Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An experimental investigation of the behaviour of connections in thin precast concrete panels under earthquake… Kallros, Mikael Kaj 1987

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

Item Metadata

Download

Media
831-UBC_1987_A7 K34.pdf [ 18.1MB ]
Metadata
JSON: 831-1.0062538.json
JSON-LD: 831-1.0062538-ld.json
RDF/XML (Pretty): 831-1.0062538-rdf.xml
RDF/JSON: 831-1.0062538-rdf.json
Turtle: 831-1.0062538-turtle.txt
N-Triples: 831-1.0062538-rdf-ntriples.txt
Original Record: 831-1.0062538-source.json
Full Text
831-1.0062538-fulltext.txt
Citation
831-1.0062538.ris

Full Text

AN EXPERIMENTAL INVESTIGATION OF THE BEHAVIOUR OF CONNECTIONS IN THIN PRECAST CONCRETE PANELS UNDER EARTHQUAKE LOADING By MIKAEL KAJ KALLROS The University of B r i t i s h Columbia, 1985 SUBMITTED IN PARTIAL FULFILLMENT OF REQUIREMENTS FOR THE DEGREE OF MASTERS OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES CIVIL ENGINEERING DEPARTMENT We accept t h i s thesis as conforming to the required standard B.A.Sc.(Civil), A THESIS THE THE UNIVERSITY OF BRITISH COLUMBIA A p r i l 1987 © Mikael Kaj Kallros, 1987 In presenting this thesis in a p a r t i a l fulfilment of the requirements for an advanced degree at the University of Br i t i s h Columbia, I agree that the Library s h a l l make i t freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his or her representatives. It i s understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of CIVIL ENGINEERING The University of B r i t i s h Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date : April 15, 1987 i i ABSTRACT Investigations of connections for precast concrete panel buildings have shown that i t i s d i f f i c u l t to design an embedded connection that w i l l perform well under earthquake loading. Some t y p i c a l connections use studs or reinforcing bars embedded in the edge of the panel. These are then welded or bolted to an adjacent panel. During earthquake loading the crushing of concrete around the embedment usually leads to premature loss of strength and s t i f f n e s s of the connection before s i g n i f i c a n t d u c t i l i t y can develop. It has been found that connection performance improves with increasing panel thickness. The behaviour of embedded connections in thin precast concrete panels was investigated. The intent was to improve connection design d e t a i l s and to develop a simple method of predicting connection strengths with panel thicknesses of 50 mm to 75 mm. Sixteen connections of six d i f f e r e n t types were tested. Three were tested monotonically and thirteen were tested under reversed c y c l i c loading. Certain types of connections can be used to transfer earthquake loads between thin concrete panels as long as they have adequate strength. Methods for predicting the strength of connections are discussed. The connections tested should not be r e l i e d on to develop d u c t i l i t y . i i i TABLE OF CONTENTS PAGE # ABSTRACT i i TABLE OF CONTENTS i i i LIST OF TABLES v i i i LIST OF FIGURES ix LIST OF PHOTOGRAPHS x i i i ACKNOWLEDGEMENTS xvi CHAPTER 1 - INTRODUCTION 1 1.1. GENERAL 1 1.2. OBJECTIVES AND SCOPE 2 CHAPTER 2 - REVIEW OF CONNECTION DESIGNS FOR PRECAST CONCRETE 4 2.1. INTRODUCTION 4 2.2. CONNECTION BEHAVIOUR DURING EARTHQUAKES 4 2.3. TYPES OF CONNECTIONS 6 2.3.1. CAST IN PLACE CONNECTIONS 6 2.3.2. DRY CONNECTIONS 8 2.3.2.1. MECHANICAL CONNECTORS 10 2.3.2.2. WELDED CONNECTIONS 10 2.3.2.2.1. EMBEDDED CONNECTIONS WITH RIGID WELD PLATES 12 2.3.2.2.2. EMBEDDED REBAR CONNECTION 15 2.3.2.2.3. EMBEDDED CONNECTIONS WITH DUCTILE WELD PLATE 24 2.4. LOCATION OF CONNECTIONS 26 iv 2.5. CONCLUDING REMARKS 26 CHAPTER 3 - EXPERIMENTAL MATERIALS AND SPECIMEN PREPARATION 27 3.1. INTRODUCTION 27 3.2. FORM 29 3.3. CONCRETE 31 3.3.1. CONCRETE MIXING AND CASTING PROCEDURE 31 3.3.2. TEST RESULTS OF CONCRETE MIXES 33 3.4. STEEL 37 3.4.1. REINFORCING MESH 37 3.4.2. SPECIMEN SUPPORT PLATE 37 3.4.3. CONNECTION STEEL 39 3.5. TYPES OF CONNECTIONS 45 3.5.1. REBAR #1 BENT TO 45 DEGREES. 45 3.5.2. REBAR #1 BENT TO 45 DEGREES AND WELDED TO AN ANGLE 47 3.5.5. ANGLE WELDED TO REINFORCING MESH 51 3.5.6. REVERSED ANGLE WELDED TO REINFORCING MESH 51 CHAPTER 4 - LABORATORY TEST DETAILS 55 4.1. INTRODUCTION 55 4.2. ASSUMPTIONS MADE FOR THE TESTS 58 4.3. TEST RIG 58 4.4. LOADING AND DATA ACQUISITION SYSTEM 63 4.5. USE OF THE DATA ACQUISITION SYSTEM 67 4.5.1. OPERATION OF MTS SERVO-CONTROLLER 67 4.5.2. OPERATION OF NEFF 620 67 V 4.5.3. OPERATION OF PDP-11/10 67 4.6. TEST MEASUREMENTS 68 4.6.1. LOAD MEASUREMENT 68 4.6.2. DISPLACEMENT MEASUREMENT 68 4.6.2.1. CONNECTION DISPLACEMENT 70 4.6.2.2. SIDEWAYS MOVEMENT 73 4.6.2.3. SPECIMEN ROTATION 73 4.6.3.4. MOVEMENT OF HYDRAULIC JACK 77 4.7. LOADING PROCEDURE 77 CHAPTER 5 - EXPERIMENTAL RESULTS 79 5.1. INTRODUCTION 79 5.2. DETAILS OF CONNECTIONS TESTED 80 5.2.1. REBAR #1 AT 45 DEGREES 80 5.2.1.1. TEST #1 81 5.2.1.2. TEST #2 87 5.2.1.3. TEST #6 90 5.2.2. REBAR #1 AT 45 DEGREES WELDED TO AN ANGLE 94 5.2.2.1. TEST #3 94 5.2.2.2. TEST #4 102 5.2.2.3. TEST #5 104 5.2.3. REBAR #2 AT 4 5 DEGREES WITH SHORT RECESS 111 5.2.3.1. TEST #7 111 5.2.3.2. TEST #8 120 5.2.3.3. TEST #13 126 5.2.4. REBAR #2 AT 45 DEGREES IN A 75 MM THICK SLAB 130 5.2.4.1. TEST #11 130 v i 5.2.4.2. TEST #12 140 5.2.4.3. TEST #14 146 5.2.5. ANGLE WELDED TO REINFORCING MESH 151 5.2.5.1. TEST #9 151 5.2.5.2. TEST #15 161 5.2.6. REVERSED ANGLE WELDED TO REINFORCING MESH 165 5.2.6.1. TEST #10 165 5.2.6.2. TEST #16 176 CHAPTER 6 - DISCUSSION, CONCLUSIONS AND FUTURE SCOPE 182 6.1. INTRODUCTION 182 6.2. DISCUSSION 183 6.2.1. BEHAVIOUR OF THE CONNECTIONS DURING CYCLIC LOADING 183 6.2.2. BEHAVIOUR OF CONNECTIONS DURING MONOTONIC LOADING 184 6.2.3. STRENGTH OF CONNECTIONS 186 6.2.3.1. MAXIMUM MEASURED STRENGTHS 186 6.2.3.2. STRENGTHS AT FAILURE 187 6.2.3.3. CALCULATED STRENGTHS 187 6.2.3.3.1. SEPARATE REBAR AT 45 DEGREES. 187 6.2.3.3.2. CONNECTIONS WELDED TO REINFORCING MESH. 191 6.2.3.3.2.1. TYPE A 192 6.2.3.3.2.2. TYPE B 192 6.2.3.3.3. CALCULATED CONNECTION STRENGTHS 195 v i i 6.2.3.4. INFLUENCE OF CONCRETE STRENGTH ON CONNECTION STRENGTHS 201 6.2.4. TYPES OF CONNECTION FAILURES 204 6.2.5. DISPLACEMENT OF THE CONNECTIONS 206 6.2.6. SIDEWAYS MOVEMENT OF THE CONNECTIONS 207 6.2.7. CONNECTION DETAILS 208 6.2.7.1. PANEL THICKNESS 208 6.2.7.2. PANEL RECESS 209 6.2.7.3. ANGLE WELDED TO CONNECTION 209 6.3. CONCLUSION 209 6.4. FUTURE SCOPE 210 REFERENCES 213 v i i i LIST OF TABLES TABLE # PAGE # 3.1. MIX PROPORTIONS 32 3.2. COMPRESSION STRENGTHS OF CYLINDERS 34 3.3. YIELD AND ULTIMATE STRESSES 41 6.1. MAXIMUM CONNECTION STRENGTHS 185 6.2. MEASURED CONNECTION STRENGTH AT FAILURE 188 6.3. CALCULATED CONNECTION STRENGTHS 196 6.4. CALCULATED CONNECTION STRENGTHS 197 6.5. MOVEMENTS OF CONNECTIONS 205 i x LIST OF FIGURES FIGURE # PAGE # 2.1. A connection t y p i c a l to those tested by Aswad Ref. [8]. 9 2.2. Typical Mechanical connection. 11 2.3. Details of the connections tested, showing the arrangements of the studs. 13 2.4. Details of connections reported in Ref. [8]. 16 2.4. Continued 17 2.5. Types of connections tested in Ref. [4]. 19 2.5. Continued 20 2.5. Continued 21 2.5. Continued 22 2.6. Model developed in Ref. [4]. 