lateral load carrying capacity. Then, this damaged as-built specimen was repaired by Hexcel Fyfe Co. with patching and fiberglass/epoxy, and re-tested applied with the same loading condition, which was the slow reversed cyclic lateral loading similar to ones used in the two tests described in the previous sections. The test results indicated that the techniques employed for repairing and retrofitting was fully effective in the following areas: • restoring the original column stiffness characteristics • transforming the brittle shear failure mode into a ductile flexural mode • providing improved displacement ductility to the systems Figure 1.1 and Figure 1.2 are load-deflection curves for the tested column before and after repair. The lateral load capacity of the original column dropped significantly after its displacement ductility had 9 Chapter 1—Introduction reached /J. = 2, while no sign of later strength degradation was found for the repair specimen even after its ductility level had reached fu = 10. Drift Ratio d/L (7.) -5 -4 -3 -2 -t 0 I 2 ? i S 250-200-150-n. 100-D 50-a o ^ 0-a -50-u a -100--IS0--200--250-_j i — i — i — i 1 1 ' • 1 — Pull M-3 Z 1.3 t / Push d - 0.467 in. V - 2 - 1 0 I 2 Deflection fin) 250-200-, ISO-' 100-i so-; o-i -so-\ -100--150--200--250 Drift Ratio a/L (7.) - 2 - 1 0 1 2 3 M-10 B 9 5 4 3 2 1 lrZgS^*^ Push A - 0.393 in. V -S -4 -3 -2 - 1 0 1 2 Deflection (in) Figure 1.1 Load-deflection Curve for Figure 1.2 Load-deflection curve for the column before repaired the column after repair (from SEQUAD Consulting Engineering, 1993) (from SEQUAD Consulting Engineering, 1993) The comparison was also made for the performance of the FRP jacket and the steel jacket retrofits. Figure 1.3 show the load-deflection envelopes for the original column and columns retrofitted with FRP jacket and steel jacket. It indicates that the FPR retrofit was equally effective as the steel jacket method, if not better, at providing improved ultimate lateral load resistance level and displacement ductility levels. Drift Ratio b/L (%) CO a o e CD T — ' — i — i — i — i — i — ' — I — ' — r - 4 - 3 - 2 - 1 0 1 2 Deflection fin) Figure 1.3 Comparison of load-deflection envelopes (from SEQUAD Consulting Engineering, 1993) 10 Chapter 1—Introduction 1.4.3 Static lateral loading test on bridge columns retrofitted with C R F P A testing project by Japan Highway Public Corporation in collaboration with Obayashi Corporation Technical Research Institute was carried out to determine the effectiveness of C R F P (Carbon Fiber Reinforced Plastics) retrofit method for seismically vulnerable concrete bridge columns ( Ohuchi et al, 1994). Several 1/3 scale column models with rebar cut off section at 1/3 height, and with various shear span ratios were wrapped with CRFP sheet, and tested under cyclic horizontal load combined with axial load. The loading conditions were similar to the previously described tests on columns retrofitted with FRP jackets, and similar test results were also generated. The retrofit significantly helped the specimens in achieving higher strength at large lateral displacements and higher ductility levels, for both short and slender columns. The improvement was more significant for shorter columns, in which high shear demand was more critical and generally responsible for the brittle shear failure. Figure 1.4 shows the comparison of load-displacement curves of a typical column with and without CFRP retrofit. Figure 1.4. Comparison of Load-displacement Curves (from Ohuchi etal, 1994) 11 Chapter 1—Introduction Before retrofitting, shorter columns with a shear span ratio of 3 experienced brittle shear failure before any yielding occurred, while ductile failure modes were obtained for the retrofitted ones. For the retrofitted columns, the displacement ductility and strength increased as the CFRP reinforcement increased as indicated in Figure 1.5-a. As for more slender columns with shear span ratio of 6, the specimens without CFRP retrofits failed in a comparatively ductile flexural mode with longitudinal steel buckling, while retrofitted slender columns behaved in a more ductile manner. In the case of slender columns, the effect of CFRP retrofit was more significant on preventing longitudinal steel buckling and concrete spalling. As shown in Figure 1.5-b, in less strengthened specimens, steel buckling and concrete spalling were not prevented because the CFRP jackets ruptured, while these depressions were prevented in the specimens with more CFRP retrofits. z oOr (c) Unstrengthened [ (PHOO) (d) Strengthened (PH Series) Figure 1.5 Load-Displacement Curves PL - Short Column, PH - Slender Column (fromOhuchi etal, 1994) 12 Chapter 1—Introduction 1.4.4 General Remarks The preliminary research reviewed above has significantly enhanced the general understanding of flexural and shear behavior of retrofitted concrete structures or structural members with composite fiber materials, such as fiberglass and carbon fiber reinforced plastic. From test results, it was evident that both FRC and C F R C were capable of providing effective tensile strength to structural members with a minimum amount of reinforcement due to their high tensile strength and light weight. Applied at desired locations, they could help structures to achieve high ultimate strength and greater ductility. However, these tests were found to have some limitations for the following reasons: • First, all these tests were carried under static loading conditions. Constant lateral loads were applied, and the specimens were brought to certain displacements repeatedly until they failed. This would certainly not be the case in reality, where most loading conditions are random and unsystematic, especially during seismic events. Seismic application is the major driving force behind this retrofit system development. • Secondly, most of these specimens tested were simple structural members, such as columns. The behavior and responses of a more complicated structure would be unique and more complex, and therefore needed to be further evaluated. • Finally, most of these columns tested were retrofitted with full height F R C or C F R C jackets. While this might be justifiable for simple members, it would be highly inefficient, unnecessary and possibly less effective for large-scale structures in actual practice. A systematic, simple, yet sensible design process was needed to be studied in detail. While these preliminary static test studies showed positive results regarding the effectiveness of this new concrete structure retrofitting technique, the necessity of further evaluating and studying the dynamic behavior and responses of a more complicated structure under a real earthquake loading condition became evident. 13 Chapter 2 — Previous Tests and Damaged Bent Models Chapter 2 PREVIOUS TESTS AND DAMAGED BENT MODELS 2.1 Introduction The Ministry of Transportation and Highways of British Columbia has conducted a seismic assessment of its major bridges in the Lower Mainland, and deficiencies have been discovered in the reinforced concrete approach bents of many important bridges, including the Oak Street Bridge and the Queensborough Bridge (Crippen International Ltd., 1993). It was decided to perform a series of tests on two-column reinforced concrete bents, typical of the approach spans of several bridges covered in the assessment. Slow cyclic tests and shake table tests were carried out on the models of the Oak Street Bridge bents. The models of the Oak Street Bridge subjected to these two different loading conditions failed in a similar manner. The cap beams of the models suffered severe damage due to sudden shear failures, and shear cracks in the joint regions were also significant, especially in the slow cyclic test models. The 27% scale model tested on the shake table suffered severe shear damage at both ends of the cap beam during the early stage of the test when the seismic load level was relatively low, and the structure disintegrated as the loading level increased. 14 Chapter 2 — Previous Tests and Damaged Bent Models 2.2 Oak Street Bridge Bent Slow Cyclic Tests A 45% scale as-built Oak Street Bridge bent S28 was tested under slow cyclic loading condition at the Structural Lab of the University of British Columbia, along with five other specimens: four 45% scale as-built Oak Street Bridge bent S28 retrofitted with various methods, and one 45% scale as-built Queensborough Bridge bent S26. The Queensborough Bridge bent was not retrofitted but a post-seismic repair was made after it was first tested (Anderson et al, 1995). The major dimensions of the Oak Street Bridge bent model are shown in Figure 2.1. r - 7 21 1 /2 -2 2 - 4 -T ' i _ r-7- T \ , R=G' R = V - 4 ' 2 - 3 ' R=r-4" ' 6 - q -21 1/2 Figure 2.1. Oak Street Bridge Bent S28, Model Dimensions The loading system consisted of constant vertical load simulating the structural dead load and a cyclic lateral load, which was applied slowly and repeatedly for many cycles with increasing imposed deflections for every cycle. The general test set-up is shown in Figure 2.2. The test program consisted of several sequences of load (displacement) cycles, and within each sequence there were three complete cycles. At first, several sequences at low load level were performed to predict the yield displacement, and the sequences after the yield point were carried out at multiples of this yield displacement. In other 15 Chapter 2 — Previous Tests and Damaged Bent Models words, for each sequence, a specimen was loaded to a certain displacement ductility level, and this level increased in successive sequences of cycles until the specimen failed. 5-4 1/2 5-7 1/2' 5-7 i/2" 5-4 1/2-5TRON0 FLOOR-Figure 2.2 Test Set-up, Slow Cyclic Test Several load cells and L V D T ' s were placed at various locations to record lateral force, vertical and horizontal displacement. There were also approximately 80 strain gauges to record strains of the reinforcing steel of each specimen. A l l signals were recorded every two seconds and stored in a personal computer. Test results indicated that the as-built specimen showed very poor ductile behavior, with peak lateral load capacity of only 60 kips. A large diagonal shear crack formed near the cut-off point of longitudinal steel in the cap beam at a very low displacement level. The crack increased in width with each cycle of loading until the specimen suddenly failed (see Photograph 1 in Appendix C). This premature cap beam brittle shear failure prevented any serious joint and column damage as the load demand on them was limited by such a failure. 16 Chapter 2 — Previous Tests and Damaged Bent Models A l l four retrofitted Oak Street Bridge bent specimens showed improvement to different extends in ultimate strength and ductility levels. The improvement introduced by applied external fiber glass wrapping jacket and cap beam post-tensioning on the specimen was especially significant. The specimen's peak lateral strength reached 105 kips, a great improvement over only 60 kips for the as-built one. In addition, a displacement ductility level of 9 was achieved before the test was terminated due to the limitation in the displacement capacity of the loading system. The comparison of the load-deflection response of this retrofitted specimen and the as-built one is shown in Figure 2.3. a. % 04 < W I V> W vi < PQ 140 100 60 20 -20 -60 -100 -140 } -5 -4 -3 -2 -1 0 1 2 3 4 5 JOINT DISPLACEMENT (in) a. g, < S Vi w VI < PQ 140 100 60 20 -20 -60 -100 -140 HA = 1 W if / t J / ii { fA 1 -5 -4 -3 -2 -1 0 1 2 3 4 5 JOINT DISPLACEMENT (in) Figure 2.3 Comparison of Load-deflection Response, Slow Cyclic Test 2.3 Shake Table Tests of 27% Scale Oak Street Bridge Bent Model The shake table tests of the 27% scale Oak Street Bridge bent model were done by E . Davey in July 1996 at the University of British Columbia, as the experimental part of her master's thesis. Most parts of the following configurations for tests were also later used for testing the repaired and re-strengthened model so that meaningful result comparisons could be made. 17 Chapter 2 — Previous Tests and Damaged Bent Models In Davey's thesis, scaling values for different physical parameters were calculated so that these parameters could be converted between the 27% scale model used in shake table tests, the 45% scale models in the slow cyclic tests, and the full size prototype. These scaling values are listed in Table2.1. Table 2.1 scaling values Parameter Scale Factor Basic Dimension Scale value Length L s Arbitrary 0.2700 Stress CTS Arbitrary 1.0000 Mass M s Arbitrary 0.0211 Force F s LS2 0.0729 Stiffness K s LS 0.2700 Moment Mom s LS3 0.0196 Period T s VMS /VLS 0.2790 Time ts VMS/VLS 0.2790 Acceleration A 5 L s / M s 3.4500 2.3.1 Test set-up The specimen tested was a 27% scale as-built Oak Street Bridge bent S28, and general set-up of the vertical loading system are shown in Figure 2.4. The total mass carried by the structure was 89kN, which consisted of three stacks of steel plates with various weights, and a concrete mass block. The weights of steel plates and the concrete block were proportioned in such a way that the center of gravity of the structure was at a height of 411mm above 18 Chapter 2 — Previous Tests and Damaged Bent Models the top of the cap beam, corresponding to 1524mm in the prototype. The weights of the concrete mass block and the steel plates were 12.3kN and 76.7kN respectively. Steel Plates Pin ^ rib Steel C«bte Concrete ll*ck S h « k e T « b l e Figure 2.4 Vertical Loading Set-up The dead load was transferred into the cap beam at five bearing locations, where five sets of rubber pads were used to simulate girder bearings in the real structure. Each set consisted of one hard 50-durometer rubber pad that gave the correct spacing between the cap beam and the concrete block, and one softer Shore A 40 polyurethane rubber pad that allowed some relative displacement with little change of applied force. However, the lateral force was transferred at two locations near two columns, where the cap beam and the concrete mass block were connected with Dywidag™ post-tensioning bars. The cap beam and joint forces were considered sufficiently accurately approximated assuming such a force transfer mechanism (Anderson et al, 1995). To restrain the out of plane motion of the mass block and achieve a fixed connection, the spaces around these bars were grouted after they had been tensioned. Each stack of 19 Chapter 2 — Previous Tests and Damaged Bent Models steel plates was placed at the top of the concrete mass block and held in place securely by 4 ready-rods embedded in the concrete block. (Figure 2.5) Ready Rod Sleel Plules= I Post Tensioning Rod ; p Cop B©Q F = 5 Concrete Block Figure 2.5 Steel Mass Block Connection To restrain the out of plane motion and at the same time allow in plane movement, the lateral support of the structure was achieved by tying down the structure at the top of the steel mass blocks using four wire ropes, as shown in Figure 2.6. The ropes were tensioned to 4.4kN, which yielded a frequency of 12Hz, which was considered sufficiently far from the structure's fundamental frequency of 7.5Hz. The tension value in one rope was monitored by an attached strain gauge. rz& a i z i Figure 2.6 Lateral Support System 20 Chapter 2 — Previous Tests and Damaged Bent Models The model bent was connected to the shake table by pin connections, which simulated the inflection points at the mid-height of both columns. 2.3.2 Instrumentation A total number of 24 strain gauges were attached to reinforcing steel bars of the bent at various locations to record strain in the reinforcements during the test. Several external instrumentation devices, such as displacement transducers and accelerometers, were placed on the system to record displacement and acceleration in different directions. A l l signals collected by these devices during tests were filtered and stored in a personal computer in the UBC Earthquake Lab. A normal video camera and a high-speed video camera were used to record the overall motion of the specimen during the earthquake motion and the development of cracks in one of the joints. 2.3.3 Input motions and test program The scaled Joshua Tree Fire Station Record E-W direction from 1992 Landers Earthquake was chosen to be the input motion signal, which drove the actuators of the shake table. Figure 2.7 is the acceleration time history of the scaled earthquake record. -300 - dt = 0.006 10 20 30 40 50 T ime (s) Figure 2.7 Scaled Acceleration Time History Record 21 Chapter 2 — Previous Tests and Damaged Bent Models A total number of 9 testing runs were performed, and their GPA (ground peak acceleration) ranged from 0.07g to 1.05g. The magnitude of shaking was increased in successive runs. It was predicted that the original record scaled to 0.7g would cause the specimen to fail; therefore, it was taken as the reference earthquake. Each testing run had a PGA value as a percentage value of 0.7g. Hammer impacting tests were performed after each run, and at the beginning and the end the all tests, in order to estimate the first mode natural frequency. 2.3.4 Test results Similar to the 45% scale as-built model tested under slow cyclic loading, this 27% scale model tested on the shake table failed at a lower seismic level due to sudden shear damage in the cap beam. Two large shear cracks first formed in the cap beam during the 10% level run at 0.07g, and the extent of the damage became severe as testing runs progressed. The test were stopped at 1.05g level when severe concrete spalling occurred and the structure became unstable. Severe structural degradation was evident at this point. 2.4 Damaged'Specimen Investigation The damaged specimen tested on the shake table was a 27% scale as-built model of the Oak Street Bridge bent S28. The main dimensions of the scaled model are shown in Figure 2.8, and the detailed drawings are shown in Appendix A. 22 Chapter 2 — Previous Tests and Damaged Bent Models The model failed because of the sudden shear failure in the cap beam, similar to the failure observed in the slow cyclic tests (Photograph 3 and Photograph 1 respectively). In both cases, the cracks in the joints did not seem to be a critical problem, and this also had been proven by the earlier slow cyclic tests (Anderson et al, 1995). However, the cracks observed on this shake table test model were more concentrated near two main openings, while the cracks on the 45% scale model were more evenly distributed, especially in the joint regions. One assumption made to explain this phenomena is that during the early stage of the shake table tests, the torque induced by an accidental jerk motion of the shake table caused premature shear cracks in the cap beam, and later increasing load levels caused more damage at these cracks but caused little damage anywhere else. These premature cracks also acted as damping devices to dissipate most energy so that at the same time the damage at these two cracks became more severe, little damage was done to the rest of the structure. 23 Chapter 2 — Previous Tests and Damaged Bent Models Two 45° angled shear openings crossed through the cap beam, and near the openings some surface concrete spalled off and the transverse stirrups as well as both top and bottom longitudinal steel bars were exposed. The shear damage near the west column was significantly more severe than the damage near the east column. The bigger shear opening was 200mm wide at both top and bottom, and slightly narrower in the middle (Figure 2.9). center, line i i 200 | 600 | • r—*r— H bent I6ng. steel / ^ " " ^ s L J ^ frbctured stirrup/_ -200 \ 180 Figure 2.9 Detailed Drawing for the West Shear Opening. 