@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix dc: . @prefix skos: . vivo:departmentOrSchool "Applied Science, Faculty of"@en, "Civil Engineering, Department of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Devine, Frank Joseph"@en ; dcterms:issued "2009-08-31T18:42:03Z"@en, "2009"@en ; vivo:relatedDegree "Master of Applied Science - MASc"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description """Within a building constructed mostly of solid tilt-up panels, the strength and inelastic response of wall systems is almost entirely controlled by the connections since solid wall panels are inherently stiff and strong. Currently system design is done without designing for either a sliding or overturning mechanism to form, and connection design to resist each of these mechanisms are done independently of one another. Designers add connections between two or more walls panels to increase the rocking resistance of the system, and add connections between wall panels and the base slab to provided horizontal sliding resistance without considering how each of these remedies may influence the other. Since vertical uplift resistance of the panel – slab connectors is ignored in design, any vertical load capacity they have will further increase the shear load transferred through the panel – panel connectors, causing them fail earlier than expected. Also if rocking does occur the vertical damage may decrease the shear carrying capacity of the panel – slab connectors. The inelastic response of the panel – slab connector and the interactions between vertical damage and horizontal response properties are key elements to the overall system response and therefore must be determined. This study seeks to experimentally determine the detailed response of the connectors. This was done by first obtaining the vertical uplift response since previous testing conducted on tilt-up connectors only focused on the horizontal shear response as well as the horizontal “tension” pullout. Next a series of test were conducted to determine the horizontal shear response at various levels of vertical damage ranging from no vertical damage to damage levels approaching failure. This allowed for detailed interaction relationships to be determined between the vertical damage level and the horizontal shear response properties, including horizontal strength, stiffness and displacement capacity. This connector response data was then used in a pushover model to predict multi-panel system responses for solid panel systems using connection configurations commonly used in industry. These responses are studied to compare the predicted levels of seismic ductility and strength achieved using the pushover model to values designers are expecting to achieve using current design techniques. It also allows for the influence the system connection configurations have on the response properties to be studied. Although further research is required which includes panels with openings, and out of plane influences, the information gained in this study is used to suggest ways to improve the efficiency of the system using current connectors, as well as suggest properties which new connectors could posses to create a more ideal system response."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/12626?expand=metadata"@en ; dcterms:extent "19223433 bytes"@en ; dc:format "application/pdf"@en ; skos:note " INVESTIGATION OF THE IN PLANE SEISMIC RESPONSE OF SOLID TILT-UP WALL SYSTEMS by FRANK JOSEPH DEVINE B.Sc. Civil Engineering, University of Alberta, 2006 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES (Civil Engineering) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) August 2009 © Frank Joseph Devine, 2009 ii ABSTRACT Within a building constructed mostly of solid tilt-up panels, the strength and inelastic response of wall systems is almost entirely controlled by the connections since solid wall panels are inherently stiff and strong. Currently system design is done without designing for either a sliding or overturning mechanism to form, and connection design to resist each of these mechanisms are done independently of one another. Designers add connections between two or more walls panels to increase the rocking resistance of the system, and add connections between wall panels and the base slab to provided horizontal sliding resistance without considering how each of these remedies may influence the other. Since vertical uplift resistance of the panel – slab connectors is ignored in design, any vertical load capacity they have will further increase the shear load transferred through the panel – panel connectors, causing them fail earlier than expected. Also if rocking does occur the vertical damage may decrease the shear carrying capacity of the panel – slab connectors. The inelastic response of the panel – slab connector and the interactions between vertical damage and horizontal response properties are key elements to the overall system response and therefore must be determined. This study seeks to experimentally determine the detailed response of the connectors. This was done by first obtaining the vertical uplift response since previous testing conducted on tilt-up connectors only focused on the horizontal shear response as well as the horizontal “tension” pullout. Next a series of test were conducted to determine the horizontal shear response at various levels of vertical damage ranging from no vertical damage to damage levels approaching failure. This allowed for detailed interaction relationships to be determined between the vertical damage level and the horizontal shear response properties, including horizontal strength, stiffness and displacement capacity. This connector response data was then used in a pushover model to predict multi-panel system responses for solid panel systems using connection configurations commonly used in industry. These responses are studied to compare the predicted levels of seismic ductility and strength achieved using the pushover model to values designers are expecting to achieve using current iii design techniques. It also allows for the influence the system connection configurations have on the response properties to be studied. Although further research is required which includes panels with openings, and out of plane influences, the information gained in this study is used to suggest ways to improve the efficiency of the system using current connectors, as well as suggest properties which new connectors could posses to create a more ideal system response. iv TABLE OF CONTENTS ABSTRACT ................................................................................................................................ ii  TABLE OF CONTENTS .............................................................................................................. iv  LIST OF TABLES ....................................................................................................................... vii  LIST OF FIGURES .................................................................................................................... viii  ACKNOWLEDGMENTS ............................................................................................................. xv  1  INTRODUCTION ................................................................................................................ 1  1.1  BACKGROUND ......................................................................................................................... 1  1.2  RESEARCH METHODOLOGY ........................................................................................................ 4  1.3  RESEARCH OBJECTIVES AND SCOPE .............................................................................................. 6  1.4  THESIS OVERVIEW .................................................................................................................... 7  2  HISTORY, CURRENT PRACTICE AND PREVIOUS TESTING ................................................... 8  2.1  HISTORY OF TILT‐UP ................................................................................................................. 8  2.2  CURRENT CONSTRUCTION TECHNIQUES ........................................................................................ 9  2.2.1  Diaphragms ................................................................................................................. 9  2.2.2  Wall Panels ................................................................................................................ 13  2.2.3  Current Connectors ................................................................................................... 14  2.3  PREVIOUS TESTING ................................................................................................................. 19  2.4  TYPICAL BUILDING PROPERTIES ................................................................................................. 27  3  EXPERIMENTAL PROGRAM ............................................................................................ 30  3.1  TESTING METHODOLOGY ......................................................................................................... 30  3.2  TEST APPARATUS ................................................................................................................... 30  3.2.1  Loading Frame .......................................................................................................... 31  3.2.2  Vertical Loading ........................................................................................................ 34  3.2.3  Horizontal Loading .................................................................................................... 36  3.3  TEST SPECIMENS .................................................................................................................... 38  v 3.4  MATERIAL PROPERTIES ............................................................................................................ 44  3.4.1  Concrete Properties ................................................................................................... 44  3.4.2  Reinforcing Steel Properties ...................................................................................... 49  3.5  INSTRUMENTATION ................................................................................................................. 52  3.6  LOADING PROTOCOL ............................................................................................................... 57  4  EXPERIMENTAL RESULTS AND DISCUSSION .................................................................... 61  4.1  SUMMARY OF RESULTS ............................................................................................................ 61  4.2  VERTICAL RESPONSE ............................................................................................................... 63  4.2.1  Monotonic Vertical Response ................................................................................... 63  4.2.2  Cyclic Vertical Response ............................................................................................ 64  4.2.3  Redesigned Connection Detail .................................................................................. 69  4.2.4  Vertical Response Trends .......................................................................................... 73  4.3  HORIZONTAL RESPONSE .......................................................................................................... 76  4.3.1  Monotonic Horizontal Response ............................................................................... 76  4.3.2  Cyclic Horizontal Response ........................................................................................ 78  4.3.3  Out‐of‐plane Movement During Testing ................................................................... 80  4.3.4  Comparison of Measured Response with Previous Testing ...................................... 82  4.4  COMBINED HORIZONTAL AND VERTICAL RESPONSE ....................................................................... 83  4.4.1  Interaction Models .................................................................................................... 84  5  ANALYTICAL STUDY OF SYSTEM RESPONSE .................................................................... 91  5.1  PUSHOVER MODEL ................................................................................................................. 91  5.1.1  Model Assumptions ................................................................................................... 91  5.1.2  Implementation of the Model ................................................................................... 93  5.2  PUSHOVER ANALYSIS WITH EXISTING CONNECTORS ...................................................................... 95  5.2.1  One Panel Pushover Results ...................................................................................... 96  5.2.2  Two Panel Pushover Results .................................................................................... 101  5.2.3  Multi ‐ Panel Pushover Results ................................................................................ 107  5.3  TILT‐UP SYSTEM STRENGTH RESULTS AND DISCUSSION ................................................................ 115  vi 5.4  TILT‐UP SYSTEM DUCTILITY DISCUSSION ................................................................................... 122  5.5  PERFORMANCE OF IDEAL CONNECTORS .................................................................................... 125  6  CONCLUSIONS AND RECOMMENDATIONS ................................................................... 131  6.1  SUMMARY OF OBSERVATIONS ................................................................................................ 131  6.2  RECOMMENDATIONS ............................................................................................................ 135  6.2.1  Recommendations for Current Design .................................................................... 135  6.2.2  Recommendations for Future Research .................................................................. 136  REFERENCES ........................................................................................................................ 138  APPENDIX A: EXPERIMENTAL PROGRAM TEST RESULTS ...................................................... 140  APPENDIX B: PUSHOVER ANALYSIS RESULTS ....................................................................... 216      vii LIST OF TABLES TABLE 2.1: CONNECTION TENSION TEST RESULTS. ..................................................................................... 20  TABLE 2.2: CONNECTION SHEAR TEST RESULTS. ........................................................................................ 22  TABLE 2.3: THE NUMBER OF CONNECTORS USED AND THEIR CONFIGURATIONS WITHIN A TYPICAL TILT‐UP BUILDING. ............................................................................................................................................... 27  TABLE 2.4: NUMBER OF EACH PANEL ‐ SLAB CONNECTOR TYPE USED IN A TYPICAL TILT‐UP BUILDING. .................. 28  TABLE 2.5: THE NUMBER OF PANELS AND THEIR WIDTHS USED IN A TYPICAL TILT‐UP BUILDING. .......................... 29  TABLE 3.1: CONCRETE CYLINDER COMPRESSION TEST RESULTS FROM 133 DAYS. ............................................ 45  TABLE 3.2: RESULTS OF SPLIT CYLINDER TESTS. ......................................................................................... 49  TABLE 3.3: MECHANICAL PROPERTIES OF 20M REINFORCING BARS FROM CSA G30.18. ................................ 51  TABLE 3.4: REINFORCING STEEL TENSION TEST RESULTS FOR 20M REINFORCING BAR. ..................................... 51  TABLE 4‐1: SUMMARY OF TESTING RESULTS. ........................................................................................... 62  TABLE 4‐2: VERTICAL DISPLACEMENT CAPACITIES. .................................................................................... 74  TABLE 5.1: SYSTEM PROPERTIES OF EACH PUSHOVER MODEL AND ITS NAME, MAXIMUM LATERAL FORCE ACHIEVED,  AND RESIDUAL STRENGTH. ............................................................................................................ 95  TABLE 5.2: LATERAL CAPACITY AND MECHANISM FOUND BY PUSHOVER MODEL AND THE NOMINAL SLIDING AND  OVERTURNING CAPACITIES EXPECTED FROM DESIGN AND MECHANISM FOR ALL PANEL SYSTEMS. ............... 117  viii LIST OF FIGURES FIGURE 1.1: GRAPHIC OF APPLIED FORCES ON TILT‐UP CONNECTORS DURING OVERTURNING. .............................. 5  FIGURE 2.1: PLAN VIEW OF ROOF DIAPHRAGM FRAMING FOR PLYWOOD ROOF DIAPHRAGM. ............................. 10  FIGURE 2.2: NAILING PATTERN FOR PLYWOOD ROOF DIAPHRAGM. ............................................................... 10  FIGURE 2.3: TYPICAL STEEL FRAMING SYSTEM. ......................................................................................... 11  FIGURE 2.4: (A) SCREW PATTERN TO OWSJ AND NAIL PATTERN TO BEAMS; (B) SIDE LAP SCREWS. ..................... 12  FIGURE 2.5: ROOF LAYER ASSEMBLY. ..................................................................................................... 12  FIGURE 2.6: WALL PANELS LAID OUT ON FLOOR SLAB. ............................................................................... 13  FIGURE 2.7: (A) FORMWORK FOR ARCHITECTURAL PATTERNS; (B) FINISHED WALL. .......................................... 14  FIGURE 2.8: EM1 JOIST SEAT. .............................................................................................................. 15  FIGURE 2.9: EM2 SHEAR PLATE. ........................................................................................................... 16  FIGURE 2.10: EM3 SHEAR PLATE. ......................................................................................................... 17  FIGURE 2.11: EM4 SHEAR PLATE. ......................................................................................................... 17  FIGURE 2.12: EM5 EDGE ANGLE. ......................................................................................................... 18  FIGURE 2.13: PREVIOUS TENSION TESTS ‐ LOAD DISPLACEMENT PLOTS. ......................................................... 21  FIGURE 2.14: PREVIOUS SHEAR TESTS ‐ LOAD DISPLACEMENT PLOTS. ............................................................ 24  FIGURE 2.15: ECCENTRICITY IN SHEAR LOADING USED IN PREVIOUS TESTING CONDUCTED ON TILT‐UP CONNECTORS. ............................................................................................................................................... 25  FIGURE 2.16: PREVIOUS TESTING TENSION TEST SETUP. ............................................................................. 26  FIGURE 2.17: PREVIOUS TESTING SHEAR TEST SETUP. ................................................................................ 26  FIGURE 3.1: RENDERING OF LOADING FRAME. ......................................................................................... 31  FIGURE 3.2: PLATE THAT SIMULATED THE WALL PANEL. ............................................................................. 32  FIGURE 3.3: VERTICAL DISPLACEMENT GUIDES. ........................................................................................ 33  FIGURE 3.4: OUT OF PLANE HORIZONTAL MOVEMENT STOP BLOCKS. ............................................................ 34  FIGURE 3.5: VERTICAL HYDRAULIC JACKS CONNECTED TO HORIZONTAL LOADING BEAM. ................................... 35  FIGURE 3.6: MANUAL VERTICAL CONTROL SETUP. .................................................................................... 36  FIGURE 3.7HORIZONTAL HYDRAULIC JACK. .............................................................................................. 37  FIGURE 3.8: ECCENTRIC HORIZONTAL LOADING. ....................................................................................... 37  FIGURE 3.9: SLOTTED PLATE TO ENSURE HORIZONTAL LOADING. .................................................................. 38  ix FIGURE 3.10: (A) FIELD CONSTRUCTED CONNECTION; (B) TEST SPECIMEN. .................................................... 39  FIGURE 3.11: ORTHOGRAPHIC DRAWING OF TEST SPECIMENS. .................................................................... 40  FIGURE 3.12: TEST SPECIMEN STEEL LAYOUT BEFORE CONCRETE CASTING. ..................................................... 41  FIGURE 3.13: TEST SPECIMEN CONCRETE FORMS BEFORE CASTING OF CONCRETE. ........................................... 41  FIGURE 3.14: PHOTO AND DETAIL DRAWING OF THE SYSTEM PREVENTING MOVEMENT OF THE SLAB. .................. 42  FIGURE 3.15: CONCRETE BEING CAST. .................................................................................................... 43  FIGURE 3.16: FINISHED CONCRETE SPECIMENS. ....................................................................................... 43  FIGURE 3.17: CONCRETE CYLINDER COMPRESSION TESTS ........................................................................... 45  FIGURE 3.18: EFFECT OF MOIST CURING ON COMPRESSIVE STRENGTH OF CONCRETE. ...................................... 47  FIGURE 3.19: SPLIT CYLINDER TEST. ....................................................................................................... 48  FIGURE 3.20: REINFORCING STEEL TENSION TEST SETUP............................................................................. 50  FIGURE 3.21: STRESS ‐ DEFORMATION CURVE FOR REINFORCING STEEL. ........................................................ 52  FIGURE 3.22: TEST SETUP INSTRUMENTATION OVERVIEW. ......................................................................... 53  FIGURE 3.23: LOAD CELL TO MEASURE HORIZONTAL FORCE. ....................................................................... 54  FIGURE 3.24: LOAD CELLS TO MEASURE VERTICAL FORCE. .......................................................................... 54  FIGURE 3.25: STRING PODS TO MEASURE HORIZONTAL DISPLACEMENTS. ...................................................... 55  FIGURE 3.26: STRING POD TO MEASURE VERTICAL DISPLACEMENTS.............................................................. 56  FIGURE 3.27: LINEAR POD TO MEASURE OUT OF PLANE MOVEMENTS. .......................................................... 56  FIGURE 3.28: LOADING PROTOCOL FOR TEST SPECIMENS SUBJECTED TO VERTICAL UPLIFT ONLY.......................... 58  FIGURE 3.29: LOADING PROTOCOL FOR TEST SPECIMENS SUBJECTED TO HORIZONTAL SHEAR ONLY. .................... 59  FIGURE 3.30: LOADING PROTOCOL FOR TEST SPECIMENS SUBJECTED TO BOTH VERTICAL UPLIFT AND HORIZONTAL  SHEAR. ..................................................................................................................................... 60  FIGURE 4.1: (A) INITIAL CONCRETE CRACKING; (B) POPPING OFF OF CONCRETE COVER; FOR SPECIMEN 2. ............ 63  FIGURE 4.2: MONOTONIC VERTICAL SHEAR RESPONSE WITH ROUND CONNECTOR TYPE FOR SPECIMEN 2. ............ 64  FIGURE 4.3: CYCLIC VERTICAL SHEAR RESPONSE WITH ROUND FILLER BAR (SPECIMEN 3). ................................. 65  FIGURE 4.4: (A) EM5 CONNECTOR BEING LOADED DOWN TO ΔV=0MM, (B): WALL – CONNECTOR WELD FAILURE  (SPECIMEN 3). ........................................................................................................................... 66  FIGURE 4.5: CYCLIC AND MONOTONIC VERTICAL SHEAR RESPONSE WITH ROUND CONNECTOR TYPE. ................... 67  x FIGURE 4.6: MONOTONIC AND CYCLIC VERTICAL SHEAR RESPONSE WITH REPAIRED WALL ‐ SLAB CONNECTION WELD. ............................................................................................................................................... 68  FIGURE 4.7: REBAR FRACTURE OF VERTICAL CYCLIC TEST (SPECIMEN 3A). ..................................................... 68  FIGURE 4.8: WALL ‐ SLAB CONNECTOR WELD, LEADING TO FAILURE (SPECIMEN 3). ......................................... 69  FIGURE 4.9: CURRENT WALL‐SLAB CONNECTION DETAIL. ............................................................................ 70  FIGURE 4.10: PROPOSED REVISED WALL‐SLAB CONNECTION DETAIL. ............................................................ 71  FIGURE 4.11: CYCLIC VERTICAL SHEAR RESPONSE WITH ROUND AND SQUARE CONNECTOR TYPE. ........................ 72  FIGURE 4.12: MONOTONIC AND CYCLIC RESPONSES WITH SQUARE CONNECTOR TYPE. ..................................... 73  FIGURE 4.13: ALL FIRST CYCLE VERTICAL RESPONSE ENVELOPES AND MODEL CURVE FIT. ................................... 74  FIGURE 4.14: AVERAGE LOAD AT VARIOUS DISPLACEMENT LEVELS FOR EACH CYCLE NUMBER............................. 75  FIGURE 4.15: MONOTONIC HORIZONTAL SHEAR RESPONSES. ...................................................................... 77  FIGURE 4.16: CONCRETE COVER DAMAGE FOR MONOTONIC TESTS (A) 5 AND (B) 12. ...................................... 78  FIGURE 4.17: MONOTONIC (SPECIMEN 12) AND CYCLIC HORIZONTAL SHEAR RESPONSES (SPECIMEN 6). ............ 78  FIGURE 4.18: HORIZONTAL SHEAR RESPONSE OF TWO EM5 CONNECTORS IN SERIES USED AS THE PANEL – PANEL  CONNECTIONS. .......................................................................................................................... 80  FIGURE 4.19: OUT‐OF‐PLANE DISPLACEMENTS DURING HORIZONTAL TEST WITH NO OUT‐OF‐PLANE RESTRICTION.  81  FIGURE 4.20: CYCLIC HORIZONTAL SHEAR RESPONSE COMPARISON WITH PREVIOUS TESTING. ............................ 82  FIGURE 4.21: COMPARISON OF THE FIRST CYCLE HORIZONTAL SHEAR RESPONSE ENVELOPE WITH THE MEASURED  RESPONSE. ................................................................................................................................ 84  FIGURE 4.22: CYCLE 1 ENVELOPES WHEN ∆V=MAX WHEN HORIZONTAL TESTING IS PERFORMED. ...................... 85  FIGURE 4.23: CYCLE 1 ENVELOPES WHEN ∆V=0MM WHEN HORIZONTAL TESTING IS PERFORMED. ..................... 86  FIGURE 4.24: STRENGTH AND VERTICAL DISPLACEMENT INTERACTION CURVE FOR FIRST CYCLE RESPONSES. .......... 87  FIGURE 4.25: EXAMPLE MODEL FIT TO FIRST CYCLE ENVELOPE OF CYCLIC HORIZONTAL RESPONSE. ...................... 88  FIGURE 4.26: EFFECTIVE STIFFNESS AND VERTICAL DISPLACEMENT INTERACTION CURVE FOR FIRST CYCLE RESPONSES. ............................................................................................................................................... 89  FIGURE 4.27:  HORIZONTAL DISPLACEMENT CAPACITY AND VERTICAL DISPLACEMENT INTERACTION CURVE FOR FIRST  CYCLE RESPONSES. ...................................................................................................................... 90  FIGURE 5.1: DRAWING SHOWING THE COMPONENTS THAT DETERMINE THE SYSTEM DRIFTS. ............................. 94  FIGURE 5.2: ONE PANEL PUSHOVER PLOTS NORMALIZED BY MAXIMUM LATERAL FORCE. .................................. 98  xi FIGURE 5.3: ONE PANEL PUSHOVER PLOTS NORMALIZED BY MAXIMUM LATERAL RESIDUAL FORCE. ..................... 98  FIGURE 5.4: TWO PANEL PUSHOVER PLOTS WITH SLIDING FINAL FAILURE MECHANISM NORMALIZED BY MAXIMUM  LATERAL FORCE. ....................................................................................................................... 101  FIGURE 5.5: TWO PANEL PUSHOVER PLOTS WITH SLIDING FINAL FAILURE MECHANISM NORMALIZED BY MAXIMUM  LATERAL RESIDUAL FORCE. .......................................................................................................... 102  FIGURE 5.6: TWO PANEL PUSHOVER PLOTS WITH OVERTURNING FINAL FAILURE MECHANISM WHICH EXPERIENCE NO  PANEL UPLIFT NORMALIZED BY MAXIMUM LATERAL FORCE. ............................................................... 103  FIGURE 5.7: TWO PANEL PUSHOVER PLOTS WITH OVERTURNING FINAL FAILURE MECHANISM WHICH EXPERIENCE NO  PANEL UPLIFT NORMALIZED BY MAXIMUM LATERAL RESIDUAL FORCE. .................................................. 104  FIGURE 5.8: TWO PANEL PUSHOVER PLOTS WITH OVERTURNING FINAL FAILURE MECHANISM WHICH EXPERIENCE  SOME PANEL UPLIFT NORMALIZED BY MAXIMUM LATERAL FORCE. ....................................................... 106  FIGURE 5.9: TWO PANEL PUSHOVER PLOTS WITH OVERTURNING FINAL FAILURE MECHANISM WHICH EXPERIENCE  SOME PANEL UPLIFT NORMALIZED BY MAXIMUM LATERAL RESIDUAL FORCE. ......................................... 106  FIGURE 5.10: ALL PANEL PUSHOVER PLOTS WITH OVERTURNING FINAL FAILURE MECHANISM WHICH EXPERIENCE NO  PANEL LIFT NORMALIZED BY MAXIMUM LATERAL FORCE. ................................................................... 108  FIGURE 5.11: ALL PANEL PUSHOVER PLOTS WITH OVERTURNING FINAL FAILURE MECHANISM WHICH EXPERIENCE NO  PANEL LIFT NORMALIZED BY MAXIMUM LATERAL RESIDUAL FORCE. ..................................................... 108  FIGURE 5.12: ALL PUSHOVER PLOTS WITH OVERTURNING FINAL FAILURE MECHANISM WHICH EXPERIENCE SOME  PANEL UPLIFT NORMALIZED BY MAXIMUM LATERAL FORCE. ............................................................... 110  FIGURE 5.13: ALL PUSHOVER PLOTS WITH OVERTURNING FINAL FAILURE MECHANISM WHICH EXPERIENCE SOME  PANEL UPLIFT NORMALIZED BY MAXIMUM LATERAL RESIDUAL FORCE. .................................................. 110  FIGURE 5.14: ALL PANEL PUSHOVER PLOTS WITH SLIDING FINAL FAILURE MECHANISM NORMALIZED BY MAXIMUM  LATERAL FORCE. ....................................................................................................................... 112  FIGURE 5.15: ALL PANEL PUSHOVER PLOTS WITH SLIDING FINAL FAILURE MECHANISM NORMALIZED BY MAXIMUM  LATERAL RESIDUAL FORCE. .......................................................................................................... 112  FIGURE 5.16: ALL PANEL PUSHOVER PLOTS WITH EXTREMELY BRITTLE SLIDING FINAL FAILURE MECHANISM  NORMALIZED BY MAXIMUM LATERAL FORCE. .................................................................................. 114  FIGURE 5.17: ALL PANEL PUSHOVER PLOTS WITH EXTREMELY BRITTLE SLIDING FINAL FAILURE MECHANISM  NORMALIZED BY MAXIMUM LATERAL  RESIDUAL  FORCE. ................................................................... 115  xii FIGURE 5.18: MAXIMUM LATERAL FORCE DETERMINED BY THE PUSHOVER MODEL COMPARED TO THE NOMINAL  OVERTURNING RESISTANCE EXPECTED IN DESIGN. ............................................................................ 118  FIGURE 5.19: RATIO OF MAXIMUM LATERAL FORCE DETERMINED BY PUSHOVER MODEL TO NOMINAL LATERAL  OVERTURNING RESISTANCE EXPECTED IN DESIGN. ............................................................................ 118  FIGURE 5.20: SHEAR RESISTANCE ENVELOPE AND LATERAL FORCE RESPONSE FROM NON‐LINEAR PUSHOVER MODEL  COMPARED WITH EXPECTED VALUES FROM CURRENT DESIGN METHODS FOR THE 3,2,4 SYSTEM. .............. 119  FIGURE 5.21: ACTUAL SLIDING RESISTANCE AT MAXIMUM LATERAL FORCE DETERMINED BY THE PUSHOVER MODEL  COMPARED TO EXPECTED NOMINAL SLIDING RESISTANCE EXPECTED IN DESIGN. ..................................... 120  FIGURE 5.22: RATIO OF THE ACTUAL SLIDING RESISTANCE AT MAXIMUM LATERAL FORCE DETERMINED BY THE  PUSHOVER MODEL TO NOMINAL LATERAL SLIDING RESISTANCE EXPECTED IN DESIGN. .............................. 120  FIGURE 5.23: COMPARISON OF THE CURRENT 4,3,4 SYSTEM RESPONSE WITH THE 4,3,4 RESPONSE USING IDEALIZED  PANEL – SLAB CONNECTIONS  NORMALIZED TO MAX FORCE. ............................................................. 126  FIGURE 5.24: COMPARISON OF THE CURRENT 4,3,4 SYSTEM RESPONSE WITH THE 4,3,4 RESPONSE USING IDEALIZED  PANEL – SLAB CONNECTIONS  NORMALIZED TO RESIDUAL FORCE ....................................................... 126  FIGURE 5.25: COMPARISON OF THE CURRENT 4,3,4 SYSTEM RESPONSE WITH THE 4,3,4 RESPONSE USING IDEALIZED  PANEL – PANEL CONNECTIONS  NORMALIZED TO MAX FORCE............................................................ 127  FIGURE 5.26: COMPARISON OF THE CURRENT 4,3,4 SYSTEM RESPONSE WITH THE 4,3,4 RESPONSE USING IDEALIZED  PANEL – PANEL CONNECTIONS  NORMALIZED TO RESIDUAL FORCE ..................................................... 128  FIGURE 5.27: COMPARISON OF THE CURRENT 4,3,4 SYSTEM RESPONSE WITH THE 4,3,4 RESPONSE USING IDEALIZED  PANEL – SLAB AND PANEL – PANEL CONNECTIONS  NORMALIZED TO MAX FORCE .................................. 129  FIGURE 5.28: COMPARISON OF THE CURRENT 4,3,4 SYSTEM RESPONSE WITH THE 4,3,4 RESPONSE USING IDEALIZED  PANEL – SLAB AND PANEL – PANEL CONNECTIONS NORMALIZED TO RESIDUAL FORCE ............................. 129  xiii ACKNOWLEDGMENTS My sincerest thanks go to my research supervisor, Dr. Perry Adebar, for his guidance and support throughout this research endeavour. Thanks also to Dr. Ken Elwood, for his insight and for completing second review duties. Support and input from Kevin Lemieux and Gerry Weiler of Weiler Smith Bowers is also greatly appreciated. I would also like to acknowledge the generous financial support of the Natural Sciences and Engineering Research Council of Canada, the Portland Cement Association and the Cement Association of Canada. 1 INTRODUCTION 1.1 Background Tilt-up construction, which is the technique of casting concrete panels on the ground and then lifting (tilting) them upright to form walls originated in California in the late 1940’s as a method of constructing solid reinforced concrete walls for industrial buildings. It has gained in popularity ever since due to its economical and quick construction. This structural system was originally designed primarily for large, single story box-type warehouse structures with very few doors and openings, and thus all the wall panels could be characterized as solid panels. The strength and inelastic response of a wall system constructed of these elements is almost entirely controlled by the connections to these elements since solid wall panels and wall panels with small openings are inherently stiff and strong. In modern days the applications in which tilt-up construction is used in Canada and the United States has shifted to include shopping centers, office buildings, and schools. Many of these applications require large windows for storefronts, unique shapes for architectural appeal as well as multiple stories. Each of these modifications to the original design creates additional complexities to the already complex and generally nonlinear seismic response of these structures. The small strips of concrete walls around and between these openings resist in-plane seismic forces like a cast-in-place reinforced concrete frame; however concrete tilt-up wall panels with large openings are constructed with minimal ductile detailing in the “beams,” “columns” and “beam-column joints.” Thus tilt-up frames have much less inelastic drift capacity than typical cast-in-place frames. Tilt-up frames are also subjected to larger drift demands than typical cast- in-place frames because of the very flexible elastic roof diaphragms as described above. Further discussion on the seismic design of tilt-up wall panels with large openings is given by Adebar et al. (2004). Since this response is not well understood even for low rise box type structures with few openings, this will be the focus of this study. Continuing research into more complex applications of tilt-up construction described above can then be conducted. 