23 2.7. Connection tested in Ref. [12]. 25 3.1. Spec imen 28 3.2. Forms for specimens 30 3.3. Compression strengths of cylinders 36 3.4. Support plate 40 3.5. Yiel d and Ultimate Stresses of Steel 43 3.6. Load-deflection curves for rebar #2 44 3.7. Rebar #1 at 45 degrees 46 3.8. Rebar #1 at 45 degrees welded to angle 48 3.9. Rebar #2 at 45 degrees with small recess 49 3.5. Rebar #2 at 45 degrees with small recess 49 X 3.10. Rebar #2 at 45 degrees with a recess in a thicker slab 50 3.11. Angle welded to reinforcing mesh 53 3.12. Reversed angle welded to reinforcing mesh 54 4.1. Test r i g 56 4.2. Test r i g separated into parts. 59 4.2. Continued 60 4.3. Locations of the LVDTs. 69 4.4. LVDT mounting for connection displacement. 71 4.5. LVDT mounting for sideways movement and specimen rotation. 74 5.1. Load-deflection curve for Test #1. 83 5.2. Load-deflection curve for Test #2. 88 5.3. Load-deflection curve for Test #6. 91 5.4. Sideways movement of connection (Test #6). 92 5.5. Load-deflection curve for Test #3. 97 5.6. Sideways movement of connection (Test #3). 98 5.7. Cracking pattern for Test #3. 99 5.8. Load-deflection curve for Test #4. 103 5.9. Load application points for Rebar #1 with angle. 106 5.10. Load-deflection curve for Test #5. 107 5.11. Sideways movement of the connection (Test #5). 108 5.12. Load-deflection curve for Test #7. 113 5.13. Sideways movement of connection (Test #7). 114 5.14. Load-deflection curve for Test #8. 122 5.15. Sideways movement of connection (Test #8). 123 5.16. Load-deflection curve for Test #13. 127 xi 5.17. Sideways movement of connection (Test #13). 128 5.18. Load-deflection curve for Test #11. 132 5.19. Enlargement of the load-deflection curve for Test #11. 133 5.20. Sideways movement of connection (Test #11). 134 5.21. Load-deflection curve for Test #12. 141 5.22. Sideways movement of connection (Test #12). 142 5.23. Load-deflection curve for Test #14. 147 5.24. Enlargement of the load-deflection curve for Test #14. 148 5.25. Sideways movement of connection (Test #14). 149 5.26. Load-deflection curve for Test #9. 153 5.27. Enlargement of the load-deflection curve for Test #9. 154 5.28. Sideways movement of connection (Test #9). 155 5.29. Load-deflection curve for Test #15. 162 5.30. Enlargement of the load-deflection curve for Test #15. 163 5.31. Sideways movement of connection (Test #15). 164 5.32. Load-deflection curve for Test #10. 167 5.33. Enlargement of the load-deflection curve for Test #10. 168 5.34. Sideways movement of connection (Test #10). 169 5.35. Load-deflection curve for Test #16. 178 5.36. Enlargement of the load-deflection curve for Test #16. 179 5.37. Sideways movement of connection (Test #16). 180 x i i 6.1. Forces in rebar connection. 190 6.2. Forces in connection welded to reinforcing mesh (perpendicular cracking). 193 6.3. Forces in connection welded to reinforcing mesh (diagonal cracking). 194 6.4. Measured versus Calculated strengths. 198 6.5. Measured versus Calculated strengths. 199 6.6. Failure versus Calculated strengths. 200 6.7. Failure versus Calculated strengths. 202 6.8. Connection strengths versus Cylinder strengths. 203 6.9. Forces perpendicular to connection. 212 x i i i LIST OF PHOTOGRAPHS PHOTOGRAPH # PAGE # 3.1. Concrete test cylinder 38 3.2 Rebar #1 at 45 degrees 46 3.3. Rebar #1 at 45 degrees welded to angle 48 3.4. Rebar #2 at 45 degrees with a short recess 49 3.5. Rebar #2 at 45 degrees with a recess in a thicker slab 50 3.6. Angle welded to reinforcing mesh 53 3.7. Reversed angle welded to reinforcing mesh 54 4.1. Test r i g 57 4.2. MTS co n t r o l l e r and NEFF 620. 64 4.3. PDP-11/10 computer terminal. 65 4.4. LVDTs measuring connection displacement. 72 4.5. LVDT measuring sideways movement. 75 4.6. LVDTs measuring specimen rotation. 76 5.1. Small crack along bottom leg (Cycle #1 +79 kN). 84 5.2. Bottom corner f a l l i n g off (Cycle #2 +91 kN). 85 5.3. Connection at f a i l u r e (Cycle #3 +74 kN). 86 5.4. Connection at f a i l u r e (Cycle #1 -44 kN). 89 5.5. Connection at f a i l u r e (Cycle #6 +59 kN). 93 5.6. Bending of top bar (Cycle #11 +40 kN). 100 5.7. Connection at f a i l u r e (Cycle #13 +32 kN). 101 5.8. Connection at f a i l u r e (Cycle #5 +50 kN). 109 5.9. Connection at f a i l u r e (Cycle #5 +50 kN). 110 5.10. Cycle #4 -50 kN (Test #7). 115 xiv 5. 1 1 . Cycle #4 -87 kN (Test #7) . 1 16 5. 12. Cycle #4 -87 kN (Test #7). 1 17 5. 13. Cycle #5 + 56 kN (Test #7) . 1 18 5. 14. Cycle #5 + 32 kN (Test #7) . 1 19 5. 15. Connection at f a i l u r e (Cycle #16 -38 kN) . 1 24 5. 16. Connection at f a i l u r e (Cycle #16 -38 kN) . 1 25 5. 17. Connection at f a i l u r e (Cycle #1 + 42 kN) . 1 29 5. 18. Cycle #10 -79 kN (Test #11). 1 35 5. 19. Cycle #12 + 65 kN (Test #11). 1 36 5. 20. Cycle #12 + 77 kN (Test #11). 1 37 5. 21 . Cycle #12 -59 kN (Test #11). 1 38 5. 22. Connection at fa i l u r e (Cycle #13 + 67 kN) . 1 39 5. 23. Tension corner f a l l i n g off (Cycl e #1 +97 kN). 143 5. 24. Top corner f a l l i n g off (Cycle #1 + 75 kN) . 144 5. 25. Connection at f a i l u r e (Cycle #1 + 75 kN) . 1 45 5. 26. Connection at f a i l u r e (Cycle #15 + 62 kN) . 1 50 5. 27. Cycle #4 -41 kN (Test #9) . 156 5. 28. Cycle #7 + 51 kN (Test #9) . 1 57 5. 29. Cycle #7 + 51 kN (Test #9) . 158 5. 30. Cycle #9 -29 kN (Test #9) . 159 5. 31 . Cycle #9 -29 kN (Test #9) . 1 60 5. 32. Cycle #2 + 42 kN (Test #10). 170 5. 33. Cycle #2 + 42 kN (Test #10) . 171 5. 34. Cycle #4 + 60 kN (Test #10) . 172 5. 35. Cycle #6 + 56 kN (Test #10) . 1 73 5. 36. Cycle #6 + 56 kN (Test #10). 174 5. 37. Cycle #7 -14 kN (Test #10) . 1 75 Cycle #11 + 42 kN (Test #16). ACKNOWLEDGEMENTS The author wishes to thank Dr. R.A. Spencer for his valuable guidance in the planning and in the investigation carried out in this thesis. The author i s proud to have been associated with Dr. Spencer in this work. Thanks are extended to the technicians in the laboratory, especially B. Merklie, G. Kirsch, D. Postgate and W. Schmit for their assistance in making the test equipment. The author is grateful for f i n a n c i a l support from the Prestressed Concrete I n s t i t u t e , Chicago, I l l i n o i s . Mikael K. Kallros 1 CHAPTER 1  INTRODUCTION 1 . 1 . GENERAL Panelized or precast concrete construction has experienced phenomenal growth in the past decade. This growth i s fueled by a combination of two factors. One i s the standardization of design and manufacturing and the second is the increase in f l e x i b i l i t y and speed of erection which i t o f f e r s . Precast buildings were i n i t i a l l y developed for use in e s s e n t i a l l y nonseismic regions. Their use has, however, spread to zones of high seismic risk in America, Europe, Japan and the Soviet Union before their seismic behaviour has been studied in depth. The expanding use of precast concrete buildings in seismic regions i s therefore presenting new challenges in the area of earthquake-resistant design. Research on precast concrete buildings during recent years has lead to s i g n i f i c a n t advances towards a better understanding of both component and system behaviour during earthquakes. Despite this progress, there s t i l l remain some questions to be answered about the seismic design procedures for precast concrete buildings. 