27% scaled model Two of the bottom longitudinal steel bars were bent slightly, and necking was observed on one of the top steel bars. Some debonding of the concrete near the cut-off points of the longitudinal steel bar were observed, especially near the west column. A l l three exposed stirrups were broken, and from the geometry of the opening, it was concluded that at least six of the total of seven stirrups in the west half of the cap beam had been broken. Only the status of the stirrup nearest to the west column was inconclusive. The shear opening near the east column was significantly smaller. The width measurement along the crack was around 5rnm~10mm, except at the top and bottom where small portions of surface concrete were spalled off. The condition of the reinforcing steels could not be observed at this location even 24 Chapter 2 — Previous Tests and Damaged Bent Models after chipping off loose concrete. Chipping off more concrete was not suitable in this case since it would cause more difficulties for the later repair operations. Based on the damage situation and the reinforcement details, it was assumed that the longitudinal reinforcing steel bars still retained their ultimate strength (no breakage), and the stirrups had lost most of their strength. Therefore, the shear capacity had to be provided by other means. More photographs of damaged specimen could be found in Appendix C. 25 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme Chapter 3 DEVELOPMENT AND DESIGN OF REPAIRING AND RETROFITTING SCHEME 3.1 Introduction The theory of capacity design is the essential guideline of the retrofitting design process in this study. The retrofitting system was designed in such a way that the repaired and re-strengthened bent would behave in a ductile manner once the damage initiated under seismic loading. More specifically, the system was designed to force plastic hinging action in the bent columns and avoid sudden shear failure in the cap beam and joints. The design process started with the seismic assessment of the bent. Once deficiencies of the cap beam, columns and joints were examined, different remedy methods for these deficiencies were studied. A newly developed fiber glass jacketing system, QuakeWrap™ was decided to be used for the retrofitting operation. A similar system, Tyfo S Fiberwrap System had been used and performed well in the slow cyclic loading tests of the 45% scale Queensborough Bridge bent and Oak Street Bridge bent models. To improve the shear strength in the joint regions, where fiber glass wrappings are difficult to apply practically, as well as to provide additional flexural strength to the cap beam, it was also decided to install an external post tensioning system on the cap beam. 26 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme 3.2 Design Objectives The design criteria were determined by certain objectives regarding the performance of the repaired and retrofitted bridge bent model, subjected to seismic loading during the shake table tests. More specifically, the retrofit system was to be designed in such a way that the behavior and the damage mechanism of the bent under the earthquake loading could be forecasted, a desirable plastic mechanism in certain regions could be developed to dissipate energy effectively, and undesirable brittle failure could be prevented. P. Paulay and M.J .N. Priestley described the concept of capacity design as following: In the capacity design of structures for earthquake resistance, distinct elements of the primary lateral force resisting system are chosen and suitably designed and detailed for energy dissipation under severe imposed deformations. The critical regions of these members, often termed plastic hinges, are detailed for inelastic flexural action, and shear failure is inhibited by a suitable strength differential. A l l other structural elements are then protected against actions that could cause failure, by providing them with strength greater than that corresponding to development of maximum feasible strength in the potential plastic hinge regions. (Seismic Design Of Reinforced Concrete And Masonry Buildings, pg.38) For this particular bent model, three main aspects concerned the design of the retrofitting system: • Since bridges should be designed to dissipate energy in the columns, two columns were identified as potential plastic hinge regions, and they should have dependable flexural strength, so that the estimated ductility demand in these regions could be reliably accommodated. • Any undesirable mode of inelastic deformation, that might be caused by shear, reinforcing steel buckling and others should be prevented in the columns, or plastic hinge regions. This could be achieved by ensuring the member strength corresponding to these failure modes exceeds the capacity of the plastic hinges at overstrength. 27 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme • The cap beam and joints, which are the potentially brittle regions, should be provided with extra shear strength so that the strength in these regions exceeds the demands originating from the overstrength of the plastic hinges. As the result, the cap beam and joint regions should remain elastic and no shear failure and other brittle failures should occur, while two columns deform plastically. The details of the retrofitting system design are stated in later sections. 3.3 Seismic Assessment of The Bent In 1993, a preliminary investigative study done by M O T H and Crippen International Ltd. indicated some serious shortcomings of reinforcing details on the Oak Street Bridge approach bents. Through a series of test projects on the bent models carried out in the University of British Columbia, the details of these deficiencies have been examined and analyzed in more physical and experimental aspects. It was believed that these deficiencies could cause the bridges to fail in a brittle manner at low seismic load levels. These deficiencies include insufficient shear reinforcement, poor anchorage of positive moment reinforcement and short cut-off locations for the longitudinal reinforcement in the cap beam, and poor shear carrying reinforcement details for the development of plastic hinges in the columns. Lack of sufficient confining reinforcement in members was also considered a problem. Among these deficiencies, poor shear reinforcement in the cap beam and questionable shear capacity required to induce plastic hinging in the columns are most critical. 28 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme 3.3.1 Analytical assessments Some work on analytical seismic assessments for the Oak Street Bridge bent prototype, 45% scale and 27% scale models has been done in earlier studies. Markus Seethaler in Oak St. Bridge Bent Test -Slow Cyclic Testing and Elizabeth Davey in Shake Table Testing Of An Oak St. Bridse Bent Model have done some studies on these models' member capacity, flexural, shear behaviors, and section demands, using code requirements, various analytical tools and computer software. The selective summaries of these study results are listed in Table 3.1, Table 3.3 and Table 3.4. Table 3.1 presents the calculated shear capacity of the beam at the edge of each haunch and the center of the beam for 27% scale, 45% scale model and full size prototype, using ACI(1992), ACI/ASCE(1978) and the Canadian Code (CSA, 1985). The results indicated that the Canadian code gives the largest capacity, while the shear values from ACI /ASCE code are most conservative. Table 3.1 Code Shear Capacities of The Cap Beam, in kN Pos. 0.27 ACI 0.27 A/A 0.27 CSA 0.45 ACI 0.45 A/A Proto. V" A/A Hnch + 164.9 142.4 185.5 - - 117 Hnch- 164.9 112.6 185.5 104.5 86.6 87.5 Centre 124.4 101.6 145.9 - - -3.3.2 Experimental assessments Both the slow cyclic testing of the 45% scale as-built model and the shake table testing of the 27% scale as-built model resulted in sudden and potentially catastrophic shear failure in the cap beam at the seismic load level significantly lower than the modem elastic design loads. Poor ductile response and ineffective energy dissipation were observed from these tests. 29 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme In the slow cyclic tests of the 45% scale as-built Oak Street Bridge bent S28 model, the model showed poor ductile behavior. A large diagonal crack was formed after the first cycle at a displacement of 0.9 inch, and the crack enlarged in width with each cycle of load until sudden failure occurred at the concrete compressive zone in the cap beam. It was estimated that the yielding occurred at a joint displacement of 0.25 inch, and at a base shear of 48 kips. The ultimate load was 57 kips at a displacement of 0.47 inch. This ultimate load of 57 kips corresponds to a base shear of 281 kips in the actual bent, which is much lower than the later load demand of 470 kips required by today's code. The displacement ductility was found to be only 1.88 at the point strength degradation set in. The bond failure was evident during the tests due to a lack of confinement reinforcement. (Seethaler, 1995) The 27% scale model in the shake table tests suffered a very similar failure at low seismic load level, even though the exact point of crack initiating and yielding were more difficult to be determined due to the nature of dynamic testing. The first crack appeared after the 20% or 0.14g run, and the inertial force achieved during this run was 68 kN (not considering accidental cracking during the 10% or 0.07g run). A peak lateral force of 93 kN was recorded at a displacement of 11.5 mm at the 80% level. The specimen failed at a displacement of 19.2 mm, and the ultimate displacement of the bent was 36mm (Davey, 1996). The bent failed due to two large diagonal cracks at both ends of the cap beam, which caused the center portion of the beam to drop about 0.5 inches (see Photograph 3, Appendix C ). One phenomenon that should be noticed is that in both slow cyclic tests and shake table tests, the columns and joints were not severely damaged. This was only because the demand on them was limited by the cap beam shear failure at the early stage of the tests, and therefore, they were not tested to their capacity. However, in the retrofitting design, sufficient shear carrying capacity has to be provided in columns to ensure the development of the ductile flexural mechanism. 30 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme 3.4 Remedy of Deficiencies As stated before, the main objectives of the retrofitting design was to provide additional strength in the cap beam and joint regions to force plastic hinging in the columns, while ensuring sufficient shear carrying capacity for the shear demand required to induce hinging in the columns. Since insufficient transverse reinforcement leads to insufficient shear resistance, the principle for the shear strength improvement is to provide extra transverse reinforcement. This could be achieved by several techniques, such as additional transverse steel reinforcing, steel jacketing, and longitudinal and vertical post-tensioning, etc. However, a technique of fiber glass jacketing was selected to retrofitted the damaged bent model for the following considerations: • It is a newly-developed technique, and it is beneficial to study and examine the effectiveness of this technique in a seismic loading condition. • The specimen retrofitted with a similar technique performed well in the 45% scale model testing, compared with other methods. • The installation of this system is relatively inexpensive compared with others, since it requires less time and manpower, and no complicated installation procedures are involved. • Unlike other isotropic materials, such as steel and concrete, the fiber glass wrapping sheets are uniaxial or unidirectional. It has a high tensile strength in the strong direction, at which glass fibers are oriented, but the strength is virtually zero in another direction. As the result, once they are wrapped around member transversely, these glass fibers act as many small stirrups or hoops in the transverse direction of members and provide extra shear strength, without increasing the strength in the longitudinal direction, or the flexural strength. This characteristic ensure the plastic hinges would developed at the design load level before the occurrence of any shear damage in the same regions (i.e. yielding before shearing), since the shear demand in members would be unchanged due to the unchanged flexural capacity. 31 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme 3.5 Design of Retrofitting System The re-strengthening design was based on the design objectives and seismic assessments stated in Section 3.2 and Section 3.3, and also considering the fact that the specimen was a repaired one and weaker than the original bent in some aspects. The fiber glass jacketing with appropriate size at certain locations was designed to provide extra strength in the cap beam and columns, and external post-tensioning system was designed to provide additional flexural capacity to the cap beam, and also extra shear strength in the joint regions. 3.5.1 F R P composite wrapping material The QuakeWrap™ system was chosen to be used in the retrofitting operation. The materials, including the fiber glass wrapping sheets and epoxy, were provided by the SRC ( Structural Rehabilitation Corporation), an Arizona based company. A unidirectional fabric of E-glass was used in the construction of the composite wraps. The fabrication of the composite wraps was described in Repair Of Earthquake-Damaged RJC Columns With Prefabricated FRP Wraps (Saadatmanesh et al, 1995). A long strip of unidirectional E-glass fabric was laid flat and then saturated with polyester resin matrix. A layer of mylar sheet was placed on the wet strip before the strip was rolled and placed in an oven to cure at 160°F for forty minutes. Photograph 10 in Appendix C shows the FRP wraps ready to be used. This material is considered to be unidirectional since the majority of the fabric fibers in the wraps were unidirectionally arranged and only a small amount of fibers were used in the transverse direction to hold the fibers together during the manufacturing. 32 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme The fiber volume ratios vary depending on the type of FRP materials, and the tensile strength increases as the fiber ratio increases. In this study, composite wraps with Vf = 50.2% were used, where Vf defines the ratio of the volume of fibers over the total volume of the wrap. The mechanical properties of this material were obtained from tensile tests, and are listed in Table 3.4. The fiber glass wrap itself is a brittle material with high tensile strength, and it has a linear stress-strain relation from initial loading to ultimate failure. Table 3. 2 Mechanical Properties of FRP Tensile Strength, MPa 532 Tensile Modulus of Elasticity, MPa 17,755 Ultimate Tensile Strain 3% 3.5.2 Wrapping design Again, the objective of the retrofitting design was to ensure that the structure would degrade in a flexural mode instead of a sudden shearing mode, and the energy would be dissipated effectively by the designed plastic hinging regions while the rest of the structure remained elastic. In this particular design, the columns were designed to be plastic hinges, with the intention that they should yield before any sheared damage start to occur. To achieve that, shear strength of all structural members, especially shear strength in the cap beam, had to exceed the shear demand required by the forming of this plastic mechanism in the columns, so that the cap beam could remain elastic during and after the forming of plastic hinges. 33 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme (3.5.2.1) Section moment capacities The first step was to determine members' flexural capacity. The standard elastic analysis (equivalent block method) was used to calculate the moment capacities of the cap beam at the different sections, and the results agreed well with the calculation done by Davey (1996) using BEAM303, in-house C A L T R A N S program. The summary results are listed in Table. 3.5. It is possible that the negative yield moment of the repaired bent in section 1 and 2 of the cap beam had increased slightly (Figure 3.1), since the yielding of one top steel over the damaged region had been noticed during the investigation of the damaged specimen. However, this would not effect the retrofitting design significantly. The columns' flexural capacities predicted by Davey (1996) using COL604 (in-house C A L T R A N S Program, Seyed, 1992), and by Seethaler (1995) using the program RESPONSE (Collins and Mitchell, 1991) are listed inTable.3.6 Figure 3.1 Analysis Points on Bent 34 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme Table 3. 3 Moment Capacity of the Cap Beam Dist. from 0.27 0.45 Prototype 0.27 0.45 Prototype col. edge Model M1" Model RT M + [kN-m] Model M" Model M" M [kN-[mm] [kN-m] [kN-m] [kN-m] [kN-m] m] 0 -259 Sec.1 29.2 47.4 45.6 119.2 113.8 112 259-390 40.7 69.1 - 100.3 100.3 -390-543 Sec.2 62.4 80.0 - 81.3 81.3 -543 - 619 81.3 100.3 - 62.4 69.1 -619 - 745 100.3 119.2 115.6 40.65 46.07 66.8 Sec.3 745 - 975 100.3 119.2 - 40.65 - -975 - CL 100.3 119.2 - 29.8 25.7 Table 3. 4 Flexural Capacity of Column Model Scale Moment Capacity (kN-m) Axial Load (kN) 27% Scale 57 0 73 132 45% Scale 83 164 Prototype 88 179 (3.5.2.2) Demands in the cap beam In a joint segment, the moment in the column is balanced by the moment and/or the shear force in the cap beam (Figure. 3.2). Three possibilities exist in this situation: beam yielding, beam shearing or 35 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme column yielding. The cap beam will not yield before the column since it has greater negative moment capacity than column. Therefore, to ensure the column yields first, the shear strength in the cap beam has to be greater than Vbeam, and the shear demand originating from yield moment of the column, ^ c^olumn. Vbeam = M c oiumn / d (Equation. 3.1) where d is the distance from the center of the column to the point of interest. M beom I V | 'column Figure 3.2 Free Body Diagram of Joint The portions of the cap beam between points 830mm and 55mm away from the edge of the column at both west and east ends were considered critical regions. Within these regions, the cap beam had suffered severe shear damage in the earlier tests and several stirrups were suspected of being broken (Figure. 3.3). The shear demand in these regions was calculated, and it ranged from 73kN at the furthest point from the column edge to 334kN nearest the column. center, line 830 mm Figure 3.3 Critical Region for Shear Demand 36 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme (3.5.2.3) FRP wrap design for the cap beam Since it was suspected that most of the stirrups in the cap beam had been damaged during the earlier tests, the shear strength provided by the transverse steel was ignored in the design. The Canadian code (CSA, 1985) and ACI (1992) all have the same formula for the total shear capacity in a reinforced concrete beam: V= V s + V c (Equation. 3.2) where V s is the shear strength provided by transverse steel V c is the shear strength provided by concrete. Since V s was ignored, at the stage after damages were repaired and before any retrofitting system installed, the shear capacity of the cap beam was provided only by V c . The C S A formula and ACI formula for V c are: where f c is concrete strength in MPa b w is section width in mm d is section depth in mm. V c calculated using CSA was 136kN, and 113kN using ACI taking f c to be 36 MPa. Therefore, the additional shear strength that was to be provided by the fiber glass wrapping is the difference between the demand and capacity, V c = 0.2 ( f c ) m b w d CSA code (Equation. 3.3) V c = 0 . 1 6 6 ( f c ) , / 2 b w d ACI code (Equation. 3.4) (Equation. 3.5) 37 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme The maximum value of Vwrap for the different sections within the critical regions of the cap beam was calculated to be 200kN using CSA (1985), or 220kN using A C I (1992). The tensile strength of the fiber glass wrapping material is 532MPa (77ksi), and the wrap sheet was 0.95mm (0.0376in) thick. Therefore the tensile strength of the material could be expressed as T^p =0.5kN/mm (2900 lb/inch). The total width of the FRP jacket, W w a p then was calculated as following: = V w a p / (2T w a p ) (Equation. 3.6) where 2 is the number of shear faces Only one wrap was needed for this retrofitting system, and in fact the strength of one wrap of FRP over the critical regions would exceed the total shear demand significantly, but as a minimum, one wrap with one non-shear face overlap was designed. (3.5.2.4) Column wrapping design The columns did not suffer any shear damage in both slow cyclic tests and earlier shake table tests, and their shear capacity was not considered problematic. However, unlike the satiation in the previous tests where the further damage in the columns was limited by the early stage shear damage in the cap beam, the retrofitted bent was expected to have greater shear strength in the cap beam, and therefore, some shear reinforcement should be provided to ensure the shear capacity in the plastic regions exceeded the capacity of the hinges at overstrength. Similar design procedures for cap beam were used to design the wrapping system for the columns and only one wrap for each column at potential plastic hinge region was needed to carry all the shear. 38 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme (3.5.2.5) Summary The following is the summary for the FRP wrapping system design procedures: • Define the top of both columns as potential plastic hinges and determine the yield moment in these sections. • Determine the shear demand in the critical sections of the cap beam originating from these potential plastic hinges. • Determine the additional shear strength that must be provided by the retrofitting system, and calculate the amount of wrapping required. • Design column wraps and wraps for the potential plastic regions to carry all the shear to ensure the development of the plastic mechanism, using the same procedures as for the cap beam. It should be noted that this wrapping system would not provide a passive confining pressure on these rectangular beam and column sections, since no pressurized epoxy was fdled in the gap between the wraps and concrete members. However, better confinement in the section corners and better bonding between longitudinal steel and concrete were expected. It was also anticipated that concrete cover spalling could be minimized by the system. 