1   There are several uncertainties and design assumptions made in current practice which this study seeks to address. Tilt-up design is covered in Chapter 13 of the 2005 CAC Concrete Design Handbook. Sections of specific importance to the current study are Sections 13.11 and 13.12 where in-plane shear design is discussed, and Section 13.16 where a design example of a shear wall with solid panels is presented in example 13.4. The assumptions and uncertainties presented in these sections, which are within the scope of this study, will be discussed further in this section. According to Weiler (Weiler, 2006), the design considerations for tilt-up panels subjected to in-plane forces are: • Resistance to sliding • Resistance to panel overturning • Concrete shear resistance • Increased localized axial forces and out-of-plane P – Δ effects • Load distribution to foundations • Frame action in panels with openings • Seismic ductility This study will focus on the resistance to both sliding and overturning as well as address seismic ductility. Currently designers add connections between two or more walls panels to increase the rocking resistance of the system, and add connections between wall panels and the base slab to provided horizontal sliding resistance. There are also connections between wall panels and the roof diaphragm to transfer force from roof diaphragm to the wall assembly. If sufficient ductility is not provided in these connectors, strong ground shaking may result in failure of any of these connections with limited energy dissipation at the connection. Many tilt-up buildings in California have timber roofs that dissipate energy when the structure is subjected to large seismic demands. Experience from recent California earthquakes has shown that the weak link in these structures is often the out-of-plane connection between the concrete tilt-up walls and the timber roof diaphragm (Hamburger and McCormick, 1994). In Canada, and more recently in the western US, steel deck diaphragms are commonly used, and these are constructed in such a way that the diaphragms will likely remain elastic during the design earthquake. When the stiff Chapter 1 Introduction 2   concrete tilt-up walls reach their lateral load capacity, the flexible elastic diaphragms cause a magnification of the drift demands (Adebar et al. 2004). It is not the intent of this study to focus on the wall – roof connection but rather to focus on the response of the wall panel only. When only panel components are considered, failure of the panel - slab connections may result and subsequently lead to either sliding or rocking of the walls on the foundation, depending on the system properties, and location of connection failure. IN an effort to explore the non-linear response of tilt-up buildings after failure of the slab-wall connection, and to investigate the preferred mode of response for the design of new tilt-up buildings, a non-linear dynamic analysis study was undertaken at the University of British Columbia (Olund 2009) considering two non-linear models relevant to this study: (a) Sliding model where the walls are allowed to slide relative to the foundation after exceeding a friction force based on the dead load at the base of the wall panel; and (b) Rocking model where by the weight of the panel and applied loads is exceeded (Olund 2009). In order for this non-linear analysis to be effectively completed for buildings including connectors, detailed connector response properties are needed. This is one main objective of this study. Currently system design is done without designing for a particular mechanism to form, and connection design to resist each of these mechanisms are done independently of one another, with panel – panel connectors added to increase overturning resistance, and panel – slab connectors added to resist sliding (see Figure 1.1). This is dangerous however as it is not known how each of these connection details will affect the other. Since sliding resistance in design is based on the sum of friction and the total factored load resistance of the panel – slab connectors, if vertical damage due to a rocking panel decreases the horizontal resistance, the system resistance may be lower than expected. Also since vertical uplift resistance of the panel – slab connectors is currently ignored in design, if they have significant vertical load capacity, they will increase the shear load transferred through the panel – panel connectors, causing them fail earlier than expected. These interactions make it virtually impossible to predict a failure mechanism, as it is unclear which connections will fail first within the system. Ultimately it is unknown what the vertical uplift resistance of the panel – slab connection is as there has never been experimental testing performed to measure its response. This is a key objective of this study. Chapter 1 Introduction 3   The amount of seismic ductility within current design of tilt-up structures is also not clear as they are currently designed in an attempt to provide enough strength, with little regard for ductility. Weiler states that typically designers try to ensure that there is adequate overturning resistance and shear strength on the basis that ductility is automatically taken care of by the selection of the appropriate R-value, however tilt-up system ductility is quite different than connector ductility (Weiler, 2006). It is assumed that if a connector shows signs of ductility, then the system also will show ductility, when in fact it may not. It is unclear if a “ductile” connector (relative to other connectors used) which still has a relatively small horizontal displacement capacity will create significant system ductility. 1.2 Research Methodology From this point, one direction this study could have taken would be to study the response of a multi - panel system through an experimental program. This however would have been very expensive, and very few tests could have been performed, or very small scale tests. As the rocking response is very dependent on the panel dimensions, as well as the configuration of the connectors, one test specimen would have only generated the response for that specific panel assembly. Because of this a more general study was desired so that results could be applied to any panel system. Since it was already obvious that for panel systems with mostly solid panels the inelastic response would be almost entirely due to deformations of the connectors, the wall panels were not necessary to achieve the objectives of this study. It was decided upon that what was needed was a detailed study into the inelastic response of the connectors, which could be incorporated into a model with rigid wall panels to get the response of a tilt-up system which should reasonably match that of a full scale test. The advantage of doing these smaller, more focused tests was that multiple full scale tests could be performed, which allowed for a detailed study of the fundamental behavior of the connector. If a panel system is studied (Figure 1.1), it can be seen that if sufficient panel – slab connectors are used to prevent sliding, panel rocking will occur. When panel rocking occurs, uplift forces (P) are applied to the panel – slab connectors. This uplift force will provide overturning Chapter 1 Introduction 4   resistance, as well as add shear demand on the panel – panel connections as the adjacent panel tends to want to lift from the ground. The inelastic response of the panel – slab connector in uplift is a key element to the overall system response and therefore must be determined. Figure 1.1: Graphic of applied forces on tilt‐up connectors during overturning.  Previous testing conducted on tilt-up connectors described in Chapter 2 only focused on the horizontal shear response as well as the horizontal “tension” pullout. The vertical uplift response was never measured, and therefore will be measured experimentally in this study. As already mentioned, it is apparent that as a tilt-up panel rocks, the panel - slab connector in the concrete slab will experience vertical uplift demands in addition to horizontal shear demands. If vertical damage has occurred within the panel – slab connectors due to panel rocking, it is conceivable that this damage could have an effect on the horizontal shear strength, stiffness and displacement capacity. This effect could cause the system failure to occur at a much lower lateral force level than is anticipated, as well provide a much less stiff system with lower displacement Chapter 1 Introduction 5   capacity. A key objective to this study is to understand these relationships through an experimental testing program. For the results of a testing program to be applicable at any location along a panel, the three interactions between vertical damage level and horizontal strength, horizontal stiffness and horizontal displacement capacity are needed. These interactions relationships for the panel – slab connector can then be used in a model of a tilt-up system with any number of panel – slab connectors in any location along the panels base to determine the system response. In order to achieve these interactions with sufficient confidence, 20 tests were conducted, each with a varying degree of vertical damage when horizontal loading was applied. 1.3 Research Objectives and Scope To summarize, the research objectives of this study are to: • Determine the vertical uplift response of the panel – slab connection used in typical tilt- up construction through experimental testing. • Confirm the horizontal shear response of the panel – slab and panel – panel connections used in typical tilt-up construction found in previous testing through experimental testing. • Determine the interaction relationships between the vertical damage level (displacement) and the residual horizontal strength capacity, horizontal stiffness and horizontal displacement capacity through experimental testing. • Develop and present a pushover model for a multi-panel tilt-up system with solid wall panels accounting for the non-linear connection response found through experimental testing. • Implement the pushover model for several tilt-up system configurations to predict multi- panel system responses and study their properties and specifically their failure mechanisms. • Investigate how to configure the system to achieve the best possible response. Chapter 1 Introduction 6   • Investigate how connection properties could be changed to achieve a better system response. 1.4 Thesis Overview The remaining chapters of this thesis will address the objectives outlined in this chapter. Chapter 2 will focus on the history and current construction techniques of tilt-up construction, as well as describe in detail the connector types used and previous research done on them. Chapter 3 will outline the experimental program including loading protocol, test and instrumentation setup and specimen material properties and construction. In Chapter 4 the results from the experimental program are presented, and from them a model curve for the vertical response of the panel – slab connectors is determined, as well as interaction curves between the vertical displacement level and the horizontal load capacity, stiffness and displacement capacity. The data obtained from experimental testing is used in the pushover model which is described and implemented for various tilt-up systems in Chapter 5. Chapter 5 also discusses how the results of this study may help address current uncertainties in the design of tilt-up structures. Chapter 6 presents the conclusions and recommendations resulting from this study. Appendix A provides a detailed document for each experimental test performed, including the measured data, and photos. Appendix B provides the individual pushover response along with the individual connector responses for the 30 systems that are investigated. The pushover responses normalized to the maximum lateral resistance achieved, and normalized to the residual capacity are also included for each pushover Chapter 1 Introduction 7 2 HISTORY, CURRENT PRACTICE AND PREVIOUS TESTING Tilt-up structures have gained wide popularity throughout many countries due to its time and cost efficient construction. Its versatility as a construction method in both building application and design has quickly made tilt-up the construction method of choice, specifically for single story box type structures. Despite its original uses, tilt-up is now being used for more architecturally pleasing structures that include multiple story buildings with numerous large openings. Research into how tilt-up responds dynamically, specifically due to earthquake loading, has been surprisingly limited considering how widely used this construction method has become. 2.1 History of Tilt-up Tilt-up construction has been used since the early nineteenth century to economically build large warehouse type structures. When reinforced concrete was introduced in the early 1900’s, contractors were able to implement the old method of tilting up wooden walls to concrete structures. Despite the ability to tilt concrete wall into place, tilt-up did not gain popularity until the late 1940’s when the first mobile cranes were introduced. This allowed quick access by a crane to virtually any job site to lift the massive panels into place. This made tilt-up a viable alternative to cast in place construction. Tilt-up quickly gained popularity as the demand for large structures increased due to the post-world war 2 boom in industrial and manufacturing industries. Tilt-up allowed contractors to efficiently keep up with the construction demand. According to the tilt-up construction association (TCA, 2007), tilt-up has since been used for structures as large as 1.7 million square feet, with individual panels as tall as 96 feet, and weighing up to 150 tons. Also it is reported that in the United States, over 15% of all industrial buildings are constructed using tilt-up construction, with an increase of 20 % each year. Over 650 million square feet of tilt-up space is constructed on an annual basis in the US, with increasing usage in countries such as Canada, Australia, and Mexico. The current construction methods will be discussed in the next section. 8   2.2 Current Construction Techniques The current construction techniques used in practice vary slightly between Canada and the United States, most notably with respect to the diaphragm construction. That being said, the American and Canadian research information in this report can still be useful to determine the response of either system as both diaphragm construction methods are both flexible, and should behave in a similar matter, even though the materials may differ. A description of each element of a tilt-up building will now be described in further detail. Much of the information was provided by Kevin Lemieux of Weiler Smith Bowers in Burnaby, BC, and photos were taken during a recent site visit. 2.2.1 Diaphragms Tilt-up construction typically has two forms of diaphragms, both considered flexible. In the United States, panelized plywood roof diaphragms are used. The plywood is supported by deep 5-1/8” wide glu-lam beams, 4” purlins that frame into the glu-lams and 2” wide sub-purlins on a 2 foot grid that frame between the purlins down the centerline of the plywood sheeting (Figure 2.1). The glu-lam beams are supported on the perimeter walls by bearing seats attached to reinforcing steel or steel stud anchor connections (see Section 2.2.3 for connection details). Heavier nailing patterns are used to resist the regions of higher diaphragm shear stresses (eg: around the outside perimeter of the building (See Figure 2.2)) (Hamburger and McCormick, 1994). Chapter 2 History, Current Practice and Previous Testing 9   Figure 2.1: Plan view of roof diaphragm framing for plywood roof diaphragm.  (Modified from Hamburger and McCormick, 1994)  Figure 2.2: Nailing pattern for plywood roof diaphragm.  (Modified from Hamburger and McCormick, 1994)    Chapter 2 History, Current Practice and Previous Testing 10   In Canada, corrugated cold-formed steel deck is utilized for tilt-up roof diaphragms. This steel decking is supported by a typical steel beams / open web steel joist (OWSJ) framing system (Figure 2.3). Figure 2.3: Typical steel framing system.    There are various ways in which to connect the steel deck to the steel framing system however screwing to the OWSJ and side laps, and nailing to the supporting beams (Figure 2.4) is becoming the norm in western Canada due to the advantageous ductile properties over the other methods (Essa, et.al., 2003).. Chapter 2 History, Current Practice and Previous Testing 11   (a)  (b)  Figure 2.4: (a) Screw pattern to OWSJ and nail pattern to beams; (b) side lap screws.  Unlike the United States, where heavier nailing provides increased shear resistance of the roof diaphragm, in Canada different gauge material is used to achieve this with nail and screw spacing remaining constant. Typically the perimeter of a tilt-up building will have 18 gauge roof decking where as near the center of the structure, 22 gauge material will be used. The shear stresses across the roof diaphragm are calculated and steel deck material will be placed as required to resist these stresses. Both the plywood system used in the United States, and the metal decking system used in Canada are then covered with a foam insulation layer, then a water tight membrane covered with a layer of smooth rock as shown in Figure 2.5. Figure 2.5: Roof layer assembly.  Chapter 2 History, Current Practice and Previous Testing 12   2.2.2 Wall Panels Tilt-up structures consists of perimeter concrete wall panels which span vertically between the floor slab and the roof diaphragm. These diaphragms provide support to the top and bottom of the wall panels to tie the system together. These panels then act as the gravity load bearing members as well as the lateral force resisting system (LFRS) in the form of shear walls. Wall panels vary dramatically in size and can be as thin as 140 mm. The maximum depth to height ratio is limited to 50 for panels with a single mat of reinforcing steel, and 65 for panels with 2 mats of reinforcing steel (CAC, 2006). Panels have been as tall as 96 feet and have weighed as much as 150 tons, however the vast majority of tilt-up projects are single story warehouse type structures with a roof height of about 26 feet to the underside of the OBSJ. In tilt-up projects, the slab on grade is required to be poured first as the wall panels are then cast horizontally directly on the concrete floor surface (Figure 2.6). Figure 2.6: Wall panels laid out on floor slab.    This requires detailed planning by the project manager so that all the wall panels fit on the floor slab, and are easily accessible for lifting. The panels are usually cast with the outside face down in order that the interior connectors (see Section 2.2.3) can be installed and temporary bracing Chapter 2 History, Current Practice and Previous Testing 13   attached before tilting occurs. Also this allows for architectural cambering and finishes to be more easily incorporated into the casting (Figure 2.7). (a)  (b)  Figure 2.7: (a) Formwork for architectural patterns; (b) finished wall.    2.2.3 Current Connectors There are five main connectors used in the construction of tilt-up buildings, each performing a specific operation. According to Lemieux, Sexsmith and Weiler (1998), these connections are mainly used for the transfer of: • Gravity loads from beams and joists to the tilt-up panels. • Shear forces from roof joists and steel deck to the tilt-up panels. • Shear forces from intermediate floors to the tilt-up panel. • Shear forces between tilt-up panels. • Shear forces from tilt-up panels to the floor slab or foundation. • Forces from mechanical equipment or architectural items to the tilt-up panels. Welded embedded steel connectors are widely used in tilt-up construction as they are an efficient and economical way of providing the connection needed. Each of the five main connection types are detailed in the Concrete Design Handbook and are described below: Chapter 2 History, Current Practice and Previous Testing 14   EM1 Joist Seat The EM1 joist seat is an L 89 X 89 X 6 X 300 angle with two 15M bent rebar anchors welded to the underside to provide embedment into the concrete wall. This connection is used mainly as a joist seat for OWSJ framing into the wall (Figure 2.8). Since this connector is not used for the base slab connection, it will not be looked at any further in this study. Side View  Top View  Joist Pocket  Figure 2.8: EM1 joist seat.  (Drawings modified from Weiler, et al., 1997)    Chapter 2 History, Current Practice and Previous Testing 15   EM2 Shear Plate The EM2 shear plate is a PL 150 X 9.5 X 200 plate with 2-16(mm) studs. This connector is usually used to as a connection point for deck angles and/or angle tie struts (Figure 2.9). This connector is also used along with the EM3 connector as part of the connection detail for the wall panel – base slab connection. The EM2/3 is placed typically in the wall panel and acts as the wall panel connection point for the EM5 edge angle to be welded to (this detail is shown in Figure 2.12). Side View  Top View  Shear Plate Connection  Figure 2.9: EM2 shear plate.  (Drawings modified from Weiler, et al., 1997)    Chapter 2 History, Current Practice and Previous Testing 16   EM3 Shear Plate The EM3 shear plate is a PL 200 x 9.5 x 200 plate with 4-16(mm) studs. This connector is commonly used to attach an angle seat for an OWSJ for floors at intermediate levels (Figure 2.10). Side View  Top View  Shear Plate Connection  Figure 2.10: EM3 shear plate.  (Drawings modified from Weiler, et al., 1997)    EM4 Shear Plate The EM4 shear plate is a PL 225 X 9.5 X 460 plate with 8-16(mm) studs. This connector is used to attach steel support beams to an exterior tilt-up wall panel (Figure 2.11). Side View  Top View  Beam Connection  Figure 2.11: EM4 shear plate.  (Drawings modified from Weiler, et al., 1997)  Chapter 2 History, Current Practice and Previous Testing 17   EM5 Edge Angle The EM5 edge angle is a L 38 X 38 X 6 X 200 angle with a long bent 20M rebar anchor welded to it. This connector is used as a shear connector between panels, or between the floor slab or foundation and a wall panel (Figure 2.12). Top View  Side View  Panel – slab Connection  Figure 2.12: EM5 edge angle.  (Drawings modified from Weiler, et al., 1997)    Since the current study is investigating the wall to base slab connections, only previous testing of the connectors relevant to these connections will be reviewed, namely the shear responses of EM2, EM3 and EM5 connectors. Although the current study does not study the tension pull out of these connectors due to out of plane movement, these movements may be an important factor in studying the overall response of a tilt-up structure, and so the previous tension test results for the relevant connectors will also be provided. Chapter 2 History, Current Practice and Previous Testing 18   2.3 Previous Testing A study conducted at the University of British Columbia by Lemieux, Sexsmith and Weiler (1998) set out to investigate the degree of ductility and the dependability of load carrying capacity of the five common connections described in Section 2.2.3. This study produced load – deformation response curves for each of the connectors subjected to both monotonic and cyclic loading through experimental testing. These results were then used to recommend the tabulated values for the shear and tensile capacities shown in the Concrete Design Handbook. Unlike other structure systems, the ductility of a tilt-up system may be dominated by the roof diaphragm. Despite this the connections may contribute to the energy dissipation and ductility of the system. This being said, for the system to behave in a ductile fashion, either the connections should be in fact ductile or they should be stronger that the other energy dissipation mechanisms such as wall rocking, wall hinging, or diaphragm plastic deformation. (Lemieux, et al., 1998). The results of the tension testing are summarized in Table 2.1. Since the minimum tilt-up panel thickness is 140 mm, this testing was performed at that thickness as to generate worst case scenarios. During tension testing, EM2 and EM3 failed by a concrete cone failure and generated capacities in tension significantly smaller (40-50%) then are allowed by code. Also upon examining the hysterisis loops for these connections (Figure 2.13), severe pinching with very little energy dissipation capacity is noticed. These results are somewhat unsettling as the code permissible values assume yielding of the steel connection (ductile) before the failure of the concrete (Brittle). Since tilt-up panels are allowed to be thin, failure of the concrete is virtually unavoidable using these connectors in tension. EM5 testing however produced a steel failure and was close to the code predicted value. The stiffness however degraded at about 50 kN tension force due to concrete spalling and permanent deformation of the exposed rebar as the bends straightened. The hysteresis loops for EM5 showed large inelastic deformations with each successive cycle. This indicates ductility, however considerable pinching of the loops makes it inefficient at energy dissipation. Chapter 2 History, Current Practice and Previous Testing 19   Table 2.1: Connection tension test  results.  (Modified from Lemieux, et al., 1998)  Specimen No. EM2 EM3 EM5 T1 83.5 113.2 174.7 T2 87.9 118.2 163.4 T3 89.6 142.4 162.8 T4 96.3 - - Average 89 125 167 Theoretical 140 202 170 Ave/Theoretical 0.64 0.62 0.98 Chapter 2 History, Current Practice and Previous Testing 20   EM2 tension test results.  EM3 tension test results.  EM5 tension test results  Figure 2.13: Previous tension tests ‐ load displacement plots.  (Modified from Lemieux, et al., 1998)  0 20 40 60 80 100 0 2 4 6 8 10 12 14 V  [k N ] ∆H [mm] Series2 Series1 0 20 40 60 80 100 120 0 1 2 3 4 5 6 7 V  [k N ] ∆H [mm] Series2 Series1 0 20 40 60 80 100 120 140 160 180 0 10 20 30 40 50 60 V  [k N ] ∆H [mm] Series2 Series1 Cyclic Monotonic Cyclic Monotonic Cyclic Monotonic Chapter 2 History, Current Practice and Previous Testing 21   The results of the shear testing are summarized in Table 2.2. As can be seen from these results, the experimental values are between 20-40% higher than the theoretical for the EM2, EM3, and EM5 connectors. Figure 2.14 shows the load-displacement relationship for the shear tests. These plots indicate that the EM2 and EM3 connectors showed some ductility. It was recommended however that this ductility not be relied on as there was a possibility of brittle concrete cone failure. EM5 on the other hand showed a ductile behavior that could be relied on and so the ductile designation for EM5 is appropriate (Lemieux, et al., 1998). However as we will see in Chapter 5, this ductility is not enough to have a noticeable effect the response of a multi-panel tilt-up system.   Table 2.2: Connection shear test  results.  (Modified from Lemieux, et al., 1998)  Specimen No. EM2 EM3 EM5 S1 155 254.2 278.8 S2 140.6 179.7 210.6 S3 136.1 215.1 215.1 S4 118.5 - - S5 110.6 - - Average 132 216 235 Theoretical 111 162 170 Ave/Theoretical 1.18 1.33 1.38 Chapter 2 History, Current Practice and Previous Testing 22   EM2 shear test results parallel to line of studs.  EM2 shear test results perpendicular to line of studs.    ‐150 ‐100 ‐50 0 50 100 150 ‐6 ‐4 ‐2 0 2 4 6 V  [k N ] ∆H [mm] Series2 Series1 ‐150 ‐100 ‐50 0 50 100 150 ‐8 ‐6 ‐4 ‐2 0 2 4 6 V  [k N ] ∆H [mm] Series2 Series1 Cyclic Monotonic Cyclic Monotonic Chapter 2 History, Current Practice and Previous Testing 23   EM3 shear test results.  EM5 shear test results.  Figure 2.14: Previous shear tests ‐ load displacement plots.  (Modified from Lemieux, et al., 1998)  ‐300 ‐200 ‐100 0 100 200 300 ‐4 ‐2 0 2 4 6 V  [k N ] ∆H [mm] Series2 Series1 ‐300 ‐200 ‐100 0 100 200 300 ‐14 ‐12 ‐10 ‐8 ‐6 ‐4 ‐2 0 2 4 6 8 10 12 14 16 V  [k N ] ∆H [mm] Series2 Series1 Cyclic Monotonic Cyclic Monotonic Chapter 2 History, Current Practice and Previous Testing 24   The values from this previous testing were used by the Committee for Tilt-up Connections to recommend design values for the tilt-up connections. Since in previous testing thin panel specimens were used, concrete cone failure occurred, hence the recommended design values proposed was ׎஼ ൈ ܮ݋ݓ݁ݏݐ ܶ݁ݏݐ ܴ݁ݏݑ݈ݐ. In 1997, ׎஼ was 0.6, so the recommended horizontal design shear strength (which is the value of interest to this study) is 0.6 ൈ 210 ݇ܰ ؆ 125 ݇ܰ (Weiler, et al, 1997). The test specimens used in the previous testing of the EM2 and EM3 connectors were made using 914 mm x 914 mm x 140 mm thick concrete panels and 1524 mm x 914 mm x 140 mm thick concrete panels for the testing of the EM5 connectors. All these specimens were constructed at off campus construction sites. These specimens were then bolted to the strong floor using steel brackets, and the hydraulic jack was bolted to a knife plate which was welded to the connector. It was noted that for the tension tests the bolting pattern created negligible eccentricities however the shear testing had eccentricities of loading which seemed more significant and would have introduced a considerable bending moment (as seen in the photo of Figure 2.15, however in this previous study justification to overlook this was presented due to uncalculated eccentricities due to fabrication and construction variations. Figure 2.15: Eccentricity in shear loading used in previous testing conducted on tilt‐up connectors.  (Photo Courtesy of Kevin Lemieux)  Chapter 2 History, Current Practice and Previous Testing 25   Figure 2.16 shows the tension test setup and Figure 2.17 shows the shear test setup for the previous testing. Figure 2.16: Previous testing tension test setup.  (Modified from Lemieux, et al., 1998)  Figure 2.17: Previous testing shear test setup.  (Modified from Lemieux, et al., 1998)  Chapter 2 History, Current Practice and Previous Testing 26   The loading protocol for the previous testing was significantly different to that of the current study. In the previous testing for both shear and tension tests, the first specimen of each connector type was loaded monotonically using a constant displacement rate. The remaining two specimens of each connector type were loaded cyclically using three cycles at six to eight constant load increments up to about 70% of the ultimate load capacity attained in the monotonic test. After this point three cycles at constant displacement increments were used till failure of the specimen. The effects of both the differed test setup and loading protocol will be discussed further in Section 4.3.4. 2.4 Typical Building Properties Information about the building properties of a typical tilt-up building on which to base the pushover model was found from two sources. The study of engineering drawings for a typical single story warehouse type tilt-up building located in Richmond, BC was conducted and the type of panel – panel connector used and the number of then supplied for a typical building are recorded in Table 2.4 and Table 2.3 respectively. Table 2.3: The number of connectors used  and their configurations within a typical tilt‐ up building.      Number of P ‐ S  Connections Total  2  3  4  N um be r o f P an el  ‐  Pa ne l  Co nn . ( 1s t e dg e‐ 2n d  ed ge )  0‐0 23 2 25 2‐0 7 7 2‐2 9 4 13 3‐0 1 8 9 3‐2 1 1 5‐0  3  3  5‐2  1  1  5‐5 2 2 Total 34 24 3 61 Chapter 2 History, Current Practice and Previous Testing 27   Table 2.4: Number of each panel ‐ slab connector type  used in a typical tilt‐up building.    Connection Type Number Used % Used  EM2 – EM5 49 80% EM3 – EM5 12 20% Total  61 100% These tables reveal that under normal circumstances a range from two to four panel – slab connectors are used, 80% of which are EM2 – EM5 connectors. They also show that a range from zero to five panel – panel connectors are typically used with most having zero to three. Also it was observed that most panels were not connected to the adjacent panel, however up to three panels were connected together within the study building. Based on this information, systems with a range of two to four panel – slab connectors, two to four panel – panel connectors and one to four panels tied together will be studied using the model presented. This generates 30 systems which will be looked at in detail in Chapter 5. The panel dimensions that were used in the pushover model presented in this study were chosen based on a survey conducted of various firms practicing in the tilt-up industry. The results of this survey indicated that a panel height of 30 feet (9.14 m) is typical as well as a height to thickness ratio of less than 50, which for a 30 foot high panel would be 7.25” (190 mm). A target panel weight of 80 kips (356 kN) was suggested, which was used to determine a panel width of 25 feet (7.62 m). Table 2.5 shows the panel widths used in the sample building. The panels in this sample building are generally narrower than was suggested from the survey results, however it still is similar to the widest panels observed in the sample building. Chapter 2 History, Current Practice and Previous Testing 28   Table 2.5: The number of  panels and their widths used  in a typical tilt‐up building.    Number Panel Width [ft] [m] 33 18'‐0\" 5.49 4 21'‐6\" 6.55 11 20'‐0\" 6.10 1 22'‐4\" 6.81 6 17'‐4\" 5.28 4 10'‐0\" 3.05 2 24'‐4\" 7.42 Average 19'‐9\" 5.81 Chapter 2 History, Current Practice and Previous Testing 29 3 EXPERIMENTAL PROGRAM This chapter details the experimental program used in this study. Detailed descriptions of the utilized loading protocol, testing specimens, test apparatus, instrumentation and material properties will be provided. 3.1 Testing Methodology As discussed in the introduction, the objectives of the current experimental program is to determine the interactions between vertical damage and the horizontal shear response (strength, stiffness and displacement). The testing protocol used to determine these interactions involved a two stage experimental program. In the first stage, a specified vertical damage level was achieved through vertical cyclic displacements. The second stage then required horizontal cyclic loading to be applied until failure, while the loading beam was held at a predetermined vertical displacement level. Although an actual tilt-up panel subjected to panel rocking would experience simultaneous vertical and horizontal shear displacements, the two part loading protocol allowed for all of the responses to be dependent only on the vertical displacement level, as opposed to being dependent on both vertical and horizontal displacements, as well as the proportions of horizontal and vertical load which in itself is dependent on number and location of panel – slab connectors as well as the aspect ratio of the panels. The elimination of these many degrees of freedom allowed for the determination of more general response data which can be applied to a tilt-up panel system with any panel dimensions, and any number of connectors at any location along the panel width. 3.2 Test Apparatus The test setup, designed to provide independent vertical and horizontal movement is detailed in this section. 30   3.2.1 Loading Frame The loading frame used for the experimental program is shown in Figure 3.1. It consisted of a horizontal loading beam (labeled A in Figure 3.1) made from an HSS 152 X 152 X 9.5 which was used to apply both the horizontal and vertical loads through a 50 mm thick loading plate which was connected to the test specimen (labeled B in Figure 3.1 & shown in Figure 3.2). Figure 3.1: Rendering of loading frame.  