2 Attempts to r e s i s t the forces within the l i m i t s of the e l a s t i c response are uneconomic for most concrete buildings. During large ground accelerations dynamic forces equal to the l a t e r a l load capacity of the structure may be induced in the structure. This can lead to y i e l d i n g or p l a s t i c deformations in some of the c r i t i c a l areas in the structure. In order to prevent serious damage and loss of l i f e during a large earthquake, i t is therefore necessary to ensure that the post-elastic deformations can occur without a complete structural collapse. A d u c t i l e connection between two precast panels can help to ensure a safe structure. The earthquake damage in precast concrete panel buildings occurs in most cases along the j o i n t s . The panels themselves usually display very l i t t l e damage. However, some of the energy generated during an earthquake can be dissipated through the panel j o i n t s . This energy dissipation can be achieved by designing the connections in such a way that they can, even during several displacement cycles, deform i n e l a s t i c a l l y without fracture while maintaining at the same time their ultimate capacity. 1 . 2 . OBJECTIVES AND SCOPE The main objectives of t h i s research are : a) To develop improved connection d e t a i l s for members with thin flanges in earthquake resistant precast concrete buildings. 3 b) To develop r a t i o n a l methods of analysis and design of connections for members with thin flanges in earthquake resistant precast concrete buildings. Chapter 2 discusses some of the existing types of connections and their locations. Chapter 3 contains a description of the experimental materials, explains how the various materials were prepared and gives a description of the di f f e r e n t types of connections. Chapter 4 describes the test r i g , the measurements taken during the test, the loading procedure and the data acquisition and control system. Chapter 5 contains the results of the laboratory tests for the various types of connections. The results are presented through the use of graphs, tables and pictures. Chapter 6 presents a discussion of the experimental results, a comparison with theore t i c a l calculations and suggestions for future research. 4 CHAPTER 2 REVIEW OF CONNECTION DESIGNS FOR PRECAST CONCRETE 2.1. INTRODUCTION Embedded concrete connections must meet an assortment of design and performance c r i t e r i a . A connection must have the strength to r e s i s t the forces to which i t w i l l be exposed during i t s design l i f e t i m e . Some of these forces are dead loads, l i v e loads, wind loads, earthquake loads, s o i l and water pressures, volume changes and loads due to i n s t a b i l i t y . At the same time, the connections must also be able to accommodate r e l a t i v e l y large deformations. Designing good connection d e t a i l s also involves balancing continuity and d u c t i l i t y with good economics and ease of construction. A review of connection behaviour during earthquakes is presented. A number of the ex i s t i n g connections are reviewed along with the locations of the connections. 2.2. CONNECTION BEHAVIOUR DURING EARTHQUAKES Constructing concrete buildings out of precast panels in earthquake prone regions of the world raises many questions in earthquake resistant design. The strength and d u c t i l i t y of the connections between the precast panels w i l l 5 s i g n i f i c a n t l y a f f e c t the behaviour of a precast concrete building during an earthquake. The earthquake damage i s in most cases observed to take place along the connection l i n e s [1] . This happens mainly because the strength of the connection i s rarely comparable to the strength of the surrounding panel. The deterioration of the region around the connection under c y c l i c loading contributes to further reduction in the resistance of the connection. The performance of the connections in the walls and in the floor and roof diaphragms w i l l therefore be c r i t i c a l in r e s i s t i n g the earthquakes. It is desirable for the diaphragms to r e s i s t earthquake forces without y i e l d i n g , which requires that the connections remain e l a s t i c . This requires a connection which can continue to carry the forces due to dead and l i v e loads, and in addition can handle the dynamic forces induced by the earthquake. Walls, on the other hand, may be assumed to y i e l d under the action of l a t e r a l loads due to earthquakes. The design procedure for a building of t h i s type i s presented in the PCI Design Handbook [2] and the CPCI Metric Design Handbook [3]. A s t a t i c analysis using equivalent l a t e r a l forces shows that the connections between adjacent panels in the shear walls and diaphragms w i l l be subject to reversing shear forces acting p a r a l l e l to the edge of the panel during the earthquake [4], Connections between wall and floor or roof diaphragm units w i l l be subject to similar forces. In addition, the connections w i l l also have to 6 r e s i s t pullout forces acting both in the plane of the panels and perpendicular to them. During an earthquake any connection might also be subject to secondary forces in addition to those predicted by the s t a t i c analysis. Spencer and Tong [5] showed that the use of connections in shear walls having an actual strength higher than the strength assumed in the design is unconservative and can lead to an unsafe building. This was i l l u s t r a t e d by the results of a non-linear analysis of the response of a one story box-type precast structure during a moderate earthquake. Designing a building with r e l a t i v e l y low equivalent s t a t i c l a t e r a l forces and incorporating connections which w i l l y i e l d in an earthquake can be made both economical and safe. This i s the case i f the connections have an actual strength similar to that assumed in the design and can undergo the necessary i n e l a s t i c displacements. 2.3. TYPES OF CONNECTIONS There are various types of connections which can be c l a s s i f i e d as follows. 2.3.1. CAST IN PLACE CONNECTIONS These connections are also c a l l e d wet connections. After the panels are placed in position, reinforced or unreinforced cast-in-place concrete is used to form the junction between the panels. The connection area w i l l often 7 be heavily congested with reinforcement so i t i s advisable to use small aggregate and mixes with high slump. The properties which influence the strength and performance of wet connections include : a) Concrete strength - of both the panels and the cast in place connections. b) Connection reinforcement - t h i s includes the location and type of steel within the connection. c) Preparation of panel surface - the surface can be p l a i n , c a s t e l l a t e d or grooved. d) Force transverse to connection - can be from gravity loads or post-tensioning. e) Shear connectors - placed in the joint to r e s i s t shear forces in the j o i n t . The shear transfer in wet connections can be achieved through adhesion, f r i c t i o n , dowel action and dir e c t bearing. However, the bond between precast and cast-in-place concrete w i l l often be destroyed during the construction process. This along with creep and shrinkage e f f e c t s e f f e c t i v e l y eliminates the adhesion and f r i c t i o n transfer mechanisms unless a clamping force, acting normal to the connection face, i s applied [6].. This clamping force can be supplied through external compressive forces, post-tensioning or transverse mild s t e e l . It i s also essential in wet connections to develop transverse s t e e l beyond both sides of the f a i l u r e plane [ 7 ] . 