3.5.3 Longitudinal post-tensioning Since it was difficult to wrap joints and the small portions of the cap beam due to the curved bottom face near both columns, a longitudinal post-tensioning system was designed to be installed externally on the cap beam for the purpose of providing extra shear strength in these regions. It was decided that the same compressive stress value as in the model OSB5 of the slow cyclic tests would be used on this 27% scale retrofitted model. A compressive stress of 340psi (2.34MPa) after 39 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme setting was to be applied to the cap beam longitudinally by one 5/8 diameter bar on each side of the cap beam. A minimum number of bars was used to ensure the accuracy of the stress applied. Located 150mm down from the top of the cap beam, at the plane level through the center lines of the cap beam end faces, each bar was to be stressed to 140kN to give the cap beam a total compressive force of 280kN. Two 470mmxl00mm, 2-inch steel plates were designed to be anchors at the two ends of the beam (Figure 3.4). Bending of the steel plates and crash of concrete were also checked and found to be satisfactory (Figure 3.5). 150 mm 411 mm 5/8" strand 327 mm 290 mm Figure 3.4 Post-tensioning System. 470 mm R=10.6 mm 40.6 mm Figure 3.5 End Anchor Plate •e-51 mm 100 mm 40 Chapter 3 — Development and Design of Repairing and Retrofitting Scheme 3.5.4 Design summary The designed retrofitting system consist mainly of FRP wraps over the critical regions of the cap beam and wraps over the potential plastic hinge regions on the columns, and in addition, a longitudinal post-tensioning system was to be installed externally on the cap beam to provide extra strength. The design summaries are listed in Table 3.7 and Table 3.8. Table 3. 5 FRP Wrapping System Section Location No. of Wrp Purpose Beam (E&W section) 55mm ~ 830mm from east and west column edges 1 provide additional shear strength to critical regions of the cap beam Column (E&W column) 390mm ~ 1000mm from column base 1 provide additional shear strength to columns Table 3. 6 Post-tensioning System Location Tendon Anchor Purpose on cap beam, one bar at each side, 150mm below beam top face Dywidag™ bar, 05/8", Tensile strength = 150ksi 2" steel plate, 470mm x 100mm provide additional flexural capacity to cap beam, and add extra shear strength to joints 41 \ Chapter 4 — Repairing and Re-strengthening Techniques Chapter 4 REPAIRING AND RE-STRENGTHENING TECHNIQUES 4.1 Introduction The repair work of the damaged specimen was done by a technician from Polycrete Restorations Ltd. Two techniques were involved in the repairing process: a patching technique to repair the larger shear opening at the west end of the cap beam, and an epoxy injecting technique to repair the small shear opening at the east end of the cap beam and other fine cracks in the beam and columns. Once the openings and cracks were repaired, the designed FRP jackets were also installed by Polycrete Restorations Ltd. The FRP sheets were saturated with epoxy before they were wrapped over the bent members, and then another coat of epoxy was put on the FRP jackets to finish the wrapping operation. Finally, the longitudinal post-tensioning system was installed after the FRP wraps were set. The post-tensioning equipment and labor were provided by Dywidag Systems International Canada Ltd. 42 Chapter 4 — Repairing and Re-strengthening Techniques 4.2 Repairing of the Damaged Bent After the earlier shake table tests, the original bent model was severely damaged. Two large diagonal shear openings near the columns at the both end of the cap beam caused the center portion of the beam to drop about 'A inch, and many smaller cracks were distributed at the joints and the upper columns (see photograph 1, Appendix C). Two repairing techniques were employed to restore the bent. A concrete patching technique was used for repairing the larger diagonal shear opening on the cap beam near the west column, and another smaller shear opening at the east end of the beam was repaired by epoxy injection technique, so were other fine cracks. The repair operation was done to bring the specimen back to its approximate original state. Even though the high strength of the repairing materials and the possible strong bonding between new and existing materials might contribute to the overall strength of the structure to some degree, the repaired bent was considered to be significantly weaker than the original one because of damage to the internal reinforcing steel. The task of strength improvement was to be achieved by the later retrofitting operations. 4.2.1 Concrete patching The entire concrete patching process included the following procedures: • surface preparing • forming • pouring and vibrating • material setting • form stripping • surface grinding 43 Chapter 4 — Repairing and Re-strengthening Techniques To ensure the strong bonding between the repair material and the existing materials, the loose concrete pieces near the damaged area were first chopped off, and the surface of the damaged area was cleaned by wiping and vacuuming. The center portion of the cap beam had dropped about Vi inch after the shake table tests, and it was not jacked up when repairing in order to avoid possible further damage to the specimen. Patching form work was built around the shear opening with plywood sheets and smooth plastic sheets. A l l connections of the form were sealed with hydraulic concrete (concrete plug), which set in about 5 minutes. Acrylic concrete was then poured into the shear opening through the notch at the top of the form (Figure 4.1). At the same time and after concrete pouring, the form was hammered repeatedly so that good vibration could be provided. The pouring and vibrating was repeated until the opening was filled with fresh concrete. The acrylic concrete set in 45 minutes, and it would have a compressive strength of 9200psi and tensile strength of 1200psi. The form was stripped once the concrete was set, and the repaired surface was ground. ACRYLIC CONCRETE CONCRETE ANCHOR Figure 4.1 Patching of Shear Opening 44 Chapter 4 — Repairing and Re-strengthening Techniques 4.2.2 Epoxy injection The technique of epoxy injection was used to repair the smaller diagonal shear opening near the east end of the cap beam as well as other fine cracks. Two different techniques were used to repair the damaged bent since it was difficult to patch small cracks, and at the same time, it was not cost effective to inject epoxy into large openings. The steps to the epoxy injection technique were: • preparing the surface • installing injecting ports and sealing cracks • Patching to restore the original shape • injecting epoxy • removing injecting ports • grinding the surface The concrete surfaces near the diagonal opening and near the joint and upper column regions were ground and vacuumed so that fine cracks could be seen better. This epoxy injection technique can be employed to repair extremely fine cracks as long as these cracks can be visually observed. Once the surface preparation was completed, along each crack, 2 - 4 plastic injecting ports were placed at appropriate intervals using epoxy as glue. The number of ports depended on the length of the individual crack. The rest of the crack uncovered by ports was sealed with epoxy at the surface. Acrylic concrete then was used to patch the relatively large spaces left by concrete cover spalling during the earlier test. 45 Chapter 4 — Repairing and Re-strengthening Techniques A l l injecting ports along each individual crack and along other cracks that might be connected to this crack were plugged with wooden sticks, except one port. The pressurized epoxy was injected into the crack by the injecting-head through this unplugged port (Figure 4.2). The pressure applied varied, and the technician adjusted the pressure depending on the size of the crack and his experience from previous repair works. Normally the starting pressure was relatively low, and once large openings were filled, the pressure was progressively increased to achieve the best penetration. The direction of epoxy flow could be monitored by observing the sequence in which surplus epoxy appeared at different ports. The pump continued injecting epoxy into the crack until a trace of epoxy leakage was observed at all other ports along this major crack and from some extremely fine and almost invisible cracks. The injecting port was then plugged with a wooden stick after the injecting head was removed. By examining the visible cracking and the evidence of debonding, combined with experience and understanding of the nature of the damage, the technician selected an injection sequence to achieve the best possible penetration throughout the voids and cracks. The injected epoxy took about 12 hours to solidify, and then these ports were removed using a chisel. The final step was to smooth the concrete by grinding off the surface seal epoxy. The finished appearance of the repaired bent would have been acceptable in a typical bridge structure, and is shown in Photograph 9, Appendix C. One advantage of this epoxy injecting technique worth mentioning is that the injected epoxy not only can fill cracks, but also may seal some existing fine voids between the concrete and reinforcing steel, i f these voids are connected to the cracks that can be observed from the concrete surface. As the result, the bonding between the concrete and steel may actually be improved. 46 Chapter 4 — Repairing and Re-strengthening Techniques PORTS Figure 4.2 Epoxy Injection 4.3 Retrofitting of the Repaired Bent with FRP Wrapping The repaired bent model was wrapped symmetrically over the critical regions of the cap beam and columns with FRP jackets according to the design. The basic steps involved in the FRP system installation were: • preparing the wrapping surface • cutting wrapping sheets into strips with appropriate width and length • satiating FPR strips with epoxy resin • wrapping strips over designated member sections • coating the jackets with epoxy 47 Chapter 4 — Repairing and Re-strengthening Techniques The west column was wrapped first. A FRP sheet of 1635mm x 610mm (1635mm was the perimeter of the column cross section plus the length of the overlap face; 610mm was the designed height of the jacket) was cut and saturated with epoxy. Some difficulty was experienced during the wrapping of the sheet. Due to the large size of the sheet, a significant amount of air bubbles were trapped under the sheet despite the repeated rolling on the wrapping surface. It was then decided to cut sheets into narrower strips for the wrapping on the cap beam and the east column, and this was justified by the following reasons: • the strength of the jacket would affected only by the total width of the wrap, but not by the number of strips used and the width of the individual strips; • it was physically less difficult to wrap the cap beam with narrower strips. Since 24 strain gauge wires came out of the beam on the south face, using narrower strips made it possible to leave gaps between wraps where these wires existed; • less air bubbles would be trapped under the wraps as they could be easily rolled out. As designed, only one wrap was installed at each designated location. The wraps were overlapped on the non-shear faces at these regions, and namely the bottom face of the cap beam and the inner faces of the columns. Since these wraps had been saturated with epoxy, once they were wrapped around, they could hold themselves in place, and no extra external pressure was required. An extra epoxy coating was put over wrapped jackets once they were installed and before the epoxy resin absorbed in the sheet was cured. This was to ensure good interaction between the fibers and resin in the finished composite, and provide the required mechanism for the transfer of load among the fibers. This coating could also protect the fibers from abrasion and other environmental and chemical attacks. 48 Chapter 4 — Repairing and Re-strengthening Techniques The finished FRP jacketing system is shown in Photograph 14, Appendix C. 4.4 Post-Tensioning System Installation The final stage of the re-strengthening operation was to install a longitudinal post-tensioning system externally on the cap beam. First, two anchor plates were held in place by the cranes in the Earthquake Lab. Then two 5/8" strands were gradually stressed to the designed load in an alternate fashion. A 25-ton C1414 jack was used to apply tension to strands to 4100psi, or 140kN in each strand, which was 88% of the strands' yield load of 160kN. The Pressure vs. Force diagram is attached in the Appendix B. The final appearance of the repaired and re-strengthened 27% scale model of the Oak Street Bridge bent S28 is shown in Photograph 16, Appendix C. 49 Chapter 5 — Retrofitted Model and Its Predicted Behavior Chapter 5 RETROFITTED MODEL AND ITS PREDICTED BEHAVIOR 5.1 Physical Description Figure 5.1 shows the re-strengthened 27% scale model of the Oak Street Bridge bent S28. The retrofitting system consisted of the FRP wraps on the critical regions of the cap beam and columns, and external post-tensioning of the cap beam to a stress of 340psi (2.34MPa) after losses. Figure 5.1 Re-strengthened Model 50 Chapter 5 —Retrofitted Model and Its Predicted Behavior The FRP jackets on the cap beam were designed and installed to provide additional shear strength, which would recover the shear capacity loss in the earlier tests due to the reinforcing steel damage, and insufficient shear strength of the cap beam in the original design. Sufficient shear strength in the cap beam is the prerequisite for the development of proper ductile flexural mechanism in the columns (plastic hinges) and the prevention of sudden shear damage of the cap beam in the potentially brittle regions. The FRP wraps on the columns were designed to be the additional shear reinforcement to ensure the shear capacity in the potential plastic hinge regions exceeded the capacity of the hinges at over strength, so that the columns would not suffer sudden shear failure before yielding flexurally. The wraps on both the cap beam and columns were also expected to provide better confinement at member corners and to prevent cover concrete spalling and therefore buckling of the tied main reinforcing steel in the fdlet region. Table. 5.1 Retrofitting System Descriptions Retrofit Location Purpose FRP jackets on the cap beam between 344mm and 730mm (from the column center) improve shear capacity of the beam, and confinement at corners. FRP jackets on the columns 570mm ~ 1180mm from the top of the bent improve shear capacity in the plastic hinge regions, corner confinement Longitudinal post-tensioning cap beam, 150mm down from the beam top Improve shear capacity of the joints and beam 51 Chapter 5 — Retrofitted Model and Its Predicted Behavior The external post-tensioning system on the cap beam was designed to provide extra shear strength to the cap beam and joint regions. 5.2 Predicted Model Behavior The detailed analytical predictions of the model's behavior regarding the members' flexural, shear responses, global frequency, and stiffness response were not carried out in this study due to the limitation of the scope of this research. Generally, under the design seismic load condition during the shake table tests, the structure was expected to undergo a more ductile damage process resulting from the retrofitting measures taken. Better energy dissipation was anticipated for this retrofitted model. More specifically, it was predicted that flexural plastic hinging would be observed at the tops of both columns, and no severe sudden shear damage should happen anywhere in the structure, specially in the cap beam and joints. In spite of the retrofitting FRP jackets and post-tensioning system installation, the yield capacity (flexural strength) of the structure, which would be mainly determined by the flexural strength of the columns, was expected to be approximately the same as the original model. More precisely, it was expected to have approximately the same flexural strength as the OSB5 in the slow cyclic tests, which was a 45% scale model having a similar retrofitting system. This was predicted considering the fact that the columns of the original 27% scale model were not severely damaged in the earlier tests due to the premature shear damage in the cap beam, and the fact that the retrofitting FRP jackets on the columns would not contribute to their flexural strength significantly. 52 Chapter 5 — Retrofitted Model and Its Predicted Behavior Because of the irregularity of the loading patterns, it is relatively difficult to find out the exact point of yielding of the original 27% scale bent by looking at the testing records from the shake table tests. However, it is easier to calculate the yield displacement and force for the re-strengthened 27% model by scaling the values obtained from the slow cyclic tests of the OSB5. Figure 5.2 shows the lateral load and displacement response curve for OSB5 in the 45% scale slow cyclic tests (English, 1996). The specimen started yielding at a base shear of approximately 82kips ( 365 kN), and the yield displacement was about 0.43 inch (10.9mm). Scaling these values to 27% scale, the yield displacement of the re-strengthened model would be 0.256 inch (6.5mm) at a base shear of 29.0kips (129 kN). Note that this displacement is not necessarily the displacement at first yield of the reinforcement, but the yield displacement of a theoretical bilinear fit to the force deflection plot (Figure. 5.3). 5.3 Scaling Factor From prototype to 27% scale, the scaling factors for some major physical parameters are listed in Table 2. 1. More analytical background and details can be found in Shake Table Testing Of An Oak Street Bridge Bent Model (Davey. 1996). 53 Chapter 5 — Retrofitted Model and Its Predicted Behavior H i a 1 2 * « 8 10 - 5 - 4 - 3 - 2 - f 0 ( 2 3 4 5 Joint Displacement (In) Figure 5.2 Lateral Load Displacement Response for OSB5 Chapter 6 — Test Arrangement Chapter 6 TEST ARRANGEMENT 6.1 Introduction One of the main objectives of this study is to examine the effectiveness of the repairing and retrofitting scheme by comparing the seismic performance of the re-strengthened and original models under seismic loading conditions. This objective could only be achieved i f there existed a consistent basis for comparison. Therefore, the re-strengthened model test arrangement was the same as that used for the previous shake table tests for the original model. The details of the test arrangement were described in Shake Table Testing Of An Oak Street Bridge Bent Model (Davey, 1996). 6.2 Test Setup 6.2.1 Loading system A concrete mass block and three stacks of steel plate were placed on the specimen to simulate the dead load from bridge girders and deck. The weight of the mass on the bent model was 89 kN, and the weight of the concrete block and steel plated was proportioned in such a way that the center of gravity of the structure would be located at a height of 411mm above the top of the cap beam, consistent with 1524mm above the cap beam in the real structure. The weight of the concrete mass block and the total weight of the steel plates were 12.3kN and 76.7kN respectively. The general set-up of the vertical loading system (dead load) is shown in Figure 2.4. 55 Chapter 6 — Test Arrangement The concrete mass block weighed 12.3 kN, and its dimensions were 500 mm x 235 mm x 4360 mm (width x depth x length); the total weight of three stacks of steel plate was 76.6 kN, and each weighted 25.6 kN. A l l plates had the same width and length of 600mm x 1500mm, but three different thicknesses of 63.5mm, 31.7mm and 9.0mm. The dead load was transferred into the cap beam at five bearing locations, where rubber bearings were used to simulate girder bearings in the real structure. Each set consisted of one hard 50-durometer rubber that gave the correct spacing between the cap beam and the concrete block, and one softer Shore A-40 polyurethane rubber that allowed some relative displacement with little change of applied force. However, the lateral force was transferred at two locations near two columns, where the cap beam and the concrete mass block were connected with Dywidag™ post-tensioning bars. To restrain the out-of-plane motion of the mass block and achieve a "pin" connection, the spaces around these bars were grouted after they had been tensioned. Each stack of steel plates was then placed at the top of the concrete and held in place securely by 4 ready-rods embedded in the concrete block. (Figure 2. 