The 50 mm thick loading plate was used to simulate the rigid concrete wall panel and wall embedded connector. This rigid plate was considered an appropriate substitute since a solid concrete wall panel can be considered completely rigid. The EM2 or EM3 connectors used as the wall embedded have been shown in previous studies to provide very little displacement capacity (2 - 4 mm ultimate displacement) when compared to the EM5 connector in the slab (~20 mm ultimate displacement), so they will contribute very little to the overall connection response. Chapter 3 Experimental Program 31   Also the stud plate connectors have the same well defined response regardless of the direction of loading, so since the EM2/3 in the wall and EM5 in the slab are connected in series, the responses can be superimposed to give the combined response. This is what is done in the pushover model presented in Chapter 4. Figure 3.2: Plate that simulated the wall panel.  In a real tilt-up system there is no significant resistance to out of plane movements of the wall panel at the base of the walls (only the connectors themselves), however out of plane rotations of the wall panel would be small due to the connection at the roof level. To simulate this, a 1200 mm long vertical guide column, labeled C in Figure 3.1, made from a HSS 102 X 102 X 6.4 was used to allow horizontal out of plane movement away from the slab at the base, but control the rotations of the horizontal loading beam by not allowing the top of the column to move out of plane. In order to restrain rotation to the exact levels that would occur in a real structure this Chapter 3 Experimental Program 32   column would have to have been as tall as an actual panel, which was not practical. Despite this a 1200 mm column was deemed to be long enough to allow reasonably small rotations so the rotational effects could be neglected. Two frames were constructed from HSS 102 X 102 X 6.4 (labeled D in Figure 3.1) to hold the vertical column guide beam as well as the hydraulic jacks used in the vertical loading. These frames were bolted to the strong floor. Attached to the threaded rods used to connect these frames to the floor were horizontal guides which were used to manually maintain the desired vertical displacement level during the horizontal loading stage of testing (labeled E in Figure 3.1 & shown in Figure 3.3). These guides were fitted with Teflon pads to ensure the only measurable horizontal resistance was from the connector. Figure 3.3: Vertical displacement guides.  During the first test the loading plate was mistakenly allowed to slip over the face of the concrete slab specimen due to a combination of insufficient length of the loading plate and the concrete damage. To ensure this did not occur again, blocks were attached to the loading frames with Teflon pads to stop horizontal movement over the face of the slab (labeled F in Figure 3.1 & shown in Figure 3.4). Chapter 3 Experimental Program 33   Figure 3.4: Out of plane horizontal movement stop blocks.  This modification worked very well to insure the out of plane movement was constrained to movements that could be experienced within a real structure. 3.2.2 Vertical Loading Vertical movement was provided manually through two small double acting hydraulic jacks mounted at either end of the horizontal loading beam as shown in Figure 3.5 and labeled G in Figure 3.1. Chapter 3 Experimental Program 34   Figure 3.5: Vertical hydraulic jacks connected to horizontal loading beam.  These hydraulic jacks were controlled manually and independent of each other. Manual movement was achieved by using two mono-block directional hydraulic control valves which allowed up and down movement of the jacks when controls were pressed up and down. The oil pressure allowed to pass to the jacks was adjusted using two pressure control valves which allowed for incremental load increases while testing progressed. Initially the pressure was set to a small value and increased in increments until the loading created vertical displacements. Once the displacement stopped for the given load level, the pressure was increased and load was again applied. This procedure was repeated until failure or until the desired displacement level was achieved. The speed of vertical movement was controlled by two flow rate valves. A level was placed on the loading beam to ensure that displacements were occurring equally in both hydraulic jack. This setup allowed for easy and accurate vertical movement of the loading beam. A photo of the manual control system is shown in Figure 3.6. Chapter 3 Experimental Program 35   Figure 3.6: Manual vertical control setup.  3.2.3 Horizontal Loading The horizontal loading was applied to the end of the horizontal loading beam by a MTS servo- controlled hydraulic jack with a maximum stroke of 24 inches as shown in Figure 3.7 and labeled H in Figure 3.1. Chapter 3 Experimental Program 36   The horizontal load was applied eccentrically to ensure the load was applied directly down the line of connection (Figure 3.8). If this was not done, an undesired bending moment would have been applied to the connector which would not be present in the actual connector of a tilt-up structure. Figure 3.8: Eccentric horizontal loading.  Figure 3.7Horizontal hydraulic jack.  Chapter 3 Experimental Program 37   During horizontal loading it was desired to load exactly horizontally. Since small vertical displacements increments were needed, a slotted plate was fabricated and inserted at the support end of the horizontal jack to allow fine adjustments of the vertical position to ensure that loading was always applied horizontally (Figure 3.9). Figure 3.9: Slotted plate to ensure horizontal loading.  3.3 Test Specimens Five concrete blocks were constructed in the UBC structures laboratory with each block having four EM5 embedded connectors slab specimens, one on each of the four long edges of the block. This allowed for a total of 20 tests specimens, of which 18 were used. Four test specimens were created in each concrete block as it made more efficient use of materials such as concrete forms. The damage caused by one failed specimen had no adverse effects on the results of the other specimens within the same concrete block because the damage from each test was localized. Each of the five concrete blocks were 60 inches long, 39 inches wide and 19 inches deep. The concrete specified was 25 MPa, type 10 concrete, with 20 mm aggregate, which is common for a typical slab on grade within tilt-up structure. A 15M reinforcing bar was included around the perimeter of the slab and was placed over the bent reinforcing bars of the EM5 connector as is done in practice. A grid of 15M reinforcing steel was laid at 500 mm on center, which is common for a typical slab on grade. Figure 3.10 shows a photo of an EM5 connector region from Chapter 3 Experimental Program 38   a tilt-up construction site and is compared with that of the test specimens. Figure 3.11 shows a detailed orthographic drawing of the test specimen and Figure 3.12 shows a photo of a specimen the concrete was cast. (a)                                                                (b)  Figure 3.10: (a) Field constructed connection; (b) Test specimen.  Chapter 3 Experimental Program 39   Figure 3.11: Orthographic drawing of test specimens.  C hapter 3 Experim ental Program40   Figure 3.12: Test specimen steel layout before concrete casting.  Figure 3.13 shows a photo of the specimen’s concrete forms. The five specimens were formed together using 2 X 10 dimensional lumber. Forming all five specimens together allowed for self- equilibrating faces which reduced lumber usage by utilizing the common form between specimens, and reduced bracing material required. Figure 3.13: Test specimen concrete forms before casting of concrete.  Chapter 3 Experimental Program 41   Four holes of 100 mm diameter were formed at each corner using 100 mm diameter PVC pipe. These holes were used in fastening the concrete specimen to the strong floor during testing. Hollow steel pins, which were machined to the diameter matching that of the strong floor holes, were placed in the strong floor which aligned with the 100 mm holes at the corners of the concrete specimens. These holes were then grouted with no-shrinking grout to create a complete connection between the concrete specimen and the strong floor. Vertical rods then passed through the center holes of the pins to restrict the concrete specimen from moving vertically. This system can be seen in the photo and detail drawing of Figure 3.14 below. Figure 3.14: Photo and detail drawing of the system preventing movement of the slab.  100 mm holes in slab Chapter 3 Experimental Program 42   Figure 3.15 shows a photo of the concrete placement which occurred on May 4, 2008. As shown in the photo, concrete was placed in lifts to ensure that no bowing of the self equilibrating form faces occurred. Figure 3.16 shows the concrete specimens after surface finishing was completed. Figure 3.15: Concrete being cast.  Figure 3.16: Finished concrete specimens.  Chapter 3 Experimental Program 43   3.4 Material Properties The material properties of interest in this study are the properties of the concrete and the strength of the reinforcing steel used in the EM5 embed connectors. Although these values will not be used at any point within this study, it is important to include these specific test details for completeness and to possibly explain any potential differences between the results presented in this study and future testing conducted. 3.4.1 Concrete Properties The concrete strength that was specified for the specimens was 25 MPa concrete. For convenience and practical reasons, concrete strength generally refers to the uniaxial compressive strength of a standard test cylinder. This is used as a quality control as it is a standardized test which allows the strength of the concrete to be compared directly without other factors which might affect the actual in-situ concrete strength. The procedure used to measure the concrete strength of the concrete used in the test specimens involved slowly loading the test cylinders in compression at a loading rate of 0.25 – 0.35 MPa/s. The cylinders, which were 100 mm in diameter by 200 mm long, were allowed to cure in their moulds for approximately the first 24 hours at a temperature of approximately 20°C. Half the concrete cylinders were then placed in a water bath saturated with lime for the remainder of the curing time and half were left beside the test specimens to be field cured. The moist curing conditions allow for optimal and consistent curing, which is important if results are to be reliable and comparable to other samples and the field cured cylinders were cured in the same conditions as the test specimens, instead of in the water bath. These results are often used in practice to determine when concrete forms can be removed, or when a structure can begin to be used, as these cylinders will more closely reflect the actual strength of the concrete (although not exactly as conditions within a large body of concrete are still different than a field cured cylinder). Table 3.1 shows the concrete cylinder compression test results for the moist cured cylinders as well as the field cured and Figure 3.17 shows photos of the cylinder compression test. It is important to note that the moist cured cylinders were tested moist, and the field cured cylinders were tested dry. Chapter 3 Experimental Program 44   Table 3.1: Concrete cylinder compression test results from 133 days.  Specimen ID Age Moist/Field Cured Diameter Length Weight Peak Load Strength Ave [Days] [mm] [mm] [g] [kN] [MPa] [MPa] 1 133 Field 101 197 3633 185.5 23.2 22.9 2 133 Field 102 199.5 3648 186.1 22.8 3 133 Field 101 199.5 3631 180.7 22.6 4 133 Moist 101 198.5 3812 298.7 37.3 37.0 5 133 Moist 102 198 3814 303.2 37.1 6 133 Moist 100 200 3828 294.2 37.5 7 133 Moist 102.5 200 3857 298.8 36.2 (a)  (b)  Figure 3.17: Concrete cylinder compression tests  According to CSA Standard A23.1, in order for the concrete to be accepted as satisfactory, the averages of all sets of three consecutive strength tests for moist cured cylinders must exceed the specified 28 day strength, and no single test (average of two cylinder values is considered a test) can be more than 3.5 MPa lower than the 28 day specified strength. During the current experimental program, 28 day cylinder tests were not preformed, which was an oversight, and so Chapter 3 Experimental Program 45   the data from the cylinder tests were collected 133 days after casting. The ACI committee 209 (ACI, 1982) has proposed a model for the strength gain of moist cured concrete. This model can be used to back predict what the 28 day strength of the moist cured cylinders would have been: ݂Ԣ௖ሺ௧ሻ ൌ ݂Ԣ௖ሺଶ଼ሻ ൬ ݐ 4 ൅ 0.85ݐ ൰ ݂Ԣ௖ሺଶ଼ሻ ൌ ݂Ԣ௖ሺଵଷଷሻ ൬ 4 ൅ 0.85ݐ ݐ ൰ ݂Ԣ௖ሺଶ଼ሻ ൌ 37.0 ܯܲܽ ൬ 4 ൅ 0.85 ൈ 133 ݀ܽݕݏ 133 ݀ܽݕݏ ൰ ݂Ԣ௖ሺଶ଼ሻ ൌ 32.6 ܯܲܽ The dry cured cylinders produced peak loads much lower than the moist cured cylinders at 133 days. The dry cured cylinders had a compressive strength of 22.9 MPa verses 37 MPa for the moist cured cylinders as shown in Table 3.1. This represents a 38% decrease in strength which was as a result of curing conditions. Figure 3.18 shows a plot from previous testing which shows how the compressive strengths of a concrete cylinder test is affected by various curing and testing conditions. Chapter 3 Experimental Program 46   Figure 3.18: Effect of moist curing on compressive strength of concrete.  (Modified from Price, 1951)  From studying Figure 3.18, it is noted that an air cured cylinder which is dry when testing produces a compressive strength about 40% lower than that of a standard moist cured cylinder which is moist when testing when testing at any time after about 3 months. In the present study, the cylinder tests occurred at 133 days (~4.4 months), and so a 38% difference between the moist and dry cylinder results is expected. The next important property of the concrete which was of interest was the tensile strength. This property is important to the current study, as the concrete cover over the bent reinforcing steel of the EM5 connector fails initially in tension when experiencing vertical displacement. The test that was used to determine the tensile strength of concrete was the split cylinder test. In a split cylinder test, a standard cylinder lying on its side (Figure 3.19) is stressed in biaxial tension and compression which creates failure in tension along the vertical diameter of the cylinder. 0 20 40 60 80 100 120 0 1 2 3 4 5 6 7 8 9 10 11 12 Re la ti ve  S tr en gt h  [% ] Time [Months] Chapter 3 Experimental Program 47   (a) Test setup for split cylinder test.  (b) Cracking of cylinder in split cylinder test.  Figure 3.19: Split cylinder test.   This Tension stress is typically between 8 – 15% of the compressive strength found using the cylinder compression test. The force (P) measured at failure can be converted to the splitting tensile strength (fct) using the following equation: ௖݂௧ ൌ 2ܲ ߨ ݈ ݀ Where: fୡ୲ ൌ Splitting tensile strngth ሺMPaሻ P ൌ Force required to split cylinder ሺkNሻ l ൌ Length of Cylinder ሺmmሻ d ൌ Diameter of Cylinder ሺmmሻ The tensile strength of the concrete ( fcr ) can then be estimated using the following equation: ௖݂௥ ൌ 0.65 ௖݂௧ Table 3.2 shows the results of the split cylinder tests conducted at 133 days and predicts the average tensile strength for the moist cured and field cured concrete. Chapter 3 Experimental Program 48   Table 3.2: Results of split cylinder tests.  ID Age (Days) Moist/Field Cured Dia. Length Weight Peak Load Split Cylinder Strength Ave Slit Cylinder Strength fct  Ave Tensile Strength fcr [mm] [mm] [g] [kN] [MPa] [MPa] [MPa] 1 133 Field 101 201 3716 74.5 2.3 2.15 1.4 2 133 Field 101 202 3721 65.3 2.0 3 133 Moist 102 204 3869 112.1 3.4 3.4 2.2 4 133 Moist 99.5 206 3889 109.7 3.4 A general equation to estimate the tensile strength of concrete without performing tests is as follows: ௖݂௥ ൌ 0.33ඥ ௖݂ᇱ If this equation is used, a moist cured tensile strength of 2.0 MPa and a field cured tensile strength of 1.6 MPa. This is in close agreement with the results presented in Table 3.2. Based on these tests, the tensile strength of the concrete used in this study was 2.2 MPa which is 6% of the compressive strength. Since the tension strength is related to the compression strength, we would expect the dry cylinder tests to produce a tensile strength about 38% lower than the moist cylinder samples just as in the cylinder compression tests. The results shown in Table 3.2 indicate that indeed a 37% lower tensile strength was observed in the dry cylinder specimens when compared to the moist samples. 3.4.2 Reinforcing Steel Properties The last material property that is of interest in this study is the steel strength of the reinforcing bars used to make the EM5 embed connector. This reinforcing steel specified is a standard 400W grade 20M reinforcing bar, which is to conform to CSA G30.18 specifications. Tension testing was conducted using a Baldwin testing machine shown in Figure 3.20. The tension testing was Chapter 3 Experimental Program 49   conducted by placing a portion of the EM5 connector reinforcing bar between the grips with a gauge length of 200 mm, as shown in Figure 3.20. Figure 3.20: Reinforcing steel tension test setup.  An error was made however in the placement of the LVDT which measured the elongation. Instead of measuring the elongation of the bar within the gauge length, the displacement of the loading machine was reported. This created an incorrect Modulus of Elasticity since initially when loading began slippage between the loading grips and the bar occurred until the grip fully locked onto the specimen. Because of this, only the measured reinforcing strengths will be reported, with confidence that the strains would have been consistent with known steel properties had the proper instrumentation been used. According to CSA G30.18, a 400W reinforcing bar is acceptable if it meets the criteria shown in Table 3.3. Chapter 3 Experimental Program 50   Table 3.3: Mechanical properties of 20M  reinforcing bars from CSA G30.18.  Weldable Low- alloy Bars Grade 400W Minimum tensile strength 540 MPa† Yield strength Minimum 400 MPa Maximum 525 MPa Minimum elongation in 200mm gauge length 20 M 13% Table 3.4 shows the strength results from the tension tests conducted on the 20M bars used in the EM5 connectors and Figure 3.21 shows the stress – deformation relationship. Since the deformation included slippage between the bars and the testing grips, the deformation values will be omitted. Table 3.4: Reinforcing steel tension test results for  20M reinforcing bar.  Test Number Yield Strength Ultimate Stress [MPa] [MPa] 1 464 587 2 480 616 3 473 611 Chapter 3 Experimental Program 51   Figure 3.21: Stress ‐ deformation curve for reinforcing steel.  † Deformations were combination of steel deformations and grip slippage, so values not provided.  It is observed from these results that all of the criteria shown in Table 3.3 from CSA G30.18 are met, and so the steel used in this study is acceptable. 3.5 Instrumentation This study was interested in determining the pure response of the EM5 embed connector subjected to pure vertical shear in the first stage of testing, followed by pure horizontal shear in the second stage of testing. To accurately measure the forces applied to the connector and the relative displacements between the concrete slab and the loading plate (simulating the wall panel) caused by those forces required careful consideration of test setup as discussed already as well as strategic placement of instrumentation. This section will focus on the instrumentation 0 100 200 300 400 500 600 700 St re ss  [M Pa ] Deformation [mm] † Steel Tension Test 1 Steel Tension Test 2 Steel Tension Test 3 Chapter 3 Experimental Program 52   used for the testing program. Figure 3.22 shows an overview rendering of all instrumentation used and the placements within the test setup. Each of these instruments sent a separate signal to the data acquisition system (Daisylab 8.0) which collected and recorded a reading from each device ten times per second, which could be edited to form the response presented later in Chapter 4. Figure 3.22: Test setup instrumentation overview.  Horizontal Force The horizontal force was measured using a MTS 100 kip load cell placed at the end of the horizontal hydraulic jack as shown in Figure 3.23. Chapter 3 Experimental Program 53   Figure 3.23: Load cell to measure horizontal force.  Vertical Force The vertical force was measured independently for each vertical hydraulic jack. Each was measured by a MTS 50 kip load cell placed at the end of each hydraulic jack as shown in Figure 3.24. The sum of the force measured in each load cell was taken as the force applied in the vertical direction. Figure 3.24: Load cells to measure vertical force.  Chapter 3 Experimental Program 54   Horizontal Displacement The horizontal displacement was measured by two 24 inch string pods. The string pod bodies were attached to the vertical column behind the loading plate and the string was connected to a reinforcing bar which was attached to the vertical threaded rod holding the slab to the strong floor (Figure 3.25). This placement of the string pods allowed for the relative movement between the loading plate and the slab to be measured, which is the amount of horizontal deformation the EM5 embed connector experienced. The horizontal hydraulic jack also had a built in actuator displacement transducer which measured the displacement of the hydraulic jack. This reading was not as accurate as those measured by the string pods because slight movements at connection interfaces as well as within the actuator clevises caused slightly larger displacement recordings. Because of this the results from this device have been discarded. Although the string pods generated values very consistent to each other, the horizontal displacements values used in this study were the averaged displacement values given by the two sting pods connected to the loading beam. Figure 3.25: String pods to measure horizontal displacements.  Chapter 3 Experimental Program 55   Vertical Displacement The vertical displacement was measured by a 10 inch string pod, whose body was connected to the underside of the horizontal loading beam and the sting to a hook on the floor (Figure 3.26). This allowed for the movement between the loading plate and the slab to be measured as there was no movement between the floor and the testing slab because of the fixed connection provided. Figure 3.26: String pod to measure vertical displacements.  Out of Plane Displacement Out of plane movements were recorded by a 100 mm linear pod (Figure 3.27). This measured the out of plane horizontal displacement between the floor and the horizontal loading beam. Figure 3.27: Linear pod to measure out of plane movements.  Chapter 3 Experimental Program 56   3.6 Loading Protocol The objectives of the test program in this study required the horizontal shear response to be measured for various levels of vertical damage. For most tests this required a two part testing program as described in Section 3.1, however the vertical response to failure is also of interest. To determine this, only vertical displacements were applied until failure. The first tests were to measure the vertical monotonic response. In these tests, vertical load was applied until failure occurred. Next the cyclic response in vertical uplift was measured by subjecting the connectors to cyclical vertical displacements. Three cycles were performed to vertical displacement steps starting at 50 mm, and continuing at 25 mm increments (eg, 75 mm, 100 mm …) until failure. Each time a cycle was reversed, the connector was forced back to zero displacement. These loading methods are shown in Figure 3.28. (a) Monotonic uplift test. 0 25 50 75 100 U pl ift  D is pl ac em en t [ m m ] Time [t] Failure Chapter 3 Experimental Program 57   (b) Cyclic uplift test. Figure 3.28: Loading protocol for test specimens subjected to vertical uplift only.  Since the horizontal shear response of the EM5 connector was also of interest, both the horizontal monotonic and cyclical responses were measured. For the monotonic horizontal test, horizontal load was applied on the connector in one direction until failure occurred. For the cyclic horizontal tests, three cycles were performed at ±5 mm increments in horizontal jack displacement. This did not however translate into ±5 mm increments in connector displacements relative to the concrete slab, as there were some movements within the interfaces of the hydraulic jack and the loading column, and in the clevises of the hydraulic jack itself. These loading methods are shown in Figure 3.29. 0 25 50 75 100 U pl ift  D is pl ac em en t [ m m ] Time [t] Failure Chapter 3 Experimental Program 58   (a) Monotonic horizontal test. (b) Cyclic horizontal test. Figure 3.29: Loading protocol for test specimens subjected to horizontal shear only.  The last type of test conducted required the horizontal response with various levels of vertical damage to be measured. For these tests the vertical damage was generated by performing three cycles to the displacement level corresponding to the amount of damage required for each test then holding the vertical displacement either at the maximum vertical displacement level, or pushing it back to zero vertical displacement depending on the test. Horizontal loading was then applied in the same manner as above with three cycles being performed to ±5 mm increments in 0 5 10 15 20 H or iz on ta l D is pl ac em en t [ m m ] Time [t] ‐20 ‐15 ‐10 ‐5 0 5 10 15 20 H or iz on ta l D is pl ac em en t [ m m ] Time [t] Failure Failure Chapter 3 Experimental Program 59   horizontal jack displacement until failure of the specimen. These loading methods are shown in Figure 3.30. (a) Part 1 of loading (vertical damage) (b) Part 2 of loading (horizontal loading) Figure 3.30: Loading protocol for test specimens subjected to both vertical uplift and horizontal shear.  Each test performed followed one of these loading protocols depending on the test. Appendix A has a schematic showing the loading protocol used for each test specimen. U pl ift  D is pl ac em en t [ m m ] Time [t] ‐20 ‐15 ‐10 ‐5 0 5 10 15 20 H or iz on ta l D is pl ac em en t [ m m ] Time [t] Held Up Pushed Down Failure ΔvMax Chapter 3 Experimental Program 60 4 EXPERIMENTAL RESULTS AND DISCUSSION In this chapter the results from the experimental testing program are presented. This includes the separate responses in the vertical and horizontal directions as well as combined horizontal and vertical responses. The measured horizontal response is compared to the results from previous testing and the similarities and differences discussed. Interaction plots for the strength, stiffness and maximum displacement are provided, as well as a discussion on shortcomings of the current connection design. 4.1 Summary of Results The testing program was divided up into three distinct testing types, namely tests with only vertical displacement, tests with only horizontal displacement, and tests where both vertical and horizontal displacements. Table 4-1 provides a summary of all the test specimens organized into these three testing types. For the vertical responses, the maximum vertical load (PMax) and maximum vertical displacement (ΔvMax) are recorded as well as whether it was a cyclic or monotonic loading protocol. For the horizontal responses, the maximum positive and negative cyclic loads (VMax+ & VMax-) are recorded as well as the maximum positive and negative cyclic displacements (ΔhMax+ & ΔhMax-) and whether it was loaded cyclically or monotonically. These positive and negative cyclic values for the loads and displacement are averaged to give the maximum average cyclic load (VMaxAve) and displacement (ΔhMaxAve). For the test specimens where vertical and horizontal responses are of interest, the vertical and horizontal responses are recorded as described above in addition to the level of vertical displacement sustained during the horizontal cycling (Δv during Δh). This chapter will focus on the typical test results determined from studying all of the test performed, however information such as response plots, photos and testing notes for all test specimen is available in Appendix A. 61   Table 4‐1: Summary of testing results.  Specimen  Name  Age at  test  date    (Days)  Loading Description Weld  Detail  (Round /  Square)  Restrained  Out of  plane  movement  Results Fracture  Location  ΔvMax  (Mono/Cyclic)  Δv During Δh  (Mono/Cyclic)  Vertical  Horizontal  PMax ΔvMax Vmax (+/‐) ΔhMax (+ / ‐) [mm]  [mm] [kN] [mm]  [kN] [mm] Uplift shear only tests  1  64  Failure (M)  N/A Round No 87 78.1  N/A  N/A  No Failure† 2  70  Failure (M)  N/A Round No 166 131.3 N/A  N/A  Angle 11  108  Failure (M)  N/A Square No 125 105.9 N/A  N/A  Angle 3  73  Failure (C)  N/A Round No 86.2 75.3 1 N/A  N/A  Weld 1 3A  73  Failure (C)  N/A Round No 127 100.9 2 N/A  N/A  Rebar 2 8  100  Failure (C)  N/A Square No 138 100  N/A  N/A  Rebar 17  175  Failure (C)  N/A Square Yes 115 103.9 N/A  N/A  Rebar Horizontal shear only tests  5  80  N/A  0 (M) Round No N/A  N/A  +279 / ‐ +32.9 / ‐ Rebar 12  108  N/A  0 (M) Square No N/A  N/A  +280 / ‐ +19.2 / ‐ Rebar 6  95  N/A  0 (C) Round No N/A  N/A  +232 / ‐209 +16.9 / ‐19.7 Rebar Uplift and horizontal shear combined tests 16  121  25.3 (C)  25 (C) Square No 70 25.3  +166 / ‐249 +13.3 / ‐11.9 Rebar 15  114  25.5 (C)  0 (C) Square No 48 25.5  +239 / ‐191 +12.5 / ‐15.4 Rebar 4  79  50.7 (C)  50 (C) Round No 62 50.7  +195 / ‐203 Error / Error Rebar 13  112  50.9 (C)  50 (C) Square No 65 50.9  +184 / ‐130 +15.1 / ‐16.4 Rebar 14  113  50.3 (C)  0 (C) Square No 64 50.3  +200 / ‐192 +10.4 / ‐14.5 Rebar 7  99  75.4 (C)  75 (C) Round No 108 75.4 1 +179 / ‐171 +17.5 / ‐18.5 Weld 1 7A  99  75.4 (C)  75 (C) Round No 108 75.4 2 +179 / ‐171 +17.5 / ‐18.5 Rebar 2 18  176  75.1 (C)  0 (C) Square Yes 77 75.1  +197 / ‐171 +20.5 / ‐19.9 Rebar 9  101  100 (C)  100 (C) Square No 105 100.1 +143 / ‐149 +14.5 / ‐13.9 Weld 10  102  100.1 (C)  0 (C) Square No 127 101  +133 / ‐152 +19.2 / ‐19.4 Rebar †  Test setup movement lead to inaccurate results, therefore test was aborted.  1 Wall – Angle weld failure caused premature connection failure (see section 4.2.3). Repaired weld and testing continued (labelled 3A,  7A).  2  Wall ‐ Angle weld repaired 3 times to achieve displacement capacities shown  C hapter 4 Experim ental R esults and D iscussion62   4.2 Vertical Response Previous testing conducted on tilt-up connectors described in Chapter 2 only focused on the horizontal shear response as well as the horizontal “tension” pullout. This section will focus on the vertical uplift response, which was not measured in previous testing. Both the monotonic and cyclic responses were measured, and the vertical tests which resulted in vertical failure will be presented separately. 4.2.1 Monotonic Vertical Response The testing program commenced with a monotonic vertical shear test, followed by a cyclic vertical shear test. A sample monotonic vertical shear response plot for specimen 2 is shown in Figure 4.2, initial concrete cracking occurred at about P = 15 kN. This is shown in the photo of Figure 4.1(a). (a)  (b)  Figure 4.1: (a) Initial concrete cracking; (b) Popping off of concrete cover; for specimen 2.  Once concrete cracking occurred the secondary slope stayed relatively linear until Δv = 36 mm. At this point the concrete cover “popped off” similar to what would occur in a punching shear failure (photo (b) in Figure 4.1).. With the cover being completely removed and no longer providing vertical resistance Because the bent reinforcing bar in the slab was still oriented at a Chapter 4 Experimental Results and Discussion 63   very shallow angle in the slab, and a longer length of bar was exposed, the vertical force dropped from P = 61 kN to P = 47 kN as the exposed bar was allowed to bend upwards until the slab edge made contact with the wall panel. As Δv increased, it forced the rebar into pure tension which increased the force again until failure of the connector at PMax = 166 kN and ΔvMax = 132 mm. Although the general trends of all vertical specimens were in close agreement, there was some variablility in the locations of the vertical force drop. Figure 4.2: Monotonic vertical shear response with round connector type for specimen 2.  4.2.2 Cyclic Vertical Response The next tests presented are the tests conducted using a cyclic vertical loading protocol. Figure 4.3 shows the vertical cyclic response plot of specimen 3, which can be used to show a typical result. During the first cycle to Δv = 50 mm, concrete cracking occurred at P150 = 33 kN (ܲ஼௬௖௟௘ #∆௩ሻ. Again, a near linear secant stiffness is measured after initial cracking. Two more cycles to Δv = 50 mm produced load levels of P250 = 69 kN and P350 = 57 kN. The first cycle to Δh = 75 mm had a load capacity of P175 = 69 kN after showing a similar drop in load level when 0 20 40 60 80 100 120 140 160 180 0 20 40 60 80 100 120 140 P  [k N ] ∆v [mm] Chapter 4 Experimental Results and Discussion 64   the concrete cover became ineffective, just like in the monotonic test. The difference was that in this test specimen, it occurred at Δv = 63 mm which is at a considerably larger vertical displacement level than the Δv = 37 mm where it occurred previously. The second cycle at Δv = 75 mm produced load levels of P275 = 58 kN before the wall to connector weld failed during the third cycle. Figure 4.4(a) shows that as the connector was cycled back down to Δv = 0 mm, a torque was produced on the wall – connector weld. This torque created rotation about the weld, and since welds are very weak and brittle when subjected to this rotational loading, it broke during the third vertical cycle (Figure 4.4(b)). Figure 4.3: Cyclic vertical shear response with round filler bar (Specimen 3).  -40 -20 0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 P  [k N ] ∆v [mm] Chapter 4 Experimental Results and Discussion 65   (a)  (b)  Figure 4.4: (a) EM5 connector being loaded down to Δv=0mm, (b): Wall – connector weld failure  (Specimen 3).  Figure 4.5 shows both the sample monotonic and cyclic uplift response plots for specimens 2 and 3 with the round filler bar. It is evident that the current connection detail does not work well in cyclic uplift, as it produces a brittle failure at only Δv = 75 mm which is less than 57% of the monotonic displacement capacity measured. Although within a real wall system this connection would have no more load carrying capacity at this point, it was not the actual EM5 connector that had failed, but only the weld attaching it to the wall panel. This is a very important discovery, however since the response of the EM5 connector itself is also of interest, if a revised panel – connector detail could be designed to prevent this weld from failing then the actual EM5 connector response could be measured. This is in fact the direction this study was required to take, and the new connection design is described in Section 4.2.3. Chapter 4 Experimental Results and Discussion 66   Figure 4.5: Cyclic and monotonic vertical shear response with round connector type.  