8 2.3.2. DRY CONNECTIONS The most common type of dry shear connections are made up of embedded steel shapes anchored into the precast members by studs or reinforcing bars. The connection i s then finished by bolting or welding a separate steel piece to the embedded steel shapes. A variety of mechanisms are used in the dry connections to transfer shear between panels. The shear forces can be transferred through bearing of the steel shapes, shear of the connecting elements, shear of the weld, shear of the bolts or through f r i c t i o n between bolted plates. Martin and Korkosz [7] point out that dry connections should be designed conservatively for earthquake-resistant construction. They say that longer weld lengths, additional anchorage of reinforc i n g bars and lower allowable stresses are j u s t i f i e d for c y c l i c loading e s p e c i a l l y with the presence of forces normal to the shear connection. Testing performed by Aswad [8] shows that the application of one kip pull-out. force normal to a connection can reduce the ultimate shear capacity by 1/3. A normal force of thi s size can e a s i l y be reached during erection or caused by effects of shrinkage and temperature. A connection similar to the one Aswad tested can be seen in Figure 2.1. 1 Figure 2.1. A connection t y p i c a l to those tested by Aswad Ref. [ 8 ] . 2.3.2.1. MECHANICAL CONNECTORS These connections are usually made up of bolts and various i n s e r t s . Post-tensioning or high tension bolts may also be used in these types of dry connections. One example of a mechanical connection i s the Drescon-Concordia system [9]. This type of connection uses steel plate or a section with s l o t t e d holes which i s f r i c t i o n bolted to the steel i n s e r t s . These inserts are anchored into the concrete panels. A t y p i c a l connection i s shown in Figure 2.2. The slotted holes are made to accommodate the manufacturing and erection tolerances with additional clearance to absorb energy by s l i p p i n g . Tests showed that with slotted holes, the f r i c t i o n a l movement could give the desired energy d i s s i p a t i o n without causing i n e l a s t i c y i e l d i n g of the materials. The main feature of such a connection i s therefore the a b i l i t y to control the slippage of the j o i n t . This can be achieved by selecting the appropriate joint surface and the appropriate clearance for the slott e d holes. Properly designed, the connection would be expected to s l i p during severe seismic excitations but not under service loads. 2.3.2.2. WELDED CONNECTIONS These connections have steel embeddments, usually bars or plates, which are well anchored into the concrete panels These embeddments are situated at points, along the panel 11 weld afltr traction 0 CD -connecting pkrte Q) red. washers U— IrtMrt - jo int ELEVATION anchors INSERT CONNECTING PLATE panel -connecting piati -bolts (ASTM A325) -ins«rt X nuts welded to Insert SECTION wall panels connecting angles bolts connecting angles insert wall panel ( A S T M A325)i insert wall panel SECTION Figure 2.2. Typical Mechanical connection. 12 edges, where the connections can e a s i l y be welded to one another. A t h i r d steel piece i s usually used in order to make the welding of the two connections easier. When the precast joi n t s are welded together, they can immediately be subjected to f u l l loading. This w i l l e f f e c t i v e l y shorten the construction period. Previous research has suggested a number of d i f f e r e n t connection designs. These include : a) Embedded connections with r i g i d weld plates. b) Embedded rebar connections. d) Embedded connections with d u c t i l e weld plate connections. 2.3.2.2.1. EMBEDDED CONNECTIONS WITH RIGID  WELD PLATES Spencer and N e i l l e [10] reported tests on thi s type of connection. They performed reversed c y c l i c tests on headed stud connectors. Typical connections are shown in Figure 2.3. The shear forces were applied in the longitudinal d i r e c t i o n of the connection angle, as shown in Figure 2.3. Connection A1 was loaded monotonically to f a i l u r e while the other fiv e connections were loaded at frequencies in the range of 0.01 to 0.02 Hz. Spencer and N e i l l e reported that f a i l u r e of the connections was preceded by progressive crushing and sp a l l i n g of the concrete above and below the angle. They also observed progressive tension cracking p a r a l l e l to the 13 CONNECTION OETAILS OF STUDS 4 PANEL REINFORCEMENT Al A2.A3 B1 B2 B3 t'x3*3/gL«lZ'LONG • a'xj'x^glxu'lONG-sVx^Lxir/lONG-3x2x3/gLx10'LONG 6' , , I LL =.... Figure 2.3. Details of the connections tested in Ref. [10], showing the arrangements of the studs. 14 angle but at the same time several inches away from i t . Failure of connections A1 and B1 occured when a large block of concrete surrounding the studs broke away. The other connections f a i l e d when one of the studs broke off close to the weld. The fact that the weld between the stud and the angle f a i l e d in four of the connectors suggests that the weld i s a weak point in the connection. Hawkins [11] suggests that f i f t y percent of the studs should be able to carry the maximum load on the unit. The following conclusions were reached by Spencer and N e i l l e [10]. a) The PCI procedures for ca l c u l a t i n g the ultimate design strength of these connections under s t a t i c loading give conservative r e s u l t s . b) The strength of the connections in the f i r s t cycles of loading up to y i e l d w i l l be approximately the same as the strength in monotonic loading. c) If c y c l i c loading i s continued above the s t a b i l i t y l i m i t , the strength of the connections w i l l f a l l with an increasing number of cycles, and the y i e l d strength envelope w i l l tend to approach the s t a b i l i t y l i m i t . d) The deflections reached before f a i l u r e were from seven to twentyfour times the theoreti c a l e l a s t i c d eflection corresponding to the design ultimate strength. e) These connections, i f properly designed and detailed, appear to be suitable for use in earthquake-resistant buildings designed as box-type systems. 