5) The shear and axial loading capacities of the mass block connections had been checked before the previous tests and were found to be satisfactory. 6.2.2 Supports To restrain the out-of-plane motion and at the same time allow in-plane movement, the lateral support of the structure was achieved by tying down the structure at the top of steel mass blocks using four wire ropes, as shown in Figure 2. 6. The ropes were tensioned to 4.4kN, which yielded a frequency of 12Hz, which was considered sufficiently far from the structure's fundamental frequency of 7.5Hz. The tension value in one rope was monitored by an attached strain gauge. 56 Chapter 6 — Test Arrangement The model bent was connected to the shake table by pin connections, which simulated the inflection points at the mid-height of both columns. After the previous shake table tests, the condition of the pins were checked and found to be satisfactory, and they were re-tightened before the re-strengthened model tests. The details of the connections are shown in Figure 6.1. •327-CQNCRETE COLUMN | >1* 1 "' 'J* r!9 12.75J — i — R=5l 1 165 C-12.75 nip \ 1 / 1 i ' i T'T ~> ' T i i r •381-- 4 -•327-CDNCRETE COLUMN 4- 4-• 1 | J L J Note: all dimensions are in mm Figure 6. 1 Pin Connection 6.3 Input Motion and Test Program 6.3.1 Input motion As in the previous shake table tests of original bent model, the Joshua Tree Fire Station record E-W direction from the 1992 Landers Earthquake was chosen to be the input earthquake motion for this test program. Three main criteria leaded to the initial selection of a specific earthquake record: • it could cause a significant amount of damage to the structure • it would satisfy the requirements of the design spectra 57 Chapter 6 — Test Arrangement • it could fit the limitations of the shake table By examining these requirements, the earthquake record mentioned above was considered to be the most appropriate for the testing program. The response spectra, with 5% damping of the Joshua Tree and Infiernillo records, along with the design response spectrum, are shown in Figure 6.2. The original acceleration time history of the Joshua Tree E-W record before scaling is shown in Figure 6.3, and its characteristics are listed in Table 6.1. 1.2 0 -I 1 1 1 1 1 1 0 0.5 1 1.5 2 2.5 3 Period [s] Figure 6. 2 Response Spectra of Different E Q Records vs. Design Spectrum, % = 5% 300 < -300 J 1 0 5 10 15 2 0 2 5 3 0 3 5 4 0 4 5 5 0 Time (s) Figure 6. 3 Time History of Joshua Tree E-W Record 58 Chapter 6 — Test Arrangement Table 6.1 Characteristics of the Joshua Tree W-E record Characteristic Symbol Unit Value Magnitude M s None 7.50 Peak Ground Acceleration PGA cm/s/s 278.00 Peak Ground Velocity PGV cm/s 42.71 Peak Ground Displacement PGD cm 15.73 Max. Incremental Velocity IV cm/s 62.52 Max. Incremental Displacement ID cm 18.51 Effective Peak Acceleration EPA cm/s/s 198.00 Effective Peak Velocity EPV cm/s 35.78 Epicentral Distance D Km 15.00 Bracketed Duration (a> 0.05g) [D] s 41.10 Once the earthquake record was chosen, some modifications to the record were required to fit the test conditions. This modification process involved two basic steps. One was to scale the actual time history record to the corresponding time history for the 27% scale using scaling factors listed in Table 2.1; another necessary modification was to condition the record by filtering out frequencies outside the frequency sensitive range of the table (1 to 30 Hz). The comparison of the actual and scaled acceleration time histories are shown in Figure 6. 4. Figure 6.5 and 6. 6 show the original and modified response spectra at 5% damping for both displacement and acceleration, in which the amplitude of the modified spectra are not scaled for acceleration or displacement. 59 Chapter 6 — Test Arrangement 300 Time (s) 40 50 300 20 30 Time (s) 40 50 Figure 6. 4 Original and Scaled Acceleration Time History 0 0.5 1 1.5 2 2.5 3 Period (s) Figure 6. 5 Displacement Response Spectra, £ = 5% 60 Chapter 6 — Test Arrangement Figure 6. 6 Acceleration Response Spectra, ^ = 5% The final input motion used to drive the actuators of the table was in the form of displacement time history, which was obtained by integrating the acceleration record twice. 6.3.2 Test program In the previous shake table tests for the original model, it had been predicted that the Joshua Tree record scaled to a PGA of 0.7g would fail the bent, and the scaled record in both time and acceleration was taken as the reference earthquake with PGA=0.7g (100%). The magnitude of the shake motion, as the percentage of this 0.7g, increased as the tests progressed: it started with 5% of 0.7g and ended at the 150% level. Even though the re-strengthened model was anticipated to be significantly stronger than the original one, the same system was followed for the sake of later data comparison. However, more runs were planned because the re-strengthened model was expected to sustain more severe shaking. Table 6.2 shows the plan for test runs and their magnitudes. The shake table would reach its displacement 61 Chapter 6 — Test Arrangement capacity at 200% of 0.7g. (Note that during the actual testing, several additional testing runs at high shaking magnitudes were conducted after all planned runs, in order to generate further damage to the specimen.) Table 6.2 Planned Testing Runs Run Acceleration, g %of0.7g 1 0.035 5 2 0.07 10 3 0.14 20 4 0.28 40 5 0.42 60 6 0.56 80 7 0.84 120 8 1.05 150 9 1.4 200 The shake table would reach its displacement capacity at the 200% level, and it was decided that i f the schedule and structure condition permitted, the post-tensioning system was to be removed at that time and further runs would be carried out on the bent without the post-tensioning. 62 Chapter 6 — Test Arrangement 6.4 Instrumentation and Data Acquisition System The general instrumentation and data acquisition system setup for this test program was consistent with the previous shake tests with a few exceptions. The detailed descriptions of the instrumentation setup can be found in Shake Table Testing Of An Oak Street Bridge Bent Model (Davey, 1996). 6.4.1 Strain gauges The internal strain and deformations were recorded by strain gauges. A total of 24 strain gauges were attached to the reinforcing steel inside the bent at different locations, and these strain gauges were connected to filtered channels on one data collection bank. Figure 6. 7 shows the layout of these strain gauges. A few of these were suspected to be out of their elastic range after the previous tests, since a significant amount of residual strain was observed from the records of a few strain gauges. The strain gauge attached to the stirrups in the cap beam were all damaged due to the possible breakage of these stirrups; however, this damage had no effect on this test since the retrofitting system was designed assuming no transverse steel in the cap beam. In general, most of these strain gauges were still functional, and were expected to produce reasonable data. T8 T7 T6 T5 T4 T3 T2 T l \l v V V K * S4 r M S3 S2 \i y • A A A si y v C4 £ g rr Jt X A A C4 E 4-/^B8 B7 C3 B6 B5 B4 B3 B2 B~i^  C2 > w Figure 6. 7 Strain Gauge Layout 63 Chapter 6 — Test Arrangement 6.4.2 Displacement transducers and accelerometers Three displacement transducers and six accelerometers were attached to the structure and the table to record the external data, namely displacement and acceleration at different locations (Figure 6. 8). These instruments were connected to another data collection bank. EXTERNAL INSTRUMENTATION ® w (P)—•displacement transducer (S) strain gauge (A)—•accelerometer Figure 6. 8 External Instrumentation 6.4.3 Hammer test data By performing the hammer test after each test run, the frequency and stiffness of the structure at different testing stages could be quickly determined on site. The layout of the sensors is shown in Figure 6. 9. The layout was different from the previous tests. Instead of placing all sensors in a longitudinal direction, only sensors 2 and 3 were in a longitudinal direction. Sensors 4 and 5 were in a transverse direction. This change ensured that any torsion vibration mode could be observed easily when performing the hammer test. The details of the hammer test setup and the complete data from all tests are attached in Appendix B andD. 64 Chapter 6 — Test Arrangement © i 2&3 4&5 - h a m m e r - Longitudinal Sensor - Transverse Sensor ® T Figure 6. 9 Hammer Test Setup 65 Chapter 7 — Testing and Overall Experimental Results Chapter 7 TESTING AND OVERALL EXPERIMENTAL RESULTS 7.1 Introduction The overall performance of the re-strengthened bent was considered to be satisfactory, and a ductile response of the structure was observed. A total number of 17 runs were conducted, with some repeated runs for configuration reasons. The first hair line cracks were observed at the top of the both columns at the end of the 20% run. The number of cracks increased and each crack propagated as the testing progressed. With the shake motion magnitude increasing, the natural frequency or the stiffness of the bent degraded gradually, but not as dramatically as for the original bent in the previous testing. It was suspected that the plastic hinges had been formed by the end of the 120% run according to visual observation. After the shake table reached its maximum displacement capacity at the 200% run level, the structure still retained its integrity. Besides cracks at the plastic hinge regions, no shear cracks were visually found in the cap beam and no major concrete spalling was observed. It was then decided at that point to remove the external post-tensioning system and try to generate further damage to the specimen, especially to the cap beam. One 120% run, one 150% run and finally one 200% were repeated. More severe concrete spalling was observed at the plastic hinge regions during these runs; however, no significant further damage was done to the cap beam. At the end of testing, the structural integrity of the specimen was well maintained as no significant permanent deformation was observed, and the bent was considered still to be serviceable under extreme (emergency) circumstances. 66 Chapter 7 — Testing and Overall Experimental Results 7.2 Runs A total number of 17 runs were conducted, and the accelerations imposed on these runs ranged from 0.035g (5%) to 1.40g (200%). The total length of each shake was 37.5 seconds, with 10 seconds of strong motion. Several runs were repeated at some stages of the testing due to various reasons. Table 7.1 shows the details of all runs. Table 7.1 Testing Runs Run Level P G A , g Remark 1 5% 0.035 No visible damage 2 5% 0.035 No visible damage 3 10% 0.07 No visible damage 4 20% 0.14 Hairline cracks at top of columns 5 20% 0.14 No significant further damage 6 40% 0.28 More hairline cracks at potential plastic hinge regions 7 60% 0.42 More hairline cracks at potential plastic hinge regions 8 80% 0.56 Incorrect motion signals, re-calibrating 9 10% 0.07 Calibrating table 10 20% 0.14 Calibrating table 11 80% 0.56 More cracks at potential plastic hinge regions 12 120 0.84 Significant width increase of main cracks, dust smoke 13 150 1.05 Closing and opening of main cracks observed during run 14 200 1.4 A crack observed at the top of each column jacket * Post-tensioning removal 15 120 0.84 No significant further damage 16 150 1.05 Spalling of small concrete pieces at plastic hinge regions 17 200 1.4 No significant further damage, structure's integrity well maintained 67 Chapter 7 — Testing and Overall Experimental Results 7.3 Test Descriptions This section describes the actual testing process and visual observations. The detailed data analysis follows in Chapter 8. To start the testing, two runs at 5% level (0.035g) and one run at 10% level (0.07g) were conducted to ensure that the shake table, all necessary instruments, and computer software were functioning properly. However, some unwanted noise was observed in the output signals, and it was suspected that the bolts connecting the upper hinges and the columns were loose. Once all bolts were tightened and the table was recalibrated, the testing started again at 10% level. At this stage, no any visible damage and cracks were found on the specimen. A 20% run at 0.14g was then carried out. A very fine hair line crack was found at the top of each column after the run, and it was marked in green. However, the stiffness of the structure was unchanged as determined by the hammer test results (see Section 7.4). The next two runs conducted were at 40% level (0.28g) and 60% level (0.42g) subsequently. No major visual changes were found after each of these runs, except that more fine hair line cracks were observed at the potential plastic hinge regions (at the tops of columns). These hair line cracks were marked in blue and pink respectively. During the next run at 80% level (0.56g), it was noticed that the computer software used to drive the shake table was not functioning properly, so that the signals received by the table were not correct. While the acceleration values in the input signal were correct, the displacement signals were 50% less 68 Chapter 7 — Testing and Overall Experimental Results than the desired values. This problem prevented the tests from progressing for several hours before it was finally resolved by re-calibrating the computer program and the table. One 10% level run and one 20% level run were done to determine the correctness of the input signals and the table calibration. Another 80% run at 0.56g was then conducted, and as the testing carried on as planned. The cracks observed after the run were marked in purple. The next run was at the level of 120% (0.84g). Near major cracks in the potential plastic hinge regions, dust smoke generated by the repeatedly opening and close of these cracks as the table moved back and forth was observed during the shaking. As the cracks in the potential plastic regions opened and closed repeatedly, the cap beam kept itself in a relatively horizontal position. The elongation of the existing cracks was more significant compared with the cracks from the previous runs. The significant stiffness drop of the structure during this run, indicated by the hammer test results, gave strong evidence for the formation of plastic hinges. No permanent deformation of the structure was observed after this 120% level run. The new cracks and elongated portions of existing cracks were marked in orange. During the next two runs, one at 150% level (1.05g) and another at 200% level (1.40g), similar phenomena as displayed in the 120% run were also observed, except that a small amount of covering concrete was spalled off at the potential plastic hinge regions near the joints during the 200% level run. During the runs, the main cracks right above the column FRP wraps became bigger, while more finer cracks formed around these main cracks. One flexural crack was found on the upper portion of each column FRP jacket near the plastic hinge region. The opening and closing of the main cracks could be visually observed during the shaking. A l l strain gauges data indicated that little force was transferred into the cap beam during these runs, and therefore, the hinging effect at the top of columns was evident. The slight natural frequency drops after each of these runs indicated the limited damage to the 69 Chapter 7 — Testing and Overall Experimental Results specimen, possibly due to the hinge effect in the columns. New cracks formed after the 150% level run and after the 200% run were marked in black line and blue " X X X " line, respectively. Since the shake table had reached its maximum displacement capacity at this point, it was then decided to remove the external post-tensioning system on the cap beam, and to repeat several higher-level runs in an effort to cause further damage to the specimen. 120%, 150%, and 200% level runs were performed after post-tensioning removal. However, the damage situation of the structure did not change significantly, except that more cover concrete in the hinge region was spalled off during runs and the specimen seemed more flexible compared with previous runs. By the end of whole testing program, no obvious shear cracks were found in the cap beam, and i f there existed any shear cracks, they were too fine to be observed visually. Two plastic hinges were fully developed at the top of two columns as anticipated. The integrity of the structure was relatively well maintained as no significant permanent deformation was noticed by visual observation, and the bent was considered still serviceable if such a usage was required by circumstance. Photograph 20 and 21 in Appendix C show the specimen after testing and the close-up photo of the plastic hinge region at the east column respectively. One phenomenon noticed in the testing was that the east column eventually suffered more severe damage during the later higher level runs, despite the fact that in the earlier testing stage the west column had been damaged more. A possible explanation for this phenomenon is that the repaired and retrofitted model was weaker at the west end initially due to more damage suffered from the previous 70 Chapter 7 — Testing and Overall Experimental Results tests at this location. As a result, during earlier stages of the re-strengthened model tests, the weaker end (the west end) might have been the first to start deteriorating at the relatively low seismic load level, when the stronger east joint was still capable of resisting such a load level. During the later runs, when the ground motion reached higher levels, the plastic hinge in the west column had probably been well-developed, and therefore more seismic energy resulting from more severe shaking was then concentrated on damaging the east column while at the west end, the energy was dissipated effectively by the well-formed plastic hinge. 7.4 Hammer Test Results One hammer impact test was carried out after each shake table run to determine the natural frequency of the bent model at that particular stage of the testing. The hammer test results are listed in Table 7.2. In a hammer impact test, both excitation and structure response signals were recorded in the time history format. By performing a FRP (Fourier transfer function) operation, a frequency-based function was obtained. This FRF was used to determine the natural frequency of the structure at that point. The FRF plot for the bent model after the 120% level run at 0.84g is shown in Figure 7.1. 71 Chapter 7 — Testing and Overall Experimental Results F r e q u e n c y ( H z ) Figure 7.1 Sample F R F Plot (120% run) Table 7.2 Hammer Test Results Test No. Run of Event prior to test 1st Natural Frequency of the model 1 Initial state 6.85 Hz 2 10%arun 6.81 Hz 3 10%b run 6.56 Hz 4 20% run 6.56 Hz 5 40% run 6.26 Hz 6 60% run 5.80 Hz 7 80%a run 5.77 Hz 8 80%b run 5.68 Hz 9 120% run 4.94 Hz 10 150%a run 4.61 Hz 11 200%a run 4.36 Hz 12 200%b run 4.30 Hz 13 Post-tensioning removal 4.18Hz 14 150%b run 3.94 Hz 15 200%c run 3.88 Hz 72 Chapter 7 — Testing and Overall Experimental Results Figure 7.2 shows the deterioration of the first mode frequency of the specimen through testing runs. The horizontal axis indicates the run level or event before one certain hammer test, and the vertical axis is the frequency value. Run Amplitude ( of 0.7g) Figure 7.2 First Natural Frequency of the Specimen Through Runs The first natural frequency of the bent model before testing was 6.85 Hz. The frequency deteriorated steadily as testing progressed, but it dropped more significantly after the 120% run at 0.84g. During several of the last-high-level runs, which included 200% runs at 1.4g and three more runs after post-tensioning removal, the hammer test results indicated little frequency changes. From the initial frequency of 6.85Hz to 5.676Hz after the second 80% run, the model's first natural frequency dropped 1.17Hz in seven runs averaging 0.167Hz per run, which is 2.4% of the initial value; however, through the 120% run, the frequency of the model dropped from 5.676Hz to 4.944Hz. The change of 0.732Hz is almost 11% of the initial value. This relatively large frequency drop indicates a significant softening of the structure, and possibly the formation of plastic hinges in the columns. The possible plastic 73 Chapter 7 — Testing and Overall Experimental Results formation in columns after the 120% level run was further demonstrated by the minor frequency change during the last high-level runs after this 120% run. The first natural frequency of the specimen after the 200% level run before post-tensioning removal was 4.303Hz, and it dropped to 4.12Hz after removing the post-tensioning system. The relatively minor frequency difference between the structure with and without the cap beam post-tensioning system shows that the stiffness of the bent at this stage was dominated mostly by the flexibility of the hinges and little by other factors. During the last test runs, little damage at most was done to the structure, as indicated by the minor frequency change. This was likely due to the fact that the hinges were fully developed at this stage of the testing and most seismic energy was dissipated by hinging action in the columns.. The first natural frequency of the bent model was 3.9Hz by the end of testing, according to the hammer test results. The comparison of stiffness degrading characteristics of the original and retrofitted models is to be studied in later sections. 