For the typical cyclic vertical test presented in this section, the weld was required to be repaired three times for testing to be continued. Figure 4.6 shows both the monotonic vertical response of specimen 2 and the cyclic response with the repaired wall – slab connector weld (specimen 3A). The connector with the repaired weld was able to achieve three cycles at Δv = 100 mm before EM5 connector failure occurred due to rebar fracture, shown in the photo of Figure 4.7. Specimen 7, which was another specimen cycled to large vertical displacements (75 mm) using the round filler bar, also required three repairs of the wall – slab connection weld in order to achieve EM5 connector failure at Δv = 100 mm. These results show that a 33% gain in displacement capacity can be gained consistently if a new weld connection design would be adopted in practice. -40 -20 0 20 40 60 80 100 120 140 160 180 0 20 40 60 80 100 120 140 P  [k N ] ∆v [mm] Specimen 2 Specimen 3 Chapter 4 Experimental Results and Discussion 67   Figure 4.7: Rebar fracture of vertical cyclic test (Specimen 3A).  Figure 4.6: Monotonic and cyclic vertical shear response with repaired wall ‐ slab  connection weld.  -40 -20 0 20 40 60 80 100 120 140 160 180 0 20 40 60 80 100 120 140 P  [k N ] ∆v [mm] Cyclic vertical response  Specimen 3(without  repaired weld) Cyclic vertical response  Specimen 3A (with  repaired weld) Monotonic vertical  response (Specimen 2) Chapter 4 Experimental Results and Discussion 68   4.2.3 Redesigned Connection Detail Based on the results from the two vertical cyclic tests, it was apparent that a new wall – slab connection detail was needed. Both of these tests needed the connection weld to be repaired three times in order to finally achieve an overall EM5 connector failure. As described above, this was due to a torque being applied to the weld due to the rotation of the connector angle as the displacement were forced back to Δv = 0 mm. Figure 4.8 shows a photo of an extreme case of this rotation from the test to specimen 7, which led to failure as soon as a positive displacements was applied. Figure 4.8: Wall ‐ slab connector weld, leading to failure (Specimen 3).  Since in current design the panel – slab connection are designed based on horizontal force with the effects of vertical movements ignored, the connector detail was designed to carry only horizontal shear force. This study shows that a new connection detail is needed which can resist rotation of the EM5 edge angle during vertical deformations (required for rocking mechanism) to prevent brittle failure at displacement capacities well short of their potential. A new weld connection detail was developed by replacing the round filler bar in the current detail with a 38 mm (1.5”) square bar. This allows for 25 mm long fillet welds to be placed on either side of the bar, and in the vertical direction on the wall connector. These small welds were effective in holding the connector angle in place so that no rotation occurred. Figure 4.9 shows a Chapter 4 Experimental Results and Discussion 69   detail drawing of the current connection along with a photo, and Figure 4.10 shows a detail drawing of the revised connection along with a photo. Figure 4.9: Current wall‐slab connection detail.  Round Bar connection detail Chapter 4 Experimental Results and Discussion 70   Figure 4.10: Proposed revised wall‐slab connection detail.  Square Bar connection detail Chapter 4 Experimental Results and Discussion 71   The first seven tests were conducted using the old weld detail, and all other tests were conducted using the new one. Monotonic and cyclic vertical shear tests were repeated to determine how the responses differ with that of the old connector detail. The monotonic response is similar except that the ultimate displacement capacity for the new connection detail fell from 132 mm in specimen 2 to 106 mm in specimen 11 (~20% difference). A small decrease was expected in this case, as the new connection detail provides a much stiffer connection, and the deformations at the connection location leading up to failure were significantly reduced (ie. rotational deformations). Figure 4.11 shows the comparison of the cyclic responses using the current connection detail (specimen 3) and the revised detail (specimen 8). It can be seen that the responses for the two are very similar, except the new connection detail achieves Δv = 100 mm whereas the current connector detail only achieves Δv = 75 mm. Figure 4.11: Cyclic vertical shear response with round and square connector type.  Figure 4.12 shows the monotonic and cyclic responses for the revised connection detail (specimen 11 and 8 respectively). These results show that the new connection detail performs in a very similar matter as the currently used connection detail except that the new connection -40 -20 0 20 40 60 80 100 120 140 160 180 0 20 40 60 80 100 120 140 P [k N ] ∆v [mm] Square connector type  (Specimen 8) Round connector type  (Specimen 3)(without  repaired weld) Chapter 4 Experimental Results and Discussion 72   detail performs drastically better in cyclic uplift, which allows the connector to reach its full vertical displacement potential. Figure 4.12: Monotonic and cyclic responses with square connector type.  4.2.4 Vertical Response Trends There is a number of trends that are apparent when the results from all fifteen test specimens with a vertical test component are compared. Figure 4.13 shows all the specimens which experienced vertical uplift plotted together. It is important to note that only six of these tests were conducted to failure and so the apparent displacement capacities of the specimens not reaching failure must be disregarded. An accurate bilinear model response trend is able to be fit to this data quite easily due to the low variability in the response shape. Figure 4.13 includes the model fit for the vertical uplift response which is used in the pushover model proposed in chapter 5 of this study. This model fit has the following equation: ܲ ൌ ൞ 20  ݇ܰ ݉݉ ൈ ∆ݒ                         ; 0 ݉݉ ൑ ∆ݒ ൏ 1.5 ݉݉ 0.86  ݇ܰ ݉݉ ൈ ∆ݒ ൅ 28.71 ݇ܰ   ; 1.5 ݉݉ ൑ ∆ݒ ൑ 100 ݉݉ -40 -20 0 20 40 60 80 100 120 140 160 180 0 20 40 60 80 100 120 140 P  [k N ] ∆v [mm] Cyclic vertical response  (Specimen 8) Monotonic vertical  response (Specimen 11) Chapter 4 Experimental Results and Discussion 73   Figure 4.13: All first cycle vertical response envelopes and model curve fit.  The displacement capacities for all the vertical uplift tests are presented in Table 4-2 . A vertical cyclic displacement capacity of 100 mm was easily chosen due to the extremely consistent displacement value which produced failure specifically within the cyclic tests. Table 4‐2: Vertical displacement  capacities.  Specimen Name Vertical Displacement Capacity [mm] Uplift shear only tests 2 131.3† 11 105.9† 3 100.8 8 100.0 17 103.9 † Monotonic Tests 0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 P  [k N ] Δv [mm] V Monotonic 1 V Monotonic 2 V Cyclic 1 V Cyclic 2 V Cyclic 3 100mm Held Up 100mm Pushed Down 75mm Held Up 75mm Pushed Down 50mm Held Up 50mm Held Up 2 50mm Pushed Down 25mm Pushed Down 25mm Held Up Model Curve Chapter 4 Experimental Results and Discussion 74   The next trend that was able to be determined through the results was how much apparent load capacity was lost as multiple cycles were performed. Figure 4.14 shows the average load P for all the valid vertical cyclic tests performed to each displacement level for each of the three cycles performed. For example, test numbers 3, 7, 8 and 18 all experienced three cycles to a displacement level of 75 mm. Figure 4.14 shows a data point representing the average of the maximum force achieved by these four tests for each of the three cycles performed. Figure 4.14: Average load at various displacement levels for each cycle number.  At Δv = 25 mm the average load P for all valid tests for cycles 1, 2 and 3 were P1 = 58.9 kN, P2 = 50.9 kN and P3 = 47.5 kN respectively. This showed a second cycle apparent load capacity loss of 13.7% and a third cycle loss of a further 5.6% for a total loss of 19.3%. At Δv = 50 mm the average load P for all valid tests for cycles 1, 2 and 3 were P1 = 71.5 kN, P2 = 55 kN and P3 = 48.8 kN respectively. This showed a second cycle apparent load capacity loss of 23.2% and a third cycle loss of a further 8.6% for a total loss of 31.8%. At Δv = 75 mm the average load P for all valid tests for cycles 1, 2 and 3 were P1 = 90.7 kN, P2 = 60.1 kN and P3 = 47.4 kN respectively. This showed a second cycle apparent load capacity loss of 33.8% and a third cycle loss of a further 14% for a total loss of 47.8%. At Δv = 100 mm the average load P for all valid 0 20 40 60 80 100 120 140 0 20 40 60 80 100 A ve ra ge P [k N ] ∆v [mm] Vertical Model Curve Ave Cycle 1 Envelope Ave Cycle 2 Envelope Ave Cycle 3 Envelope Chapter 4 Experimental Results and Discussion 75   tests for cycles 1, 2 and 3 were P1 = 124.1 kN, P2 = 95.9 kN and P3 = 91.1 kN respectively. This showed a second cycle apparent load capacity loss of 22.8% and a third cycle loss of a further 3.9% for a total loss of 26.6%. These results also show that until Δv = 75 mm ~70% of the total vertical load capacity loss happened in the first cycle with the remaining ~30% happening due to the second cycle. However when displacements were extremely large (Δv = 100 mm) the first cycle contributed a much larger ~86% of the total capacity loss with the second cycle accounting for the remaining ~14% loss. It is important to note that this is apparent load capacity loss, not actual capacity loss. This is because if displacements proceed past the previous maximum displacement level, it will continue on the first cycle path, even if it had completed multiple cycles to a lower displacement level. This result may be important in post-earthquake analysis, as once the connectors are cycled to a specific maximum vertical displacement level, they will only provide a fraction of the expected resistance during a subsequent event with similar displacement demands. It also shows that this fraction will decrease as the connector’s displacement capacity is approached during the initial event. 4.3 Horizontal Response The horizontal response test results for the EM5 connectors showed that the revised connector detail proposed in section 4.2.3 had no impact on the response. For both monotonic and cyclic tests it behaved the same as the currently used connection detail. This is due to the fact that in both details, the wall is connected to the slab connector by two – six inch fillet welds. The horizontal response results therefore are not be presented separately as they are for the vertical responses. 4.3.1 Monotonic Horizontal Response The first horizontal monotonic test was conducted with the currently used connection detail in test number 5. It was loaded to a maximum load of V = 279 kN and failed due to rebar fracture at a displacement of Δh = 33 mm. The second monotonic test was conducted with the proposed Chapter 4 Experimental Results and Discussion 76   revised connection detail in test number 12 and this test produced a maximum load capacity of V = 280 kN and failed due to rebar fracture at a displacement of Δh = 21 mm. The load – displacement plots for both of these are shown in Figure 4.15. From observation of the failure results it was noticed that the discrepancy in the monotonic displacement capacities came from the amount of damage to the concrete surrounding the bar that ultimately fractured. In test number 5, the concrete cover was erroneously damaged, which exposed much of the bent rebar (Figure 4.16(a)). This resulted in a much less stiff response after cracking, which allowed significantly higher displacements before the rebar failed in tension. This accidental damage was due to an error in the testing setup for test number 5 in which the loading beam was not held in the horizontal position. This allowed a rotational moment to be applied to the connector, which created uplift in the concrete and damaged the concrete cover of a large section shown in the photo of Figure 4.16(a). In test number 12 this accidental damage did not occur (shown in Figure 4.16(b)). This resulted in a much lower displacement capacity. Despite this, the strength capacities are in agreement, which shows consistency within results, however, test number 5 will be disregarded because of the erroneous displacement capacities, and test number 12 will be used from this point on as the monotonic horizontal envelope. Figure 4.15: Monotonic horizontal shear responses.  0 50 100 150 200 250 300 0 5 10 15 20 25 30 35 V  [k N ] ∆h [mm] Specimen 5 Specimen 12 Chapter 4 Experimental Results and Discussion 77   (a)  (b)  Figure 4.16: Concrete cover damage for monotonic tests (a) 5 and (b) 12.  4.3.2 Cyclic Horizontal Response The cyclic horizontal shear response measured in test number 6 is shown in Figure 4.17 along with the monotonic horizontal shear envelope from specimen 12. Figure 4.17: Monotonic (Specimen 12) and cyclic horizontal shear responses (Specimen  6).  -250 -200 -150 -100 -50 0 50 100 150 200 250 300 -25 -20 -15 -10 -5 0 5 10 15 20 25 V  [k N ] ∆h [mm] Cyclic Monotonic Premature Damage Chapter 4 Experimental Results and Discussion 78   The first observation is that the measured cyclic response envelope is 15% lower in load capacity than that of the monotonic response, and 25% lower in displacement capacity. For the cyclic response, in the positive loading direction at an average displacement of Δh = 1.4 mm, the connector produced load capacities of V11.4 = 193 kN, V21.4 = 190 kN and V31.4 = 189 kN for the three cycles, while in the negative direction at an average displacement of Δh = -3.2 mm the connector produced load capacities of V1-3.4 = -182 kN, V2-3.4 = -180 kN and V3-3.4 = -175 kN for the three cycles respectively. The connector was then loaded in the positive direction to an average of Δh = 5.5 mm which produced load capacities of V15.5 = 229 kN, V25.5 = 196 kN and V35.5 = 177 kN for the three cycles, while in the negative direction at an average displacement of Δh = -8.1 mm the connector produced load capacities of V1-8.1 = -209 kN, V2-8.1 = -182 kN and V3-8.1 = -166 kN for the three cycles respectively. The next displacement step was to a positive average displacement of Δh = 15.3 mm in which the connector produced load capacities of V115.3 = 232 kN, V215.3 = 198 kN and V315.3 = 170 kN for the three cycles, while in the negative direction at an average displacement of Δh = -18.5 mm the connector produced load capacities of V1-18.5 = -197 kN, V2-18.5 = -156 kN and V3-18.5 = -123 kN for the three cycles respectively. The EM5 connector is not only used in the panel – slab connection, but also in the panel – panel connection. This is important for modeling a multi-panel system response, as a panel – panel connection is made of two EM5 connectors attached in series. Because they are attached in series, the displacement at a particular force level will be two times as much as if there were only one connector. This means that as the panel rocks, horizontal shear forces will be applied to the EM5 connector and failure will occur after a horizontal displacement equal to twice the horizontal displacement capacity of one EM5 connector. Figure 4.18 shows a plot of what the panel – panel response would be if two specimen 6 responses were connected in series. From this the tri-linear response plot used in the model was chosen as: ܸ ൌ ە ۖ ۔ ۖ ۓ 74  ݇ܰ ݉݉ ൈ ∆݄                         ; 0 ݉݉ ൑ ∆݄ ൏ 0.54 ݉݉ 56.4  ݇ܰ ݉݉ ൈ ∆݄ ൅ 9.5 ݇ܰ   ; 0.54 ݉݉ ൑ ∆݄ ൑ 3.82 ݉݉ 225                                           ; 3.82 ݉݉ ൑ ∆݄ ൑ 30 ݉݉ Chapter 4 Experimental Results and Discussion 79   The method for choosing this model curve is found in Section 4.4.1. Figure 4.18: Horizontal shear response of two EM5 connectors in series used as the panel – panel  connections.  4.3.3 Out-of-plane Movement During Testing As discussed in Chapter 3, the initial testing setup allowed for out-of-plane movements of the loading beam which are caused from the EM5 connector as testing progressed. Although in a real line of wall panels there would be some resistance to out-of-plane movements at the base of the wall panels from friction, as well as continuity between adjacent wall panels, this is thought to provide little resistance. Also as load reversals occur, it is conceivable that damage to the connectors and concrete surrounding them would not allow panels to rest on the foundation, but be bearing entirely on the connectors, effectively eliminating friction at the base of the panels. As this is somewhat speculation because the true resistance to out-of-plane movements is unclear, two cases were studied to see what the effect of out-of-plane movements were on the overall response. Sixteen tests were conducted while allowing free out-of-plane movement and two tests were repeated while restricting out-of-plane movements. There was no distinguishable difference between the load or displacement capacities of these two test setups. This means that -250 -200 -150 -100 -50 0 50 100 150 200 250 300 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 V  [k N ] ∆h [mm] P‐P Connector Response Model Response Chapter 4 Experimental Results and Discussion 80   as far as connector response properties are concerned, the true out-of-plane restriction conditions have no bearing on the results of the system, and therefore can be neglected. Figure 4.19 shows a sample of the typical out-of-plane movements which occurred during a cyclic horizontal test to failure in test number 6. Test number 6 is shown as an example, however the out-of-plane results for all tests can be found in Appendix A. Positive values indicate movement away from the slab edge, and negative values indicate movements toward the slab edge. Figure 4.19: Out‐of‐plane displacements during horizontal test with no out‐of‐plane restriction.  Figure 4.19 shows a gradual ratcheting of out-of-plane displacements as the number of cycles is increased. Also a clear arc shape is observed as the cycles range from positive to negative horizontal displacements. This is expected because when there are horizontal displacements, it will draw in the wall panel into the slab edge as the bars experience tension, however as it passes zero horizontal displacement, the bent bars which have been slightly straightened, will force the wall panel out and away from the slab edge. This arc shape is not present in the test specimens where out-of-plane movements were restricted. This is because as the horizontal displacements approached zero, the bars were forced to deform, instead of pushing the wall panel out. 0 2 4 6 8 10 12 ‐20 ‐15 ‐10 ‐5 0 5 10 15 20 O ut ‐o f‐ pl an e  D is p  [m m ] ∆h [mm] Chapter 4 Experimental Results and Discussion 81   4.3.4 Comparison of Measured Response with Previous Testing Previous testing conducted at the University of British Columbia of EM5 shear connectors is described in detail in Chapter 2. Figure 4.20 shows the horizontal cyclic shear responses from test number 6 measured in this study compared with that of the previous testing. Figure 4.20: Cyclic horizontal shear response comparison with previous testing.  (Modified from Lemieux et al., 1998)  A few observations can be made from Figure 4.20. First of all, it is apparent that the ultimate strength level as well as stiffness is quite consistent between the two test specimens. The current test recorded an ultimate force level of +232 kN / -209 kN where as the previous testing recorded ultimate force levels of +215 kN / -203 kN. This corresponds to only a 7.9% / 3.0% increase in ultimate force levels for positive and negative cycles respectively. This is quite remarkable since there is a large difference in testing setup and loading protocol between the two tests, with -250 -200 -150 -100 -50 0 50 100 150 200 250 300 -20 -15 -10 -5 0 5 10 15 V  [k N ] ∆h [mm] Cyclic horizontal response  from current study  (Specimen 6) Cyclic horizontal response  from previous testing Chapter 4 Experimental Results and Discussion 82   minimal discrepancy in results. Also as discussed in chapter 2, the previous testing allowed for some bending moment to be applied to the connector, which could account for the small discrepancy in strength capacities. This displays extreme predictability in the connector behavior which allows for confidence when modeling the strength behavior of the EM5 connector. Where the two tests differ substantially is in the displacement capacities. The current test showed a displacement capacity of 19.7 mm whereas the previous testing only showed a displacement capacity of 11.5 mm. This discrepancy is a direct result of the difference in loading protocol. In both tests, three cycles at each cycle step were used. The difference was in the choice of cycle step. In the current study, cycle steps in the order of 5mm increments were used. This resulted in only nine complete cycles before failure occurred. In the previous study, six to eight constant load increments were used up to 70% of the ultimate load value achieved in the monotonic test, followed by constant displacement increments until the failure of the specimen (Lemieux et al., 1998). This resulted in 28 complete cycles being performed before failure. Because of the large variance in the number of completed cycles, the previous testing specimen would have experienced fatigue failure from repeated cycling of the connector bar causing failure much earlier than the current study where fatigue played a much smaller role in the failure. 4.4 Combined Horizontal and Vertical Response One main objective of this study is to determine the interaction between the vertical damage state and the horizontal shear strength, effective stiffness and displacement capacity. In order to determine this, several tests were conducted in which the specimen underwent vertical displacement damage before the cyclic shear response to failure was measured. These horizontal shear tests were conducted with vertical damage caused by cyclic uplift to Δv = 25 mm, 50 mm, 75 mm and 100 mm displacement levels then held at those displacement levels, as well as with vertical damage caused by cyclic uplift to Δv = 25 mm, 50 mm and 100 mm displacement levels then forced back to Δv = 0 mm. This was done as it is of interest how the connector behaves in horizontal shear when the wall panels rock to larger vertical displacements as well as when it returns back to zero displacement when a load reversal occurs. Chapter 4 Experimental Results and Discussion 83   Since it is quite difficult to compare cyclic hysteretic plots with one another on a single plot due to clutter, response envelopes of the data were extracted that plots the response envelopes of the three cycles. This will make it quickly apparent the maximum load levels that a test specimen achieved at a specific displacement level for a specific cycle number. This will be useful for comparing the differences amongst the different tests. Figure 4.21 shows an example, using specimen number 7, of how these envelopes were fitted to the hysteretic plots. Figure 4.21: Comparison of the first cycle horizontal shear response envelope with the measured  response.  4.4.1 Interaction Models The procedure above was used to develop the cyclic envelopes for the horizontal shear response plots. When the envelopes for each of the five tests at the different vertical displacement levels (Δv = 0 mm, 25 mm, 50 mm, 75 mm, 100 mm) are plotted together, it was possible to determine the interaction between vertical displacement (Δv) and maximum horizontal shear capacity ‐200 ‐150 ‐100 ‐50 0 50 100 150 200 ‐20 ‐15 ‐10 ‐5 0 5 10 15 20 V  [k N ] ∆h [mm] Chapter 4 Experimental Results and Discussion 84   (VMax). Figure 4.22 compares the first cycle envelopes when the ΔvMax is sustained as horizontal loading was applied. Figure 4.22: Cycle 1 envelopes when ∆V=Max when horizontal testing is performed.  Figure 4.23 compares the cyclic envelopes for the first cycle when ΔvMax was previously reached, but when the horizontal loading was conducted at a sustained displacement level of Δv = 0 mm. ‐300 ‐200 ‐100 0 100 200 300 ‐20 ‐15 ‐10 ‐5 0 5 10 15 20 V  [k N ] ∆h [mm] H Monotonic H Cyclic 25mm Held Up 50mm Held Up 75mm Held Up 100mm Held Up Previous Testing Chapter 4 Experimental Results and Discussion 85   Figure 4.23: Cycle 1 envelopes when ∆V=0mm when horizontal testing is performed.  Figure 4.24 shows the interaction between the cyclic horizontal shear capacity and the vertical damage level in which it was tested for the first cycle. This interaction was produced from plotting VMax of each envelope in Figure 4.22 and Figure 4.23 against its respective ΔvMax. The values found in the previous testing program described in Chapter 2 is also presented on this interaction curve, and once again is in good agreement with the results found in the current study. ‐300 ‐200 ‐100 0 100 200 300 ‐25 ‐15 ‐5 5 15 25 V  [k N ] ∆h [mm] H Monotonic H Cyclic 25mm Pushed Down 50mm Pushed Down 75mm Pushed Down 100mm Pushed Down Previous Testing Chapter 4 Experimental Results and Discussion 86   Figure 4.24: Strength and vertical displacement interaction curve for first cycle responses.  From the data scatter presented in Figure 4.24 a best fit linear curve can be determined for use in the pushover model presented in this study. This modeled interaction has a strength capacity of 225 kN when Δv = 0 mm and degrades at a slope of -0.8 as the vertical damage level is increased giving an equation of: ெܸ௔௫ ൌ 225 ݇ܰ െ 0.8  ݇ܰ ݉݉ ൈ ∆ݒெ௔௫ ; 0 ݉݉ ൑ ∆ݒெ௔௫ ൑ 100 ݉݉ Where VMax is the maximum force in kilo-Newton’s and ΔvMax is the maximum vertical displacement in millimeters. An interaction to adjust the horizontal effective stiffness of the connection due to vertical damage level is also of interest. When all of the first cycle envelopes are plotted together it is apparent that they all have a similar horizontal stiffness until ~40 kN of force is applied. After 40 kN the stiffness becomes softer as the vertical damage level is increased until the maximum 0 50 100 150 200 250 300 0 20 40 60 80 100 H or iz on ta l F or ce [k N ] ∆v [mm] Model Curve Positive Held Up Positive Pushed  Down Positive (Previous  Testing) Negative Held Up Negative Pushed  Down Negative (Previous  Testing) Chapter 4 Experimental Results and Discussion 87   horizontal force capacity is reached and plastic deformations occur. Model curves for all of the various vertical damage levels were fit using tri-linear curves with the initial stiffness for all curves being 146.5 kN/mm. This was determined by averaging the initial stiffness between the origin and 40 kN force points of the first positive cycles. The third portion of the model curve was a plastic line through the value of maximum force. The slope of the second portion was determined by fitting a line between the first and third portions of the model curve which most closely fit the data. This was done by fitting a line which created equal areas inside and outside the envelope. Figure 4.25 shows an example plot of a model fit curve for the connector test number 15 which experienced ΔvMax = 25 mm and Δv = 0 mm when horizontal cycling occurred. Figure 4.25: Example model fit to first cycle envelope of cyclic horizontal response.  The effective stiffness’s of all the tests were determined in this manner and Figure 4.26 shows these secondary slopes values plotted for the positive and negative directions against the vertical damage level. This was used to determine the interaction for the slope of this secondary portion of the horizontal response. ‐100 ‐50 0 50 100 150 200 250 ‐5 0 5 10 15 V  [k N ] ∆h [mm] First Cycle  Envelope Initial Slope Plastic Region Effective Stiffness Chapter 4 Experimental Results and Discussion 88   Figure 4.26: Effective stiffness and vertical displacement interaction curve for first cycle  responses.  The model curve to fit to this data is a bilinear curve with an initial effective stiffness of 100 kN/mm when ΔvMax = 0 mm and degrades at a slope of -1.5 kN/mm2 as the vertical damage level is increased to ΔvMax = 50 mm after which the effective stiffness remains constant at a value of 25 kN/mm until ΔvMax = 100 mm. This produces an equation of: ܧ݂݂݁ܿݐ݅ݒ݁ ܵݐ݂݂݅݊݁ݏݏ ൌ ൞ 100  ݇ܰ ݉݉ െ 1.5  ݇ܰ ݉݉ଶ ൈ ∆ݒ  ; 0 ݉݉ ൑ ∆ݒெ௔௫ ൑ 50 ݉݉ 25  ݇ܰ ݉݉                                 ; 50 ݉݉ ൑ ∆ݒெ௔௫ ൑ 100 ݉݉ Where effective stiffness is the secondary slope in kilo-Newton’s per millimeter and ΔvMax is the maximum vertical displacement in millimeters. An interaction between the total cyclic horizontal displacement capacity and the vertical damage level for first cycle envelopes can be determined if ΔhMax for all the envelopes from Figure 4.22 0 20 40 60 80 100 120 140 0 20 40 60 80 100 Ef fe ct iv e St iff ne ss [k N /m m ] ∆vMax [mm] Model Curve Positive Held Up Positive Pushed  Down Positive (Previous  Testing) Negative Held Up Negative Pushed  Down Negative (Previous  Testing) Chapter 4 Experimental Results and Discussion 89   and Figure 4.23 are plotted against its respective ΔvMax (Figure 4.27). The cyclic displacement capacity from the results of previous testing is also shown, even though it is disregarded when choosing a model fit to this data because of its extremely low value due to its loading protocol. Figure 4.27:  Horizontal displacement capacity and vertical displacement interaction curve for  first cycle responses.  Figure 4.27 shows no distinguishable trend between vertical damage level and horizontal displacement capacity. The displacement capacities appeared to be somewhat random, and no explanation can be given to the scatter in results. However, since the horizontal displacement capacity results lie within a fairly narrow band (ΔhMax = 13.3 – 20.5 mm) a linear best fit line is chosen to fit the data. The Displacement capacity which is used in the model presented in this study is ΔhMax = 15 mm. In design practice, it would be recommended that a lower bound value be chosen to be conservative (ie 13 mm) however since the objective of the model in this study is to accurately model a tilt-up system, a value closer to the average displacement value is used. 6 8 10 12 14 16 18 20 22 0 20 40 60 80 100 ∆ h M ax [m m ] ∆v [mm] Model Curve Held Up Pushed  Down Previous  Testing Chapter 4 Experimental Results and Discussion 90 5 ANALYTICAL STUDY OF SYSTEM RESPONSE One objective of this study was to invest the in plane behavior of solid concrete tilt-up wall panels. It is important to understand the behavior of these structures before progressing to more complex structures such as wall panels with large openings. An in-plane shear pushover model was developed to determine the response of a multi panel tilt-up system with various configurations of number of panels, connectors and dimensions. These pushover curves can be studied to determine how the design affects the energy dissipation mechanism, as well as the ductility and load capacity of the system as a whole. Design variables such as number of panel to panel connectors, panel to slab connectors, and number of panels tied together will be analyzed to attempt to determine the effect these variables have on the overall system response. Once these effects are understood, the model can then be used to suggest how to optimally place connectors to achieve the predetermined energy dissipation mechanism. The chapter will conclude with the model being used to compare the system response using connectors used in current practice with a couple of improved connector details. 5.1 Pushover Model It is important that when using a model it is understood what it is intended for, and therefore what its limitations and assumptions are. This section will present the assumptions and limitations of the model used in this study, as well as how it may be implemented. 5.1.1 Model Assumptions The model that was developed assumed solid concrete panels without openings. The model assumes completely rigid panels with the exception of the compression strain region (c) at the corner of the concrete panel where rocking occurs. This rigid panel assumption results in all the inelastic deformations to occur in the connectors, except in the compression strain region described. It is acknowledged that the distance from the extreme compression edge to the neutral 91   axis will increase as the panel is rocked to larger rotational displacements, however since this has a negligible effect on the total response a constant value of c = 75 mm is used in the model for this study, however this can be adjusted by the user to accommodate different conditions. Since this model only allows for in-plane shear forces, all out of plane movements are neglected except for out of plane movements at the panel to slab connectors. In sixteen of eighteen test specimens, the experimental testing allowed for out of plane movements as the connector was loaded therefore the non-linear model response for these connectors includes these movements. The two specimens where out of plane movements were restricted (specimen 17 & 18) showed no distinguishable difference in response when compared to the specimens free to move out of plane; therefore at the connector level this omission is valid. These movements could potentially add to the demands on the panel to panel connections, and would definitely affect the panel to roof diaphragm connection, however these connections is not the main focus of this study. The next assumption within the model is a rigid slab. When discussing the testing setup, it was noted that the slab response could potentially play a large role in the overall response. The physical deformations that the slab would have to undergo in order to lift off the ground were then considered. As the panels begin to rock and lift away from the foundation, one of two things would happen. Either the slab would lift up along with the panel rotation, or the connector would have to fail. After comparing the force carrying capacity with the weight of the slab it was apparent that each of the connectors could lift a significant area of slab. However when the corner locations where the panel to panel connectors undergo significant displacement were looked at, one panel corner is lifted off the ground, and the other remains on the foundation. Since the physical connections between these two points are typically only about 6 feet apart, the slab would have to flex into an “S” shape, or the slab would have to fail in shear or flexure to make this possible. Either of these conditions would require a much larger force than is available in the connectors, therefore it is safe to assume that the connectors would fail before movement in the slab would occur. A constant friction factor of μ=0.5 is assumed in this model. This value is assumed rigid- perfectly plastic and is independent of rocking. As the panel rocks, less surface area is in contact Chapter 5 Analytical Study of System Response 92   with the foundation. This along with damage to the panel may affect the actual coefficient of friction, however these occurrences are neglected in this model. The last major assumption of the model is that the lateral force is assumed to be fully transmitted into the wall panels. Currently in practice, a deck angle attached to embedded plates in the wall is used to transfer this lateral load. However as the panels begin to rock, the deck angle must deform to match the shape of the wall panels. This will undoubtedly affect its ability to transfer lateral loads, and will also induce vertical loads into the panels. Both of these affects are neglected in the current model, however it is acknowledged that these affects will be significant and therefore more research is needed in this area. 5.1.2 Implementation of the Model The model is set up to be implemented within Microsoft Excel with most of the processes being carried out in Visual Basic coding language. The model is a displacement controlled pushover model which increments the rotational displacement at the top of the wall in step sizes defined by the user. In this study a step size of 1 mm was used. The pushover begins by determining whether or not the panel – panel connectors have enough strength capacity to lift the adjacent panels. If there is not enough strength, the panel – panel connectors will yield, ensuring the adjacent panels remain on the ground. If the adjacent panel is able to be lifted, the panel – panel connector is assumed to be rigid as a 1 mm rotational drift is applied at the top of the panel. Beginning from the compression side of the panel system, the deformations in the panel – panel connectors are determined. Iterations are needed until the system converges on the balanced condition. Once the wall panel systems internal displacements converge on the solution for the specified rotational displacement, the lateral force at the top of the wall panels required to cause the displacement is determined. Once the total lateral force is known, the additional horizontal movement (sliding) caused by horizontal displacements in the panel to slab connectors can be Chapter 5 Analytical Study of System Response 93   determined. This sliding displacement is the solution to a system of equations of N dimensions, where N is the total number of panel to slab connectors in the wall panel system. Each connector has a different stiffness and maximum force capacities, which are prescribed based on the interaction diagrams detailed in Chapter 4, due to the varying vertical damage levels of each connector at the base of the panel. The sum of the horizontal force in the panel to slab connectors must equal the total lateral force, and based on compatibility, all the displacements must be equal. Once this sliding displacement is found the total displacement (ΔTotal) at the top of the wall corresponding with the current load is the sum of the rotational displacement (ΔR) and the sliding displacement (Δh) as shown in Figure 5.1. Figure 5.1: Drawing showing the components that determine the system drifts.  