15 2.3.2.2.2. EMBEDDED REBAR CONNECTION This type of connection i s made up of steel r e i n f o r c i n g bars which are cast into the edge of the panels. There are several design advantages with t h i s type of connection compared to those using studs. Some of these are : a) A larger surface area i s available for welding the reinforcing bars to the angle. This means lower weld stresses and a reduction in the chance of p u l l i n g out the connection. b) There i s a longer development of the bars into the precast members. This reduces the chance of p u l l i n g out the connection. Aswad [8] tested a series of connections using f u l l -size double tee shapes. The tests did not cover a wide range of c y c l i c tests because the main objectives were to discover the f a i l u r e c a p a b i l i t i e s of the connections and to measure the r e l a t i v e s l i p between adjacent precast elements. In most cases the load was applied monotonically during several steps. Details of some of the connections can be seen in Figure 2.4. The c y c l i c tests, performed by Aswad [8], were not carri e d far enough to give much insight into the dynamic behaviour of the connections. The major observation was that the connections subjected to the c y c l i c loads showed no major s t i f f n e s s deterioration after three cycles. 16 Figure 2.4. Details of connections reported in Ref. [8], Figure 2.4. Continued 18 Another investigation performed by Spencer [4] consisted of c y c l i c shear tests on a variety of connections anchored into concrete panels. The types of connections tested are shown in Figure 2.5. Spencer [4] discussed design methods for c a l c u l a t i n g the strength of these connections under both monotonic and c y c l i c loading. The models used for the calculations can be seen in Figure 2.6. A comparison was made between the measured strength values and the calculated strength values. The following conclusions were reached by Spencer [4]. a) Loading cycles in the e l a s t i c range do not reduce the strength of the connections. b) The nominal strength of the connections can be found using the models shown in Figure 2.6. c) The strength of the connections, with the rebar running into the connection at 45 degrees, f a l l s to about 50 % of the nominal strength under c y c l i c loading into the i n e l a s t i c range. d) The strength of the connections, with the rebar running into the connection at 90 degrees, f a l l s to less than 50 % of the nominal strength under c y c l i c loading into the i n e l a s t i c range. These connections were not recommended for use in situations where they might be loaded past their e l a s t i c l i m i t . e) The connection with a recess in the panel edge and a straight embedded bars appears to perform best under simulated earthquake loading. Figure 2 . 5 . Types of connections tested in Ref. [4], Figure 2.5. Continued 21 Figure 2.5. Continued 22 Figure 2.5. Continued Figure 2.6. Model developed in Ref. [4]. 24 f) Panel thickness and concrete qu a l i t y can have a marked effect on the behaviour of the connections. 2.3.2.2.3. EMBEDDED CONNECTIONS WITH DUCTILE WELD  PLATE Saxena [12] performed tests on t h i s type of connection. He tested a system which used a d u c t i l e steel connector to replace the r i g i d plate or bar normally used to connect adjacent connections. In his connection a steel pipe with a longitudinal s l i t was welded between embeddments in adjacent precast panels. A s p l i t pipe connector can be seen in Figure 2.7. The following conclusions were reached by Saxena [12] : a) The s p l i t pipe connection i s able to accommodate the r e l a t i v e movement between panels due to shrinkage and temperature changes. b) The s p l i t pipe connection l i m i t s the forces that develop during dynamic loading which leaves the panels largely undamaged. c) Slight inaccuracies in the dimensions of the precast panels during casting can be accommodated when the pipe is welded in position on s i t e . 25 /• -pre-cast Panel Figure 2.7. Connection tested in Ref. [12]. 26 2.4. LOCATION OF CONNECTIONS In precast panel construction there are b a s i c a l l y five locations in which the connections can dissipate energy during an earthquake. These locations are : a) Connection between floor panels. b) Horizontal connections between wall panels. c) V e r t i c a l connections between wall panels. d) Connections between floor and wall panels. e) Connections between foundations and wall panels. 2.5. CONCLUDING REMARKS It should be noted that the types of connections used in the industry vary widely. Most of them are designed empirically and have not been studied experimentally or t h e o r e t i c a l l y . Their ultimate strength and seismic loading behaviour would therefore generally not be known. More research in the area of dynamic properties of connections, u t i l i z i n g embedded steel sections, would therefore be desi rable. 27 CHAPTER 3 EXPERIMENTAL MATERIALS AND SPECIMEN PREPARATION 3.1. INTRODUCTION A t o t a l of sixteen specimens were prepared in the Materials Laboratory of the C i v i l Engineering Department at the University of B r i t i s h Columbia. The sixteen specimens were prepared from materials obtained from various l o c a l suppliers. Each of the specimens was cast in a reusable plywood form. The forms were b u i l t to model as closely as possible, a section of a double-tee flange (Figure 3 . 1 . ) . The section was poured out of concrete mixed in the laboratory. A l l specimens were reinforced with a welded wire mesh and had a s o l i d s t e e l support plate cast into one side. The purpose of the steel support plate was to hold the specimen in the testing frame during the testing procedure. The connections were cast into the opposite side of the specimens. The aim of the testing program was to examine the behaviour of six di f f e r e n t types of connections during c y c l i c loading. In order to study a number of d i f f e r e n t connections, only two or three specimens were cast with the same 28 Figure 3.1. Specimen 29 connection type. It i s therefore not possible to conduct a s t a t i s t i c a l comparison of the connections. 3.2. FORM The formwork for the specimens consisted of 19 mm plywood and 50 mm x 100 mm (nominal) lumber bolted together for easy and e f f i c i e n t s t r i p p i n g . Two i d e n t i c a l forms were b u i l t in order to speed up the casting process. One of the forms was modified for specimens 11, 15 and 16. These were 25 mm thicker than the other specimens. Details of the forms can be seen in Figure 3.2. The side opposite the connection was made 19 mm thicker than the rest of the specimen. The width of this thicker section was 100 mm. The thicker part was made in an attempt to model the t r a n s i t i o n between the flange and the web in a real double-tee beam. The wider part also provided more cover for the steel plate which was used to hold the specimen in the testing frame. The forms were thoroughly o i l e d before the concrete was placed. This was done in order to make the stripping process easier. Care was also taken to prevent any leakage through the j o i n t s during the casting procedure. 