74 Chapter 8 — Data Interpretation Chapter 8 DATA INTERPRETATION 8.1 Introduction The performance of the repaired and retrofitted model was studied quantitatively by examining the data obtained from several critical testing runs. The acceleration and displacement time history studies, frequency and stiffness studies, hysteresis curve studies, and ductility studies were conducted to assess the effectiveness of the retrofitting design. The strain gauge data were also examined but in less detail due to the fact that a few strain gauges were suspected to have been damaged during the previous original model tests, and their reliability was considered to be questionable. Nevertheless, the results from the above studies clearly indicated significant performance improvement of the re-strengthened model over the original one. The comparative studies of both the original and re-strengthened models will be discussed in Chapter 9. 8.2 Displacement and Acceleration Time History During testing runs, both displacement and acceleration of the model at several locations were recorded as the time history data. These data and strain data were converted into actual physical terms before analysis, as they were originally stored in terms of electric voltage. The sample sheets of both un-calibrated and calibrated data were attached in Appendix D. 75 Chapter 8 — Data Interpretation 8.2.1 Displacement time history The displacement time history data were recorded by three displacement transducers connected to data collection channels. One displacement transducer was located at the top of the dead load mass block, one at the cap beam level and on the shake table and another on the shake table (Figure 6. 8). During testing, the absolute displacement values of the mass block, the cap beam and the shake table were recorded by these three displacement transducers. Responding to a table displacement, the bent displacement in a testing run was magnified to a certain degree, since the chosen earthquake response spectrum was intended to amplify the shaking motion. The degree of magnification was depending on the magnitude of the ground motion, but more importantly depending on the stiffness of the bent at that certain point. Generally, the less stiff the structure was, the great its relative displacement would be, unless the structure was too stiffness or too flexible so that its first natural period was outside of the peak region in the displacement response spectrum. For the Joshua Tree record, this peak region ranges approximately from T=0.2 second to T=1.0 second. In other words, as long as the bent first natural period (or frequency) stayed within this range (T=0.2-1.0 second, o r / =1.0-5.0 Hz), the amplification of the motion would be significant. The hammer test results indicated that, the specimen's first natural frequencies were greater than 5.0 Hz during the earlier low-level runs, and after the first 120% level run its natural frequencies were located within the above frequency range during the later high-level runs. Figure 8.1 shows the time history response plots of the bent's relative displacement during the 10%c run at 0.07g and 150%b run at 0.84g (after post-tensioning system removal). In the 10%c run, the bent relative displacement had a maximum value of 1.63mm, which was about 70% of the maximum table displacement of 2.21mm. However, such high bent displacement values occurred only at points where there were significant peaks in the excitation signal, and the displacement amplification 76 Chapter 8 — Data Interpretation anywhere else was not significant compared with these peak values. For the 150% level run after post-tensioning removal (150%b), a maximum table displacement of 33.88mm generated a maximum relative displacement of 36.42mm at the cap beam level, which yielded the magnification factor of 1.07. Unlike the displacement response during the low-level runs, the whole time history of the bent displacement response was amplified significantly. Displacement Time History Response for 10%c run, 0.07g Displacement Time History for 150%b run, 1.05g (after post-tension removal) Figure 8.1 Table and Bent Relative Displacement for 10%c run and 150%b run Another observation from the above plots worth mentioning is that, even at the very late stages of the testing when the shaking level was high, no permanent residual displacement of the bent was found by the end of 150%b run at 1.05g (after post-tensioning removal). This indicates that the bent had no 77 Chapter 8 — Data Interpretation permanent deformation at this point, and is strong evidence of good maintenance of the structure's integrity. The relative displacement time history plot for the final testing run at 200% level (1.40g) still shows no sign of permanent deformation of the bent (Figure 8.2) Displacement Time History for 200%b run, 1.40g (after post-tension removal) -50 4 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Time, second Figure 8.2 Table and Bent Relative displacement for 200%b run, 1.40g More Displacement time history plots are shown in Appendix D. Table 8.1 lists the maximum shake table displacement recorded during some critical runs, calculated bent relative displacement, and their corresponding percentage of magnification. A complete set of maximum displacement and acceleration values during each run is attached in Appendix D. During the first low-level runs before the table re-calibration, the cap beam's relative displacement was significantly greater than the displacement excitation and was about 200% to 300% of the table displacement. These extremely high values might be the result of insensitivity of the recording devices to small displacement values, and the inaccuracy of the table and computer calibration. After the table re-calibration, both relative bent displacement and its corresponding magnification factor increased 78 Chapter 8 — Data Interpretation gradually as the shaking magnitude increased, and as the first natural frequency of the specimen moved towards the peak range in the response spectrum. The magnification factor of the bent relative displacement compared with the table displacement was about 0.73 during the 80%b run at 0.56g. It increased to approximately 0.95 during the 200%a run at 1.40g. For the last three runs after post-tensioning removal, it reached the values of 1.0 and greater as the structure had been softened significantly due to the increasing of the hinging effect in columns and the removing of the post-tensioning system. This phenomena was possibly the result of the fact that as the structure was softening, its first natural period increased, approaching the highest peak in the earthquake displacement spectrum, which occurs approximately at T=0.3 second o r / = 3.3 Hz. These high values of the relative bent displacement were essential for the structure to achieve high global displacement ductility levels. Table 8.1 Maximum Displacements Run/Event 10% 20% 60% 80% 10% 20% 80% Designed acc. level, g 0.07 0.14 0.42 0.56 N / A 0.07 0.14 0.56 Actual max acc.value, g 0.16 0.30 0.46 0.39 N / A 0.069 0.13 0.50 Table disp., mm 0.84 3.21 4.50 6.98 N / A 2.21 4.22 17.06 Cap beam rel. disp, mm 2.37 7.39 13.63 13.14 N / A 1.63 3.52 13.59 Magnification % 282 230 303 188 N / A 73 83 80 Run/Event 120% 150% 200% i i i i i i 120% 150% 200% Acceleration level 0.84 1.05 1.40 N / A 0.84 1.05 1.40 Actual max acc.value, g 0.89 1.30 2.50 N / A 0.90 1.24 2.67 Table disp., mm 26.36 33.67 39.65 N / A 23.83 33.88 45.15 Cap beam rel.disp, mm 22.76 32.14 37.38 N / A 27.34 36.42 46.66 Magnification % 86 95 94 N / A 115 107 103 * - table re-calibration ** - post-tensioning removal 79 Chapter 8 — Data Interpretation The trend of the specimen's maximum displacement changing through the testing was also studied. Figure 8.3 shows that the absolute displacements (or relative displacements) at both the cap beam level and the top of mass block level increased as the shaking magnitude increased. Figure 8.3 Maximum Displacements of the Cap Beam and Mass Block through Runs By comparing displacement values at these two locations, one could find that even though the displacement at the cap beam level and at the mass block level were approximately the same, the latter was about 20% higher than the former during the early testing runs. This difference reduced gradually as the testing progressed, and by the time testing reached the first 150% level run at 1.05g (before post-80 Chapter 8 — Data Interpretation tensioning removal), the displacements at the cap beam and at the top of the mass block had become almost the same. Such a displacement change of the cap beam and the mass block was probably influenced by the nature of the plastic hinge forming in columns. Before the formation of the plastic hinges, the columns and joints were relatively rigid, and the cap beam and the mass block were not in a same elevation. As the result, the displacement at the top of the mass block was greater than that at the cap beam level since the cap beam along with the mass block was tilted by a small amount. Once the plastic hinges in columns started to form, the top of the column and joint regions became more flexible and little rotation was generated, and therefore the tilt of the cap beam and mass block lessened. By the time the plastic hinges were fully developed, ideally the free rotation was allowed in the hinge regions, and consequently the cap beam and the mass block remained horizontal to the ground during shaking, and the displacements at these two locations became approximately the same (Figure 8.4 ). In other words, during the early runs, the cap beam and the mass block were not only moving left and right as the shake table moved, but also rotating; however, during the later runs when the plastic hinges were formed, the cap beam and mass block shifted only horizontally as the plastic hinges rotated. Lmass Figure 8. 4 Bent Displacement and Mass Block Displacement 81 Chapter 8 — Data Interpretation However, the sudden drop to almost zero displacement indicated in the above maximum mass block displacement plot, during runs after removing the post-tensioning system, was not anticipated and the equipment failure of the devices recording displacement at the mass block top was suspected in this case. 8.2.2 Acceleration time history and inertia force The purpose behind acceleration time history studies is to eventually analyze the inertia force generated during an earthquake event. During the shake table testing of this re-strengthened bent model, the acceleration values at different locations of the structure were recorded by several accelerometers (Figure 6. 8). As to the bent displacements, the bent absolute accelerations were also magnified to different degrees. A similar explanation for the displacement amplification could be applied here for the acceleration magnification. The only differences came from the difference between the peak locations in the displacement and acceleration spectra. The bent absolute accelerations during testing runs were typically magnified 1 to 3 times during the earlier and middle stages of the testing, but the extent of magnification was reduced as the testing progressed and the structure softened, and in the final high-level runs the recorded bent absolute accelerations were even smaller than the table accelerations. The sample plots of acceleration time history for 80%b run at 0.56g and 150%b run at 1.05 are shown in Figure 8.5. 82 Chapter 8 — Data Interpretation 2.0 1.5 03 1.0 o 0 5 o < 0.0 o S -0.5 S -1.0 -1.5 -2.0 A c c e l e r a t i o n T i m e H i s t o r y f o r 8 0 % b r u n , 0 .56g 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Time, second 2.0 1.5 1.0 0.5 0.0 -0.5 -1.0 -1.5 -2.0 , - - 1 -.JiAnilAflft/lllnW -vAA^^AAA^AriA/ll J iJIrUI( innA11K - ,- * , - A M r\ A - - - - -1 1 1 1 1 1 1 1 1 1 1 5 6 7 8 9 10 11 12 13 14 15 Time, second 17 18 19 20 21 22 23 24 2.0 1.5 1.0 0.5 0.0 -0.5 -1.0 -1.5 -2.0 liii - H H A / U I JV^_ir^irT._AtijuAiAV^^l('^Wfi- -i _ r ~ i • - /yum.—Vim III.T.- r r 4 S 6 7 8 9 10 11 12 13 14 15 16 Time, second 17 18 19 20 21 22 23 24 Figure 8.5 Acceleration Time History Plots for 80%b run and 150%b run As shown in the above plots, the bent absolute acceleration had a peak value about three times greater than the peak excitation acceleration during the run at 80% level (0.56g) before the post-tensioning removal; however, during the 150% level run at 1.05g after removing the post-tension system, the bent absolute acceleration values were lower than the excitations and their peak value was only about 93% of the table peak acceleration. 83 Chapter 8 — Data Interpretation Table 8.2 lists maximum table accelerations and their bent absolute acceleration responses during some critical test runs. A complete table for all testing runs is attached in Appendix D. To better illustrate the acceleration response of the bent, a plot for the maximum bent acceleration changes was generated and is shown in Figure 8.6. Table 8.2 Maximum Accelerations Run/Event 10% 20% 60% 80% l l l l l l l 10% 20% 80% Designed acc. level, g 0.07 0.14 0.42 0.56 N / A 0.07 0.14 0.56 Actual max table acc. g 0.16 0.30 0.46 0.39 N / A 0.069 0.13 0.50 Bent absolute acc, g 0.28 0.58 1.06 1.45 N / A 0.21 0.45 1.68 Bent acc./ Table acc. 1.75 1.93 2.30 3.71 N / A 3.04 3.46 3.36 Run/Event 120% 150% 200% l l l l l l l l l l l l l l 150% 200% Acceleration level 0.84 1.05 1.40 N / A 0.84 1.05 1.40 Actual max table acc. g 0.89 1.30 2.50 N / A 0.90 1.24 2.67 Bent absolute acc, g 1.70 1.70 1.50 N / A 0.95 1.16 1.29 Bent acc./ Table acc. 1.91 1.31 0.60 N / A 1.06 0.94 0.48 - table re-calibration - post-tensioning removal Figure 8.6 shows that the degree of acceleration magnification increased significantly after the 60% level run at 0.42g until the acceleration response of the bent reached the maximum value at about 120%a run before post-tensioning removal. In fact, the bent maximum absolute acceleration level remained virtually the same for the 120% level and 150% level runs before removing the post-tensioning system, despite the significant change of the table acceleration. The maximum bent absolute accelerations during these two runs reached 1.70g and 1.70g respectively. (Note that the table acceleration drop during the 80% level run before post-tensioning removal should be ignored since the input driving signals during this run were not correct). As the structure softened further, the bent absolute acceleration was only about 60% of the table excitation acceleration value during the 200% 84 Chapter 8 — Data Interpretation level run at 1.40g before removing the post-tensioning. After the post-tensioning system was removed, the structure was softened more, and the maximum acceleration response of the bent decreased to only around 0.95g during the 120% level run at 0.84g. It increased slightly during the next two runs despite the significant increase in excitation acceleration. 3.0 Figure 8. 6 Max. Bent Absolute Acceleration vs. Max. Table Acceleration During the early stage low-level runs, the first natural frequency of the structure was 6.0Hz or higher, which is located at the ascending arm of the first peak of the acceleration response spectrum. As a result, the acceleration response of the bent increased as the tests progressed. When the structure softened to a certain point, when its natural frequency was close to the peak frequency in the response spectrum, the bent acceleration reached its maximum. This explains the reason why the acceleration was amplified the most around the 80% level at 0.56g, when the first natural frequency of the bent was approximately 5.8 Hz. Once the structure softened further and its natural frequency passed this peak, the acceleration magnification of the bent started to decrease as the structure's frequency moved further and further from this peak value as indicated in Figure 8.7. 85 Chapter 8 — Data Interpretation Run Figure 8. 7 Acceleration Magnification Changes through Runs The inertia forces generated in each run were obtained from the mass acceleration values recorded during these runs. Figure 8.8 shows the maximum inertia forces generated during some critical runs. Run Figure 8. 8 Maximum Inertia Force Changes through Runs The maximum inertia force was observed in the first 120% level run before removing the post-tensioning, and it reached a value of 192kN. Similar to the bent acceleration response, the maximum inertia force values reached a plateau after the 80% level run at 0.56g, and the forces dropped 86 Chapter 8 — Data Interpretation significantly after the post-tensioning system was removed. They were much lower during the last three runs, which had the same shaking magnitudes as the three runs before the post-tensioning removal. It is important to note that at the 80% level, the inertia force drop is not a indication of the structure's load resisting level decreasing, and instead, it is simply the result of incorrect input acceleration signals. As shown in Figure 8. 6, the maximum value of the input acceleration signals during the 80% level run was even less than that during the 20% runs. This unreasonable shake table acceleration value generated the incorrect acceleration response of the structure during the 80% level run. The similar reason also caused the incorrect acceleration response during the 20% level run. To better understand the structure's global degrading mechanism and plastic behavior, a force-displacement curve was generated by relating the maximum bent relative displacements and maximum inertia forces through runs. This was justified because of the fact that both quantities were in ascending order corresponding to the sequence of the testing progress. In other words, each pair of displacement and inertia force values represented a force-displacement relation in that certain run. Figure 8.9 shows the westward force-displacement relationship of the specimen through the testing process. The last three runs after removing the post-tensioning system were not considered in the plot, since the removal of the post-tensioning itself changed the stiffness of the structure with no earthquake load applied, and it was not logical to use values from two "different" structures for plotting the force-deformation curve for one specimen. 87 Chapter 8 — Data Interpretation 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 Displacement, mm Figure 8. 9 Inertia Force - Displacement Plot (westward motion) As motioned before, the unexpected load droppings at displacement of approximately 7mm and 13mm were again the result of incorrect acceleration shake table signals during the 20% and 80% level runs. Figure 8. 10 shows that the bent displacement during the 20% and 80% level runs were about 7mm and 13mm, respectively. Run Figure 8.10 Maximum Bent Relative Displacement Change (westward motion) Chapter 8 — Data Interpretation A more reasonable load-deflection curve was generated by eliminating these incorrect signals, and it is shown in Figure 8.11. 180 ; ; 1 1 ; : r" 1 160 1 \ I ! i \ ] 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 Displacement, mm Figure 8. 11 Load-deflection Curve for The Retrofitted Bent This load-deflection curve shows that the structure started to yield globally at around a displacement of 6.6mm and a base shear of 120kN. This is in good agreement with the prediction made prior to the testing (see Chapter 5, section 5.2 — Predicted Model Behavior). Once the load reached its maximum value of nearly 160kN at a lateral displacement of close to 20mm, the structure was in the stage of plastic deformation, as indicated by the plateau of the curve. At this stage, the structure continued to deform without any increase in the load applied, and this was time when the plastic hinges in the columns were fully developed. The fact that the load level did not drop after the structure underwent plastic deformation indicated a sufficient load-resistant ability in the process of deformation. This ability enabled the bent to maintain its structural integrity at the end of the testing. Combining the load-deflection curve above and the maximum bent relative displacement vs. run plot shown in Figure 8.10, it was concluded that the plastic hinges in the column and joint regions started to form at 20% 89 Chapter 8 — Data Interpretation level run at 0.14g, and were fully developed around 120% level at 0.84g before the post-tensioning removal. This conclusion agreed well with the visual observations during the actual tests, and it supported the assumption made by examining the results from hammer impacting tests (see Chapter 7, section 7.3 — Tests Description). The inertia force - displacement plot for the eastward motion gave similar results, except that the load level of the plastic "plateau" was about 12% higher than for the westward motion plot. The eastward plot gave a base shear of around 180kN and displacement of near 20mm at the point when the plastic hinges were fully developed. 