ΔTotal ΔR Δh Chapter 5 Analytical Study of System Response 94   5.2 Pushover Analysis with Existing Connectors The model described above was used to study the different wall systems with various different connector combinations and configurations typical in current construction of single story tilt-up building. Section 2.4 outlines how the typical building parameters were determined in this study. Table 5.1 shows the name and system properties for each of the 30 panel systems studied as well as the maximum force and residual force found. Table 5.1: System properties of each pushover model and its name,  maximum lateral force achieved, and residual strength.  System  Name  #  Panels  # Panel ‐  Panel # Panel‐ Slab VMax  [kN] Residual Lateral  Force [kN]  1,0,4  1  0  4  289  132  1,0,3  3  253  132  1,0,2  2  217  132         2,2,4  2  2  4  799  265  2,2,3  3  758  265  2,2,2  2  735  265        2,3,4  3  4  1077  265  2,3,3  3  996  318  2,3,2  2  848  318        2,4,4  4  4  1145  318  2,4,3  3  995  318  2,4,2  2  847  318         3,2,4  3  2  4  1386  397  3,2,3  3  1325  397  3,2,2  2  1281  397        3,3,4  3  4  1720  397  3,3,3  3  1596  476  3,3,2  2  1471  476        3,4,4  4  4  1911  476  3,4,3  3  1786  476  3,4,2  2  1627  476         4,2,4  4  2  4  1974  530  4,2,3  3  1892  530  4,2,2  2  1828  530        4,3,4  3  4  2495  530  4,3,3  3  2350  635  4,3,2  2  2205  635        4,4,4  4  4  2874  635  4,4,3  3  2728  635  4,4,2  2  2404  635  Chapter 5 Analytical Study of System Response 95   The convention used to name each of the pushover systems is: ሺ# ݋݂ ݈ܲܽ݊݁ݏ, # ݋݂ ݈ܲܽ݊݁ െ ݈ܲܽ݊݁ ሺܲܲሻܥ݋݊݊݁ܿݐ݋ݎݏ, # ݋݂ ݈ܲܽ݊݁ െ ݈ܾܵܽ ሺܲܵሻܥ݋݊݊݁ܿݐ݋ݎݏሻ 5.2.1 One Panel Pushover Results The first step is to study the response of a one panel system with panel – slab connectors. It is assumed in all cases that lateral loading occurs at the top of the panels, and that the coefficient of friction is μ = 0.5, and so since the height to width ratio of a single panel is greater than one, rocking will always occur first, and in the single panel case will always be the final failure mechanism, as the applied force required to overturn the panel is less than the force required to overcome friction. Figure 5.2 shows the three pushover plots for the single panel systems which are with two, three, and four panel – slab connectors normalized to the maximum lateral force achieved and Figure 5.3 shows the same set of plots normalized to the residual lateral force after all panel – slab connectors are failed. Normalizing to the maximum lateral force allows for the shapes of the plots and important events within the pushover analysis to be easily compared and allows the percent of maximum strength which is lost due to each event to be easily seen relative to other systems. Normalizing the plots to the final residual lateral force shows the multiple of strength gain which adding connectors to the system achieves since all systems with the same number of panels and the same failure mechanism are normalized by the same value. This allows for direct comparison of the differences in actual strength between the various systems with the same number of panels. Since this method factors each system by its own residual strength, a two panel system will have a lower residual strength than a four panel system (different only by a multiple of the number of panels), and thus does not allow for the actual strength of systems with different number of panels to be directly compared, however since the shapes are the same as the actual pushover response plots, it allows for the shapes to be compared even amongst systems with different number of panels. What is valuable about this normalizing method is that for systems with panels of the same aspect ratio, you can just multiply the values of the plots shown by the residual strength of the system (given in Table 5.1) to get actual strength values for the system. In the case of this study, the panel weight is given as: Chapter 5 Analytical Study of System Response 96   ௉ܹ௔௡௘௟ ൌ ܪ ൈܹ ൈ ݐ ൈ ߛ஼௢௡௖௥௘௧௘ ௉ܹ௔௡௘௟ ൌ 9.14݉ ൈ 7.62݉ ൈ 0.19݉ ൈ 24.0 ݇ܰ ݉ଷ ௉ܹ௔௡௘௟ ൌ 317.6݇ܰ Therefore the residual strength for a system with an overturning final failure mechanism would be: ோܸ௘௦௜ௗ௨௔௟ ൌ ௉ܹ௔௡௘௟ ൬ ܹ 2ܪ ൰ ൈ ݊௉௔௡௘௟௦ ோܸ௘௦௜ௗ௨௔௟ ൌ 317.6݇ܰ ൬ 7.62݉ 2 ൈ 9.14݉ ൰ ൈ ݊௉௔௡௘௟௦ ோܸ௘௦௜ௗ௨௔௟ ൌ 132.4 ൈ ݊௉௔௡௘௟௦ And the residual strength for a system with a sliding final failure mechanism would be: ோܸ௘௦௜ௗ௨௔௟ ൌ ௉ܹ௔௡௘௟ ൈ ߤ ൈ ݊௉௔௡௘௟௦ ோܸ௘௦௜ௗ௨௔௟ ൌ 317.6݇ܰ ൈ 0.5 ൈ ݊௉௔௡௘௟௦ ோܸ௘௦௜ௗ௨௔௟ ൌ 159݇ܰ ൈ ݊௉௔௡௘௟௦ Where: VResidual = Residual lateral force W = Width of panel H = Height of panel NPanels = Number of panels μ = Coefficient of friction Chapter 5 Analytical Study of System Response 97   Figure 5.2: One panel pushover plots normalized by maximum lateral force.    Figure 5.3: One panel pushover plots normalized by maximum lateral residual force.  The first portion of each of these plots is the force required to overcome the dead load of the panel. This is calculated as follows for all systems: V ൌ WeightPୟ୬ୣ୪ ൬ W 2H ൰ Where: V = Lateral force W = Width of panel H = Height of panel 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 1 2 3 4 5 6 7 La te ra l F or ce  (R at io  o f M ax  F or ce ) Panel Drift (%) 1,0,2 1,0,3 1,0,4 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 0 1 2 3 4 5 6 7 La te ra l F or ce  (R at io  o f R es id ua l F or ce ) Panel Drift (%) 1,0,2 1,0,3 1,0,4 Chapter 5 Analytical Study of System Response 98   After the dead load is overcome, the panel begins to rock. This value is 1.0 for the systems which experience an overturning final failure mechanism on the plots normalized by the residual strength, and 0.83 for the systems which experience sliding. As it rocks, the panel-slab connectors begin to engage in the vertical direction following the response curve given in Figure 4.13. Initially, before the concrete cracks, the panel – slab connectors are very stiff, accounting for the very stiff pushover response lasting until approximately 70% max force in Figure 5.2. As the concrete begins to crack, the pushover response softens, with the bulk of the stiffness change occurring when the connectors furthest from the corner of rotation crack the concrete. This softer response continues until the vertical displacement capacity of the panel – slab connectors is reached, which fail in succession, producing a series of stepped drops. Figure 5.3 shows that the stiffness loss is greater in systems with fewer panel – slab connectors as shown by the smaller slopes. This is simply due to the fact that there is less connectors contributing to the system response. The first failure in each of the one panel cases occurred at approximately 2% drift, and created a drop in lateral strength back down to 71% maximum strength for the 1,0,2 system, 72% maximum strength for the 1,0,3 system and 74% maximum strength for the 1,0,4 system. This pattern shows that in the overturning systems, the more panel – panel connectors the system has, the smaller the percentage of strength loss will be with each panel – slab connector failure. Figure 5.3 shows the capacities of the systems relative to this final residual lateral capacity. The 1,0,2, 1,0,3 and 1,0,4 systems have a lateral strength capacities of 1.63, 1.91 and 2.19 times the residual strengths respectively. This represents a 34% increase in lateral strength when the system has 4 panel – slab connectors as opposed to 2. Since the EM5 connector has a vertical displacement capacity of 100 mm and the connectors in each case are evenly distributed along the base of the wall panel, the drift where each of the panel – slab connectors break can easily be predicted as in the following example: Chapter 5 Analytical Study of System Response 99   One panel, two panel – slab connectors: For first connector: θ ൌ ∆vMୟ୶ h୬ െ c ൌ 100mm ሺ5080mmെ 75mmሻ ൌ 2.0% For second connector: θ ൌ ∆vMୟ୶ h୬ െ c ൌ 100mm ሺ2540mmെ 75mmሻ ൌ 4.06% Where: θ = Panel drift ΔVMax = Vertical displacement capacity of EM5 hn = Length from the corner of panel to nth panel – slab connector c = Panels compression zone (75 mm) In this case since after the second connector fails, the panel is completely disconnected from the base slab, and is free to overturn without restriction, thus the residual lateral capacity is simply the overturning resistance of a single panel. It is noted from both Figure 5.2 and Figure 5.3 that the panel system with more panel – slab connectors will have higher drift capacities. Since there are more connectors, they are spaced closer to the corner of rotation, hence the panel would have to rotate further in order to achieve the connectors displacement capacity in order to fail. This however this may be deceiving since this pushover response is only a result from pushing in one direction, and in an earthquake situation, rocking would occur in both directions, hence in ΔVMax Chapter 5 Analytical Study of System Response 100   practice, having connectors closer to the center of the panel would in fact allow for the largest drifts (This will be discussed more when ideal connector are discussed in Section 5.5). 5.2.2 Two Panel Pushover Results The next logical step in the progression of studying the pushover plots is to study two panel systems. Figure 5.4 and Figure 5.5 show a series of two panel systems each normalized to the maximum lateral force achieved and normalized to the final residual capacity respectively. Figure 5.4: Two panel pushover plots with sliding final failure mechanism normalized by maximum  lateral force.  0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.0 1.0 2.0 3.0 4.0 La te ra l F or ce  (R at io  o f M ax  F or ce ) Panel Drift (%) 2,4,2 2,4,3 2,4,4 2,3,2 2,3,3 Chapter 5 Analytical Study of System Response 101     Figure 5.5: Two panel pushover plots with sliding final failure mechanism normalized by maximum  lateral residual force.  The first observation made is that the shapes of these plots are very similar to the single panel plots of Figure 5.2. The first portion of each of these plots is the force required to overcome the dead load of the panel just as in the single panel pushover. After the dead load is overcome, the panel begins to rock. As it rocks, the panel-slab connectors begin to engage in the vertical direction following the model response curve given in Figure 4.13. The initial stiff pushover response again lasts until approximately 70% of the maximum force until the concrete begins to crack and the pushover response softens. Figure 5.5 shows the extent that it softens again depends on the number of panel – slab connectors the system has just as in the one panel systems. This softer response continues until the vertical displacement capacity of the panel – slab connectors is reached, which fail in succession, producing a series of stepped drops. This is where the first difference between the responses of one and two panel systems arises. The first panel – slab failure in the two panel system occurs at a drift of approximately 0.8%, compared to approximately 2.0% in the one panel systems. Since in these two panel systems the panel – panel connections are strong (three and four connectors) this connection does not fail and the second panel gets lifted off the ground, which effectively doubles the panel width. This wider panel width causes the first panel – slab connector to fail at a smaller drift due to the larger lever arm. This wider effective panel width also increases the overturning resistance which makes it larger than the sliding resistance thereby changing the final failure mechanism from overturning to 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0.0 1.0 2.0 3.0 4.0 La te ra l F or ce  (R at io  o f R es id ua l F or ce ) Panel Drift (%) 2,4,2 2,4,3 2,4,4 2,3,2 2,3,3 Chapter 5 Analytical Study of System Response 102   sliding. Once enough panel – slab connectors fail that the lateral force exceeds the sum of the friction and the horizontal shear capacity of the remaining connectors the panels will begin to slide. At the very end of the response plots in Figure 5.4, there is a little yield plateau which are the result of the small yielding of the remaining connectors in horizontal shear. It is apparent that this yielding is insignificant to the overall response of the system. Figure 5.5 shows that the maximum lateral strength of the system again depends partially on the number of panel – slab connectors as the systems with two, three and four panel – slab connectors reached a lateral force capacity of 2.67, 3.13, and 3.60 times the residual lateral strength respectively, corresponding again to a 34% increase in lateral force capacity for the four panel – slab connector system over the two connector system, just as in the single panel case. The next two panel systems shown in Figure 5.6 experience an overturning failure mechanism, however these systems responses with two panel – panel connectors look quite different than the one panel systems, and the two panel systems which experience sliding. Figure 5.6: Two panel pushover plots with overturning final failure mechanism which experience no  panel uplift normalized by maximum lateral force.  0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 1 2 3 4 5 6 7 La te ra l F or ce  (R at io  o f M ax  F or ce ) Panel Drift (%) 2,2,3 2,2,4 Chapter 5 Analytical Study of System Response 103     Figure 5.7: Two panel pushover plots with overturning final failure mechanism which experience no  panel uplift normalized by maximum lateral residual force.  The force required to overcome the dead load is the same as in the previous systems, however at the point of initial softening the softening is not a result of softening in the panel – slab connectors but of the panel – panel connection response reaching its yield plateau, which renders it unable to increase its load carrying capacity, hence making the whole system quite soft. The significant drop experienced at 0.39% drift is due to the displacement capacity of the panel – panel connectors being exceeded, thus completely disconnecting the two panels from each other, which drastically reduces the overturning capacity of the system to 51% the maximum lateral force capacity for the system with three panel – slab connectors and 53% of the maximum lateral force capacity for the system with four panel – slab connectors, which again shows a slight variance in the system degradation based on the number of panel – slab connectors, as seen in Figure 5.6. This drift of 0.39% which causes the panel – panel connection to fail in a system where no panel uplift occurs can easily be determined as follows: 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0 1 2 3 4 5 6 7 La te ra l F or ce  (R at io  o f R es id ua l F or ce ) Panel Drift (%) 2,2,3 2,2,4 Chapter 5 Analytical Study of System Response 104   θ ൌ 2 ൈ ∆hMୟ୶ W െ c θ ൌ 2 ൈ 15 mm ሺ7620 mmെ 75 mmሻ θ ൌ 0.39 % Where ∆hMୟ୶ ൌ Horizontal shear capacity of EM5 connector W = Width of panel C = Panel compression zone (75mm) Figure 5.7 shows the maximum strength achieved by the 2,2,3 and 2,2,4 systems is 2.86 and 3.02 times the final overturning residual capacity respectively, and that the first residual stiffness is dependent on the number of panel – slab connectors used. These results are consistent with the observations of the single panel systems. The final two panel systems shown in Figure 5.8 and Figure 5.9 also experience an overturning failure mechanism; however unlike the overturning systems of Figure 5.6 and Figure 5.7, these experience some initial uplift of the second panel and therefore initially behave similar to the Chapter 5 Analytical Study of System Response 105   systems which experience sliding. Some panel uplift will occur when there is enough panel – panel connection strength to overcome the dead load of the panel plus the tie down capacity of the panel – slab connectors. Figure 5.8: Two panel pushover plots with overturning final failure mechanism which experience some  panel uplift normalized by maximum lateral force.    Figure 5.9: Two panel pushover plots with overturning final failure mechanism which experience some  panel uplift normalized by maximum lateral residual force.  0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 1 2 3 4 5 6 7 La te ra l F or ce  (R at io  o f M ax  F or ce ) Panel Drift (%) 2,2,2 2,3,4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0 1 2 3 4 5 6 7 La te ra l F or ce  (R at io  o f R es id ua l F or ce ) Panel Drift (%) 2,2,2 2,3,4 Chapter 5 Analytical Study of System Response 106   In both of these systems the softening in region c is due to the softening of the panel – slab connections just as in the sliding cases of Figure 5.4. In the 2,2,2 system, the panel – panel connection fails in a extremely brittle manor at 0.39% drift when the force capacity of the panel – panel connectors is reached. The yield plateau for these connectors is not utilized in this case because the vertical displacement of the panel lifted from the ground is greater than the displacement capacity of the panel – panel connectors and so when the yield strength is reached, the panel which was lifted drops violently to the ground. The 2,3,4 system behaves in much the same manor, except that the panel – panel connection is stronger due to having three connectors instead of two. This strength increase allows the drift to extend beyond 0.39% to a drift of 0.65% before the panel – panel connectors fail. This drift ratio is more complicated to show mathematically due to the iterations required to converge on the solution for the deformation of the panel – panel connectors (or panel slip) due to the panel weight and the vertical hold down force created by the panel – slab connectors due to the vertical location of the panel. In both systems of Figure 5.8, after the panel – panel connection has failed, the system behaves just as the other overturning systems of Figure 5.6. Figure 5.9 again shows that with more panel – slab connectors on each panel of the system, the stiffer the response will be, which is consistent with the other overturning systems, however Figure 5.8 shows the system with four panel – slab connectors having a larger initial drop in percentage of maximum capacity to 42% when compared to the two connector system which dropped only to 47% its maximum capacity. This is inconsistent with the other overturning systems, and shows that adding more panel – panel connectors to an overturning system will have a larger effect on the initial capacity loss. 5.2.3 Multi - Panel Pushover Results The next step in studying the array of pushover responses outlined at the beginning of this chapter is to study the systems with more than one panel. Figure 5.10 shows all of the panel systems, including the two panel systems of Figure 5.6, which experience overturning as a final failure mechanism and experiences no panel lift normalized to the maximum lateral force achieved. Figure 5.11 shows these plots normalized to the final residual capacity. Chapter 5 Analytical Study of System Response 107   Figure 5.10: All panel pushover plots with overturning final failure mechanism which experience no  panel lift normalized by maximum lateral force.    Figure 5.11: All panel pushover plots with overturning final failure mechanism which experience no  panel lift normalized by maximum lateral residual force.  The first observation made from Figure 5.10 is that the response curves for all systems regardless of the number of panels look very similar in shape and initially, until the point when the panel – panel connections break at 0.39% drift, they follow the exact same response path when normalized to the maximum lateral force achieved. The percentage of capacity drop from the maximum capacity on an overturning system is slightly dependent on the number of panel – slab 0.0 0.2 0.4 0.6 0.8 1.0 0 1 2 3 4 5 6 7 La te ra l F or ce  (R at io  o f M ax  F or ce ) Panel Drift (%) 2,2,3 2,2,4 3,2,3 3,2,4 4,2,3 4,2,4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0 1 2 3 4 5 6 7 La te ra l F or ce  (R at io  o f R es id ua l F or ce ) Panel Drift (%) 2,2,3 2,2,4 3,2,3 3,2,4 4,2,3 4,2,4 Chapter 5 Analytical Study of System Response 108   connectors as has been discussed earlier, with more panel – slab connectors helping to lower the percentage of capacity loss, however Figure 5.10 shows that the number of panels has a larger effect on the percentage of residual strength. It is noted that systems with less panels experience a lower percent capacity loss when compared to a system with more panels. This can be easily explained by studying Figure 5.11, which has the same plots normalized to the final residual capacity, and since all panels are the same in this study, this normalization effectively factors out the number of panels in the system. Because of this Figure 5.11 compares the demand on each panel of the system. This figure shows that the in the initial part of the response before the panel – panel connector failures, the systems with more panels reach a higher capacity when compared to systems with fewer panels. After the panel – panel connector failure, each system is simply single panels rocking independently of each other. The result of this, and the fact that number of panels is factored out, is normalized plots of the systems with identical connection configurations coinciding with each other regardless of the number of panels, which is indeed what we see in Figure 5.11. So initially before panel – panel connector failure, systems with more panels reach a higher lateral capacity per panel, and after panel – panel connector failure they coincide with systems with the same configuration regardless of the number of panels, and so systems with a larger number of panels experience a larger percentage capacity loss when the panel – panel connection fails. Figure 5.12 is a plot of all the pushover systems which experience overturning as the final failure mechanism, but experience some initial uplift of panels normalized to the maximum lateral force, including the two panel systems from Figure 5.8, and Figure 5.13 shows the same plots normalized by the final residual capacity. As mentioned earlier, in order to achieve an overturning failure mechanism, the final overturning residual capacity must be lower than the sliding resistance when all of the panel – slab connectors have been disconnected. For this to occur in the study systems, the panel – panel connectors must not be able to lift the panels adjacent to it off the ground, otherwise the height to width ratio decreases which increases the overturning capacity and produces a sliding mechanism. In the systems shown in Figure 5.12, initially the panel – panel connectors are strong enough to lift the outside panel off the ground, however eventually the capacity is reached and it fails, allowing for the overturning failure mechanism. Chapter 5 Analytical Study of System Response 109   Figure 5.12: All pushover plots with overturning final failure mechanism which experience some panel  uplift normalized by maximum lateral force.    Figure 5.13: All pushover plots with overturning final failure mechanism which experience some panel  uplift normalized by maximum lateral residual force.  Figure 5.12 show the identical trends to what was seen in the two panel systems of Figure 5.8. The systems with two panel – panel and 2 panel – slab connectors initially experience softening due to panel – slab connector softening along with uplift of the outermost panel. At a drift level of 0.39% the panel – panel connectors fail and the system response follows that of the other overturning systems of Figure 5.10. The only systems with three panel – panel connectors which 0.0 0.2 0.4 0.6 0.8 1.0 0 1 2 3 4 5 6 7 La te ra l F or ce  (R at io  o f M ax  F or ce ) Panel Drift (%) 2,2,2 3,2,2 4,2,2 3,3,4 4,3,4 2,3,4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0 1 2 3 4 5 6 7 La te ra l F or ce  (R at io  o f R es id ua l F or ce ) Panel Drift (%) 2,2,2 3,2,2 4,2,2 3,3,4 4,3,4 2,3,4 Chapter 5 Analytical Study of System Response 110   can experience overturning within the study setup are the systems with four panel – slab connectors. Fewer than four panel – slab connectors make it possible for the panel – panel connectors to stay connected which will allow uplift of a panel and create a sliding mechanism. Even though the panel – panel connection still eventually fails, the stronger connection allows for the connection to remain intact longer than 0.39% drift, which is when the panel – panel connection of the systems with two panel – panel connectors fail. As mentioned earlier, the 2,3,4 system experiences panel – panel failure at a drift level of 0.65%. The systems with 3 panel – panel connectors and more than two panels experience panel – panel failures at both of these drift ratios. The three and four panel systems have more than one panel – panel connection interface. The three panel – panel connection configuration (and in the four panel – panel connection configurations which will be discussed later) do not have sufficient capacity to allow more than one panel to be lifted from the ground under any circumstance. Because of this, all panel – panel connection interfaces other than the last will experience no panel lift and will fail at 0.39% drift just as in the previous cases looked at. This accounts for the first initial drop of load capacity in the 3,3,4 and 4,3,4 systems in Figure 5.12 and Figure 5.13 (as well as all the sliding cases which will be looked at next). After this initial capacity loss, there is another capacity drop at a drift level of 0.65% where the panel – panel connection of the final panel – panel interface fails just as in the 2,3,4 system. All of the previous observations about the characteristics of the response for overturning systems hold true in this case. The percentage of the capacity drop decreases with a smaller number of panel – slab connectors, the system with more panel – slab connectors has a stiffer initial residual response shown by the steeper slope of the plot, and a larger number of panel – panel connectors increases the percent capacity loss when systems with otherwise identical configurations are compared. Figure 5.13 shows that once again after the panel – panel connections have failed, each system follows the same response curve dependent only on the number of panel – slab connectors, and independent on number of panels. The next step is to study all the systems which experience sliding as the final failure mechanism. As described earlier, a sliding failure mechanism will occur only if the overturning capacity of the system is greater than that of sliding capacity due to friction. Using the panel dimensions Chapter 5 Analytical Study of System Response 111   chosen for this study, which reflect typical panel dimensions used in single story tilt-up building in practice, sliding will only occur if multiple panels remain connected together with sufficient strength to lift an adjacent panel after all the panel – slab connectors have failed. Figure 5.14 shows all such cases normalized to the maximum lateral capacity achieved and Figure 5.15 shows the same cases normalized to the residual capacity of the system, which in these cases is the sliding resistance of the system due to friction. This is calculated as: ோܸ௘௦௜ௗ௨௔௟ ൌ ௉ܹ௔௡௘௟ ൈ ߤ ൈ ݊௉௔௡௘௟௦ Figure 5.14: All panel pushover plots with sliding final failure mechanism normalized by maximum  lateral force.    Figure 5.15: All panel pushover plots with sliding final failure mechanism normalized by maximum  lateral residual force.  0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 1 2 3 4 5 6 7 La te ra l F or ce  (R at io  o f M ax  F or ce ) Panel Drift (%) 3,3,2 3,3,3 3,4,3 3,4,4 4,3,2 4,4,3 4,3,3 4,4,4 2,4,2 2,4,3 2,4,4 2,3,2 2,3,3 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0 1 2 3 4 5 6 7 La te ra l F or ce  (R at io  o f R es id ua l  Fo rc e) Panel Drift (%) 3,3,2 3,3,3 3,4,3 3,4,4 4,3,2 4,4,3 4,3,3 4,4,4 2,4,2 2,4,3 2,4,4 2,3,2 2,3,3 Chapter 5 Analytical Study of System Response 112   The first trend that is obvious from looking at Figure 5.14 and Figure 5.15 is that all systems with three and four panels have a severe drop in lateral force capacity at 0.39% drift just as in the overturning cases studied above. This again is due to the panel – panel connectors between all panel – panel interfaces except the last reaching their displacement capacity and failing. This occurs because even four panel – panel connectors does not have enough shear capacity to lift more than one adjacent panel from the ground, and so all but the outermost panel remains on the ground as initial overturning occurs. Panel – panel failure occurs once the deformations between panels exceeds the displacement capacity of the panel – panel connection, which just as in the overturning cases above, occurs at 0.39% drift. This capacity drop does not occur in the two panel sliding systems because there is only one panel – panel interface, and so the second panel does get lifted from the ground. Overturning continues until enough panel – slab connectors fail through vertical displacements to allow the lateral applied force to overcome the friction force of the system and the horizontal shear capacity of the remaining panel – slab connectors. The sliding final failure mechanism is created due to the one panel – panel interface that does not fail. In these systems the panel – panel connection stays intact where in the overturning systems it fails. This is because of the stronger panel – panel connection created by increased number of panel – panel connectors. This increases the overturning resistance which becomes larger than the friction resistance, and so sliding occurs first. It is also noted that the three and four panel systems do not have the small yield plateau at the end of the pushover response that the two panel systems experience. This is because for these systems it is not until all of the panel – slab connectors have failed in the vertical direction that sliding occurs, whereas in the two panel systems, there is still one panel – slab connector intact when the lateral force overcomes the sum of the friction resistance and the horizontal shear capacity of that lone connector. This is due to there being multiple panels in the three and four panel systems experiencing multiple simultaneous panel – slab connector failure, as there is multiple panels on the ground experiencing the same vertical displacement demands (three panel systems have two panels experiencing simultaneous panel – slab connection failures and four panel systems have three panels experiencing simultaneous panel – slab connection failures). Chapter 5 Analytical Study of System Response 113   This shows that these systems do not even utilize the small amount of ductility that the EM5 connector has. The last two systems which experience sliding as a final failure mechanism are shown in Figure 5.16 and Figure 5.17. These systems experience an extremely brittle sliding failure due to a combination of the strong panel – panel connection and the weak panel – slab connection. These systems fail due to sliding before 0.39% drift which in every other system would induce panel – panel failure. These systems fail before any of the panel – slab connectors fail in uplift therefore there is a small yield plateau just before failure which is caused by the yielding of the EM5 panel – slab connectors in horizontal shear. Figure 5.16: All panel pushover plots with extremely brittle sliding final failure mechanism normalized  by maximum lateral force.  0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 La te ra l F or ce  (R at io  o f M ax  F or ce ) Panel Drift (%) 3,4,2 4,4,2 Chapter 5 Analytical Study of System Response 114     Figure 5.17: All panel pushover plots with extremely brittle sliding final failure mechanism normalized  by maximum lateral  residual  force.  This extremely brittle failure mechanism would be the most undesirable failure mechanism of all the systems presented, and shows the importance of proper selection of connection configuration. It shows the importance of having enough horizontal shear capacity, which is partially considered in current design, however it also shows how it is equally important to consider how the panel – panel strength and displacement properties will influence the response and overall failure mechanism. 5.3 Tilt-up System Strength Results and Discussion In current design practice, sliding is designed to be resisted by the panel – slab connections in combination with friction between the panels and the foundations. Currently it is recommended that ductile connections, such as the so-called ductile EM5 be used. This recommendation is made without consideration of how vertical damage will affect the horizontal shear capacity. Once the sliding force is calculated, friction is subtracted, and the number of panel – slab connectors needed is determined by dividing the net sliding force by the full horizontal shear capacity of a panel - slab connector. Chapter 4 of this study shows that vertical damage to the panel – slab connectors has a significant degrading effect on the horizontal strength capacity. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 La te ra l F or ce  (R at io  o f R es id ua l F or ce ) Panel Drift (%) 3,4,2 4,4,2 Chapter 5 Analytical Study of System Response 115   The design sliding capacity is the nominal expected horizontal shear strength that design engineers are expecting to achieve, which is the sum of the friction in the system and the nominal horizontal shear capacity of an EM5 connector with no vertical damage. This is calculated as follows: ܧݔ݌݁ܿݐ݁݀ ݈ܵ݅݀݅݊݃ ܴ݁ݏ݅ݏݐܽ݊ܿ݁ ൌ  ߤ ൈ ௉ܹ௔௡௘௟ ൈ ݊௉௔௡௘௟ ൅ ெܸ௔௫ ൈ ݊௉ିௌ ஼௢௡௡௘௖௧௢௥௦ Where: ߤ ൌ ܥ݋݂݂݁݅ܿ݅݁݊ݐ ݋݂ ܨݎ݅ܿݐ݅݋݊ ሺ0.5ሻ ௉ܹ௔௡௘௟ ൌ ܹ݄݁݅݃ݐ ݋݂ ܽ ݌݈ܽ݊݁ ሺ݇ܰሻ  ݊௉௔௡௘௟ ൌ ܰݑܾ݉݁ݎ ݋݂ ݌݈ܽ݊݁ݏ ெܸ௔௫ ൌ ܯܽݔ݅݉ݑ݉ ݄݋ݎ݅ݖ݋݊ݐ݈ܽ ݏ݄݁ܽݎ ݎ݁ݏ݅ݏݐܽ݊ܿ݁ ݋݂ ܧܯ5 ܿ݋݊݊݁ܿݐ݋ݎ ݊௉ିௌ ஼௢௡௡௘௖௧௢௥௦ ൌ ܰݑܾ݉݁ݎ ݋݂ ݌݈ܽ݊݁ െ ݏ݈ܾܽ ܿ݋݊݊݁ܿݐ݋ݎݏ Currently the method for determining the overturning resistance is to determine the overturning equilibrium of the panel system without including the contributions of the panel – slab connectors in uplift. This is a very conservative method to determining maximum overturning resistance as the panel – slab connectors have been shown in Chapter 4 to provide significant vertical force capacity, which provides significant overturning resistance. However it will be shown later that it is still only conservative for very small drift levels due to the extremely brittle system behavior. Table 5.2 presents both the expected sliding resistance and the overturning resistance for each of the systems using the method outlined by Weiler in Chapter 13 of the Concrete Design Handbook (Weiler, 2006). The results obtained from implementing the non- linear pushover model are also presented. The maximum lateral load, Vmax, is the maximum lateral load resistance achieved by the system during the pushover analysis. It is also indicated what the mechanism was at the point of maximum lateral force as well as the drift level it occurred at. Chapter 5 Analytical Study of System Response 116   Table 5.2: Lateral capacity and mechanism found by pushover model and the nominal sliding and overturning  capacities expected from design and mechanism for all panel systems.  System  ID  Non‐Linear (N.L) Pushover Model  Simplified Model Used in Design  VMax N.L/  VMax Design  Sliding  Resistance  N.L / VMax  Design  VMax  [kN]  Drift at  VMax  [%]  Mechanism  at VMax  Sliding  Resistance  at VMax [kN] Design  Sliding  Capacity  [kN]  Design  Overturning  Capacity  [kN]  Governing  Mechanism  1,0,2  217  1.99  Overturning  449  609  132  Overturning  1.64  0.74  1,0,3  253  1.77  Overturning  621  834  132  Overturning  1.92  0.74  1,0,4  289  1.65  Overturning  685  1059  132  Overturning  2.19  0.65  2,2,2  735  0.39  Overturning  1116  1218  530  Overturning  1.39  0.92  2,2,3  758  0.39  Overturning  1575  1668  530  Overturning  1.43  0.94  2,2,4  799  0.39  Overturning  1989  2118  530  Overturning  1.51  0.94  2,3,2  848  0.84  Overturning  994  1218  530  Overturning  1.60  0.82  2,3,3  996  0.80  Overturning  1347  1668  530  Overturning  1.88  0.81  2,3,4  1077  0.65  Overturning  1761  2118  530  Overturning  2.03  0.83  2,4,2  847  0.82  Overturning  995  1218  530  Overturning  1.60  0.82  2,4,3  995  0.79  Overturning  1349  1668  530  Overturning  1.88  0.81  2,4,4  1145  0.77  Overturning  1689  2118  530  Overturning  2.16  0.80  3,2,2  1281  0.40  Overturning  1694  1826  1191  Overturning  1.08  0.93  3,2,3  1325  0.39  Overturning  2363  2501  1191  Overturning  1.11  0.94  3,2,4  1386  0.39  Overturning  2983  3176  1191  Overturning  1.16  0.94  3,3,2  1471  0.40  Overturning  1692  1826  1191  Overturning  1.24  0.93  3,3,3  1596  0.40  Overturning  2300  2501  1191  Overturning  1.34  0.92  3,3,4  1720  0.40  Overturning  2900  3176  1191  Overturning  1.44  0.91  3,4,2  1627  0.27  Sliding†  1627  1826  1191  Overturning  1.37  0.93  3,4,3  1786  0.40  Overturning  2298  2501  1191  Overturning  1.50  0.92  3,4,4  1911  0.40  Overturning  2897  3176  1191  Overturning  1.60  0.91  4,2,2  1828  0.40  Overturning  2272  2435  1964  Overturning  0.93  0.93  4,2,3  1892  0.39  Overturning  3151  3335  1964  Overturning  0.96  0.94  4,2,4  1974  0.40  Overturning  3977  4235  1964  Overturning  1.01  0.94  4,3,2  2205  0.41  Overturning  2270  2435  2118  Overturning  1.04  0.93  4,3,3  2350  0.40  Overturning  3088  3335  2118  Overturning  1.11  0.93  4,3,4  2495  0.40  Overturning  3894  4235  2118  Overturning  1.18  0.92  4,4,2  2404  0.10  Sliding†  2404  2435  2118  Overturning  1.14  0.98  4,4,3  2728  0.40  Overturning  3086  3335  2118  Overturning  1.29  0.93  4,4,4  2874  0.40  Overturning  3891  4235  2118  Overturning  1.36  0.92  † 3,4,2 and 4,4,2 systems have overturning capacities of 2518 kN and 2909 kN respectively at maximum lateral force level   Figure 5.18 shows a plot of the actual lateral capacity values compared to the lower of the two design capacities shown above for the systems studied in this chapter. As it turns out the failure mechanism is always predicted to be the overturning mechanism for the configurations studied. Figure 5.19 shows a plot of the ratio of maximum lateral force achieved using the non-linear pushover model to the nominal overturning resistance expected from design. This is used to determine if current design strength calculations are accurate and conservative. Chapter 5 Analytical Study of System Response 117   Figure 5.19: Ratio of maximum lateral force determined by pushover model to nominal lateral  overturning resistance expected in design.  0.0 0.5 1.0 1.5 2.0 2.5 A ct ua l V M ax /  N om in al  E xp ec te d  O ve rt ur ni ng   Re si st an ce 2 Panel‐Slab Conn. 3 Panel‐Slab Conn. 4 Panel‐Slab Conn. 3 Panels 4 Panels1 Panel 2 Panels Figure 5.18: Maximum lateral force determined by the pushover model compared to the nominal  overturning resistance expected in design.    0 500 1000 1500 2000 2500 3000 1 0 2 1 0 3 1 0 4 2 2 2 2 2 3 2 2 4 2 3 2 2 3 3 2 3 4 2 4 2 2 4 3 2 4 4 3 2 2 3 2 3 3 2 4 3 3 2 3 3 3 3 3 4 3 4 2 3 4 3 3 4 4 4 2 2 4 2 3 4 2 4 4 3 2 4 3 3 4 3 4 4 4 2 4 4 3 4 4 4 La te ra l F or ce  (k N ) Actual Vmax Nominal Overturning Resistance Expected in Design # Panels # Panel-Panel Conn. # Panel-Slab Conn. Chapter 5 Analytical Study of System Response 118   These plots show that when predicting the overturning capacity of the system, the current method used in design is conservative, however it is only conservative for very small drift levels. Figure 5.20 shows the response of the 3,2,4 system with the residual sliding capacity, as well as the expected overturning and sliding capacities using current design methods. It is apparent from this figure that the expected overturning capacity is only conservative until there is panel – panel connection failure at 0.39% drift ratio. Since current design methods do not consider the ductility of the system, it fails to predict this significant drop in lateral force capacity due to the panel – panel failure. Figure 5.20: Shear resistance envelope and lateral force response from non‐linear pushover model  compared with expected values from current design methods for the 3,2,4 system.  When the model presented in this chapter was implemented, the residual sliding resistance at the maximum lateral force level was determined and is presented in Table 5.2. These values have taken into account the effect that load reversals would have on the residual sliding resistance. The degrading effect due to vertical damage produced tilt-up system strength capacities which fall below the expected sliding resistance values using current design methods. Figure 5.21 compares the residual sliding resistance at the maximum lateral force with the expected sliding resistance from design and Figure 5.22 plots the ratios. 0 500 1000 1500 2000 2500 3000 3500 0 0.5 1 1.5 2 2.5 3 La te ra l F or ce  [k N ] Drift [%] Shear Resistance from  N.L Pushover Model Actual Response of  3,2,4 System from N.L  Pushover Model Sliding Resistance  from Design Method Overturning  Resistance from  Design Method Chapter 5 Analytical Study of System Response 119   Figure 5.21: Actual sliding resistance at maximum lateral force determined by the pushover model  compared to expected nominal sliding resistance expected in design.    Figure 5.22: Ratio of the actual sliding resistance at maximum lateral force determined by the  pushover model to nominal lateral sliding resistance expected in design.  0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 1 0 2 1 0 3 1 0 4 2 2 2 2 2 3 2 2 4 2 3 2 2 3 3 2 3 4 2 4 2 2 4 3 2 4 4 3 2 2 3 2 3 3 2 4 3 3 2 3 3 3 3 3 4 3 4 2 3 4 3 3 4 4 4 2 2 4 2 3 4 2 4 4 3 2 4 3 3 4 3 4 4 4 2 4 4 3 4 4 4 La te ra l F or ce  (k N ) Nominal Sliding Resistance Expected in Design Actual Sliding Resistance at Vmax 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 Sl id in g  Re si st an ce  a t V M ax /  N om in al  E xp ec te d  Sl id in g  Re si st an ce 2 Panel‐Slab Conn. 3 Panel‐Slab Conn. 4 Panel‐Slab Conn. 3 Panels 4 Panels1 Panel 2 Panels # Panels # Panel-Panel Conn. # Panel-Slab Conn. Chapter 5 Analytical Study of System Response 120   These plots show the effect that the degradation of the EM5 has on the predicted sliding resistance. These figures however only include one point on the sliding resistance envelope for each system. In reality there is an envelope for each system which will decay differently depending on the system response. Figure 5.20 shows an example sliding resistance envelope for the 3,2,4 system, and the point on that envelope used in Figure 5.21 and Figure 5.22 is indicated at 0.39% drift. The point that was chosen was the point of maximum lateral force as this is the location when an overturning mechanism would be the closest to the sliding resistance envelope (excluding when panel – slab connectors begin to fail). This is significant because if the actual response were to exceed the sliding resistance envelope, a sliding failure would occur. If this occurred at the point of maximum lateral force, a brittle sliding failure would occur which would not be detected using current design methods, as seen in the 3,4,2 and 4,4,2 systems. Designers attempt to resist a sliding mechanism by ignoring any overturning and adding panel – slab connectors with the assumption that they will always retain their full horizontal shear capacity. They then attempt to resist an overturning mechanism by adding panel – panel connectors to tie panels together while ignoring the connectors they added to resist sliding. This method of trying to resist these mechanisms independently and without consideration of how one remedy will affect the other leads to uncertainty of what the failure mechanism will be. It is shown in this study that both the horizontal shear strength (resistance to sliding) and the stiffness of the EM5 connector used in the panel – slab connection assembly are significantly affected by its vertical damage level. This degradation accounts for the fractional values shown in Figure 5.21 and Figure 5.22. It is also shown that the EM5 connector has significant resistance in vertical shear (up to 115 kN), which will add significantly more shear force to the panel – panel connectors being used to resist overturning than what is designed for. These results show that the solutions to resist sliding are likely always under designed and overturning resistance is under designed after relatively small drift levels with respect to the values a designer might expect to achieve. This leads to uncertainties of the prediction of the capacities and failure mechanism. Chapter 5 Analytical Study of System Response 121   5.4 Tilt-up System Ductility Discussion The final consideration to make when studying the tilt-up systems presented in this chapter is with respect to seismic ductility. Currently there is no guidance provided by CSA A23.3 for choosing a system R-value or for designing tilt-up panel systems where seismic ductility is achieved only through panel overturning and edge connections. Many designers have assumed that seismic ductility would be achieved through panel rocking, plastic yielding of the panel edge connections and, to some extent, deformation of the roof diaphragm (WEILER, 2006). It would be great if seismic ductility could be found in these places, however with perhaps the exception of the roof diaphragm, nothing in the tilt-up system is ductile and so it is almost impossible to expect system ductility. Achieving ductility and energy dissipation within a solid concrete tilt-up panel is not usually a consideration, and this study has shown that even though it has slightly more ductility than other connectors, the EM5 connector does not have enough displacement capacity to provide significant system ductility. Currently in design, often EM2 wall connectors are used in series with the EM5 connector in the slab. The EM2 connector has been shown in Chapter 2 to have an average shear strength capacity of 132 kN, which is substantially lower than the 225 kN shear capacity of the EM5 connector found in this study. When these connectors are used in series, the ductility is governed by the extremely brittle EM2 instead of the EM5 which has at least some ductility. In a situation such as this when two connectors are in series, it is important that the strength capacity of the ductile component be achieved in order to take advantage of its ductile properties otherwise once the strength capacity of the brittle connector is exceeded, failure occurs. This can potentially be resolved by using the EM3 as the wall connector which has a similar strength to the EM5 connector and can for sure be resolved if the EM4 connector is used in the wall panel. Because of these uncertainties, it is recommended in the concrete design handbook (Weiler, 2006) that panel overturning moments be computed based on Rd = 1.5 and Ro = 1.3 forces and the lateral sliding forces are computed for Rd = 1.0 and Ro = 1.3 based on the extremely non-ductile EM2/3 stud plate within the wall panel. These low R- values are justified, as the system ductility is very small. One of the desirable features of using tilt-up panels as the SFRS is their potential to re-distribute seismic shear forces within a line of panels. All panels can contribute to the seismic resistance, Chapter 5 Analytical Study of System Response 122   but some more than others. If the overturning resistance of a narrow panel is exceeded, the excess forces can be transferred to the wider panels by means of the connecting drag struts. Similarly, excess shear force on panel – panel connections do not necessarily mean that total failure will occur at this point. Deflection compatibility within the panel line ensures that the forces are re-distributed to the other panels. If the connections have reasonable ductility, it is possible to have most or all of the panels working together (Weiler, 2006). However for this to occur, the load path must remain intact, and for the load path to stay intact, the panel – panel connection must be ductile. As seen earlier in this chapter, the problem currently is that if few panel – panel connectors are used, such that the strength capacity is not enough to lift the adjacent panel, a drift level of only θ ൌ 2 ൈ ∆hMୟ୶ W െ c is achieved (0.39% for a 25 foot panel) before this connection reaches its ultimate displacement capacity. This is because a panel – panel connection, consisting of 2 – EM5 connectors in series, only has a horizontal displacement capacity of 2 X 15 mm before failure is reached. On the other hand, if many panel – panel connectors are used such that the adjacent panel is able to be lifted from the ground, the panel – slab connectors, which are designed to resist sliding, will fail prematurely due to the increased vertical displacement demand from initial panel rocking. Once enough of the panel – slab connectors fail such that the sum of friction and the shear capacity of the remaining panel – slab connectors is overcome, sliding will occur. Also, even if this strength design approach is used, it is shown in this chapter that under no circumstances can even four panel – panel connectors lift more than one adjacent panel off the ground. This means that the load path is lost, and at no time will more than two panels work together to resist overturning at drift levels greater than 0.39% using the panel dimensions in this study. It is seen earlier in this chapter that none of the systems really have much system ductility, as most of the systems loose 40 – 60% of their load carrying capacity after 0.39% drift. Even though none of the systems displayed much system ductility, some of the system configurations Chapter 5 Analytical Study of System Response 123   produced more favorable performances than others. Consideration for the buildings geometry must be made when deciding which designed mechanism is more desirable. In reality, buildings are not always perfectly rectangular and can have re-entrant corners, or some of the walls may not be parallel to each other. When these complications in building geometry are considered, the practicability of the sliding mechanism becomes questionable. Also due to the residual displacement inherent in the sliding mechanism, earthquake damage to a building designed to fail in this manner would be difficult to repair. The rocking mechanism is more practicable for solid panels since it inherently does not involve a residual displacement; rocking panels always return to their original position (Olund 2009). This is important as in the pushover analysis conducted in this study, the two panel systems of Figure 5.4 which ultimately experience a sliding mechanism appear to have the most ductile responses, however if sliding is not practical, than these systems cannot be considered. If the panel systems are designed to overturn, even when all of the panel – slab connectors have failed, then there will never be a risk of residual displacements, and the building will more easily be repaired. In order to achieve this, the panels must be designed to have a aspect ratio which is conducive to overturning when only friction acting to prevent sliding, and the panel – panel connections must not be stronger than the sum of the weight of the adjacent panel and the uplift capacity of the panel – slab connectors. The panel – panel connectors must also not have more displacement capacity than the panel – slab connectors in uplift, which is always the case with current connectors, however must still hold true if now connectors are developed. All these measures will ensure that the panel – panel connection fails before the panel – slab connections and overturning will result. It could be argued that when looking at the pushover results that the drifts would never reach levels which would cause all of the panel – slab connectors to fail in uplift, and so the final residual failure mechanism is not important, however for the pushover analysis, the displacement capacities reached are based on lateral loading in only one direction. In reality, load reversals will occur, which would cause failure of the panel – slab connectors at lower drift levels than predicted in the pushover analysis due to panel rocking in the opposite direction. This lowers the drift capacity to levels which could conceivably be achieved in reality (approximately 1 – 3.5% drift) which makes the selection of the residual mechanism relevant. Another simple design modification which could maximize the rotational drift capacity when a system experiences Chapter 5 Analytical Study of System Response 124   panel rocking would be to located the connectors as close to the center of the panel as possible. This will allow for the largest rotational drift possible without affecting the overturning or sliding resistances. 5.5 Performance of Ideal Connectors It is not the intent of this study to suggest specific solutions to all the issues discussed above, but to study in detail the current connectors being used, and show the panel system response produced by using these connectors. Through this study it has become clearer the type of properties which ideal connectors would possess to allow for a more favorable system response. This is what will be presented in this section with the knowledge that more research would be needed to develop connectors which more closely fit these properties. Figure 5.23 and Figure 5.24 below show an example pushover using the current panel – panel connectors with redesigned panel – slab connectors. The only property that was changed was that the uplift displacement capacity. It was increased from 100 mm to 300 mm. The vertical connector strength was left unchanged. These responses plots are also shortened to include the impact load reversals would have, so the displacement capacities can be directly compared. Chapter 5 Analytical Study of System Response 125   Figure 5.23: Comparison of the current 4,3,4 system response with the 4,3,4 response using idealized  panel – slab connections  Normalized to Max Force.    Figure 5.24: Comparison of the current 4,3,4 system response with the 4,3,4 response using idealized  panel – slab connections  Normalized to Residual Force  Figure 5.23 and Figure 5.24 above show how improvements to the panel – slab connectors could affect the system response. An ideal panel – slab connector would include connectors to resist horizontal shear and vertical uplift independently. For the example plots shown above, horizontal 0.0 0.2 0.4 0.6 0.8 1.0 0 1 2 3 4 5 6 7 8 9 10 La te ra l F or ce  (%  o f M ax  F or ce ) Panel Drift (%) Currently Used  4,3,4 System Redesigned Panel ‐ Slab 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 0 1 2 3 4 5 6 7 8 9 10 La te ra l F or ce  (%  o f M ax  F or ce ) Panel Drift (%) Currently Used 4,3,4 System Redesigned Panel ‐ Slab Chapter 5 Analytical Study of System Response 126   shear is taken by dowels cast in the foundation which fit into sleeves cast into the wall panels, whereas vertical uplift is resisted by connectors with the same strength capacity as the EM5 connector in uplift (115 kN), located at near the center of the wall panel with an increased vertical displacement capacity of 300 mm. It is apparent from the plots that this improvement alone does not influence the ductility, as the initial capacity loss at 0.39% experienced by the system using current connectors still occurs. It does substantially increase the rotational drift capacity at which all of the panel – slab connectors fail, however this is still far from ideal as there is still a large capacity loss at a very small drift level. This capacity loss as discussed earlier is due to the premature failure of the panel – panel connectors due to their lack of displacement capacity. Next Figure 5.25 and Figure 5.26 show the effects on the system response if only the panel – panel connection is improved. If the strength capacity is kept the same, but the displacement capacity of two EM5 connectors in series is increased from 30 mm to 150 mm, the result is as shown in the plots below. Figure 5.25: Comparison of the current 4,3,4 system response with the 4,3,4 response using idealized  panel – panel connections  Normalized to Max Force  0.0 0.2 0.4 0.6 0.8 1.0 0 1 2 3 4 5 6 7 8 9 10 La te ra l F or ce  (%  o f M ax  F or ce ) Panel Drift (%) Redesigned Panel ‐ Panel Currently Used  4,3,4 System Chapter 5 Analytical Study of System Response 127     Figure 5.26: Comparison of the current 4,3,4 system response with the 4,3,4 response using idealized  panel – panel connections  Normalized to Residual Force  It can be seen that this makes a significant improvement to the drift level at which the load capacity drops. Instead of occurring at 0.39% drift as before, the system now yields until 1.99% drift when the panel – panel connection fails. This is a significant improvement in system ductility, and it is obvious from this result that the panel – panel connector design is vitally important to system ductility. From this result it is obvious that the more ductile the panel – panel connections are, the more ductile the system response will be. The panel – panel ductility must be limited however by the panel – slab uplift ductility to ensure it will fail before the panel – slab connectors to allow overturning, therefore to further increase the system ductility, both the panel – panel and panel – slab connectors must be redesigned. When both the redesigns described above are combined, the response generated is shown in Figure 5.27 and Figure 5.28. If the panel – slab vertical displacement capacity is increased to 300 mm, it allows for the panel – panel displacement capacity to be increased to approximately 500 mm to still ensure overturning. If a panel – panel displacement capacity of 500 mm and a panel – slab vertical displacement capacity of 300 mm is used, the resulting pushover response is shown below. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 0 1 2 3 4 5 6 7 8 9 10 La te ra l F or ce  (%  o f M ax  F or ce ) Panel Drift (%) Redesigned Panel ‐ Panel Currently Used 4,3,4 System Chapter 5 Analytical Study of System Response 128   Figure 5.27: Comparison of the current 4,3,4 system response with the 4,3,4 response using idealized  panel – slab and panel – panel connections  Normalized to Max Force    Figure 5.28: Comparison of the current 4,3,4 system response with the 4,3,4 response using idealized  panel – slab and panel – panel connections Normalized to Residual Force  This result is a remarkable improvement over any other system. This system displays large system ductility, and the panel – panel connection remains intact until a drift level of 6.6%. These are just examples of how increasing the ductility of the connectors would influence the system response, however if connectors could be developed which display these properties, 0.0 0.2 0.4 0.6 0.8 1.0 0 1 2 3 4 5 6 7 8 9 10 La te ra l F or ce  (%  o f M ax  F or ce ) Panel Drift (%) Currently Used  4,3,4 System Ideal Redesigned Connectors 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 0 1 2 3 4 5 6 7 8 9 10 La te ra l F or ce  (%  o f M ax  F or ce ) Panel Drift (%) Currently Used 4,3,4 System Ideal Redesigned Connectors Chapter 5 Analytical Study of System Response 129   finally multiple panels could be connected together and continue to work together in dissipating seismic energy through the inelastic deformations of the connectors during panel rocking. Chapter 5 Analytical Study of System Response 130 6 CONCLUSIONS AND RECOMMENDATIONS This study was conducted to determine the vertical response of the base slab connectors used in current tilt-up design, as well as how the presence of vertical displacement due to panel rocking affects the horizontal shear strength, stiffness and displacement capacity properties. All this information was then used to predict the pushover response of tilt-up systems with solid wall panels and with system configurations similar to what would be constructed in practice. This chapter summarized the findings of this study, and provides recommendations for improved system design as well as further research needs. 6.1 Summary of Observations A review of the current design practice of tilt-up structures was conducted and it was discovered that design is done without designing for a particular mechanism to form. For a single story tilt- up structure with solid wall panels the two mechanisms which could form are sliding and overturning. Since solid wall panels would remain rigid during a seismic event, most of the seismic energy dissipation would be due to inelastic deformations of the connectors. Currently overturning is resisted by adding panel – panel connectors to tie multiple panels together to increase overturning resistance, and sliding is resisted by adding panel – slab connectors, without considering how vertical displacements on the panel – slab connectors will affect their horizontal shear capacity. Previous testing conducted at the University of British Columbia only measured the horizontal capacity of the EM5 connector so the affects of vertical uplift were left unknown. This study has found that: • The current EM5 – EM2/3 weld detail does not allow for multiple cycles to be performed at large vertical displacements. When load reversals occur, the weld breaks due to rotations of the EM5 edge angle. This substantially decreases the displacement capacity of the panel – slab connector, and causes it to fail prematurely. A revised design is given in Section 4.2.3 which was used for testing purposes and remedied this problem. 131   • The EM5 slab connector has on average 115 kN of uplift capacity and a uplift displacement capacity of approximately 100 mm. These relatively large uplift values show that the effects they have on overturning and on the shear transfer to the panel – panel connectors should not be neglected, as it can be a significant source of overturning resistance, but also can significantly increase the shear force being transferred through the panel – panel connectors. • The EM5 horizontal shear capacity has an average cyclic strength capacity of 220 kN which is in close agreement with the values found in the previous testing discussed in Chapter 2. Weiler recommended for the CAC Concrete Design Handbook design values of ׎௖ ெܸ௜௡ ൌ 0.6 ൈ 210 ݇ܰ ؆ 125݇ܰ. The results of this study confirm this recommendation. • Vertical damage due to uplift displacement significantly affects the horizontal shear capacity of the EM5 panel – slab connector. This study has shown in Figure 4.24 that the horizontal shear capacity degrades somewhat linearly (െ0.8∆ݒெ௔௫) as the vertical displacement increases. Whether or not the connector remains at the maximum vertical displacement does not noticeably affect the strength degradation, as the trend was consistent for both cases. • Vertical damage due to uplift displacement significantly affects the secondary horizontal shear stiffness of the EM5 panel – slab connector. Until approximately 40 kN is applied the initial stiffness is virtually unaffected by vertical displacement, however after this initial portion of the horizontal response, Figure 4.26 shows that the secondary stiffness degrades from a slope of 100 kN/mm when there is no vertical damage to a slope of approximately 25 kN/mm when there is 50 mm of vertical displacement damage. After 50 mm of vertical damage the stiffness is not affected further. Again these results are consistent for both the case when the maximum vertical displacement is held, and when the vertical displacement is pushed back to zero. Chapter 6 Conclusions and Recommendations 132   • Vertical damage due to uplift displacement does not noticeably affect the displacement capacity of the EM5 panel – slab connector. Figure 4.27 shows that the displacement capacity is not dependent on the vertical damage. When the results from this study are compared to previous testing in Figure 4.20 it is observed that displacement capacity is more dependent on number of cycles than anything else. • Restricting out of plane movements at the base of the wall does not affect any of the results. Test specimens 17 and 18 were conducted while limiting the amount of out of plane movement allowed, and the results were consistent with the results when out of plane movements at the base of the wall were not restricted. This does not mean that out of plane movements are not important, as they obviously will significantly affect the system response (which is a recommendation for future research), however this study shows that these out of plane movements do not affect the panel – slab connection response. These response properties of the EM5 panel – slab connector found through experimental testing were then used in the multi-panel pushover model outlined in Chapter 5 to predict multi-panel tilt-up system responses for various system configurations found in common practice. Comparisons of these results can be found in Chapter 5, and individual system results are found in Appendix B. The following results were discovered through this study. • All systems experienced panel rocking however to varying degrees. The only systems which didn’t experience much panel rocking were systems with multiple panels ( ≥ 3 panels) tied together using very strong panel – panel connections (4 panel – panel connectors per panel) and very weak panel – slab connections (< 2 panel – slab connections per panel). In these systems, the lateral load exceeded the capacity of the panel – slab connectors plus friction almost immediately, causing an extremely brittle sliding failure. This type of design is not recommended as it provides very little energy dissipation before a very brittle failure. All other systems experienced panel rocking with two possible final failure mechanisms. If all of the panel – panel connectors fail, the system will result in all panel overturning independently as the final residual capacity Chapter 6 Conclusions and Recommendations 133   (provided the aspect ratios of the individual panels are narrow enough that the overturning resistance is smaller than the sliding resistance due to friction). If any of the panel – panel connections do not fail before all the panel – slab connections fail, the overturning capacity may be increased, and a sliding mechanism form. • In all systems which had more than two panels (and two panel systems with weak panel – panel connection), panel – panel failure occurred at very small drift levels (θ ൌ ଶൈଵହ୫୫ Wିୡ  ). This generated system capacity losses of between 35 and 62% depending on the system configuration. This capacity loss is due to insufficient strength of the panel – panel connection to lift the adjacent panel from the ground and the small inelastic displacement capacity of the EM5 – EM5 panel – panel connectors. • The overturning calculations made in design to determine the overturning resistance of the panel system is a conservative estimate of the maximum lateral load that may be applied, as shown in Figure 5.19, however this in most cases is only for very small drift levels. As the panel – panel connectors have very little ductility, the overturning capacity drops well below predicted values when they fail, which in many cases in this study was only at a drift level of 0.39%. • Only a fraction of the expected horizontal sliding design strength is achieved by the systems, as shown in Figure 5.22. Currently designers use the full factored horizontal strength of the panel – slab connectors plus friction to design for sliding resistance, however due to the horizontal shear resistance of the panel – slab connectors being degraded when uplift damage is introduced, the actual shear capacity degrades also. This means that as the panel system rocks, the sliding resistance assumption becomes less and less conservative. • Increasing the number of panel – slab connectors increases the systems lateral load carrying capacity and stiffness, and generates a smaller lateral capacity drop as each of the panel – slab connectors fail in uplift. Chapter 6 Conclusions and Recommendations 134   6.2 Recommendations 6.2.1 Recommendations for Current Design It was stated in the introduction that one of the objectives of this study was to give some guidance in the design of tilt-up buildings for in-plane shear. Although the pushover model used in this study modeled only solid panel systems with constant panel widths, it has helped in understanding how a panel system behaves so that some guidance in modifying connector configuration can be given. Ultimately connectors with properties more conducive to a ductile response should be designed as none of the systems studied show a very desirable response, so this section serves to try to suggest ways of making current connectors more efficient in generating the most desirable response possible until more research is done. • The revised EM5 – EM3 weld detail shown in Section 4.2.3 should be incorporated into new design in order to maximize the vertical displacement capacity of the connector. This design has been shown through the experimental program of this study to provide a significant improvement in uplift displacement capacity to allow the connection to achieve its maximum uplift displacement. • The EM2 wall embed plate should not be used for the panel – slab connection as it has a lower shear strength than the EM5 connector and so the ductility of the connector is governed by the more brittle EM2 connector. Although the EM5 does not have much ductility, it does have a more desirable response than the EM2, and so only the EM3 connector should be used. • Tilt-up panels and systems should be designed in such a way that should the panel – slab connectors become disconnected that overturning still occur instead of sliding. By doing so it is ensured that the panels will have no residual displacements as the rocking panels will come to rest in the same position as they started, unlike a sliding mechanism. This is done by ensuring that all the panel – panel connections fail before the panel – slab connections do, and that the individual panel dimensions are conducive to overturning. Chapter 6 Conclusions and Recommendations 135   Using current connectors, the easiest way to ensure panel – panel failure occurs first would be to ensure the strength of the panel – panel connection is less than the sum of the weight of the adjacent panel and the uplift capacity of the panel – slab connectors (115 ݇ܰ  ൈ  ܰݑܾ݉݁ݎ ݋݂ ݌݈ܽ݊݁ –  ݏ݈ܾܽ ܿ݋݊݊݁ܿݐ݋ݎݏ). • Panel – slab connectors should be located close to the center of the panel. A seismic event will have load reversals, which is neglected in a pushover analysis, thus having the connectors in the center of the panel will allow for the greatest rotational drift before uplift failure of the panel – slab connectors occur. 