30 50 mm inn? I I I T777 I 1 1 1 19 mm V'Y'Y.''''''''' |*— 50 mm 700 mm 1500 mm 1 70 mm 100 mm 4 * 100 mm Opening for support plate Figure 3.2. Forms for specimens 31 3.3. CONCRETE 3.3.1. CONCRETE MIXING AND CASTING PROCEDURE The concrete mixed in the laboratory was designed for a compressive strength of 30 MPa. The following materials were used for the mixture (the mix proportions can be found in Table 3.1.) : a) Cement - Type 10 normal strength cement supplied in 40 kg paper bags. b) Sand - Industrial fine sand supplied in 36 kg paper bags. c) Gravel - Assortment of sizes, up to 10 mm in s i z e . The gravel came in 36 kg paper bags. However, bagged gravel was not available for mixes 11, 15 and 16. The gravel for these mixes was obtained from a separate p i l e . The gravel in t h i s p i l e also had a maximum aggregate size of 10 mm but had a s l i g h t l y higher moisture content. d) Water - Ordinary tap water. The cement, aggregates and water were weighed and stored in separate containers. The ingredients for the concrete were then mixed together in a pan-type mixer, in the following order : a) Coarse and fine aggregates were poured into the mixer. b) The aggregates were batched together. c) Cement was added to the aggregates. d) The cement was batched together with the aggregates. e) Water was added to the mix. TABLE 3.1. MIX PROPORTIONS MIX # TEST # CEMENT (lb) SAND (lb) GRAVEL (lb) WATER (lb) 1 2 40 80 1 40 20 2 1 40 80 1 40 20 3 4 40 80 1 40 20 4 3 40 80 1 40 20 5 6 40 80 1 40 20 6 5 40 80 1 40 20 7 8 40 80 1 40 20 8 7 40 80 1 40 20 9 13 40 80 1 40 20 10 9 40 80 1 40 20 1 1 1 4 60 120 210 30 1 2 15 40 80 1 40 20 1 3 10 40 80 1 40 20 14 1 6 40 80 1 40 20 15 1 2 60 120 210 30 16 1 1 60 1 20 210 30 33 f) The whole combination was mixed for approximately 10 minutes. After the mixing was completed, the concrete was placed in the form and was compacted by means of a vibrating table. Care was taken not to disturb the reinforcing mesh and the connection during the placing of the concrete. Three test cylinders were also cast from each mix in order to check the 28 day compressive strength of the concrete mixtures. The surfaces of the specimens were fini s h e d with a trowel and both the form and the cylinders were covered with a p l a s t i c sheet to prevent moisture losses during the f i r s t day. The form was disassembled during the second day and the specimen was moved into the moisture room for further curing. 3.3.2. TEST RESULTS OF CONCRETE MIXES Before the specimens were cast, the slump and the a i r content of the mixture was recorded. A summary of these results are available in Table 3.2. Unfortunately, the a i r meter did not work properly for the f i r s t seven mixes. It can also be seen in the table that the slump and a i r content for mixes 11, 15 and 16 were quite d i f f e r e n t from the other mixtures. This i s most l i k e l y caused by the d i f f e r e n t coarse aggregate which was used for them. The results of the 28 day compressive strengths of the cylinders can be seen in Table 3.2. In Figure 3.3. the relationships between the average compression strengths of TABLE 3.2. COMPRESSION STRENGTHS OF CYLINDERS MIX CYLI- STRE- STRE- AVRG. SLUMP AIR PIC. NDER NGTH NGTH STRE. CONTENT # # (lb) (MPa) (MPa) (mm) (%) # 1 1 92300 50.64 50.64 63.5 (1 ) 2 1 91 300 50.09 2 2 78000 42.80 49.00 63.5 (1) 2 3 98600 54. 1 0 3 1 89800 49.27 3 2 95000 52. 1 2 51 .37 57.2 (1) 3 3 96100 52.73 4 1 75600 41.48 . 4 2 83600 45.87 42.92 63.5 (1) 4 3 75500 41 .42 5 1 85900 47. 1 3 2-A 5 2 77400 42.47 42.63 63.5 (1 ) 1 -A 5 3 69800 38.30 3-A 6 1 85300 46.80 4-A 6 2 96100 52.73 49.31 57.2 (1 ) 5-A 6 3 88200 48.39 6-A 7 1 93600 51 .36 7 2 86800 47.62 49.89 50.8 (1) 7 3 92400 50.70 8 1 68200 37.42 8 2 76000 41 .70 42.43 44.5 1.9 8 3 87800 48. 1 7 9 1 92000 50.48 9 2 85000 46.64 49.84 44.5 1 .8 9. 3 95500 52.40 10 1 87800 48.17 10 2 91000 49.93 51 .72 31 .8 1.7 10 3 104000 57.06 4-B TABLE 3.2. CONTINUED MIX CYLI- STRE- STRE- AVRG. SLUMP AIR PIC. NDER NGTH NGTH STRE. CONTENT # # (lb) (MPa) (MPa) (mm) (%) # 1 1 1 99600 54.65 5-B 1 1 2 101000 55.42 51 .43 19.1 2.7 1 1 3 80600 44.22 (2) 1 2 1 81000 44.44 12 2 90400 49.60 50.70 44.5 1 .7 1-B 1 2 3 105800 58.05 13 1 77000 42.25 13 2 68000 37.31 38.81 44.5 1.9 3-B 13 3 67200 36.87 14 1 101000 55.42 14 2 89000 48.83 53.26 38. 1 1.7 14 3 101200 55.53 15 1 82800 45.43 6-B 15 2 90300 49.54 49.84 19.1 2.3 15 3 99400 54.54 (2) 16 1 88800 48.72 2-B 16 2 110000 60.35 52.71 19.1 2.2 16 3 89400 49.05 (2) (1) A i r meter not working. (2) Gravel taken from separate p i l e . Compression strengths of cylinders Average compression strength i i i i i i i . i i i i i • • • • > • 1 2 6 3 4 5 7 8 9 11 15 16 10 12 13 14 Mix # 37 the d i f f e r e n t mixtures are shown. This figure shows that the average compression strengths are f a i r l y uniform but well above the design strength of 30 MPa. A l l of the compression cylinders f a i l e d in an explosive f a i l u r e , causing small pieces of concrete to f l y o f f . About 80 % of the cylinders f a i l e d in a conical shaped f a i l u r e mode. The remaining ones f a i l e d in a diagonal mode. Photograph 3.1. i l l u s t r a t e some ty p i c a l compression r e s u l t s . 3.4. STEEL 3.4.1. REINFORCING MESH The main reinforcement for the specimens consisted of a welded wire mesh. The mesh was produced from 8.89 mm diameter plai n wire pieces which were welded together making 152 mm x 152 mm squares. The location of the wires was kept the same for a l l specimens. A tension test performed on a set of wire pieces produced a y i e l d strength of 530 MPa and an ultimate strength of 660 MPa. 3.4.2. SPECIMEN SUPPORT PLATE The support plate was made of ordinary f l a t steel 100 mm wide and 8 mm thick. 75 mm of the plate was cast into the back of the specimen and the wire mesh was welded to the plate. In the middle and at both ends of the plate, short 38 Photograph 3.1. Concrete test cylinder 39 pieces were welded to the edge of the long plate. These pieces had 22 mm bolt holes through the middle. A sketch of the support plate can be found in Figure 3.4. The support plate was chosen as the means for mounting the specimens in the test apparatus to avoid clamping the specimens at the ends. Compression on the ends of the specimens, during testing, might have prevented cracks from propagating from the connections along the edge of the test spec imen. 3.4.3. CONNECTION STEEL Ordinary 1OM G40.21-300W weldable reinforcing steel was used for four of the connections. The f i r s t six connections were made from one 6.0 meter length of rebar. A d i f f e r e n t length of rebar, obtained from a d i f f e r e n t supplier, was used for the other connections. The rebar used for the f i r s t six connections w i l l be referred to hereafter as Rebar #1 and the rebar used for the other connections as Rebar #2. The y i e l d and ultimate strengths of the two rebars were quite d i f f e r e n t as can be seen in Table 3.3. The measured values were also higher than the expected values for a 300W st e e l . The rebar used in the connections was bent to a desired shape (discussed l a t e r ) . A loading plate was welded to the connection before the specimen was tested. A set of tension tests were performed on the following rebar systems : 40 22 mm d i a . 150 mm 200 mm 100 mm Figure 3.4. Support plate TABLE 3.3. YIELD AND ULTIMATE STRESSES REBAR SYSTEMS YIELD STRESS (MPa) ULTIMATE STRESS (MPa) REBAR #2 STRAIGHT BAR 409. 2 689.5 REBAR #2 BAR BENT ONCE 387. 0 676. 1 REBAR #2 BAR BENT FIVE TIMES 347. 0 680.6 REBAR #2 P-IECES WELDED TO THE BAR 471 . 5 688.6 REBAR #2 PIECES WELDED TO THE BAR AND BENT ONCE 489. 3 698.4 REBAR #2 MILL SPECIFICATIONS 445. 0 674.0 REBAR #1 STRAIGHT BAR 596. 4 971 .5 WIRE IN REINFORCING MESH 530. 