8.3 Hysteresis Curves and Energy Dissipation Mechanism To examine the energy absorption characteristics of the bent model at different structural conditions during the seismic loading, hysteresis curves were plotted for all 17 runs of testing. These hysteresis curves were obtained by plot load values against their corresponding displacement values. A plot consisted of many continuous load-displacement loops as the structure was loaded and unloaded continuously in opposite directions. Figure 8. 12 shows hysteresis loops for six typical testing runs. In the first three plots of relatively low-level runs, the areas surrounded by the outer boundaries of the loops are relatively lean as the load-deformation relations at these stages were mainly elastic. The load-deformation relation was dominantly linear with a relatively large load increasing rate (large slope). As a result, the load level experienced by the structure during these runs increased significantly while the displacement level changed only minimally, and relatively less energy was dissipated within loops compared with later 90 Chapter 8 — Data Interpretation runs. As the shaking severity increased and the structure softened, the deformation became less elastic and more plastic. The loops grew fatter as the loading increasing rate decreased gradually. Once the plastic hinges in the structure had been fully developed, the plastic deformation became the dominant mechanism in determining the load-deformation characteristics of the structure. As indicated in the last three plots, the areas within these loops were much fatter and the pinching effect was more obvious, and significantly more seismic energy was absorbed during the plastic deformation process. 1 0 % o f 0 . 7 g 4 0 % o f 0 . 7 g 2 0 0 % o f 0 . 7 g ( b e f o r e P T r e m o v a l ) 2 0 0 % o f 0 . 7 g ( a f te r P T r e m o v a l ) Bent Relative Displacement, mm Bent Relative Displacement, mm Figure 8.12 Hysteresis Plot for Six Typical Runs 91 Chapter 8 — Data Interpretation After the plastic hinges in the column and joint regions were formed, the structure behaved in a plastic ductile manner in response to the ground motion. Despite the shaking magnitude increase, the force in the cap beam remained unchanged, and the most of the seismic energy was dissipated by this plastic mechanism. This process of energy dissipation was physically achieved by the rotation of the plastic hinges. As noticed during visual observation, the closing and opening of the cracks in the plastic hinge regions during the last several high-level runs were the manifestations of such a hinge rotation. The ground energy transferred through columns was mostly dissipated by this hinge rotation, and therefore little force could travel to the cap beam to cause any damage. This explains the fact that no damage was found in the cap beam through the entire testing process, and little strain was recorded by strain gauges during the last several high-level runs. 8.4 Ductility Level To evaluate the plastic deformation ability of the specimen during the seismic loading conditions, the global displacement ductility levels were computed for each individual run. The displacement ductility levels were obtained using the following formula, ju = A t / A y (Equation. 8.1) where ju = displacement ductility A t = total imposed displacement Ay = yield displacement Both the response prediction and the data analysis showed that the structure reached its yield point at a displacement of 6.5mm, and therefore the global yield displacement of the bent was 6.5mm. Table 8.3 92 Chapter 8 — Data Interpretation lists the maximum bent relative displacements recorded through runs and their corresponding displacement ductility levels. Table 8.3 Max Bent Relative Displacements and Ductility Levels Run 5%a 5%b 10%a 20%a 20%b 40% 60% 80%b Max rel. disp, mm 0.69 0.77 2.73 5.77 7.39 10.52 13.63 13.66 Disp. ductility 0.11 0.12 0.37 0.89 1.14 1.62 2.09 2.10 Run 120%a 150%a 200%a 1111 120%b I50%b 200%b Max rel. disp, mm 22.76 32.14 39.29 / 27.64 36.42 46.66 Disp. ductility 3.50 4.95 6.05 / 4.25 5.61 7.18 - post-tensioning removal The displacement ductility values for some critical runs were plotted in Figure 8. 13. The fact that the bent reached a ductility of 1 at the 20% level run at 0.14g indicated that the structure started to yield at that moment. This supports with the conclusion made earlier in the load - deformation analysis. By the time the 200% level run at 1.40g had been done just prior to the removal of the post-tensioning system, the bent had a displacement ductility of 6, which indicated that for a reinforced concrete structure, the bent's post-yielding deformation ability was quite significant. During the three high-level runs after post-tensioning removal, the ductility level changed and followed a similar pattern as in the three runs before removing the post-tensioning. Each run, however, generated a higher ductility value than the corresponding run did before post-tensioning removal. The structure's displacement ductility during the final 200% level run (1.40g) subsequently passed the level of 7. 93 Chapter 8 — Data Interpretation Figure 8.13 Displacement Ductility Changes through Runs It should be noted that this ductility value of 7.18 is not the ultimate displacement ductility level for the structure. The limitations of the shake table, not the structural condition of the specimen, put an end to the testing. In fact, the structural integrity of the bent was relatively well maintained by the end of the testing program, judging from both visual observations and the load - deformation characteristics study. While a high displacement ductility level of a structure is desirable, it becomes meaningless i f the structure's basic integrity cannot be maintained in the process of plastic deformation. The reason behind improving the ductility level of a structure is to ensure that the structure is able to give enough warning before it ultimately fails, and at the same time most seismic energy is dissipated by the designated plastic deformation mechanism. 94 Chapter 8 — Data Interpretation 8.5 Frequency Responses Besides the hammer impact tests performed between each adjacent run, FRF (Fourier transfer function) operations were also carried out for the excitation and response signals recorded during the shaking tests to examine the frequency responses at the various stages of the testing. A computer program named FRF (Ventura et al, 1995) was used to carry out this operation. The names of all calibrated input data files for the computation are listed in Appendix D. Significant but not random differences were found between the frequencies obtained from the hammer tests results and the frequencies determined by the FRF operations on shake table data (Figure 8. 14). The frequency from the shake test data of the 10% level run was the same as that obtained from the hammer test data, and for the runs after this 10% level run the frequencies determined from the shake table data were lower than those from the hammer test data. The typical differences between these two were about 10%~30% of the frequency from shake table data , with the exception of the last high-level runs, where the differences were as high as 50%. Figure 8.14 Natural Frequencies from Hammer Test Data and Shake Table Test Data 95 Chapter 8 — Data Interpretation Similar phenomena, even similar frequency changing patterns, had also been noticed in the previous original bent shake table tests. A possible explanation for this phenomena, as stated by Davey, is that the hammer impact was not strong enough to excite the specimen into its full motion, and the friction within the cracks would contribute to the structure's stiffness when the structure was subjected to the hammer impacts (see Shake Table Testing of an Oak Street Bridge Bent Model by E. Davey, 1996. Page 95). It was also noticed from Figure 8. 14 that these frequency differences increased significantly during the last high-level testing runs. The reason for this phenomena was considered to be that at the late stages of the testing, the specimen had suffered some local damages, and at the same time the hammer impact test sensors recorded the global vibration signals of the bent, they also recorded the signals of those local vibrations, such as the vibration of loose concrete, loose mass block, etc.. As a result, the peaks of the first mode vibration of the specimen became less obvious in the FRF plots, and the accuracy of the frequency determination decreased. Figure 8. 15 shows the FRF plots for the acceleration response of the cap beam in the 20%b run and 200%a run. 4 8. Frequency (Hz) 11. 20D%a run Frequency (Hz) Figure 8.15 Sample FRF Plots for the 20%b Run and 200%a Run 96 Chapter 8 — Data Interpretation Despite the differences of the frequencies obtained from the two methods, the changing patterns of these two groups of frequency values agreed with each other, in regards to the possible time of plastic hinge formation. As stated in earlier sections, the hammer test results indicated a significant drop in frequency after the 120%a run at 0.84g. These results gave the strong evidence of plastic hinge formation during that run. Similarly, FRF results on the shake table test data showed that the frequency obtained by analyzing the data recorded during the 150% run was significantly smaller than the ones from previous runs. However, the condition of the specimen during a run was not the same as the condition after that run, and the frequency was more influenced by the initial condition before that run started. In other words, the low frequency recorded during the 150% level run indicated, to a great extent, that the frequency had dropped significantly after the previous run ( 120% level run). 97 Chapter 9 —Test Result Comparison of Re-strengthened and Original Models CHAPTER 9 TEST RESULT COMPARISON OF RE-STRENGTHENED AND ORIGINAL MODELS 9.1 Introduction A comparative study on the performances of the restrengthened versus the original bent model was carried out quantitatively following the tests. The study results indicated that the repaired and retrofitted model outperformed the original one, and showed more desirable seismic response characteristics in many aspects, such as a more ductile response, a better energy dissipation mechanism, a slower stiffness degrading rate, higher load resistance levels, and a more desirable damage mode. 9.2 Damage to Specimens The original Oak Bridge Bent S28 27% scale as-built model had been severely damaged by the shear in the cap beam at very low seismic load level during the previous testing, and the failure had been sudden and brittle. No major flexural cracks and plastic behavior of the columns were observed during the test due to the early stage shear damage in the cap beam. By the end of the testing, it appeared to be the concrete mass block that held the badly damaged bent together, and prevented it from collapsing. The damage mode of the as-built specimen, denoted OSB-O, is illustrated in Photograph 2 and 3 in Appendix C. 98 Chapter 9 —Test Result Comparison of Re-strengthened and Original Models The repaired and retrofitted model, denoted as OSB-R, suffered far less damage than the as-built bent, and it showed ductile behavior as it underwent plastic deformation during the shake table tests. The formation of plastic hinges at the tops of the two columns was responsible for this desirable behavior. The bent model was able to keep its structural integrity and adequate load-resisting capability at high seismic load levels by the end of the testing program. No major shear damage was found anywhere in the structure throughout the tests. The specimen after testing is shown in Photograph 19. Photograph 20 and 21 are close-up photos of the plastic hinge in the west and east column. Compared with the OSB-O, the re-strengthened specimen had significantly less severe concrete spalling and debonding problems. This verified the fact that even though the fiberglass wraps were not stressed to provide active confinement pressure, the FRP jackets indeed helped to achieve better confinement in section corners. Figure 9. 1 is the time history comparison plot for the OSB-O and OSB-R at 150% level. It shows that the OSB-O had a permanent residual displacement by the end of the run, and such a residual displacement was not found on the OSB-R through out all testing runs. 40 OSB-R OSB-O reference -40 5 10 15 Time, second 20 25 Figure 9. 1 Comparison of Displacement Time History, 150% run 99 Chapter 9 —Test Result Comparison ofRe-strengthened and Original Models To determine when the OSB-0 started losing its lateral strength, a plot for the maximum values of the shake table acceleration and the responding acceleration of the structure at the mass block level was generated (Figure 9.2). 120% 150% Figure 9. 2 Max. Mass Acceleration vs. Max Table Acceleration, OSB-O As indicated by the above plot, the acceleration experienced by the OSB-0 had dropped dramatically after the 40% level run, while the table acceleration kept increasing consistently. In fact, the acceleration or inertia force in the structure during the 60% level run was only about 60% of that during the previous 40% level run. Therefore, it was concluded that the bent had probably failed during the 60% level run at 0.42g ( 0.12g in prototype ). After the specimen lost its strength, the concrete mass block was probably responsible for keeping the structure stable and taking the most of the lateral load in the later testing runs. This conclusion is supported by the fact that, in the above plot, the structure regained its acceleration level or inertia force level during the 120% mn, as the concrete mass block was taking the load and acting as a beam. However, the acceleration in the structure dropped again during the final 150% level run due to the possible damage to the mass block. In fact, a crack was found on the block after the testing. 100 Chapter 9 —Test Result Comparison of Re-strengthened and Original Models The same plot was also generated for the OSB-R (Figure 9. 3). As stated in the previous chapter (Section 8.2.2 — Acceleration Time History and Inertia Force), the decreasing of the acceleration or inertia force in the structure during the 80% level run was the result of the incorrect low acceleration excitation of the shake table. In fact, the bent acceleration never dropped throughout all testing runs, except the times when the shake table acceleration decreased. In other words, the retrofitted model still maintained an adequate lateral strength, even at high seismic load levels. 3.0 Figure 9. 3 Max. Mass Acceleration vs. Max Table Acceleration, OSB-R 9.3 Stiffness Degradation Frequencies of OSB-0 and OSB-R at different testing stages were calculated by carrying out the FRF operation for data from both hammer impact tests and shake table tests. For both specimens, the frequencies obtained from hammer test data were generally higher than those from the shake table test data. The possible explanation for this phenomena are stated in Chapter 8. The frequency degradation for the OSB-0 and OSB-R is shown in Figure 9. 4 and Figure 9. 5. 101 Chapter 9 —Test Result Comparison of Re-strengthened and Original Models Run amplitude or event Figure 9. 5 Comparison of Frequency Degradation, Shake Table Tests The hammer test results indicated that the initial first natural frequency of the OSB-O was 7.7Hz. It dropped to 3.5Hz after being subjected to 10 testing runs with the maximum amplitude of 150% (1.05g). The OSB-R started with a natural frequency of 6.9Hz; its frequency after 14 testing runs was 4.3Hz, and it further dropped to 3.9Hz after post-tensioning removal and the additional three runs. The shake table testing data showed similar results but with lower frequency values. According to these data, the final frequency for the OSB-O was 2.3Hz, while it was 3.5Hz and 2.6Hz for the OSB-R before post-tensioning removal and at the end of the testing respectively. 102 Chapter 9 —Test Result Comparison of Re-strengthened and Original Models Both Figure 9. 4 and 9. 5 show that the OSB-R had a slow frequency degrading rate, even though the OSB-R was slightly more flexible in its initial state. After the 40% level run, the natural frequencies of the O S B - 0 were significantly lower than those of the OSB-R. The stiffness degradation accelerated after the 80% level run for OSB - 0 until the its last test at 150% level (1.05g). However, for the specimen OSB-R, the rate of stiffness degradation slowed down significantly after its plastic hinges were fully developed at 120% level, and little frequency changes were observed during the late stages of testing. The significant frequency drop of the OSB-R after the first 120% level run before post-tensioning removal indicated the possibility of the plastic hinge forming at this point, while the large frequency drop of the OSB - 0 after its 80% level run might have been the sign of development of the plastic action of the shear damage in the cap beam. It is important to note that the frequency or stiffness characteristics of the OSB - 0 shown in the above plots may not correctly represent its actual damage situation, especially during the later high level runs. During these runs, the concrete mass block possibly acted as a beam, and provide the most of the stiffness to the whole structure, since as stated in the previous section, the bent itself was considered failed by the 60% level run. Therefore, the actual damage situation of the OSB - 0 was likely much more severe than was indicated in the above plots. 9.4 Hysteresis loops The hysteresis curves for both OSB - 0 and OSB-R did not show significant differences during the earlier low-level runs. However, as specimens got softer as ground motion increased, the hysteresis curves for OSB-R became much fatter than those for OSB-O, and consequently more ground energy was dissipated in the process. In addition, the OSB-R showed higher strength than the OSB-O, 103 Chapter 9 —Test Result Comparison of Re-strengthened and Original Models especially during the later testing runs with large shake magnitudes. The comparison of hysteresis plots for both specimens during the 60% level run at 0.42g and the 150% level run at 1.05g are shown in Figure 9. 6. As the OSB-R softened further as the tests progressed, the area within its hysteresis loops grew even fatter and the pinching effects became more obvious. 200 60% of0.7g, OSB-O 200 150 100 50 0 -50 -100 -150 4 -200 60%of0.7g, OSB-R H h •40 -20 20 40 200 150 100 - 50 01 u I ° n 1 "50 c -100 -150 -200 150% of 0.7g, OSB-O 150% of 0.7g, OSB-R —\ 1 1 1 --40 -20 0 20 Bent Relative Displacement, mm 40 -40 -20 0 20 Bent Relative Displacement, mm Figure 9. 6 Comparison of Hysteresis Loops The figure above indicates that the stiffness of the OSB-0 and OSB-R were about the same at the 60% level. However, the area within the outer boundary of the OSB-R loops is much larger than that of the OSB-O, which indicates that more seismic energy was dissipated by the OSB-R at this stage. In addition, the peak load level experienced by the OSB-R, which exceeded 150kN, was significantly 104 Chapter 9 —Test Result Comparison of Re-strengthened and Original Models higher than that resisted by the OSB-O. During the 150% level run, the strength of the OSB-O deteriorated significantly as indicated by the hysteresis plot, while the OSB-R still kept adequate load resistant ability. Comparing with the 60% level run, the strength of the OSB-R during this run increased slightly at the same time it reached a higher ductility level. In fact, the specimen still held its strength at this level (about 160kN) even after the final run of the testing. Again, the hysteresis characteristics shown by the OSB-O during the runs conducted after its failure at the 60% run are likely the characteristics of the system consisting of the damaged bent and the concrete mass block, instead of the actual hysteresis responses of the bent itself. 9.5 Load - Deformation Characteristics As indicated by the hysteresis plots, the specimen OSB-R showed significantly higher strength than the OSB-O throughout the testing, and even after its final test, the OSB-R still showed no sign of strength deterioration. In contrast, the strength of the OSB-O had started to degrade at a very low seismic load level. Figure 9. 