6.2.2 Recommendations for Future Research This study has helped to begin to understand the seismic behavior of tilt-up systems. It has also served to highlight the need for further research in this area. Some recommendations for future research which this study has neglected include: • Wall systems with panels with openings. This study focused only on solid panels which can be considered rigid, and so all the energy dissipation occurred in the connectors. Since many tilt-up applications do not have solid wall panels, systems where the panel may yield and contribute to the energy dissipation should be studied. • The affects of out of plane movements of the walls during in-plane rocking. This study only considered out of plane movement at the base location for the EM5 connector response, however the pushover model did not consider the affects of out of plane movements on the system response. It is conceivable that during a seismic event, the in- plane response would be affected by the out of plane movements. • Development of new and more ductile connectors and configurations. This study has shown that none of the current connectors display ductile properties, as well as very low displacement capacities which limit their usefulness for the intended purpose. The horizontal shear capacity of the current connectors are significantly affected by the uplift Chapter 6 Conclusions and Recommendations 136   displacement (as shown in Chapter 4), which is not ideal. A panel – slab connection detail should be developed to decouple the horizontal and vertical responses. Chapter 6 Conclusions and Recommendations 137 REFERENCES ACI Committee 209, (1982) “Prediction of Creep, Shrinkage and Temperature Effects in Concrete Structures,” Designing for Creep and Shrinkage in Concrete Structures, ACI Publication SP-76, American Concrete Institute, Detroit, 1982, pp. 193-300. Adebar, P., Guan, Z., and Elwood, K., (2004) “Displacement-Based Design of Concrete Tilt-up Frames Accounting for Flexible Diaphragms”, Proceedings of the 13th World Conference on Earthquake Engineering, Vancouver, BC, Canada. Canadian Standards Association (CSA), CSA A23.1-94, Concrete Materials and Methods of Concrete Construction, Rexdale, Ontario, 1994, pp 1-148. Canadian Standards Association (CSA), CSA G30.18-M92 (R 2002), Billet-Steel Bars for Concrete Reinforcement, Rexdale, Ontario, 2002. Essa, H.S., Tremblay, R., and Rogers, C.A. (2003), “Behaviour of Roof Deck Diaphragms under Quasi-Static Cyclic Loading,” Journal of Structural Engineering, 129(12), 1658-1666. Hamburger, R.O., and McCormick, D.L. (1994) “Implications of the January 17, 1994, Northridge Earthquake on Tilt-up and Masonry Buildings with Wood Roofs.” Structural Engineers Association of Northern California (SEAONC), Seminar Papers. San Francisco, CA, pp. 243-255. Lemieux, K., Sexsmith, R., and Weiler, G. (1998), “Behavior of Embedded Steel Connectors in Concrete Tilt-up Panels,” ACI Structural Journal, V. 95, No. 4, July-August, pp. 400-411. Olund, O. (2009) “Analytical Study to Investigate the Seismic Performance of Single Story Tilt- up Structures,” M.A.Sc. Thesis, Department of Civil Engineering, University of British Columbia, Vancouver, Canada. 138 Price, W., (1951) “Factors Influencing Concrete Strength,” ACI Journal, Proceedings, Vol. 47, No. 6, December 1951, pp. 417-432. Tilt-up Concrete Association (TCA) (2007), “Building with Tilt-up.” Retrieved April 21, 2007 from http://www.tilt-up.org/build/index.htm Weiler, G., (2006), “Chapter 13 - Concrete Design Handbook, 3rd Edition”, Cement Association of Canada (CAC) CAC, Ottawa, Ontario, Canada. Weiler, G. et al, (1997), “Memo to SECBC dated September 30, 1997,” Committee for Standard Tilt-up Connections, Vancouver, BC, Canada Weiler Smith Bowers, Sample Drawings for Tilt-up Building, 2008, Vancouver, B.C., Canada. 139 APPENDIX A: EXPERIMENTAL PROGRAM TEST RESULTS 140  1 TEST NUMBER 1 Test Description: Monotonic vertical uplift test number one with round weld detail. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 1 07/03/08 Round 78 N/A Monotonic N/A Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 78 87 N/A N/A N/A N/A N/A N/A No Failure Loading Protocol: Testing Comments: The first attempt to test this specimen on July 3, 2008 failed due to the insufficient load applying capacity of the vertical loading jacks. After obtaining larger jacks, testing resumed on July 9, 2008. The specimen was clamped to the strong floor to prevent vertical movement, however horizontal movement of the loading beam occurred due to the loading plate sliding over the top edge of the specimen (shown in Figure A - 1.8). Since in an actual tilt-up building the panel would be restricted from sliding over the edge of the slab, the test was not relevant after this point, hence the specimen was not completely failed and only the first part of the load- displacement plot is reliable. This was corrected by adding stop blocks to the testing frame that prevented inward horizontal movement of the loading beam. This is discussed more in chapter 4. Appendix A Experimental Program Test Results 141   Vertical Response Plots: Figure A ‐ 1.1: The monotonic vertical response of specimen 1.  Figure A ‐ 1.2: Out‐of‐plane movement during monotonic vertical test number 1.  Appendix A Experimental Program Test Results 142   Photos: Figure A ‐ 1.3  Figure A ‐ 1.4  Figure A ‐ 1.5  Figure A ‐ 1.6  Figure A ‐ 1.7  Figure A ‐ 1.8  A ppendix A Experim ental Program Test R esults 143   2 TEST NUMBER 2 Test Description: Monotonic vertical uplift test number two with round weld detail. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 2 07/14/08 Round 132 N/A Monotonic N/A Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 132 166 N/A N/A N/A N/A N/A N/A Angle Failure Loading Protocol: Testing Comments: This test was conducted on July 14, 2008 and was a repeat of the monotonic uplift conducted in test number one. The testing rig was repaired with stops so that the steel plate simulating the wall panel was not allowed to move over the edge of the slab which resulted in the second half the test number one to be inaccurate. This repair worked properly as the steel plate stayed outside the edge of the concrete as shown in Figure A - 2.5. The concrete slab was fastened to the strong floor by use of vertical rods to ensure there was no vertical movement, however, near the end of this test, slight inward horizontal movement of the slab was observed since it was not anticipated that the force level would be as high as was observed, which thus was slightly greater than the friction force. This movement was deemed insignificant to the results as this movement was very small (~5mm). Appendix A Experimental Program Test Results 144   Vertical Response Plots: Figure A ‐ 2.1: The monotonic vertical response of specimen 2.  Figure A ‐ 2.2: Out‐of‐plane movement during the monotonic vertical test 2.  Appendix A Experimental Program Test Results 145   Photos: Figure A ‐ 2.3  Figure A ‐ 2.4  Figure A ‐ 2.5  Figure A ‐ 2.6  Figure A ‐ 2.7  Figure A ‐ 2.8  A ppendix A Experim ental Program Test R esults 146   3 TEST NUMBER 3 Test Description: Cyclic vertical uplift with round weld detail. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 3 07/17/08 Round 75 N/A Cyclic N/A Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 75 86 N/A N/A N/A N/A N/A N/A Weld Failure Loading Protocol: Testing Comments: This test was conducted on July 17, 2008 and was conducted in cyclic uplift. To avoid horizontal movement of the concrete slab, the slab was grouted to the strong floor to ensure no movement. Three cycles at Δv = 50 mm followed by two cycles at Δv = 75 mm were applied before the wall to slab connector weld fractured as shown in Figure A - 3.7. This weld fracture was caused by the connector angle being torque about the weld in its weak direction as shown slightly in Figure A - 3.6 (This torque is more obvious in following tests). At this point it is noted that this was not the failure of the connector itself, but the failure of the wall to slab connection weld. Although in a real structure this weld failure is significant because there is no longer any load carrying capacity, it is not the intention of this study to test this weld, but the connector. This weld will be repaired and the test will continue until failure of the connector itself. Appendix A Experimental Program Test Results 147   Vertical Response Plots: Figure A ‐ 3.1: The cyclic vertical response of specimen 3.  * Out of plane response not available. Appendix A Experimental Program Test Results 148   Photos: Figure A ‐ 3.2  Figure A ‐ 3.3  Figure A ‐ 3.4  Figure A ‐ 3.5  Figure A ‐ 3.6  Figure A ‐ 3.7  A ppendix A Experim ental Program Test R esults 149   TEST NUMBER 3A Test Description: Cyclic vertical uplift test continued with repaired round weld detail. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 3A 07/17/08 Round 101 N/A Cyclic N/A Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 101 131 N/A N/A N/A N/A N/A N/A Rebar Fracture Loading Protocol: Testing Comments: This test was conducted on July 17, 2008 and was the continued test in cyclic uplift after the wall to slab connection was repaired. In test number three the connector was cycled to three cycles at Δv = 50 mm followed by two cycles at Δv = 75 mm were applied before the wall to slab connector weld fractured. It was repaired by re-welding over where it had previously failed, which can be seen in Figure A - 3A.2. The test then continued to a third cycle at Δv = 75 mm followed by three cycles at Δv = 100 mm. The wall to connector weld failed once again at Δv = 65 mm after the third cycle at Δv = 100 mm but was repaired once again and the test continued. The rebar finally fractured as Δv = 86 mm on the first cycle which was anticipated to reach Δv = 125 mm (Figure A-3A.7). Appendix A Experimental Program Test Results 150   Vertical Response Plots: Figure A ‐ 3A.1: The cyclic vertical response of specimen 3A.  * Out of plane response not available. Appendix A Experimental Program Test Results 151   Photos: Figure A‐3A.2  Figure A‐3A.3  Figure A‐3A.4  Figure A‐3A.5  Figure A‐3A.6  Figure A‐3A.7  A ppendix A Experim ental Program Test R esults 152   4 TEST NUMBER 4 Test Description: Cyclic uplift to Δv = 50 mm, held at Δv = 50 mm then cycled in horizontal shear. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 4 07/23/08 Round 51 50 Cyclic Cyclic Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 51 62 Error 195 Error -203 Error 199 Rebar Fracture Loading Protocol: Testing Comments: This test was conducted on July 23, 2008 and was to find the horizontal cyclic shear capacity while being held at Δv = 50 mm after cycling three time to Δv = 50 mm. The horizontal displacement capacity was not known before hand and so the step sizes were not well defined, so an initial horizontal jack displacement of ±25 mm was used. After observing the cycles at this level of displacement (Figure A - 4.8) it was apparent that the increase to a horizontal jack displacement of ±50 mm as planned would be too high, so it was reduced to ±37 mm, which caused the rebar to fracture during the second positive cycle shown in Figure A - 4.13. It was noticed that the jack displacement was significantly higher than the relative displacement Appendix A Experimental Program Test Results 153   between the connector and the concrete slab and so the Δh measurements recorded during this test were not accurate and have not been used in this report. It was noted that instrumentation to measure this relative displacement was required for accurate Δh results. It was not clear at the time of test the cause this significant discrepancy in Δh because it was not found out until analysis of the results. Special attention was given to this during the following tests. Appendix A Experimental Program Test Results 154   Vertical Response Plots: Figure A ‐ 4.1: The cyclic vertical response to 50 mm displacement of specimen 4.  Figure A ‐ 4.2: Out‐of‐plane movement during cyclic vertical test number 4.  Appendix A Experimental Program Test Results 155   Horizontal Response Plots: Figure A ‐ 4.3: The cyclic horizontal response of specimen 4.  Figure A ‐ 4.4: Out‐of‐plane movement during cyclic horizontal test number 4.  Appendix A Experimental Program Test Results 156   Photos: Figure A ‐ 4.5  Figure A ‐ 4.6  Figure A ‐ 4.7  Figure A ‐ 4.8  Figure A ‐ 4.9  Figure A ‐ 4.10  Figure A ‐ 4.11  Figure A ‐ 4.12  Figure A ‐ 4.13  Appendix A Experimental Program Test Results 157   5 TEST NUMBER 5 Test Description: Monotonic horizontal shear test with round weld detail. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 5 07/24/08 Round N/A 0 N/A Monotonic Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description N/A N/A 33 279 N/A N/A N/A N/A Rebar Fracture Loading Protocol: Testing Comments: This test was conducted on July 24, 2008 and was conducted in monotonic horizontal shear. At first it was decided to not restrain the upward movement of the loading beam. As the test was progressing it was noticed that only one end of the loading beam was lifting, causing a moment to be applied to the connector. As this is not desired, it was decided to stop loading and to clamp the loading beam in place at zero vertical displacement so it was restrained from any vertical movement. This caused a slight jump in the load noticed at Δh = 11 mm because the loading beam needed to be forced back down. Loading was continued and rebar fracture occurred. Again a significant discrepancy was noticed between the relative movement between the slab and connector and the horizontal jack movement. It was also observed that no perceivable movement Appendix A Experimental Program Test Results 158   occurred in the slab which was grouted to the strong floor or the loading column. It was then concluded that this movement was between the interfaces of loading plates, movement in bolts and loading jack clevises. Appendix A Experimental Program Test Results 159   Horizontal Response Plots: Figure A ‐ 5.1: The monotonic horizontal response of specimen 5.  Figure A ‐ 5.2: Out‐of‐plane movement during monotonic horizontal test number 5.  Appendix A Experimental Program Test Results 160   Photos: Figure A ‐ 5.3  Figure A ‐ 5.4  Figure A ‐ 5.5  Figure A ‐ 5.6  Figure A ‐ 5.7  Figure A ‐ 5.8  A ppendix A Experim ental Program Test R esults 161   6 TEST NUMBER 6 Test Description: Cyclic horizontal shear test with round weld detail. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 6 08/08/08 Round N/A 0 N/A Monotonic Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description N/A N/A 17 232 -20 -209 18 220 Rebar Fracture Loading Protocol: Testing Comments: This test was performed on August 8, 2008 and was conducted in cyclic horizontal shear. This test went according to plan with rebar fracture happening after three cycles at Δh+ = 15mm. This fracture occurred as the bend in the rebar is worked from bent (Figure A - 6.6) to straight (Figure A - 6.7). It was also noticed that the plot is skewed about 5 mm to the negative side. This is due to the discrepancy between the relative movement between the slab and connector and the horizontal jack movement. It was apparent that there was about 3 mm more discrepancy when the testing system was in compression in the initial positive cycle. This accounts for this plot skew, however it was still unclear exactly why it occurred. That being said, the relative Appendix A Experimental Program Test Results 162   displacement between the slab and the loading beam was accurately measured, and so this skew does not adversely affect the results. Appendix A Experimental Program Test Results 163   Horizontal Response Plots: Figure A ‐ 6.1: The cyclic horizontal response of specimen 6.  Figure A ‐ 6.2: Out‐of‐plane movement during cyclic horizontal test number 6.  Appendix A Experimental Program Test Results 164   Photos: Figure A ‐ 6.3  Figure A ‐ 6.4  Figure A ‐ 6.5  Figure A ‐ 6.6  Figure A ‐ 6.7  Figure A ‐ 6.8  A ppendix A Experim ental Program Test R esults 165   7 TEST NUMBER 7 Test Description: Cyclic uplift to Δv = 75 mm, held at Δv = 75 mm then cycled in horizontal shear. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 7 08/12/08 Round 75 75 Cyclic Cyclic Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 75 108 19 179 -17 -171 18 175 Rebar Fracture Loading Protocol: Testing Comments: This test was performed on August 12, 2008 and was conducted to determine the cyclic horizontal shear strength when the connector was cycled and held at Δv = 75 mm. During the vertical loading component of this test the wall to connector weld fractured during the third loading cycle as it did during the initial cyclic uplift test (Test #3). This weld was repaired just as in test 3A and the third cycle was completed. This horizontal test went according to plan with failure occurring once again due to rebar fracture during the second negative horizontal cycle of Δh = -17 mm (Figure A - 7.12). It was also noted during this test there was not the plot skew that was present in the previous horizontal tests. Although the discrepancy between the horizontal Appendix A Experimental Program Test Results 166   jack movement and relative movement between the slab and loading beam still existed, it was not skewed which suggests that this skew was dependent on the initial conditions of the testing setup. If the system was in tension before loading began (due to inserting the jack pin) the interface gaps would be open slightly which would have to be closed during the first cycle in compression, resulting in a skew to the negative side (as in test #6). However if the initial system was in compression before loading began, these gaps would be closed when loading began, which would skew the plot to the positive side when the first load reversal occurred and the gaps open up. If the system was not in tension of compression before loading began, and the gaps were perfectly half open, I would suspect then the plot would have no skew. Appendix A Experimental Program Test Results 167   Vertical Response Plots: Figure A ‐ 7.1: The cyclic vertical response to 75 mm displacement of specimen 7.  * Out of plane response not available. Appendix A Experimental Program Test Results 168   Horizontal Response Plots: Figure A ‐ 7.2: The cyclic horizontal response of specimen 7.  Figure A ‐ 7.3: Out‐of‐plane movement during cyclic horizontal test number 7.  Appendix A Experimental Program Test Results 169   Photos: Figure A ‐ 7.4  Figure A ‐ 7.5  Figure A ‐ 7.6  Figure A ‐ 7.7  Figure A ‐ 7.8  Figure A ‐ 7.9  Figure A ‐ 7.10  Figure A ‐ 7.11  Figure A ‐ 7.12  8 Appendix A Experimental Program Test Results 170   TEST NUMBER 8 Test Description: Cyclic vertical uplift with square weld detail. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 8 08/13/08 Square 100 N/A Cyclic N/A Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 100 138 N/A N/A N/A N/A N/A N/A Rebar Fracture Loading Protocol: Testing Comments: This test was performed on August 13, 2008 and was conducted in cyclic uplift. Due to the loading beam to connector weld failures during previous uplift testing which involved larger Δv it was clear that a new connection detail was needed. The cause of failure was due to a torque being applied to the connector angle about the failing welds weak direction. A new connection was developed using a 1.5”X 1.5” square connector. This would allow for small welds to be placed along the edges of the square bar which resisted the torque in the welds strong direction. This revised connection detail performed exactly as expected as the connector angle was not allowed to rotate (Figure A - 8.5) as it previously was. Failure of the connector occurred due to rebar fracture at the location where it is welded to the underside of the connector angle as shown in Figure A - 8.8. Appendix A Experimental Program Test Results 171   Vertical Response Plots: Figure A ‐ 8.1: The cyclic vertical response of specimen 8.  Figure A ‐ 8.2: Out‐of‐plane movement during cyclic vertical test number 8.  Appendix A Experimental Program Test Results 172   Photos: Figure A ‐ 8.3  Figure A ‐ 8.4  Figure A ‐ 8.5  Figure A ‐ 8.6  Figure A ‐ 8.7  Figure A ‐ 8.8  A ppendix A Experim ental Program Test R esults 173   9 TEST NUMBER 9 Test Description: Cyclic uplift to Δv = 100 mm, held at Δv = 100 mm then cycled in horizontal shear. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 9 08/14/08 Square 100 100 Cyclic Cyclic Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 100 105 15 143 -14 -149 14 146 Rebar-Angle weld Fracture Loading Protocol: Testing Comments: This test was performed on August 14, 2008 and was used to determine the cyclic horizontal shear response after being cycled and held at Δv = 100 mm. This test went according to plan and failure occurred due to the weld connecting the bent rebar to the connection angle failing as shown in Figure A - 9.12 and Figure A - 9.13. This occurred during the first negative cycle at Δh = -14 mm. It was expected that this test would produce a failure at lower displacement levels as the Δv was approaching Δv which causes failure. Also it was expected that it would produce a lower failure load because of the extreme vertical damage causing a softer connection. Appendix A Experimental Program Test Results 174   Vertical Response Plots: Figure A ‐ 9.1: The cyclic vertical response to 100 mm displacement of specimen 9.  Figure A ‐ 9.2: Out‐of‐plane movement during cyclic vertical test number 9.  Appendix A Experimental Program Test Results 175   Horizontal Response Plots: Figure A ‐ 9.3: The cyclic horizontal response of specimen 9.  Figure A ‐ 9.4: Out‐of‐plane movement during cyclic horizontal test number 9.  Appendix A Experimental Program Test Results 176   Photos: Figure A ‐ 9.5  Figure A ‐ 9.6  Figure A ‐ 9.7  Figure A ‐ 9.8  Figure A ‐ 9.9  Figure A ‐ 9.10  Figure A ‐ 9.11  Figure A ‐ 9.12  Figure A ‐ 9.13  Appendix A Experimental Program Test Results 177   10 TEST NUMBER 10 Test Description: Cyclic uplift to Δv = 100 mm, held at Δv = 0 mm then cycled in horizontal shear. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 10 08/15/08 Square 101 0 Cyclic Cyclic Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 101 127 21 133 -19 -152 20 142 Rebar Fracture Loading Protocol: Testing Comments: This test was performed on August 15, 2008 and was used to determine the cyclic horizontal shear response after being cycled to Δv = 100 mm three times then held at Δv = 0 mm. This test went according to plan and failure occurred due to fracture of the rebar (Figure A - 10.13) at the location of the bend due to bending (figure A-10.6) and straightening (Figure A - 10.12). This occurred during the second positive cycle at Δh = 21 mm. It was expected that this test would produce a failure at higher displacement levels than tests such as test 9 where the Δv was held at relatively large values, as the Δv would not contribute to the resultant failure displacement. Also Appendix A Experimental Program Test Results 178   it was expected that it would still produce a lower failure load because of the extreme vertical damage causing a softer connection. Appendix A Experimental Program Test Results 179   Vertical Response Plots: Figure A ‐ 10.1: The cyclic vertical response to 100 mm displacement of specimen 10.  Figure A ‐ 10.2: Out‐of‐plane movement during cyclic vertical test number 10.  Appendix A Experimental Program Test Results 180   Horizontal Response Plots: Figure A ‐ 10.3: The cyclic horizontal response of specimen 10.  Figure A ‐ 10.4: Out‐of‐plane movement during cyclic horizontal test number 10.  Appendix A Experimental Program Test Results 181   Photos: Figure A ‐ 10.5  Figure A ‐ 10.6  Figure A ‐ 10.7  Figure A ‐ 10.8  Figure A ‐ 10.9  Figure A ‐ 10.10  Figure A ‐ 10.11  Figure A ‐ 10.12  Figure A ‐ 10.13  Appendix A Experimental Program Test Results 182   11 TEST NUMBER 11 Test Description: Monotonic vertical uplift test with square weld detail. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 11 08/21/08 Square 106 N/A Monotonic N/A Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 106 125 N/A N/A N/A N/A N/A N/A Angle Failure Loading Protocol: Testing Comments: This test was performed on August 21, 2008 and was conducted in monotonic vertical uplift to determine the uplift envelope using the improved connection. The test was proceeding as expected until a weld failure between the square bar and the connection angle due to insufficient weld penetration into the angle which can be seen in Figure A - 11.7 due to the welding of this connector being done by a new lab technician who was unfamiliar with the welding equipment. However upon closer inspection it was also apparent that the angle was also beginning to tear, which is also seen in Figure A - 11.7. When the weld was repaired and testing continued, the previous load was not able to be achieved as the tear continued until angle failure at extremely Appendix A Experimental Program Test Results 183   small load levels. This means that the entire response plot was captured and the results found for ultimate load and displacement capacities are accurate. Appendix A Experimental Program Test Results 184   Vertical Response Plots: Figure A ‐ 11.1: The monotonic vertical response of specimen 11.  Figure A ‐ 11.2: Out‐of‐plane movement during monotonic vertical test number 11.  Appendix A Experimental Program Test Results 185   Photos: Figure A ‐ 11.3  Figure A ‐ 11.4  Figure A ‐ 11.5  Figure A ‐ 11.6  Figure A ‐ 11.7  Figure A ‐ 11.8  A ppendix A Experim ental Program Test R esults 186   12 TEST NUMBER 12 Test Description: Monotonic horizontal shear test with square weld detail. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 12 08/21/08 Square N/A 0 N/A Monotonic Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description N/A N/A 21 280 N/A N/A N/A N/A Rebar Fracture Loading Protocol: Testing Comments: This test was performed on August 21, 2008 and was conducted in monotonic horizontal shear to determine the response envelope using the improved connection detail. The test proceeded as expected with failure occurring due to rebar fracture (Figure A - 12.8) in exactly the same place as in test #5. It was observed that there was significantly less concrete spalling damage when compared to test #5, however in that test there was some initial uplift that occurred due to the loading beam being allowed to move vertically, which was restricted part was through the testing as it was creating a moment on the connection that was not desired. This uplift could have caused more concrete to spall off. It was also observed that the Δh at failure with the square connector was about 12 mm less than with the round connector. Appendix A Experimental Program Test Results 187   Horizontal Response Plots: Figure A ‐ 12.1: The monotonic horizontal response of specimen 12.  Figure A ‐ 12.2: Out‐of‐plane movement during monotonic horizontal test number 12.  Appendix A Experimental Program Test Results 188   Photos: Figure A ‐ 12.3  Figure A ‐ 12.4  Figure A ‐ 12.5  Figure A ‐ 12.6  Figure A ‐ 12.7  Figure A ‐ 12.8  A ppendix A Experim ental Program Test R esults 189   13 TEST NUMBER 13 Test Description: Cyclic uplift to Δv = 50 mm, held at Δv = 50 mm then cycled in horizontal shear. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 13 08/25/08 Square 51 50 Cyclic Cyclic Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 51 65 15 184 -16 -130 15 157 Rebar Fracture Loading Protocol: Testing Comments: This test was performed on August 25, 2008 and was used to determine the cyclic horizontal shear response after being cycled and held at Δv = 50 mm with a square connection detail. This test went according to plan and failure occurred due to fracture of the rebar (Figure A - 13.13) at the location of the bend due to bending (Figure A - 13.12) and straightening (Figure A - 13.11).This occurred after three complete positive cycles to Δh = 15 mm. This test produced an ultimate average load of 157 kN compared to 199 kN the same test with the round connection type produced. This value seems low, although there were no indications during testing which could account for this lower value. Appendix A Experimental Program Test Results 190   Vertical Response Plots: Figure A ‐ 13.1: The cyclic vertical response to 50 mm displacement of specimen 13.  Figure A ‐ 13.2: Out‐of‐plane movement during cyclic vertical test number 13.  Appendix A Experimental Program Test Results 191   Horizontal Response Plots: Figure A ‐ 13.3: The cyclic horizontal response of specimen 13.  Figure A ‐ 13.4: Out‐of‐plane movement during cyclic horizontal test number 13.  Appendix A Experimental Program Test Results 192   Photos: Figure A ‐ 13.5  Figure A ‐ 13.6  Figure A ‐ 13.7  Figure A ‐ 13.8  Figure A ‐ 13.9  Figure A ‐ 13.10  Figure A ‐ 13.11  Figure A ‐ 13.12  Figure A ‐ 13.13  Appendix A Experimental Program Test Results 193   14 TEST NUMBER 14 Test Description: Cyclic uplift to Δv = 50 mm, held at Δv = 0 mm then cycled in horizontal shear. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 14 08/26/08 Square 50 0 Cyclic Cyclic Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 50 64 11 200 -15 -192 13 196 Rebar Fracture Loading Protocol: Testing Comments: This test was performed on August 26, 2008 and was used to determine the cyclic horizontal shear response after being cycled three time to Δv = 50 mm then held at Δv = 0 mm. This test went according to plan and failure occurred due to fracture of the rebar (Figure A - 14.12 & Figure A - 14.13) at the location of the bend due to bending (Figure A - 14.11) and straightening (Figure A - 14.12).This occurred during the third positive cycle to Δh = 11 mm and at a load of 196 kN. The displacement was significantly lower than what was expected since in previous tests the specimen that was held at a larger Δv failed at a lower Δh then when Δv was held at zero. Upon inspection of Figure A - 14.11 however it was observed that due to the limited vertical Appendix A Experimental Program Test Results 194   damage and that Δv was held to zero, the bending curvature in the rebar during cycling was higher than when more vertical damage was present which may have caused fatigue failure to occur sooner in this case. Appendix A Experimental Program Test Results 195   Vertical Response Plots: Figure A ‐ 14.1: The cyclic vertical response to 50 mm displacement of specimen 14.  Figure A ‐ 14.2: Out‐of‐plane movement during cyclic vertical test number 14.  Appendix A Experimental Program Test Results 196   Horizontal Response Plots: Figure A ‐ 14.3: The cyclic horizontal response of specimen 14.  Figure A ‐ 14.4: Out‐of‐plane movement during cyclic horizontal test number 14.  Appendix A Experimental Program Test Results 197   Photos: Figure A ‐ 14.5  Figure A ‐ 14.6  Figure A ‐ 14.7  Figure A ‐ 14.8  Figure A ‐ 14.9  Figure A ‐ 14.10  Figure A ‐ 14.11  Figure A ‐ 14.12  Figure A ‐ 14.13  Appendix A Experimental Program Test Results 198   15 TEST NUMBER 15 Test Description: Cyclic uplift to Δv = 25 mm, held at Δv = 0 mm then cycled in horizontal shear. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 15 08/27/08 Square 26 0 Cyclic Cyclic Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 26 48 13 239 -16 -191 14 215 Rebar Fracture Loading Protocol: Testing Comments: This test was performed on August 27, 2008 and was used to determine the cyclic horizontal shear response after being cycled three time to Δv = 25 mm then held at Δv = 0 mm. This test went according to plan and failure occurred due to fracture of the rebar (Figure A - 15.12 & Figure A - 15.13) at the location of the bend due to bending (Figure A - 15.11) and straightening (Figure A - 15.10).This occurred during the second positive cycle to Δh = 13 mm. Upon inspection of Figure A - 15.11 it was again observed that due to the limited vertical damage and that Δv was held to zero, the bending curvature in the rebar during cycling was higher than when more vertical damage was present which may have caused fatigue failure to occur sooner in this Appendix A Experimental Program Test Results 199   case. It was also noted during this test that slab movements (approximately ±5 mm) occurred. When the grout was used to secure the slab to the strong floor, small amounts of sand was poured into the grouting holes first to ensure the liquid grout did not seep out underneath the slab. In this test too much sand was placed in the holes which allowed small rotations of the grouted pins when loading occurred. Despite this, the relative movement between the slab and the loading beam was always measured, and so this did not have an effect on the results of this test. Appendix A Experimental Program Test Results 200   Vertical Response Plots: Figure A ‐ 15.1: The cyclic vertical response to 25 mm displacement of specimen 15.  Figure A ‐ 15.2: Out‐of‐plane movement during cyclic vertical test number 15.  Appendix A Experimental Program Test Results 201   Horizontal Response Plots: Figure A ‐ 15.3: The cyclic horizontal response of specimen 15.  