0 660.0 42 a) Straight rebar. b) Rebar bent once to 30 degrees and then straightened. c) Rebar bent f i v e times to 30 degrees and then straightened. d) Two 25 mm rebar pieces were welded 50 mm apart to the side of a straight rebar. e) Two 25 mm rebar pieces were welded 50 mm apart to the side of a rebar bent once to 30 degrees and then stra ightened. The effect on the rebar's y i e l d stress and ultimate stress, due to bending and heating, can be seen in Figure 3.5. and Figure 3.6. The figures show that bending of the rebar decreases the y i e l d strength s l i g h t l y , while heating the rebar increases the y i e l d strength. The y i e l d strength increases further when the rebar i s both bent and heated. However, no noticeable change i s observed in the ultimate strength. Figure 3.5. also displays the d i f f e r e n t shapes of the load-deflection curves for the five cases. The differences in the y i e l d strengths can be attributed to aging. The aging process in low-carbon steels, such as rebar, can be divided into two types: quench aging and s t r a i n aging. Quench aging occurs when low-carbon steels are rapidly cooled down from temperatures s l i g h t l y below the lower c r i t i c a l temperature of the s t e e l . The result of the rapid cooling down i s an increase in strength and hardness and a decrease in the d u c t i l i t y of the s t e e l . | \ I Yield stress \//A Ultimate stres 45 Strain aging occurs when low-carbon s t e e l i s deformed or cold worked at room temperature. When the steel i s deformed or cold worked, i t s strength and hardness i s increased while the d u c t i l i t y i s decreased. 3.5. TYPES OF CONNECTIONS Six di f f e r e n t types of connections were prepared. Four of them were made out of bent rebar. Some had an angle welded to the rebar. Two types used an angle welded d i r e c t l y to the reinforcing mesh. Some of the connection designs for specimens 7-16 were based on the results of tests 1-6. These tests were conducted before specimens 7-16 were fabricated. 3.5.1. REBAR #1 BENT TO 45 DEGREES. This connection consisted of a pl a i n rebar placed along the edge of the specimen with two legs bent in at 45 degrees. Each leg was 300 mm long and the section of the connection running along the edge was 200 mm long. Figure 3.7. and Photograph 3.2. show this connection. A drawback with this connection type i s that part of the specimen's edge had to be chiseled off in order to expose the rebar, before the loading plate could be successfully welded to the rebar. This connection would therefore be d i f f i c u l t to weld to another connection. 46 1400 mm 1 Figure 3.7. Rebar #1 at 45 degrees Photograph 3.2. Rebar #1 at 45 degrees 47 3.5.2. REBAR #1 BENT TO 45 DEGREES AND WELDED TO AN  ANGLE Similar to the previous connection, except that a 50 mm x 50 mm x 200 mm long angle was welded to the rebar. Figure 3.8. and Photograph 3.3. show t h i s connection. The major benefit of thi s type of connection i s the ease with which the connection can be welded to an adjacent connection. 3.5.3. REBAR #2 BENT TO 45 DEGREES WITH SHORT RECESS Similar to connection #1, except that a 20 mm x 20 mm x 200 mm long recess was made in order to expose the side of the rebar. Figure 3.9. and Photograph 3.4. show thi s connection. Since the side of the bar, along the edge of the specimen, i s exposed in thi s type of connection i t w i l l be easier to weld two flanges together. 3.5.4. REBAR #2 BENT TO 45 DEGREES WITH A RECESS IN A  THICKER SLAB Similar to connection #1, except that this connection had a 25 mm x 25 mm recess along the length of the specimen. The thickness of the slab was also 75 mm, compared to 50 mm for the other specimens. Figure 3.10. and Photograph 3.5. show thi s connection. Photograph 3.3. Rebar #1 at 45 degrees welded to angle 49 Photograph 3.4. Rebar #2 at 45 degrees with small recess 50 1400 mm F i g u r e 3.10 Rebar #2 at 45 degrees with a r ecess i n a t h i c k e r s l a b Photograph 3.5. Rebar #2 at 45 degrees with a r e cess in a t h i c k e r s l a b 51 The major benefit derived from using t h i s connection i s the added concrete cover over the connection. The recess also makes i t easier to weld two connections together. 3.5.5. ANGLE WELDED TO REINFORCING MESH This connection had a 20 mm x 20 mm x 200 mm long angle welded to the outermost wire of the reinforcing mesh. The outer wire, running p a r a l l e l with the side of the specimen, was also cut close to the ends of the angle. The reinforcing wire had to be cut off since i t was too close to the edge of the specimen. Removing the wire w i l l however reduce the shear capacity of the specimen. Figure 3.11. and Photograph 3.6. show thi s connection. One benefit with t h i s connection i s that no extra piece of rebar i s needed. The connection i s also easy to weld to another one. 3.5.6. REVERSED ANGLE WELDED TO REINFORCING MESH This connection had a 20 mm x 20 mm x 200 mm long reversed angle welded to the outermost wire of the reinforcing mesh. Since the angle was reversed, the outermost wire was 50 mm into the slab and was therefore not cut o f f . Figure 3.12. and Photograph 3.7. show t h i s connection. This connection would be preferable over the previous one since a l l of the reinforcing strands can be kept intact. Two flanges with this connection can also be e a s i l y welded together. 53 25 mm 600 mm 50 mm 1 50 mm Figure 3.11. Angle welded to r e i n f o r c i n g mesh Photograph 3.6. Angle welded to r e i n f o r c i n g mesh 54 55 CHAPTER 4  LABORATORY TEST DETAILS 4.1. INTRODUCTION A t o t a l of sixteen connections were tested in the Structural Engineering Laboratory of the C i v i l Engineering Department at the University of B r i t i s h Columbia. During the testing, the specimens were held v e r t i c a l l y in the test r i g (see Figure 4.1. and Photograph 4.1.). A steel plate with a narrow spacer was welded to the connections. A displacement controlled hydraulic jack, mounted with i t s axis v e r t i c a l in the test r i g , applied either a monotonic or a reversed c y c l i c load to the steel plate. A l l connections were loaded into the i n e l a s t i c range, and testing continued u n t i l each connection f a i l e d . The v e r t i c a l displacement and the sideways movement of the connections were measured r e l a t i v e to the concrete and load-displacement curves were recorded for each t e s t . The data was recorded with a NEFF 620 data a c q u i s i t i o n system which was controlled by a PDP-11/10 minicomputer. The data f i l e s were transferred at the end of each test, through the UBC Amdahl computer, to a PC computer where they were processed. 56 i: \Y\4 R e a c t i o n frame i Top control rod Figure 4 . 1 . Test r i g Photograph 4.1. Test r i g 58 4.2. ASSUMPTIONS MADE FOR THE TESTS It i s assumed that the connections in a precast panel system w i l l behave in more or le s s the same fashion as they did in the t e s t . During earthquake excitation, i n e l a s t i c action w i l l occur in the c r i t i c a l regions around the connections. A du c t i l e connection w i l l then help to ensure that the i n e l a s t i c deformation i s limited to the connection zone. It i s therefore assumed that the panels w i l l remain in the e l a s t i c range while the non-linear behaviour i s limited to the connections. 