7 and Figure 9. 8 are comparison plots for maximum load and bent relative displacement of the specimen OSB-O and OSB-R through some critical runs. These plots were generated by computing westward inertia forces and displacements from recorded data; values in the eastward direction gave similar results. 105 Chapter 9 —Test Result Comparison of Re-strengthened and Original Models Figure 9 . 7 Comparison of Maximum Inertia Force Figure 9 . 7 shows that inertia forces induced in the OSB-R were larger than those on the OSB-O in all testing runs. During the earlier low-level runs, the load difference was not as significant as during later runs, where the inertia forces experienced by the OSB-R were almost 7 0 % higher than OSB-O. At the 1 2 0 % level run, the load plot of the OSB-R reached a plateau, and never dropped through later runs. Again, as stated in the earlier sections, the load drop experienced by the OSB-R at the 8 0 % level was the result of the incorrect table input acceleration signals; however, the load drop of the OSB-O after the 4 0 % o level run was the indication of the bent failure. During the last several runs with high ground motion, the load levels of the OSB-O shown in the above plot were actually provided by the concrete mass block acting as a beam. The load drop at the 150% level run on the OSB-O curve indicates the possible damage in the mass block at that certain time. The bent relative displacements generated by ground motion were slightly higher for the retrofitted specimen OSB-R than those for the OSB-O during the earlier low-level runs due to the fact that the OSB-R was a slightly more flexible structure at the start of the testing. After the 8 0 % run, the bent 1 0 6 Chapter 9 —Test Result Comparison of Re-strengthened and Original Models displacement of the OSB - 0 increased rapidly as the OSB - 0 softened significantly due to the possible plastic deformation in the shear-damaged regions of the cap beam. The displacement of the OSB - 0 became much larger than the displacement of the OSB-R after the 120% level run at 0.84g. To better understand the load-deformation characteristics of the specimen, load-displacement envelopes were generated for both OSB - 0 and OSB-R. The comparison plot is shown in Figure 9.9. The OSB-R started yielding at a displacement of about 6.5mm and a base shear of around 130kN. This result matched the values scaled down from the result of the slow cyclic test for the 45% scale model (see Section 5.2). After the displacement reached around 20mm, the bent began to deform plastically, and the load resistant level was kept at near 160kN without a significant drop as the structure was deforming. It was evident that the plastic hinges in columns were fully developed by this point (A»20mm, F«160kN), which proved the earlier suspicion that the hinge formed during 120% level run, since a maximum displacement of around 20mm was reached during this run. The OSB-O, however, 107 Chapter 9 —Test Result Comparison of Re-strengthened and Original Models had a much lower ultimate load resistant level at near lOOkN, and the bent failed after it reached its peak strength at the 40% level. It showed little signs of plastic deformation before its ultimate failure. The ultimate load-resistant level of the retrofitted specimen OSB-R was 68% higher than that of the original model OSB-O. Again, this value is in a good agreement with the results obtained from the earlier slow cyclic testing of the 45% scale bent model, in which the ultimate strength of the fiberglass wrapped bent was shown to be 67% higher than that of the un-retrofitted specimen (D. L . Anderson et al, 1995). 108 Chapter 9 —Test Result Comparison of Re-strengthened and Original Models 9.6 Displacement Ductility Level To evaluate the plastic deformation ability of the specimens, the global displacement ductility levels were computed for OSB-O and OSB-R for each testing run. The displacement ductility changes through runs for both specimens are plotted in Figure 9. 10. Run Ampl i tude (of 0.7g) ^ a. Figure 9. 10 Comparison of Displacement Ductility By the time it had failed at 60% level run, the OSB-O had a ductility of only 1.8, which was its ultimate ductility. The displacement ductility for OSB-R was 7.2 after its final run at 1.4g after the post-tensioning removal. Its ultimate ductility should have been significantly higher than this value, since the testing still could have been continued and higher seismic load could have been applied to the structure i f the table had not reached its maximum capacity. Again, it should be noted that relatively large ductility levels of the OSB-O during the later stages of the testing (after the 60% level run) did not indicate the real ability of the bent to deform plastically 109 Chapter 9 —Test Result Comparison of Re-strengthened and Original Models while retaining an adequate load resistant level. These ductility values were simply results of the bent's relatively large displacements, generated by the "shear hinges" rotation in the cap beam instead of deformation of the columns. They were only the indications for the severity of it shear damages but not the column flexure ability at these stages. 1 1 0 Chapter 10 — Summary and Conclusions Chapter 10 SUMMARY AND CONCLUSIONS 10.1 Summary As one part of a series of testing and analytical programs performed at the University of British Columbia on the as-built and retrofitted concrete bridge bent of the Oak Street Bridge, this shake table testing project of a post-seismic repaired and retrofitted 27% scale Oak Street Bridge bent S28 model was carried out successfully. The performance of an as-built 27% scale bent model during the previous shake table test, and the performance of an as-built 45% scale model under slow cyclic loading condition were reviewed, and their seismic deficiencies were studied, before a repairing and re-strengthening scheme was developed and a detailed retrofitting system was designed accordingly. A newly-developed glass fiber reinforced plastic sheet wrapping system, QuakeWrap™ , was employed to retrofit the severely damaged 27% scale bent model, with the goal being to repair the specimen and upgrade its seismic performance under severe earthquake loading conditions. The repaired and retrofitted 27% scale model was then subjected to the same earthquake loading on the shake table as the original model had been. The test was set up and carried out successfully, and reasonably reliable testing data was generated and collected. Ill Chapter 10 — Summary and Conclusions Overall, the performance of the re-strengthened bent model was satisfactory, out-performing the original as-built model in all aspects regarding structural seismic responses. Both visual observation and data analysis have indicated that, even though the retrofitted model was subjected to a more severe seismic loading condition compared with the original as-built model, the repaired and retrofitted model bent had significantly higher strength, and behaved in a more ductile manner under severe ground motion because plastic hinges formed in the designated regions. The ultimate load-resistance level of the retrofitted model was about 70% higher than that of the original as-built bent model. A good energy dissipation mechanism was noticed by analyzing the data, and high displacement ductility levels of the structure were achieved without sacrificing the bent's load resistant level or its structural integrity. Compared with the original as-built model, the re-strengthened bent suffered far less damage. Throughout all 17 testing runs, no sudden shear damage was generated anywhere in the specimen, and no major shear crack was visually found in the cap beam. The specimen was considered still in a serviceable condition by the end of the test when the shake table had reached its maximum displacement capacity at a PGA of 1.4g, since the structure did not show any signs of strength decreasing at this point, and its structural integrity was still well-maintained at the end of the test. The FRP wrapping system installed was considered to be an effective and reliable means of providing extra shear strength to members. It also minimized the cover concrete spalling in the plastic hinge regions and longitudinal steel buckling due the confinement provided to member corners by this wrapping system. The wrapping jacket did not suffer any damage besides flexural cracks in the plastic hinge regions, and no debonding between the wrapping sheets and the surface concrete was noticed by the end of the test. 112 Chapter 10 — Summary and Conclusions 10.2 Conclusions The research results have indicated that the FRP wrapping system could provide the structure members with sufficient and reliable shear strength so that brittle shear damage could be prevented or rninimized during an actual seismic event. Better confinement in the section comers can reduce the severity of cover concrete spalling and minimize the steel buckling. Installed at designated locations of the structure, the system would help the structure to deform in a plastic ductile manner and to dissipate seismic energy in the process, and eventually to reach large displacements and ductility levels without a significant reduction of the structure's load resisting ability and without losing its structural stability. During a post-seismic restoration period, it is essential to bring some damaged important structures, such as vital bridges, dams, and other infrastructure, back to service as soon as possible. The fiberglass sheet wrapping technique presented in this research requires less time and manpower to install, and the installation requires little special skills. In addition, few materials and procedures are involved in this installation operation compared with other conventional concrete structure retrofitting methods. Consequently, this technique is considered to be more economical than most of the other retrofitting alternatives. The research not only presented a fast and effective method to restore seismically damaged bridge structures in the shortest time possible, but also provided an economical and reliable retrofitting alternative for upgrading many seismically vulnerable existing structures to meet the demands of today's structure restoration practice. 113 References R E F E R E N C E S ACI, "Building Code Requirements for Reinforced Concrete", and Commentary - A C I 318R-89, American Concrete Institute, Detroit, MI. , 1992. Anderson, D.L. , Sexsmith, R.G. , English, D.S., Kennedy, D.W., and Jennings, D.B., "Oak Street & Queensborough Bridges Two Column Bent Tests", Earthquake Engineering Research Facility, Department of Civil Engineering, University of British Columbia, Canada, Technical Report 95-02, July, 1995. Chai, Y . H . , Priestly, M.J .N. , and Seible, F., "Seismic Retrofit of Circular Bridge Columns for Enhanced Flexural Performance", ACI StmcturalJoumal, Volume 88, No. 5, October 1991. Chopra, A . K . , Dynamics of Structures - Theory and Applications to Earthquake Engineering, New Jersey, Prentice Hall, 1995. Collins, M.P. , and Mitchell, D., Prestressed Concrete Structures, New Jersey, Prentice Hall, 1991. Crippen International Ltd., "Oak Street Bridge No. 1380 - Two-Column Bent Test Program -Comparison of Five Major bridges", Report Q l 17C/7.2.0., March 1993. CSA, " C S A Standard CAN3-A23.3-M84" - Design of Concrete Structures for Buildings, Canadian Standards Association, Roxdale, Ontario, 1985. Davey, E. , "Shake Table Testing of An Oak St. Bridge Bent Model", Thesis for M. A. Sc. Degree, Department of Civil Engineering, University of British Columbia, Canada, August 1996. Design Specifications for Highway Bridges, Department of Public Works of British Columbia, 1995. Earthquake Design Code - Section 4 and Commentary J, National Building Code of Canada, National Research Council of Canada, Institute for Research on Construction, Ottawa, 1995. English, D.S., "Comparison of Non-linear Analytical and Experimental Curvature Distributions in Two-Column Bridge Bents", Thesis for M.A.Sc. Degree, Department of Civil Engineering, University of British Columbia, Canada, January 1996. English, D.S., Cigic, T., Sexsmith, R.G. , and Anderson, D.L. , "Test of a Bridge Bent Seismic Retrofit Using Fiberglass Jackets", Canadian Society for Civil Engineering Annual Conference Proceedings, Ottawa, June 1995, Vol. 3, pp. 410-427. Handbook of Steel Construction, Canadian Institute of Steel Construction, Universal Offset Limited, Markham, Ontario, 1989. MacRae, G.A., Priestly, M.J .N. , and Seible, F., "Shake Table Tests of Twin-Column Bridge Bents", CALTRANS Third Annual Seismic Research Workshop Proceedings, Sacramento, California, June 1994. 114 References Mitchell, D., Sexsmith, R.G. , and Tinawi, R., "Seismic Retrofitting Techniques for bridges - A State of the Art Report", Canadian Journal of Civil Engineering, Volume 21, No.5, 1994. Ohuchi, H . , Ohno, S., Katsumata, H . , Kobatake, Y . , Meta, T., Yamagata, K . , Inokuma, Y . , and Ogata, N . , "Seismic Strengthening Design Technique for Existing Bridge Columns with CFRP", Seismic Design and Retrofitting of Reinforced Concrete Bridges - Proceedings of the Second International Workshop, Queenstown, New Zealand, August 9-12, 1994, pp. 495-514. Paulay, T., and Priestly, M.J .N. , Seismic Design of Reinforced Concrete and Masonry Buildings, New York, John Wiley & Sons, 1992. Pillai, S.U., and Kirk D.W., Reinforced Concrete Design, McGraw-Hill Ryerson Limited, 1988. Priestly, M.J .N. , and Seible, F., "Seismic Assessment and Retrofit of Bridges", University of California, San Diego, La Jolla, California, Report No. SSRP-91/03, 1991. Seethaler, M.F . , "Cyclic Response of Oak Street Bridge Bents", Thesis for M.A.Sc. Degree, Department of Civil Engineering, University of British Columbia, Canada, April 1995. Saadatmanesh, H , Ehsani, M.R. , and Jin, L . 1995. "Repair of Earthquake-damaged R/C Columns with Prefabricated FRP Wraps", ACI Structural Journal, MS No. 8847, June, American Concrete Institute, Detroit. SEQUAD Consulting Engineering Inc., "Repair of Shear Column Using Fiberglass / Epoxy Jacket and Epoxy Injection", Report 93-04, July 1993. SEQUAD Consulting Engineering Inc., "Seismic Retrofit of Bridge Columns Using High-Strength Fiberglass / Epoxy Jackets -Design Recommendations", Report 93-07, August, 1993. Williams, M.S. , "Inelastic Damage Analysis of the Oak Street and Queensborough Bridge Bents", Earthquake Engineering Research Facility, Department of Civil Engineering, University of British Columbia, Canada, Technical Report 94-03, 1994. 115 Appendix A — Drawings APPENDIX A DRAWINGS 116 Appendix A — Drawings 117 Appendix A — Drawings SYM. ABOUT PIER LOWER ROW I UPPER ROW 860 MK. 13 r 290 L 1265 MK. 11 819 MK. 7. 383 -MK. 10 1E65 MK. 11 - 66E MK. 8 —i 601 MK. 9 1 REINFDRCING PLAN 1 4 - 5 Go.. STIRRUPS 390.0 1573 13 GQ. TIES e 105 n n SPACING. 118 Appendix A — Drawings 119 Appendix A — Drawings 120 Appendix A — Drawings 121 Appendix B — Data Information and Technical Specifications APPENDIX B DATA INFORMATION AND TECHNICAL SPECIFICATIONS Appendix B — Data Information and Technical Specifications B-l Data Storage Information There are two main sets of data collected during the testing: Shake table test data and hammer impact hammer test data. The shake table test data are data collected during each testing run, which includes acceleration, displacement, and strain data. Hammer test data are input and output vibration signals recorded during hammer impact tests performed after each individual testing run. In addition, all testing runs were videotaped with a S-VHS camera, and a high-speed camera was used to record the development of plastic hinges at a speed of 1000Hz during some critical runs. Copies of these data files are available upon request to the Earthquake Engineering Research Facility, Department of Civil Engineering, UBC, 2324 Main Mall , Vancouver, B. C , Canada, V6R 2Y3. B - l . l Shake Table Data There are two sets of data from the shake table tests: (1) Labview This is the main data base, which contains the information on the acceleration and displacement of the structure and table, table power output, and strain gauge readings. (2) Global This is a secondary data acquisition set-up. The Global is a back-up for some of the information on Labview. In addition, it also contains information from the out of plane and vertical accelerometers on the structure. Table B . l lists the shake table data files for all testing runs, both calibrated and un-calibrated. Data column keys for each data file are listed in Table B.2, and Table B.3 presents channel setting and calibration information for Labview data files. 123 Appendix B — Data Information and Technical Specifications Table B . l Shake Table Data Files Run# Level / PGA Labview File (un-calibrated) Labview File (calibrated) Global File 1 5% / 0.035g Runl.dat Run01-5a.xls Runlgl.dat 2 5% / 0.035g Run2.dat Run02-5b.xls Run2gl.dat 3 10%/0.07g Run3.dat Run03-10.xls Run3gl.dat 4 20%/0.14g Run4.dat Run04-20.xls Run4gl.dat 5 20%/0.14g Run5.dat Run05-20.xls Run5gl.dat 6 40% / 0.28g Run6.dat Run06-40.xls Run6gl.dat 7 60% / 0.42g Run7.dat Run07-60.xls Run7gl.dat 8 80%/0.56g Run8.dat Run08-80.xls Run8gl.dat 9 10%/0.07g Run9.dat Run01-5a.xls Run9gl.dat 10 20%/0.14g Runl0.dat RunlO-ca.xls RunlOgl.dat 11 80% / 0.56g Runll.dat Runll-80.xls Runl lgl.dat 12 12'0%/0.84g Runl2.dat Runl2120.xls Runl2gl.dat 13 150%/ 1.05g Runl3.dat Runl3150.xls Runl3gl.dat 14 200%/ 1.40g Runl4.dat Runl4200.xls Runl4gl.dat 15 120% / 0.84g Runl5.dat Runl5120.xls Runl5gl.dat 16 150%/ 1.05g Runl6.dat Runl6150.xls Runl6gl.dat 17 200%/ 1.40g Runl7.dat Runl7200.xls Runl7gl.dat 124 Appendix B — Data Information and Technical Specifications Table B . 2 Data Column Keys and Calibration Factors Data Column Information (Labview) Channel Cali. factor, lv = Direction Information (Global) 1 Table Disp. 0 0.5 inch west + Bent Acc. 2 Bent Disp. 1 1.1875 inch west + Table Acc. (E/W) 3 Delta Pressure 2 Table Disp. (E/W) 4 Mass Disp. 3 0.8906 inch west + Table Acc. #1 (N/S) 5 Mass Acc. 4 lg east + Table Acc. #4 (N/S) 6 Plate Base Acc. 5 lg west + Table Disp. #1 (N/S) 7 Bent Acc. 6 lg east + Table Disp. #4 (N/S) 8 S4 Strain 7 -0.224 tension + Bent Disp. 9 S3 Strain 8 0.801v tension + 10 T5 Strain 9 0.982v3 tension + 11 T4 Strain 10 0.17v tension + 12 B4 Strain 11 -35.6mv tension + 13 T6 Strain 12 0.195v tension + 14 T3 Strain 13 0.055v tension + 15 B6 Strain 14 0.008v tension + 16 B3 Strain 16 5.399v tension + 17 T7 Strain 17 0.404v tension + 18 T2 Strain 18 5v tension + 19 B7 Strain 19 5.801v tension + 20 B2 Strain 20 tension + 21 T8 Strain 21 5.002v tension + 22 T l Strain 22 5v tension + 23 B8 Strain 23 4.998 tension + 24 C4 Strain 25 -3.012v tension + 25 C2 Strain 26 -2.383v tension + 26 C3 Strain 27 7.056v tension + 27 CI Strain 28 -4.817 tension + 28 Actuator #1 Acc. (N/S) 29 1.025g south + 29 Actuator #4 Acc. (N/S) 30 lg south + 30 Actuator #5 Acc. (W/E) 31 1.025g east + 125 Appendix B — Data Information and Technical Specifications B - l . l Hammer Test Data The storage information of Hammer impacting test data are listed in Table B. 3. Table B.3 Hammer Test Data Storage Information Test No Level (after) Input file name 1 initial test 1001-1005.bbb 2 10% bentl001-1005.bbb 3 20%a bnet2001-2005.bbb 4 20%b ben20b01-20b05.bbb 5 40% ben4001-4005 .bbb 6 60% ben6001-6005.bbb 7 80%a ben8001-8005 .bbb 8 80%b ben80b01-80b05.bbb 9 120%a benl2001-12005.bbb 10 150%a benl5001-15005.bbb 11 200%a ben20001-20005 .bbb 12 post-tension removal tnsrmvO 1-05.bbb 13 (transverse impact) post-tension removal trmvtrO 1-05. bbb 14 150%b bel50r01-05.bbb 15 200%b be200r01-05.bbb 16 (transverse impact) 200%b be200rt01-05.bbb 126 Appendix B — Data Information and Technical Specifications B-2 Technical Specifications B-2.1 Hammer Test Equipment Hammer Specifications: Model Range Sensitivity Maximum Input Stiffness Sensor DYTRAN/model 5803A 12 pound impulse hammer 50001bs (nominal range for +5 volts) l.Omv/lb 10,0001bs 1101b///in Resonant Frequency 75kHz (sensor with no impact cap) Sensor Specifications: Model Full Scale Range Output Range Dynamic Range Kinemetrics F B A -11 ±0.5g ± 2.5 volts 130 dB from 0 to 50 Hz 140 dB from 0 to 10 Hz Natural Frequency 50Hz (damping: 70% critical) B-2.2 Shake Table Data Recording Device Specifications See Appendix B-2 in Davey, 1996. B-2.3 Post-tensioning Equipment Specifications The specifications for post-tensioning equipment were provided by Dywidag Systems International Canada Ltd., and are attached in this appendix. 