Figure A ‐ 15.4: Out‐of‐plane movement during cyclic horizontal test number 15.  Appendix A Experimental Program Test Results 202   Photos: Figure A ‐ 15.5  Figure A ‐ 15.6  Figure A ‐ 15.7  Figure A ‐ 15.8  Figure A ‐ 15.9  Figure A ‐ 15.10  Figure A ‐ 15.11  Figure A ‐ 15.12  Figure A ‐ 15.13  Appendix A Experimental Program Test Results 203   TEST NUMBER 16 Test Description: Cyclic uplift to Δv = 25 mm, held at Δv = 25 mm then cycled in horizontal shear. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 16 09/03/08 Square 25 25 Cyclic Cyclic Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 25 70 14 166 -12 -249 13 207 Rebar Fracture Loading Protocol: Testing Comments: This test was performed on September 3, 2008 and was used to determine the cyclic horizontal shear response after being cycled and held at Δv = 25 mm with a square connection detail. This test went according to plan and failure occurred due to fracture of the rebar (Figure A - 15.25) at the location of the bend due to bending (Figure A - 15.23) and straightening (Figure A - 15.24).This occurred on the first negative cycle to Δh = -12 mm. This test produced a maximum positive load of 166 kN and a maximum negative load of -249 kN. It is unclear why there was such an uneven skew in the maximum loads; however it could be explained by the uneven vertical damage created on the two sides as shown clearly in Figure A - 15.20. This uneven damage was due to the fact that Δv = 25 mm is a relatively small value, and the extent of the Appendix A Experimental Program Test Results 204   vertical damage was found to vary greatly at this displacement level. This being said, the average maximum load of 207 kN was exactly the value that was expected based on the trends from the other tests conducted. Appendix A Experimental Program Test Results 205   Vertical Response Plots: Figure A ‐ 15.14: The cyclic vertical response to 25 mm displacement of specimen 16.  Figure A ‐ 15.15: Out‐of‐plane movement during cyclic vertical test number 16.  Appendix A Experimental Program Test Results 206   Horizontal Response Plots: Figure A ‐ 15.16: The cyclic horizontal response of specimen 16.  Figure A ‐ 15.17: Out‐of‐plane movement during cyclic horizontal test number 16.  Appendix A Experimental Program Test Results 207   Photos: Figure A ‐ 15.18  Figure A ‐ 15.19  Figure A ‐ 15.20  Figure A ‐ 15.21  Figure A ‐ 15.22  Figure A ‐ 15.23  Figure A ‐ 15.24  Figure A ‐ 15.25    Appendix A Experimental Program Test Results 208   TEST NUMBER 17 Test Description: Cyclic uplift to failure while restricting large out of plane movements. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 17 10/27/08 Square 104 N/A Cyclic N/A Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 104 115 N/A N/A N/A N/A N/A N/A Rebar Fracture Loading Protocol: Testing Comments: This test was performed on October 27, 2008 and was used to determine the cyclic vertical shear response of the connector with a square connection detail when out of plane movements were restricted. This test went according to plan and failure occurred due to fracture of the rebar (Figure A - 15.31) at the location of the bend due to bending and straightening. This occurred on the first positive cycle past Δv = 100 mm at 104 mm. This test produced a maximum uplift resistance of 115 kN. As these results are in good agreement with the previous cyclic uplift test where out of plane movements were allowed, it can be concluded that out of plane resistance at the base of the wall panel does not influence the vertical response. Appendix A Experimental Program Test Results 209   Vertical Response Plots: Figure A ‐ 15.26: The cyclic vertical response of specimen 17.  * Out of plane response not available. Appendix A Experimental Program Test Results 210   Photos: Figure A ‐ 15.27  Figure A ‐ 15.28  Figure A ‐ 15.29  Figure A ‐ 15.30  Figure A ‐ 15.31  Figure A ‐ 15.32  A ppendix A Experim ental Program Test R esults 211   16 TEST NUMBER 18 Test Description: Cyclic uplift to Δv = 75 mm, held at Δv = 0 mm then cycled in horizontal shear while restricting large out of plane movements. Test Properties: Name Date Tested (mm/dd/yy) Weld Detail (Round/Square) ∆vMax [mm] ∆v During ∆h [mm] ∆v (Mono/Cyclic) ∆h (Mono/Cyclic) 18 10/27/08 Square 75 0 Cyclic Cyclic Summary of Results: Uplift Shear Horizontal Shear Positive Direction Negative Direction Average ∆vMax [mm] PMax [kN] ∆hMax+ [mm] VMax+ [kN] ∆hMax- [mm] VMax- [kN] ∆hMaxAve [mm] VMaxAve [kN] Failure Description 75 77 20 197 -20 -171 20 184 Rebar Fracture Loading Protocol: Testing Comments: This test was performed on October 27, 2008 and was conducted to determine the cyclic horizontal shear strength when the connector was cycled to Δv = 75 mm and held at Δv = 0 mm when out of plane movements were restricted. This test went according to plan and failure occurred due to fracture of the rebar (Figure A - 16.11) at the location of the bend due to bending and straightening. This occurred on the second negative cycle to Δh = -20 mm. This test produced a maximum positive load of 197 kN and a maximum negative load of -171 kN. As these results are in good agreement with the previous cyclic uplift test where out of plane movements were allowed, it can be concluded that out of plane resistance at the base of the wall panel does not influence the vertical response. Appendix A Experimental Program Test Results 212   Vertical Response Plots: Figure A ‐ 16.1: The cyclic vertical response to 75 mm displacement of specimen 18.  Figure A ‐ 16.2: Out‐of‐plane movement during cyclic vertical test number 18.  Appendix A Experimental Program Test Results 213   Horizontal Response Plots: Figure A ‐ 16.3: The cyclic horizontal response of specimen 18.  Figure A ‐ 16.4: Out‐of‐plane movement during cyclic horizontal test number 18.  Appendix A Experimental Program Test Results 214   Photos: Figure A ‐ 16.5  Figure A ‐ 16.6  Figure A ‐ 16.7  Figure A ‐ 16.8  Figure A ‐ 16.9  Figure A ‐ 16.10  Figure A ‐ 16.11  Figure A ‐ 16.12    Appendix A Experimental Program Test Results 215 APPENDIX B: PUSHOVER ANALYSIS RESULTS 216   1 PANEL, 0 PANEL – PANEL CONNECTORS, 2 PANEL – SLAB CONNECTORS (1,0,2)     Figure B ‐ 1,0,2 ‐ 1: System pushover response of the 1,0,2 system.    Appendix B Pushover Analysis Results 217     Panel 1  Connector 1  Connector 2  Figure B ‐ 1,0,2 ‐ 2: Panel 1 connector uplift demands during system pushover.    Normalized Response Plots  Figure B ‐ 1,0,2 ‐ 3: System pushover response of the 1,0,2 system normalized to maximum achieved  capacity.  Figure B ‐ 1,0,2 ‐ 4: System pushover response of the 1,0,2 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 218     1 PANEL, 0 PANEL – PANEL CONNECTORS, 3 PANEL – SLAB CONNECTORS (1,0,3)       Figure B ‐ 1,0,3 ‐ 1: System pushover response of the 1,0,3 system.    Appendix B Pushover Analysis Results 219     Panel 1  Connector 1  Connector 2  Connector 3    Figure B ‐ 1,0,3 ‐ 2: Panel 1 connector uplift demands during system pushover.    Normalized Response Plots  Figure B ‐ 1,0,3 ‐ 3: System pushover response of the 1,0,3 system normalized to maximum achieved  capacity.  Figure B ‐ 1,0,3 ‐ 4: System pushover response of the 1,0,3 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 220     1 PANEL, 0 PANEL – PANEL CONNECTORS, 4 PANEL – SLAB CONNECTORS (1,0,4)       Figure B ‐ 1,0,4 ‐ 1: System pushover response of the 1,0,4 system.    Appendix B Pushover Analysis Results 221     Panel 1  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 1,0,4 ‐ 2: Panel 1 connector uplift demands during system pushover.    Normalized Response Plots  Figure B ‐ 1,0,4 ‐ 3: System pushover response of the 1,0,4 system normalized to maximum achieved  capacity.  Figure B ‐ 1,0,4 ‐ 4: System pushover response of the 1,0,4 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 222     2 PANEL, 2 PANEL – PANEL CONNECTORS, 2 PANEL – SLAB CONNECTORS (2,2,2)       Figure B ‐ 2,2,2 ‐ 1: System pushover response of the 2,2,2 system.  Appendix B Pushover Analysis Results 223     Panel 1  Connector 1  Connector 2  Figure B ‐ 2,2,2 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 2,2,2 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Figure B ‐ 2,2,2 ‐ 4: Panel 2 connector uplift demands during system pushover.    Appendix B Pushover Analysis Results 224     Normalized Response Plots    Figure B ‐ 2,2,2 ‐ 5: System pushover response of the 2,2,2 system normalized to maximum achieved  capacity.    Figure B ‐ 2,2,2 ‐ 6: System pushover response of the 2,2,2 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 225     2 PANEL, 2 PANEL – PANEL CONNECTORS, 3 PANEL – SLAB CONNECTORS (2,2,3)       Figure B ‐ 2,2,3 ‐ 1: System pushover response of the 2,2,3 system.  Appendix B Pushover Analysis Results 226     Panel 1  Connector 1  Connector 2  Connector 3  Figure B ‐ 2,2,3 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 2,2,3 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3    Figure B ‐ 2,2,3 ‐ 4: Panel 2 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 227     Normalized Response Plots    Figure B ‐ 2,2,3 ‐ 5: System pushover response of the 2,2,3 system normalized to maximum achieved  capacity.    Figure B ‐ 2,2,3 ‐ 6: System pushover response of the 2,2,3 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 228     2 PANEL, 2 PANEL – PANEL CONNECTORS, 4 PANEL – SLAB CONNECTORS (2,2,4)       Figure B ‐ 2,2,4 ‐ 1: System pushover response of the 2,2,4 system.  Appendix B Pushover Analysis Results 229     Panel 1  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 2,2,4 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 2,2,4 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 2,2,4 ‐ 4: Panel 2 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 230     Normalized Response Plots    Figure B ‐ 2,2,4 ‐ 5: System pushover response of the 2,2,4 system normalized to maximum achieved  capacity.    Figure B ‐ 2,2,4 ‐ 6: System pushover response of the 2,2,4 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 231     2 PANEL, 3 PANEL – PANEL CONNECTORS, 2 PANEL – SLAB CONNECTORS (2,3,2)       Figure B ‐ 2,3,2 ‐ 1: System pushover response of the 2,3,2 system.  Appendix B Pushover Analysis Results 232     Panel 1  Connector 1  Connector 2  Figure B ‐ 2,3,2 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 2,3,2 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Figure B ‐ 2,3,2 ‐ 4: Panel 2 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 233     Normalized Response Plots    Figure B ‐ 2,3,2 ‐ 5: System pushover response of the 2,3,2 system normalized to maximum achieved  capacity.    Figure B ‐ 2,3,2 ‐ 6: System pushover response of the 2,3,2 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 234     2 PANEL, 3 PANEL – PANEL CONNECTORS, 3 PANEL – SLAB CONNECTORS (2,3,3)       Figure B ‐ 2,3,3 ‐ 1: System pushover response of the 2,3,3 system.  Appendix B Pushover Analysis Results 235     Panel 1  Connector 1  Connector 2  Connector 3  Figure B ‐ 2,3,3 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 2,3,3 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3  Figure B ‐ 2,3,3 ‐ 4: Panel 2 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 236     Normalized Response Plots    Figure B ‐ 2,3,3 ‐ 5: System pushover response of the 2,3,3 system normalized to maximum achieved  capacity.    Figure B ‐ 2,3,3 ‐ 6: System pushover response of the 2,3,3 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 237     2 PANEL, 3 PANEL – PANEL CONNECTORS, 4 PANEL – SLAB CONNECTORS (2,3,4)       Figure B ‐ 2,3,4 ‐ 1: System pushover response of the 2,3,4 system.  Appendix B Pushover Analysis Results 238     Panel 1  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 2,3,4 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 2,3,4 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 2,3,4 ‐ 4: Panel 2 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 239     Normalized Response Plots    Figure B ‐ 2,3,4 ‐ 5: System pushover response of the 2,3,4 system normalized to maximum achieved  capacity.    Figure B ‐ 2,3,4 ‐ 6: System pushover response of the 2,3,4 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 240     2 PANEL, 4 PANEL – PANEL CONNECTORS, 2 PANEL – SLAB CONNECTORS (2,4,2)   Figure B ‐ 2,4,2 ‐ 1: System pushover response of the 2,4,2 system.  Appendix B Pushover Analysis Results 241     Panel 1  Connector 1  Connector 2  Figure B ‐ 2,4,2 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 2,4,2 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Figure B ‐ 2,4,2 ‐ 4: Panel 2 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 242     Normalized Response Plots    Figure B ‐ 2,4,2 ‐ 5: System pushover response of the 2,4,2 system normalized to maximum achieved  capacity.    Figure B ‐ 2,4,2 ‐ 6: System pushover response of the 2,4,2 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 243     2 PANEL, 4 PANEL – PANEL CONNECTORS, 3 PANEL – SLAB CONNECTORS (2,4,3)   Figure B ‐ 2,4,3 ‐ 1: System pushover response of the 2,4,3 system.  Appendix B Pushover Analysis Results 244     Panel 1  Connector 1  Connector 2  Connector 3    Figure B ‐ 2,4,3 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 2,4,3 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3    Figure B ‐ 2,4,3 ‐ 4: Panel 2 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 245     Normalized Response Plots    Figure B ‐ 2,4,3 ‐ 5: System pushover response of the 2,4,3 system normalized to maximum achieved  capacity.    Figure B ‐ 2,4,3 ‐ 6: System pushover response of the 2,4,3 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 246     2 PANEL, 4 PANEL – PANEL CONNECTORS, 4 PANEL – SLAB CONNECTORS (2,4,4)   Figure B ‐ 2,4,4 ‐ 1: System pushover response of the 2,4,4 system.  Appendix B Pushover Analysis Results 247     Panel 1  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 2,4,4 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 2,4,4 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 2,4,4 ‐ 4: Panel 2 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 248     Normalized Response Plots    Figure B ‐ 2,4,4 ‐ 5: System pushover response of the 2,4,4 system normalized to maximum achieved  capacity.    Figure B ‐ 2,4,4 ‐ 6: System pushover response of the 2,4,4 system normalized to the residual capacity.    Appendix B Pushover Analysis Results 249     3 PANEL, 2 PANEL – PANEL CONNECTORS, 2 PANEL – SLAB CONNECTORS (3,2,2)   Figure B ‐ 3,2,2 ‐ 1: System pushover response of the 3,2,2 system.  Appendix B Pushover Analysis Results 250     Panel 1  Connector 1  Connector 2  Figure B ‐ 3,2,2 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 3,2,2 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Figure B ‐ 3,2,2 ‐ 4: Panel 2 connector uplift demands during system pushover.    Appendix B Pushover Analysis Results 251     Panel 2 – Panel 3 Interface  Figure B ‐ 3,2,2 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.    Panel 3  Connector 1  Connector 2  Figure B ‐ 3,2,2 ‐ 6: Panel 3 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 252     Normalized Response Plots    Figure B ‐ 3,2,2 ‐ 7: System pushover response of the 3,2,2 system normalized to maximum achieved  capacity.    Figure B ‐ 3,2,2 ‐ 8: System pushover response of the 3,2,2 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 253     3 PANEL, 2 PANEL – PANEL CONNECTORS, 3 PANEL – SLAB CONNECTORS (3,2,3)   Figure B ‐ 3,2,3 ‐ 1: System pushover response of the 3,2,3 system.  Appendix B Pushover Analysis Results 254     Panel 1  Connector 1  Connector 2  Connector 3    Figure B ‐ 3,2,3 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 3,2,3 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3    Figure B ‐ 3,2,3 ‐ 4: Panel 2 connector uplift demands during system pushover.    Appendix B Pushover Analysis Results 255     Panel 2 – Panel 3 Interface  Figure B ‐ 3,2,3 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.    Panel 3  Connector 1  Connector 2  Connector 3    Figure B ‐ 3,2,3 ‐ 6: Panel 3 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 256     Normalized Response Plots    Figure B ‐ 3,2,3 ‐ 7: System pushover response of the 3,2,3 system normalized to maximum achieved  capacity.    Figure B ‐ 3,2,3 ‐ 8: System pushover response of the 3,2,3 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 257     3 PANEL, 2 PANEL – PANEL CONNECTORS, 4 PANEL – SLAB CONNECTORS (3,2,4)   Figure B ‐ 3,2,4 ‐ 1: System pushover response of the 3,2,4 system.  Appendix B Pushover Analysis Results 258     Panel 1  Connector 1  Connector 2  Connector 3  Connector 4    Figure B ‐ 3,2,4 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 3,2,4 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3  Connector 4    Figure B ‐ 3,2,4 ‐ 4: Panel 2 connector uplift demands during system pushover.    Appendix B Pushover Analysis Results 259     Panel 2 – Panel 3 Interface  Figure B ‐ 3,2,4 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.    Panel 3  Connector 1  Connector 2  Connector 3  Connector 4    Figure B ‐ 3,2,4 ‐ 6: Panel 3 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 260     Normalized Response Plots    Figure B ‐ 3,2,4 ‐ 7: System pushover response of the 3,2,4 system normalized to maximum achieved  capacity.    Figure B ‐ 3,2,4 ‐ 8: System pushover response of the 3,2,4 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 261     3 PANEL, 3 PANEL – PANEL CONNECTORS, 2 PANEL – SLAB CONNECTORS (3,3,2)   Figure B ‐ 3,3,2 ‐ 1: System pushover response of the 3,3,2 system.  Appendix B Pushover Analysis Results 262     Panel 1  Connector 1  Connector 2  Figure B ‐ 3,3,2 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 3,3,2 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Figure B ‐ 3,3,2 ‐ 4: Panel 2 connector uplift demands during system pushover.    Appendix B Pushover Analysis Results 263     Panel 2 – Panel 3 Interface  Figure B ‐ 3,3,2 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.    Panel 3  Connector 1  Connector 2  Figure B ‐ 3,3,2 ‐ 6: Panel 3 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 264     Normalized Response Plots    Figure B ‐ 3,3,2 ‐ 7: System pushover response of the 3,3,2 system normalized to maximum achieved  capacity.    Figure B ‐ 3,3,2 ‐ 8: System pushover response of the 3,3,2 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 265     3 PANEL, 3 PANEL – PANEL CONNECTORS, 3 PANEL – SLAB CONNECTORS (3,3,3)   Figure B ‐ 3,3,3 ‐ 1: System pushover response of the 3,3,3 system.  Appendix B Pushover Analysis Results 266     Panel 1  Connector 1  Connector 2  Connector 3  Figure B ‐ 3,3,3 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 3,3,3 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3  Figure B ‐ 3,3,3 ‐ 4: Panel 2 connector uplift demands during system pushover.    Appendix B Pushover Analysis Results 267     Panel 2 – Panel 3 Interface  Figure B ‐ 3,3,3 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.    Panel 3  Connector 1  Connector 2  Connector 3  Figure B ‐ 3,3,3 ‐ 6: Panel 3 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 268     Normalized Response Plots    Figure B ‐ 3,3,3 ‐ 7: System pushover response of the 3,3,3 system normalized to maximum achieved  capacity.    Figure B ‐ 3,3,3 ‐ 8: System pushover response of the 3,3,3 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 269     3 PANEL, 3 PANEL – PANEL CONNECTORS, 4 PANEL – SLAB CONNECTORS (3,3,4)   Figure B ‐ 3,3,4 ‐ 1: System pushover response of the 3,3,4 system.  Appendix B Pushover Analysis Results 270     Panel 1  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 3,3,4 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 3,3,4 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3  Connector 4    Figure B ‐ 3,3,4 ‐ 4: Panel 2 connector uplift demands during system pushover.    Appendix B Pushover Analysis Results 271     Panel 2 – Panel 3 Interface  Figure B ‐ 3,3,4 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.    Panel 3  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 3,3,4 ‐ 6: Panel 3 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 272     Normalized Response Plots    Figure B ‐ 3,3,4 ‐ 7: System pushover response of the 3,3,4 system normalized to maximum achieved  capacity.    Figure B ‐ 3,3,4 ‐ 8: System pushover response of the 3,3,4 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 273     3 PANEL, 4 PANEL – PANEL CONNECTORS, 2 PANEL – SLAB CONNECTORS (3,4,2)   Figure B ‐ 4,2,2 ‐ 1: System pushover response of the 3,4,2 system.  Appendix B Pushover Analysis Results 274     Panel 1  Connector 1  Connector 2  Figure B ‐ 4,2,2 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 4,2,2 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Figure B ‐ 4,2,2 ‐ 4: Panel 2 connector uplift demands during system pushover.    Appendix B Pushover Analysis Results 275     Panel 2 – Panel 3 Interface  Figure B ‐ 4,2,2 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.    Panel 3  Connector 1  Connector 2  Figure B ‐ 4,2,2 ‐ 6: Panel 3 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 276     Normalized Response Plots    Figure B ‐ 4,2,2 ‐ 7: System pushover response of the 3,4,2 system normalized to maximum achieved  capacity.    Figure B ‐ 4,2,2 ‐ 8: System pushover response of the 3,4,2 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 277     3 PANEL, 4 PANEL – PANEL CONNECTORS, 3 PANEL – SLAB CONNECTORS (3,4,3)   Figure B ‐ 3,4,3 ‐ 1: System pushover response of the 3,4,3 system.  Appendix B Pushover Analysis Results 278     Panel 1  Connector 1  Connector 2  Connector 3    Figure B ‐ 3,4,3 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 3,4,3 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3    Figure B ‐ 3,4,3 ‐ 4: Panel 2 connector uplift demands during system pushover.    Appendix B Pushover Analysis Results 279     Panel 2 – Panel 3 Interface  Figure B ‐ 3,4,3 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.    Panel 3  Connector 1  Connector 2  Connector 3  Figure B ‐ 3,4,3 ‐ 6: Panel 3 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 280     Normalized Response Plots    Figure B ‐ 3,4,3 ‐ 7: System pushover response of the 3,4,3 system normalized to maximum achieved  capacity.    Figure B ‐ 3,4,3 ‐ 8: System pushover response of the 3,4,3 system normalized to the residual capacity.  Appendix B Pushover Analysis Results 281     3 PANEL, 4 PANEL – PANEL CONNECTORS, 4 PANEL – SLAB CONNECTORS (3,4,4)   Figure B ‐ 3,4,4 ‐ 1: System pushover response of the 3,4,4 system.  Appendix B Pushover Analysis Results 282     Panel 1  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 3,4,4 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 3,4,4 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 3,4,4 ‐ 4: Panel 2 connector uplift demands during system pushover.    Appendix B Pushover Analysis Results 283     Panel 2 – Panel 3 Interface  Figure B ‐ 3,4,4 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.    Panel 3  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 3,4,4 ‐ 6: Panel 3 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 284     Normalized Response Plots    Figure B ‐ 3,4,4 ‐ 7: System pushover response of the 3,4,4 system normalized to maximum achieved  capacity.    Figure B ‐ 3,4,4 ‐ 8: System pushover response of the 3,4,4 system normalized to the residual capacity.    Appendix B Pushover Analysis Results 285     4 PANEL, 2 PANEL – PANEL CONNECTORS, 2 PANEL – SLAB CONNECTORS (4,2,2)     Figure B ‐ 4,2,2 ‐ 9: System pushover response of the 4,2,2 system.  Appendix B Pushover Analysis Results 286     Panel 1  Connector 1  Connector 2  Figure B ‐ 4,2,2 ‐ 10: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 4,2,2 ‐ 11: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Figure B ‐ 4,2,2 ‐ 12: Panel 2 connector uplift demands during system pushover.    Panel 2 – Panel 3 Interface  Figure B ‐ 4,2,2 ‐ 13: Panel 2 – panel 3 interface connector shear demands during system pushover.  Appendix B Pushover Analysis Results 287     Panel 3  Connector 1  Connector 2  Figure B ‐ 4,2,2 ‐ 14: Panel 3 connector uplift demands during system pushover.    Panel 3 – Panel 4 Interface  Figure B ‐ 4,2,2 ‐ 15: Panel 3 – panel 4 interface connector shear demands during system pushover.    Panel 4  Connector 1  Connector 2  Figure B ‐ 4,2,2 ‐ 16: Panel 4 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 288     Normalized Response Plots    Figure B ‐ 4,2,2 ‐ 17: System pushover response of the 4,2,2 system normalized to maximum achieved  capacity.    Figure B ‐ 4,2,2 ‐ 18: System pushover response of the 4,2,2 system normalized to the residual  capacity.  Appendix B Pushover Analysis Results 289   4 PANEL, 2 PANEL – PANEL CONNECTORS, 3 PANEL – SLAB CONNECTORS (4,2,3)     Figure B ‐ 4,2,3 ‐ 1: System pushover response of the 4,2,3 system.    Appendix B Pushover Analysis Results 290     Panel 1  Connector 1  Connector 2  Connector 3  Figure B ‐ 4,2,3 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 4,2,3 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3  Figure B ‐ 4,2,3 ‐ 4: Panel 2 connector uplift demands during system pushover.    Panel 2 – Panel 3 Interface  Figure B ‐ 4,2,3 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.  Appendix B Pushover Analysis Results 291     Panel 3  Connector 1  Connector 2  Connector 3  Figure B ‐ 4,2,3 ‐ 6: Panel 3 connector uplift demands during system pushover.    Panel 3 – Panel 4 Interface  Figure B ‐ 4,2,3 ‐ 7: Panel 3 – panel 4 interface connector shear demands during system pushover.    Panel 4  Connector 1  Connector 2  Connector 3  Figure B ‐ 4,2,3 ‐ 8: Panel 4 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 292     Normalized Response Plots    Figure B ‐ 4,2,3 ‐ 9: System pushover response of the 4,2,3 system normalized to maximum achieved  capacity.    Figure B ‐ 4,2,3 ‐ 10: System pushover response of the 4,2,3 system normalized to the residual  capacity.  Appendix B Pushover Analysis Results 293   4 PANEL, 2 PANEL – PANEL CONNECTORS, 4 PANEL – SLAB CONNECTORS (4,2,4)   Figure B ‐ 4,2,4 ‐ 1: System pushover response of the 4,2,4 system.  Appendix B Pushover Analysis Results 294     Panel 1  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 4,2,4 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 4,2,4 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 4,2,4 ‐ 4: Panel 2 connector uplift demands during system pushover.    Panel 2 – Panel 3 Interface  Figure B ‐ 4,2,4 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.  Appendix B Pushover Analysis Results 295     Panel 3  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 4,2,4 ‐ 6: Panel 3 connector uplift demands during system pushover.    Panel 3 – Panel 4 Interface  Figure B ‐ 4,2,4 ‐ 7: Panel 3 – panel 4 interface connector shear demands during system pushover.    Panel 4  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 4,2,4 ‐ 8: Panel 4 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 296     Normalized Response Plots    Figure B ‐ 4,2,4 ‐ 9: System pushover response of the 4,2,4 system normalized to maximum achieved  capacity.    Figure B ‐ 4,2,4 ‐ 10: System pushover response of the 4,2,4 system normalized to the residual  capacity.  Appendix B Pushover Analysis Results 297   4 PANEL, 3 PANEL – PANEL CONNECTORS, 2 PANEL – SLAB CONNECTORS (4,3,2)   Figure B ‐ 4,3,2 ‐ 1: System pushover response of the 4,3,2 system.  Appendix B Pushover Analysis Results 298     Panel 1  Connector 1  Connector 2  Figure B ‐ 4,3,2 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 4,3,2 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Figure B ‐ 4,3,2 ‐ 4: Panel 2 connector uplift demands during system pushover.    Panel 2 – Panel 3 Interface  Figure B ‐ 4,3,2 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.  Appendix B Pushover Analysis Results 299     Panel 3  Connector 1  Connector 2  Figure B ‐ 4,3,2 ‐ 6: Panel 3 connector uplift demands during system pushover.    Panel 3 – Panel 4 Interface  Figure B ‐ 4,3,2 ‐ 7: Panel 3 – panel 4 interface connector shear demands during system pushover.    Panel 4  Connector 1  Connector 2  Figure B ‐ 4,3,2 ‐ 8: Panel 4 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 300     Normalized Response Plots    Figure B ‐ 4,3,2 ‐ 9: System pushover response of the 4,3,2 system normalized to maximum achieved  capacity.    Figure B ‐ 4,3,2 ‐ 10: System pushover response of the 4,3,2 system normalized to the residual  capacity.  Appendix B Pushover Analysis Results 301   4 PANEL, 3 PANEL – PANEL CONNECTORS, 3 PANEL – SLAB CONNECTORS (4,3,3)   Figure B ‐ 4,3,3 ‐ 1: System pushover response of the 4,3,3 system.  Appendix B Pushover Analysis Results 302     Panel 1  Connector 1  Connector 2  Connector 3  Figure B ‐ 4,3,3 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 4,3,3 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3  Figure B ‐ 4,3,3 ‐ 4: Panel 2 connector uplift demands during system pushover.    Panel 2 – Panel 3 Interface  Figure B ‐ 4,3,3 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.  Appendix B Pushover Analysis Results 303     Panel 3  Connector 1  Connector 2  Connector 3  Figure B ‐ 4,3,3 ‐ 6: Panel 3 connector uplift demands during system pushover.    Panel 3 – Panel 4 Interface  Figure B ‐ 4,3,3 ‐ 7: Panel 3 – panel 4 interface connector shear demands during system pushover.    Panel 4  Connector 1  Connector 2  Connector 3  Figure B ‐ 4,3,3 ‐ 8: Panel 4 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 304     Normalized Response Plots    Figure B ‐ 4,3,3 ‐ 9: System pushover response of the 4,3,3 system normalized to maximum achieved  capacity.    Figure B ‐ 4,3,3 ‐ 10: System pushover response of the 4,3,3 system normalized to the residual  capacity.  Appendix B Pushover Analysis Results 305   4 PANEL, 3 PANEL – PANEL CONNECTORS, 4 PANEL – SLAB CONNECTORS (4,3,4)   Figure B ‐ 4,3,4 ‐ 1: System pushover response of the 4,3,4 system.  Appendix B Pushover Analysis Results 306     Panel 1  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 4,3,4 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 4,3,4 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 4,3,4 ‐ 4: Panel 2 connector uplift demands during system pushover.    Panel 2 – Panel 3 Interface  Figure B ‐ 4,3,4 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.  Appendix B Pushover Analysis Results 307     Panel 3  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 4,3,4 ‐ 6: Panel 3 connector uplift demands during system pushover.    Panel 3 – Panel 4 Interface  Figure B ‐ 4,3,4 ‐ 7: Panel 3 – panel 4 interface connector shear demands during system pushover.    Panel 4  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 4,3,4 ‐ 8: Panel 4 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 308     Normalized Response Plots    Figure B ‐ 4,3,4 ‐ 9: System pushover response of the 4,3,4 system normalized to maximum achieved  capacity.    Figure B ‐ 4,3,4 ‐ 10: System pushover response of the 4,3,4 system normalized to the residual  capacity.  Appendix B Pushover Analysis Results 309   4 PANEL, 4 PANEL – PANEL CONNECTORS, 2 PANEL – SLAB CONNECTORS (4,4,2)   Figure B ‐ 4,4,2 ‐ 1: System pushover response of the 4,4,2 system.  Appendix B Pushover Analysis Results 310     Panel 1  Connector 1  Connector 2  Figure B ‐ 4,4,2 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 4,4,2 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Figure B ‐ 4,4,2 ‐ 4: Panel 2 connector uplift demands during system pushover.    Panel 2 – Panel 3 Interface  Figure B ‐ 4,4,2 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.  Appendix B Pushover Analysis Results 311     Panel 3  Connector 1  Connector 2  Figure B ‐ 4,4,2 ‐ 6: Panel 3 connector uplift demands during system pushover.    Panel 3 – Panel 4 Interface  Figure B ‐ 4,4,2 ‐ 7: Panel 3 – panel 4 interface connector shear demands during system pushover.    Panel 4  Connector 1  Connector 2  Figure B ‐ 4,4,2 ‐ 8: Panel 4 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 312     Normalized Response Plots    Figure B ‐ 4,4,2 ‐ 9: System pushover response of the 4,4,2 system normalized to maximum achieved  capacity.    Figure B ‐ 4,4,2 ‐ 10: System pushover response of the 4,4,2 system normalized to the residual  capacity.  Appendix B Pushover Analysis Results 313   4 PANEL, 4 PANEL – PANEL CONNECTORS, 3 PANEL – SLAB CONNECTORS (4,4,3)   Figure B ‐ 4,4,3 ‐ 1: System pushover response of the 4,4,3 system.  Appendix B Pushover Analysis Results 314     Panel 1  Connector 1  Connector 2  Connector 3  Figure B ‐ 4,4,3 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 4,4,3 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3  Figure B ‐ 4,4,3 ‐ 4: Panel 2 connector uplift demands during system pushover.    Panel 2 – Panel 3 Interface  Figure B ‐ 4,4,3 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.  Appendix B Pushover Analysis Results 315     Panel 3  Connector 1  Connector 2  Connector 3  Figure B ‐ 4,4,3 ‐ 6: Panel 3 connector uplift demands during system pushover.    Panel 3 – Panel 4 Interface  Figure B ‐ 4,4,3 ‐ 7: Panel 3 – panel 4 interface connector shear demands during system pushover.    Panel 4  Connector 1  Connector 2  Connector 3  Figure B ‐ 4,4,3 ‐ 8: Panel 4 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 316     Normalized Response Plots    Figure B ‐ 4,4,3 ‐ 9: System pushover response of the 4,4,3 system normalized to maximum achieved  capacity.    Figure B ‐ 4,4,3 ‐ 10: System pushover response of the 4,4,3 system normalized to the residual  capacity.  Appendix B Pushover Analysis Results 317   4 PANEL, 4 PANEL – PANEL CONNECTORS, 4 PANEL – SLAB CONNECTORS (4,4,4)   Figure B ‐ 4,4,4 ‐ 1: System pushover response of the 4,4,4 system.  Appendix B Pushover Analysis Results 318     Panel 1  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 4,4,4 ‐ 2: Panel 1 connector uplift demands during system pushover.    Panel 1 – Panel 2 Interface  Figure B ‐ 4,4,4 ‐ 3: Panel 1 – panel 2 interface connector shear demands during system pushover.    Panel 2  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 4,4,4 ‐ 4: Panel 2 connector uplift demands during system pushover.    Panel 2 – Panel 3 Interface  Figure B ‐ 4,4,4 ‐ 5: Panel 2 – panel 3 interface connector shear demands during system pushover.  Appendix B Pushover Analysis Results 319     Panel 3  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 4,4,4 ‐ 6: Panel 3 connector uplift demands during system pushover.    Panel 3 – Panel 4 Interface  Figure B ‐ 4,4,4 ‐ 7: Panel 3 – panel 4 interface connector shear demands during system pushover.    Panel 4  Connector 1  Connector 2  Connector 3  Connector 4  Figure B ‐ 4,4,4 ‐ 8: Panel 4 connector uplift demands during system pushover.  Appendix B Pushover Analysis Results 320     Normalized Response Plots    Figure B ‐ 4,4,4 ‐ 9: System pushover response of the 4,4,4 system normalized to maximum achieved  capacity.    Figure B ‐ 4,4,4 ‐ 10: System pushover response of the 4,4,4 system normalized to the residual  capacity.    Appendix B Pushover Analysis Results 321"@en ; edm:hasType "Thesis/Dissertation"@en ; vivo:dateIssued "2009-11"@en ; edm:isShownAt "10.14288/1.0063140"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Civil Engineering"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "Attribution-NonCommercial-NoDerivatives 4.0 International"@en ; ns0:rightsURI "http://creativecommons.org/licenses/by-nc-nd/4.0/"@en ; ns0:scholarLevel "Graduate"@en ; dcterms:title "Investigation of the in plane seismic response of solid tilt-up wall systems"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/12626"@en .