127 Appendix B — Data Information and Technical Specifications Rename: 25T1404.WK4 DYWIDAG SYSTEMS International Canada Ltd. In-house jack calibration record Master Gauge Reading Field Gauge Reading (psi) Load Cell Reading Load reading at load cell (kips) PSI run#1 run #2 run #3 Average run#1 run #2 run #3 Average SCO 500 500 500 500 44 44 44 44 3.7 1000 1000 1000 1000 1000 90 90 89 90 7.6 2000 2000 2000 2000 2000 179 180 180 180 15.3 3000 3000 .3000 3000 3000 271 273 273 272 23.2 4000 4000 4000 4000 4000 363 361 363 362 30.8 5000 5000 5000 5000 5000 453 453 453 453 38.5 6000 6000 6000 6000 6000 542 540 541 541 46.0 7000 7000 7000 7000 7000 633 631 632 632 '- 53.8 DYWIDAG SYSTEMS International Canada Ltd. Customer: UBC DSI job No: piston area: 7.796 sq in Stressing jack: 25tonC1414 <j-Field Gauge No: GA104 Calibrated by: CS* Date: 02-Oct-96 Master Gauge No. Enerpac Digital DGP1 Load Cell No.: SN C830 Strain gauge indicator calibrated by R S Technical June 12,1996 Project Field Load Load Gauge PSI Kips KN 500 3.7 17 1000 7.6 34 2000 15.3 68 3000 23.2 103 4000 30.8 & 5000 38.5 171 6000 46.0 205 7000 53.8 239 16 if' <? 128 Appendix B — Data Information and Technical Specifications 7000 6000 5000 4000 Pr'-ssii'r** —V<; Fnrrprl , 1 ; • • ' • . . . . I . . , • , - • M • ' ( . M : r ' i : "l r M M ' H44-4-I ' M • M l M M M M y-A i . -m-1 1 • . , i 1 • — H - ' ' M l : • i : M i l <. i i 11 ; ; 1 . ; ; : i — • . i i i i • l i i i • 1 : , .A, M ' 1 1 1 • > • '• i i ! J 1 L 1 • t I I • , 1 1 / ' ! : / / ' . 1 1 ! i . _ : — | 1 • : • i • M | • 1 . • • i / t . • ? • > ; J 1 1 • • i i . i i M / ' ; 1 . 1 — 1 . : ! ' ! —¥ ^ • : : ' ' i i . 1 i 1 i i 1 1 1 .. 1 M M 1 11 • • \ • •• . M • . J 1' —7~7 . : ' i ; i 1 i ! ' • • M l 1 i i '//• . . 1 M M : I 1 1 i • ! ' M • ' ; it • . . . , M M ' 1 i •• 1 • fr* : • 1 1 ( I 1 i i 1 i T — T , 1 M M ~ ~ ' 1 . . . 1 . . • 1 • • 1 1 M l . . . . . . . ! • ' | _ . i 1 1 • • ! ' ' 1 ' • ' • M M •—: 1 ! _ M • | . 1 • • 1 • 1 1 . . 1 . • : • 1 ' 1 . / / 1 • ' : i 1 : • • M l : ! I i • 1 . ' ' f 1 1 ' -; i • •  ; 1 : ; i / l i i i i 1 f • ' M l ' r M ;:: = l i ! ~ M . I , i i • , : | • i i : 1 • : • • . . M l : • • • M | M ! 1 ! ' | 1 ; i - . 1 - 1 • • ! , | , , . i 1 1 • . | - • , 1 ^ i i i i : • M | \ * . ; ; : 1 : : ; • ——'•—i—•—I :——— i1 • ! 1 ' i l l i i : \ i ; i • 1 : ! . • 1 ; . • : M M - 1 ! . | . i i • • 1 I . 1 ' ' M M ' :———1 :— : • : : i . . y > . • ' : M . • M | — U — • : • . | i ' M • . M I : • ! ' I ' : ? i • • : M . , . 1 . | ; • ' : 1 - • U : i . i i : . ; H ^ H — H 1 • • M "T • . i , • ' M „ • • ' ' ' • 1 ' : » MM I 1 1 1 1 i i • 1: / M : : •: \ ! 1 M M 1 ' • ( 1 1 • r : . \ i • • •  \ : . 1 ; M : < ! VA : 1 • ; M : . • • M . . . 1 : I 3000 ZOOO 3 « 1000 _2U.. s o : M i l ' - • M l • i D Y W I D A G - S Y S T E M S I N T E R N A T I O N A L C A N A D A L T D . #103-19433 - 961h Avenue Surrey. aC. V 4 N 4C4 Phone: (604) 888-8818 Fax: (604) 888-5008 25 Mp ( ' t o n s ' ) D S I j a c k : C T e n s i o n a r e a : 7 . 7 9 6 S q . i n . / - 5 0 . 3 cm Max. c a p a c i t y : 5 5 . 0 8 K i p s / 2 4 5 - KN F i e l d g a u g e : ^ yQ^_ D a t e c a l i b r a t e d : 1 k i p = 4 . 4 4 8 KN _I_L I . ' I • i 3iT .10- -15-"FO"r"ce:-i•n""K^f" - 2 0 — L -+--30-i • . pore 'e i n l i p s _ i M I I | | -40-_200-' 1 i .240--50-_ U _ i i / A O W { Q l l N7,OP^ ^ 129 Appendix C — Photographs APPENDIX C PHOTOGRAPHS 130 Appendix C — Photographs Photograph 1 Damaged Oak St. Bent S28 after Slow Cyclic Tests, as-built 45% scale Photograph 2 Damaged Oak St. Dent S28 after Shake Table Tests, as-built 27% scale 131 Appendix C — Photographs Appendix C — Photographs Appendix C — Photographs Appendix C — Photographs Appendix C — Photographs Appendix C — Photographs Appendix C — Photographs Appendix C — Photographs Appendix C — Photographs Appendix C — Photographs Appendix C — Photographs Appendix D — Data Samples APPENDIX D DATA SAMPLES 143 Appendix D — Data Samples D . l Hammer Test Data D . l . l Sensor Setting © i 2&3 4&5 - hammer - Longitudinal Sensor - Transverse Sensor @ W T D.1.2 Datasheet Attenuation: 54, for channel # 1 to 5 Filter: 50, for channel # 1 to 5 Activation: h avda 500 32 5 2 m 0 Test No. Level (after) Input file name 1*' Natural Frequency 1 initial testl001-05.bbb 6.85 Hz 2 10% bentl001-05.bbb 6.81Hz 3 20%a bnet2001-05.bbb 6.56 Hz 4 20%b ben20b01-05.bbb 6.56 Hz 5 40% ben4001-05.bbb 6.26 Hz 6 60% ben6001-05.bbb 5.80 Hz 7 80%a ben8001-05.bbb 5.77 Hz 8 80%b ben80b01-05.bbb 5.68 Hz 9 120%a benl2001-05.bbb 4.94 Hz 10 150%a benl5001-05.bbb 4.61Hz 11 200%a ben20001-05.bbb 4.36 Hz 12 p-t removal tnsrmv01-05.bbb 4.18 Hz 13 p-t removal trmvtr01-05.bbb 4.20 Hz 14 150%b bel50r01-05.bbb 3.94 Hz 15 200%b be200r01-05.bbb 3.88 Hz 16 200%b be200rt01-05.bbb 3.90 Hz 144 Appendix D — Data Samples D.2 Shake Table Data D.2.1 Sample Data Sheet (Calibrated 60% level run) N N s s n c o c o s r t s s o o a s s C ; T - ( D ' - < D i r ) r N < N ( D O ( D > - M ( N * rat*- ^ M ^ S N Q O ) T - T r N ( 0 ( D ^ O ' l t f i o i o i n r f o g o J s N w i n O J I ^ Q Q Q Q O f N i e O O O r ^ i S O Q r N Q o o o o o o o o o o o o o o o o o P P d H d P d P P r i r S P P H H P r S £ H « <-» u i d p o o O O O 1 9 9 ) TT •<*• eo i s 5 o (0 o o p to o 5 P o o 5 c o 5 o O T - e O T - i y T - o i e g e o - •• <p CN co y5 cp S S CO CN T- o i to N M ( 0 < - O O Q t O O w v . , . O O O O O O O ' - O O O g § o CO CN CN u w w w w w w - w O O Q w o o o o ~ ~ ~ 1 ~ ~ o o o i ( D ( p ( D i - i - ( D ' - ( D ( D ( 0 ( p ( p ^ - ( Q O : - ^ ? 0 ) r - - ^ - i - T - ^ O O O r - CN CN CN o o o o d o o o S S C O t O t O S S t O N S t O C O C O t O » - M g S i - < p i i ) r i - i r ) r N N N . . M i ^ a > « C H ! ^ ? c O W C ^ » e o C N g C N | < N T- (O N CM S N L . _ . » » • ' W W W » ' ' l_( l_> l_P «_« W W l_« WW C O C O C N C N C N C N C N C N C N C N O O C O C O C O C N C N C ^ S K n 5 c o 5 5 5 5 c o c o 5 o Q Q 5 e o c o 5 5 c » c o r C N ^ e o ^ t o i n ^ c p t t t o t o c o ^ r r ^ s s s s 8 i (N N N . . . o Q e O Q Q t t O P l o o o o o o o o o o o o o o o o — o o o o o o d d d d d o b d d ^ ^ ^ ^ P d d d d . . . ~ ~ ~ o o o o o o o o ~ ~ ~ ~ o o o s $ r ^  s § | § | 8 . 8 8 " d o tr\ rr\ tr\ . ..s CO r -O o 8 5 r S 0 0 9 0 0 ' ! p p p p c o ^ - r ^ r ^ O ) i n i n c N ( N t o i r > ( (D CO CO O) ) U ) t n i n u i ' i A i i ) U ) i n - » u ) i n i n i n u ) u i ' i r > i eo CN m ot in - v i n I D i n 10 > in in in m in • 8 8 8 8 8 9 9 P 9 9 8 8 CM 1Q N N » • O) (D IT) W U> W tf) ^  ^ in in in in in in in ( O r t i f l t O T M n N a K O t O C M M n t N T - ^ C O -C CO N M ' i C O t N M S S t O » - i - O i • CO O T- y- <D CN CO T T c o c N o ? ' « - g > q ) r ^ a > ^ L n < x > r ^ u o o ^ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 9 W 9 9 9 g ^ o o o p o o p p p s r - c N 8 U J U I 2 3 8 8 2 2 2 8 8 8 < D C ^ ^ ^ ^ ^ m ^ m ^ t e i C 4 t o < ^ ^ o c o c 4 t ^ c o c o ^ ^ m r ^ m ^ t o ^ c > c ^ y - t n ^ m ^ c o t e j ^ O T - p Q P p Q p t - C N ^ - i - Q O p r r C O r r C O Q t « « - Q O P Q C N r ( N ^ O Q Q r ! : 5 ~ ~ 8 8 w § 2 2 8 K 8 8 8 8 § § K S 8 8 8 8 8 w 8 § g K 8 8 8 8 8 § 8 S S 8 8 o o o d ^ ° ^ ^ d - ? d d d d ° ° - ? d d d d d o 1 1 ? © 0 0 5 ? © © d d d ^ d T d d d i n u ) 0 ) ( \ ^ ^ « ) ' r i r j ' - < o i r j S 3 t 2 ( o < 0 3 c O ' - 3 S 0 0 0 0 , ; , ,T, 0 0 0 0 0 0 0 0 0 0 ~ O O O O W U J O O O O O O Q O O Q -C T-<b o IA O ,r d d d 0 0 * CD T 1- O ( N C N O T - c n ' J L D ^ O ' - C N ^ ' T N r t " — Q Q Q O o o o r o o o g , , ' , Q O O ' o o o o o o . _ . . . . . . ) LO O CO ' t O O g . / . Q O d O O Q Q O Q O , 0 0 0 ^ 0 0 0 0 0 0 0 ^ 0 , - , o o d d d • o o d d o d d S CD T- S 5 C N ( CI Tt CO CO O 10 1 o o o < 8 8 8 d d d o • o o o o 0 0 0 ^ d d ^ o o d d d V d d d d o d d d d o o d d d d d d d d C O < N h - C N C O r C ^ C N C O C O T r c O C O r - - - U l C O r ^ ^ ( D c o c o ^ o c o r o ^ ^ s i - 0 ( 0 ( o n o ) n « - 0 ( O c o w o i / ) « - f f l « - i - T - ( D s v i n s s o ^ O i r ) N i n O ( N < D ( N O O Q n r N r O ( N Q Q r i n ( 3 r r l N O ( N O O ( O r ^ Q t N ( 0 0 O O p O O O O O O O O O O O O O O O O O d Q p O O O O O O O O O Q O Q O O O O O Q O O O Q O O O Q O d O O 0 O O O O O 0 O O O O O O 0 0 O Q O 0 0 0 0 O Q O O O O O O 0 O 0 O O 0 O O 0 O O 0 O 0 O W I CN § 5 S S § 8 D ^ C D C D C N C N C O C O C O T J - C O r c n o o x o o i w - r r r T ) ( D r r l M ( N I O O O O O 3 0 0 0 0 0 0 0 0 0 0 J O O O O O O O O — «"» j d o ' d d d d d d J O O O O O O O O O O O O O O O O O O O O O O W W W W W ( J ( J O O O O O O O Q 0 0 0 0 0 0 0 0 0 0 0 0 0 d o d d o o d d d d o d d o o o o o ' 0 9 0 0 0 ' d ' d ' d d d d d d d d • O r r r 1 ( N r ( N r ( O O O O r C D ; § 8 8 8 8 8 8 8 8 8 § 8 8 8 8 T o n r i r i n n n n n T o o d d ; OI Oi 0) 3 in in in ^ CN CN CN CN CO J^" V CO O CN w o r - 1-5 0 0 0 -O o o o « d d o" CO CN CN CN o> o o o in ( N S N S t O f C O in CM co g 0 0 0 0 0 0 0 0 0 d d d CN CN CN 0 0 0 ^j- TT N- r-~- r-~-O) CN O S N CN CN i -co cO CO CO cO CN Oi Oi Oi 0> 0> O in in in in in CN CN CN (D N T CO CN T-p p p 1 8 tN i n m 0 0 0 £ £ W UJ S0 1 T- N N (\| OI O o o in O O f . C N C O C O c O C O c O C N t N C N C N C N C N C O C O C O C O f N t N C N i n * i n i n i n i n i n T r ^ ^ n ^ ^ i n i n u i i n ^ ^ ^ . T i i C N ^ C N C N | ( N N N N N S N ( N C N ( N t N N N N L y n n ( N © n c n N O ) i - T - ^ i o w r t N N c r i c » £ p i g 2 8 S 2 8 & 8 S ? $ £ 2 8 8 & i 8 S & ; 8 8 8 8 8 8 8 S 8 8 8 8 8 8 8 8 8 8 « i d d d d d p ' d o ' o ' o ' o d d d d d d d * - < 3 eo CO t ) in in 1 i) CN CN I ) r IO I 5 CO CO CN 9 0 ) 0 ) 0 ) in i n T S i > co eo to cp CN CN 1 ) o) S 35 5 o o ) i n m i n m Tf J CO CO (N O) ^ N I ? 8 8 S 8 S § i > d 0 0 0 d 0 1 3 8 8 8 8 ? 9 9" ° ' ° M S S N C O C O C O S N C O C O N N N N C O C O S S C O S N N N N C O C O i n i n S C O N C O S S C O C O C O C O N S N C O S N N S C O C O S ^ i n c o t o t D i n i n t n c D C D i n i n c o t o t o c D i n i n c o t o i n o c o c o c o c o i n i n o o c o u ^ — — — 3 O V C 0 5 a c 0 < 0 S r 5 - t 0 N N N r C N i - r ( 0 { 0 r N C 0 O ( N O C 3 ) y j S 3 C 0 ^ T - r s n 5 V T - ( o i n n o r i - o ) i f t i n s f « i v o N t o N O i - ^ ^ 2 n r t i n ^ . _ ™ p C 0 3 t N Q O O « O p O ^ ^ < 0 O O O O ^ Q O Q T - ( D T - O O C 0 C 0 i n O Q P O O O O O O O O O O O O O O O O O O Q O O O O O W W o O O O O O O O O P Q Q S d . 9 ° . d d d c ? P d o ' , P , 0 . d d 0 0 ' " " g ' 0 0 0 " " ' 0 0 " ' 0 0 0 0 " ' ( o o 1 o o o d d d d d 9 9 « o o 0 0 0 0 0 0 0 0 0 0 0 o o o o ^ CO CN : O CN : TT r -: *r r -w - ^ CO < m m CN co r*- i n in in to CN co m m •«- 1- CN co r*. co co — ' CN CN » CN -• CM to m m * 1 in cn . CN CO o _ o co to to h-CO CO T CO 0 0 0 0 d d o" d S m m CD Oi O T-co O O 0 0 0 d d d to CN co r-- CD co CN in co CN in 3 - 0 0 co CD cn . CN O) CN 1- Oi in T i n 5 co in in to CO CO O) o S CJ CN CO 0 0 0 i d d d o" o" _ 3 h- O) O) CO to o o m co to ' CN 5 ^ ^ [ T CO -<T < moscocnScoS'TCD r V ( T r C D O i n O O ) C O O h~ 10 o " ' — — O O 1- o g CO CO g J g> o o o o o o o o t 0 0 0 0 9 0 0 0 0 0 ' 9 0 0 0 0 d o c o i - c N c o g j ^ i n ' i T - r o i i o c o c N f N - t N S ^ t O Q C a O l t J J S S i n - j i n i n o N c N K c o t D i n ^ c o c o i n ^ r a i n ^ f N t o ' T c p i o a i l 8 a § 8 S S 8 8 8 8 8 8 5 8 8 8 ? d d d d d d d d d d d d o o o o o o C O C N C O ' - ^ T l L n O ) ( O T - t - c O ( 0 0 ) S V N C O L O < O N C O ' - t - f N T - 0 ) C O r - N V t N C O O l ' - I N ' - C O i n N T - T O C O L O T - T - T j L n C N C D ' - T - N ^ C O N C O O I O N S C O _ « - r O N i - O r ^ O ! N I N N T - C S ( N O ' - T - ^ ( N W O ^ N T - ^ r ^ ^ ^ ^ N T - ( N ^ . - ( N . 0 0 o 0 0 , S 0 0 C > 0 0 0 0 0 0 o ' O O O O O r-j O O O O O O O O O O O O O O O O cocotncocginscoin* C O C N O > C N C S t O C O C O * d ' d d o o o o o o o o o o d d o o o o o o o o o d d d d d d o d d d d d d d d d d d o d d o d d d d d d d d d d d d d d d d d ; o o o v T~ CO CO T-N N r CO : O) i n CN CO • O! CO 1- N ' V O CD I V N O Q O r C O O 1 . ~ p d i o o o ~ 0 0 0 0 0 0 0 0 0 O O O O O CO T- CO CO CO CO CN T- o) 1 CD in o i r-tn 0 m a T- o i co to co oi tn in o to M O CM (N CO IO ID CN O CO T- O O O O O O T- H O 0 0 0 0 r C O C O C O i - C O C O C O * -N 0 ' - 0 1 N ' - 1 - 0 ) . . _ . . . . . . N O C O ^ ^ C t ) O S C O C O t N r N i - C O O ) 0 ( 0 0 ) C O Q 3 l O ' T ( O C N . < o i o ( N u ) g ) f i < o i - N ^ N ( o c » o i n o ) o u > c Q c 5 d i - ^ ^ T f t N c O f N i i r i c o N m c N c o n c o i n ' - C N i n r t o ^ a c o ' J O ' i U ) S C D < 0 T - ( 0 i n C N t N Q ^ N ^ < 0 S i n Q r N f f l ( 0 t 0 F i « - O N U l i n r O r O r O O N Q r C N r d O O O O C j f i r C O r i T - o o o O p - q j O Q - Q - o o o o o O O O O O O O O O O 9 9 9 9 9 0 0 0 0 0 0 0 0 0 0 0 0 0 8 d O r - d d 0 0 0 0 0 0 0 0 ^ 2 S w S S S ^ J f t ^ , . . ; ^ t t a i n ^ C N C O r ^ u 5 c 5 t 5 ^ u i c d S I T J T - T - C O C D C O C O C O C O i n O ) T - T - T - ^ ( N l O l D l O C O N C J ) S C O i n • C 0 l f ) C N r t N 3 3 a 5 O ( 0 0 ) ( 0 C 0 t 5 C 0 O Q Q r r C ) ) r i - i n - c o i ^ c g i n c o c o o c o c o c o c o c N t o c o c o C N c o r t i n r - c O O ( O i n ^ ( N r N O ( D Q C O N C D t O ( O C N S ( O i n T - C D C N O ' - g ( N f 1 J t N C O ' S o c t 5 K ^ o o t - o o o o c o o g > ^ o N o ^ o o c » c N i n ^ S T - o o O ' r- Oi tO O CD CO : o  o   o o 1 9 9 9 J o o o o d d 0 0 o o o o o o ' i n c o c o c o u o c o i n c D ^ c o c o i p c o ( o i n u o < - « c o f O O ) c o Q T - i - c D L O ^ o c M M ^ ' r o c o ^ i n i n a c O ^ o j c o o ) C 0 l O C O « - O ) i - ^ ( O ( O i - i - g > ^ i - f N f N C N C O ( D C 0 C O C D T - C D N S ^ C O O ) l 7 ) C O n c O ^ N C O C v C O C N 3 Q M O O > _ V O O C O r O C O O O O O ^ C O O ® S r O O N O l O o o o ^ r i 0 ^ r i N ' o ' d d r 9 d ° ° d o ' d P b r 1 0 0 O O ' ' o • o f r t t N O o 5 s N c o ( o n c » s o n t n t N c » c o g ) T f < i t 3 E t o c o n ^ o i n u i i n i n n o Q 3 t o c o c o o ) O o r , o i o 3 Q « ) ^ Y O O O O ( D r r O O C 0 r C 0 O 0 ) N f i r ( D r ( \ | ( 0 T f O 8 C N N C 0 N O , 2 r - S { N t N C N ^ ^ ^ T - r - g C D N r t O C O O ) C N O N O N C N n 7 5 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 o o o o o _ co 1- in CN i n 1 •Q- O O r O N I « P d P d P d 1 W - O c -c o i n ^ i n i n i n c D i O N i n c o i o o i i n T - i n r - i n c N i n c o i n < r i n i ^ O C 0 O 5 O l D O C D O N O C 0 O 5 r i O r i - i - C ^ ^ C 0 r ^ i - i n r ( 0 r N r C 0 r f f i ' O ( N r N ^ P d O d . d ° d ° o , d . r d r d . d r d r d r d r d r d r d " . d 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3847 1437 CN O m 0 CN CN O 9 9 9 CO m CO Oi d d d i c o c O c o c o c o c o c o s s c o c o c o c o c o c o r o N N C f l t f S N 3 0 ) N i r ) 0 ) 0 i c o m o o m o N c o c o c o I t N i - ^ f f i N r i n f i f i f f l r - r - h- h- CO CO c o i n ^ c > > c ^ § c § t S S J t o C N c o o ^ r m c o t o t o t o o 145 Appendix D — Data Samples D.2.2 Displacement and Acceleration Time History Plots (samples from 10%c, 80%b, 150%a, and 200%b run) Note: longitudinal direction of the specimen is W-E; all displacements shown in plots are in W-E direction, West "+"; all accelerations shown in plots are in W-E direction unless specified (1) 10%cRun Time, second 146 Appendix D — Data Samples 0.5 < o -0.5 A f— H 1 1 1-1 1 1 1 1 1 1 1 1 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 0.5 K 0 -0.5 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Time, second 147 Appendix D — Data Samples (2) 80%bRun Time, second 148 Appendix D — Data Samples Time, second 149 Appendix D — Data Samples (3) 150%aRun 50 -50 -I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 50 -i 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Time, second 150 Appendix D — Data Samples <o Time, second 151 Appendix D — Data Samples (4) 200%bRun 152 Appendix D — Data Samples Appendix D — Data Samples D.2.3 Max Values form Tests east + actuator5 (g) | 0.046903395| -0.046945605| 0.0504478681 -0.041832882| 0.1654088871 | -0.1802416131 0.34637605| -0.320529951 0.3069297731 -0.2842697271 | 0.437683256| | -0.374977744| i 0.463604467| | -0.448532783| | 0.392038792| | -0.2958182081 | 0.064690248| | -0.069185002| | 0.1324834911 | -0.135595009| | 0.607213403| | -0.500432347| | 0.906877304| | -0.889875946I \ 1.574463875| | -1.3036643751 | 2.508465362| | -2.617467388| | 0.974029043| | -0.756406707| | 1.273768781| [ -1.241806719| | 1.730000497| | -2.676484753| south + at •G o « 0.01003062| -0.009189381 0.009783171 -0.01097683| 0.020475095| -0.0176649051 0.05004337| -0.04486663| 0.068666469| -0.0607335311 0.099369017| -0.0944209831 0.125048124| -0.101091876| [ 0.069759559| ] -0.0797704411 0.014063721| -0.012786279| 0.045648365| | -0.038271635| 1 0.084328375| | -0.077111625| | 0.094506981| | -0.103553019| ] 0.121149067] | -0.104380933| | 0.181227894] | -0.127002106 | 0.066762963 | -0.054697037 f 0.082978938 | -0.085171063 | 0.122413088 | -0.098836913 south + actuatorl (g) | 0.009586611 -0.00980639| 0.0105055921 -0.010455658| 0.016640911| -0.015267339| 0.049265222| -0.058646778| | 0.058817656| -0.058801094| 0.106373657| -0.130739593| 0.093406636| | -0.124610864| | 0.090486507| | -0.100327493] | 0.013858997| | -0.012729503| | 0.034665212| | -0.038222538| | 0.123338137| | -0.097815863| | 0.095263356| | -0.125265394| | 0.205706849] | -0.217208151| | 0.160529414| | -0.200762586] | 0.09690514] | -0.10391286 | 0.097320432] | -0.090992568 | 0.159364719 | -0.146249281 betal(g) | 0.0692471 -0.06076| 0.0597311 -0.05929| 0.2840511 -0.28633| 0.6093931 | -0.62108| 0.5874821 -0.55357| 0.886834| -0.85176| 1.067335| -1.03411| | 1.455344| j -1.43681| | 0.173357| | -0.21513| | 0.4462681 | -0.44393| |T. 563016] | -1.67794| | 1.69663| | -1.69998) r 1.72072| | -1.68199] | 1.388969] | -1.51051] [ 1.038718] | -0.93821 | 1.170304 | -1.16064] | 1.371663 | -1.29283 plate base a(g) | 0.07304924| -0.06184076| 0.061700893| -0.062199107| 0.232845517| -0.227964483| 0.4737043631 -0.4369356371 0.473593059| -0.443156941| 0.564050632| -0.5596093681 0.649432598| -0.675337402| 0.038324483| -0.045295517| 0.0549637511 I -0.0460462491 | 0.085941444| | -0.080378556| | 0.461176108| | -0.344483892| | 0.5206338821 | -0.349116118| | 1.044799168| | -0.985840832| | 1.417220742| | -1.754769258] | 0.890033688] | -0.648366313] | 1.048978988 | -0.811071013 | 1.388685013 | -1.450674988 mass alu(g) | 0.1068377| -0.0973323| 0.1068898| -0.0984902| 0.5079397| -0.5214203| 1.0756485| -1.0611915| 0.9474648| -0.9391352| 1.5505688| -1.4401512| 1.8419447| | -1.6333953| | 1.6383143| | -1.5193457| | 0.2354101| | -0.2269299| | 0.4374123| | -0.5006977| | 1.8725844| | -1.5450756| | 2.1592393| | -1.7250407| | 2.0182299] | -1.7442801] | 1.9448919] | -1.7431581 | 1.3291124 | -1.2230776 | 1.5015264 | -1.4687536 | 1.7480158 | -1.4834942 west + | mass dis(mm) | 0.772690861 -0.884361484| 0.872471944| -0.832538775| 3.409701509| -2.658616459| 6.713345225| -8.005351525| 7.530416718| | -9.204115313| | 12.88384441| j -14.31669809| j 18.73852225| | -14.09531956) | 19.19278652| | -13.97246576| | 3.310132586| | -3.317398133| | 6.581898744| | -6.831968256| | 23.28624339J | -28.1805582| | 32.73641981| | -41.88522816 | 46.63670772] | -50.36634471 | 62.01123691 | -59.38249494 | 1.844527438 | -1.797752563 | 2.05562155 | -2.32639845 | 2.50625695 | -3.04885305 west + | rel dis (mm) | 0.57489269] -0.69151956| 0.77337827| -0.51433823| 2.37270933| -1.99434442| 5.76614258| -5.75369405| 7.39011796| -5.639034291 9.36641445| -10.5227222| 13.6273148| -12.1768151| 13.1415676| | -12.5323873| | 2.03560152| | -1.63179335| | 3.78160609| | -3.51638541| | 13.6626176| | -13.5989801| | 17.64116461 | -22.76453631 | 25.386156| | -32.1442725] | 37.3848783] | -39.290197] | 27.3379704] | -27.6370594 | 30.0098267 | -36.423794] | 35.9805064 -46.6593144 bnt dis(mm) 1 0.799118941 -0.949702811 1.01008676| -0.98727399| 3.002798491 -2.15197276| 5.665273231 -6.7523264| 6.575880011 -8.01402287| 11.2598234| -12.3233305| 16.3520382| -11.6122188| 1 16.97958911 1 -12.99168061 | 3.00449994| | -2.997234311 | 5.87968231| i -6.25227844| 1 21.24370051 1 -26.29028821 | 31.6008086| | -39.9024139] | 46.6264178] | -48.5703534] | 61.2185073] | -57.9200498] | 40.4342873 | -39.693902] | 45.2069861 | -51.3060766i | 56.1072057 -67.6154482 tbl dls(mm) 1 0.694618371 -0.7355286I 0.67100087| -0.73209511 0.835575161 -0.9976698] 1.677845021 -1.9110481 3.211979551 -3.5900135| 3.04012371J 1 -3.57187731 4.496897271 I -5.4055467I j 6.977334521 1 -9.1107715] i 1.61609041| I -2.2130866] 1 3.4394661J | -4.2228249| 1 14.4527512| 1 -17.056965| I 23.1933854] | -26.361761| | 30.2363861| | -33.670522| | 39.6518759| I -45.156946] I 23.8297855| I -27.248472] | 29.6323604| | -33.883134| | 39.561557| -45.150364 max 1 min I max 1 min 1 max 1 min 1 max 1 min 1 max 1 min 1 max I min 1 Imax 1 | UILU max 1 min I max 1 Imax |min J |max |min | max 1 u!iu| Imax UILU | Imax Imin Imax Imin Imax |min Imax min Irun 1 |5%a 1 |5%b 1 Il0%a| |20%a 1 |20%b 1 O o CO |80%a| 10%c |20%c |80%b 120% 1150% |200% 1120% 1150% 1200% 154 Appendix D — Data Samples D.2.4 Shake F R F Files Run Excitation Signal File (table) Response Signal File (bent) 10%a 10at.dat 10ab.dat 20%a 20at.dat 20ab.dat 20%b 20bt.dat 20bb.dat 40% 40t.dat 40b.dat 60% 60t.dat 60b.dat 80%a 80at.dat 80ab.dat 80%b 80bt.dat 80bb.dat 120%a 120at.dat 120ab.dat 150%a 150at.dat 150ab.dat 200%a 200at.dat 200ab.dat 120%b 120bt.dat 120bb.dat 150%b 150bt.dat 150bb.dat 200%b 200bt.dat 200bb.dat 155 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0050184/manifest

Comment

Related Items