4.3. TEST RIG The test r i g which was used in the laboratory can be seen in Figure 4.1. and Photograph 4.1. This test r i g can be divided into several separate parts : a) Reaction frame - This frame consisted of two 5.5 m long columns which were placed 2.5 m apart (Figure 4.2.(a)). These columns were bolted to the concrete f l o o r and had a large cross beam between them. The purpose of the cross beam was to hold the hydraulic jack in a v e r t i c a l posit ion. b) Support column - This 1.6 m high column was bolted to the concrete floor and served as the support for the (a) Reaction frame Spacer Angle C li > o 0) o oo 0 o (b) Support column Top control rod Mild steel s t r i p 1 (c) side control rod (d) Load plate Figure 4.2. Test r i g separated into parts. Figure 4.2. Continued 61 specimens, holding them in a v e r t i c a l position (Figure 4.2.(b)). The three control rods were also bolted to the column. c) Hydraulic jack - The hydraulic jack had a 100 Kip capacity (Figure 4.2.(a)). A l o a d c e l l with the same capacity was attached to one end of the jack. The other end of the jack was bolted to the bottom of the cross beam through a c l e v i s which allowed the jack to swing about an axis p a r a l l e l to the cross beam. The jack was therefore restrained from moving sideways but was free to move towards or away from the specimen. d) Yoke - The yoke consisted of two side arms which were welded to a top cross beam (Figure 4.2.(e)). The hydraulic jack was connected to the yoke with a bolt through the cross beam. The two side arms of the yoke were connected to an inverted T-shape. e) Inverted T-shape - This piece consisted of a cross beam with a v e r t i c a l hollow control arm welded to i t (Figure 4.2. ( f ) ) . A short arm was attached to each end of the cross beam. One control rod was attached to each of the short arms and one rod was connected to the top of the hollow control arm. These three control arms formed a parallelogram linkage that constrained the T-shape to remain v e r t i c a l when i t moved up and down. The cross beam had six bolt holes through the middle. These were used to transfer the load from the cross beam to the loading plate which was attached to the connection. The 62 inverted T-shape was attached to the yoke by a 19 mm diameter pin through each of the short arms. f) Loading plate - A 300 mm x 125 mm x 20 mm plate was welded to each connection. A mild steel s t r i p was welded down the middle in order to act as a spacer between the connection and the loading plate (Figure 4.2.(d)). The plate was welded to the connection before the specimen was l i f t e d into the test r i g . Prior to the start of each test, the cross beam of the inverted T-shape was bolted to the loading plate with six bolts . A small aluminum angle was bolted to the top of the loading plate and served as the reference point for the LVDTs. g) Control rods - These three rods were pinned between the inverted T-shape and the specimen support column (Figure 4.2.(c)). These rods formed a parallelogram linkage which prevented the inverted T-shape's cross beam from rotating. This prevented the application of moments to the connections and ensured that they were subjected only to dire c t shear loading. h) Spacer - This piece served as a spacer between the specimen and the support column (Figure 4.2.(b)). A large angle with bolt holes matching the specimen support plate was attached to one side of the spacer. 4.4. LOADING AND DATA ACQUISITION SYSTEM 63 The loading and data acqu i s i t i o n system consisted of four major components : a) MTS servo-controlled load system. b) A NEFF 620 data acqu i s i t i o n system. c) A PDP-11/10 minicomputer. d) UBC Amdahl computer. The MTS load system was used to apply the load. The applied load and the corresponding displacements were recorded with the NEFF and the PDP. The data for each test was transferred with the Amdahl computer to a PC computer where i t was subsequently processed. Photographs 4.2. and 4.3. show part of the system. Some of the features of the data acqu i s i t i o n system can be seen below : The MTS servo-controlled load system controls : a) Speed of loading. b) Direction of loading. c) Movement of the hydraulic jack. It can be moved under either stroke or load c o n t r o l . d) Type of o s c i l l a t o r y motion. A function generator can produce either a sawtooth or a sine function for use as the command si g n a l . Photograph 4.2. MTS controller and NEFF 620. 65 Photograph 4.3. PDP-11/10 computer terminal. 66 The NEFF 620 provides : a) Signal conditioning for up to 64 transducers. b) Selection of the channels to be measured. c) Measurement of the voltage for the selected channels. d) D i g i t a l output representing the voltage and the channel selected. The PDP-11/10 provides : a) Control of the NEFF, including selection of channel to be read and the command to read the channel. b) Control to stop and start the data acqu i s i t i o n system at any time during the test. c) Processing of data from the NEFF using a program written in BASIC. The program can compute and print results for pre-selected channels during the te s t . d) Storage of measured voltages on a disk. e) F i l e transfer to the UBC Amdahl computer. The Amdahl computer provides : a) Capability to process the data transferred from the PDP-11/10 minicomputer. b) F i l e transfer to a PC computer. 67 4.5. USE OF THE DATA ACQUISITION SYSTEM A short description of the operations which are required to obtain data for the tests follows. For more comprehensive descriptions, please see the appropriate system manuals. 4.5.1. OPERATION OF MTS SERVO-CONTROLLER a) Set the appropriate c a l i b r a t i o n factors. b) Connect the leads to the NEFF. c) Select appropriate function on function generator. d) Set speed and d i r e c t i o n of loading. 4.5.2. OPERATION OF NEFF 620 a) Connect the leads from the LVDTs and the MTS c o n t r o l l e r to the front panel. b) Set the front panel switches to their appropriate settings. 4.5.3. OPERATION OF PDP-11/I 0 a) Prepare one RK05 disk with the necessary software. b) Load and run the BASIC software language processor. c) Prepare and run the data acquisition program for the pa r t i c u l a r a p p l i c a t i o n . d) Store the measured data on the disk. 68 e) Transmit the data f i l e s to the UBC Amdahl computer for further processing. 4.6. TEST MEASUREMENTS The load and the displacement voltages were recorded with the NEFF data a c q u i s i t i o n system and the PDP-11/10 computer during the tests. During the tests the load-deflection curves were also plotted on a X-Y recorder. This was done in order to have a v i s u a l control over the load and the def l e c t i o n of the connection. 4.6.1. LOAD MEASUREMENT The load applied to the connection was recorded with a 100 Kip load c e l l which was situated between the hydraulic jack and the yoke. The load c e l l was c a r e f u l l y c a l i b r a t e d before the testing program started. The obtained c a l i b r a t i o n factor was then used to calculate the applied load during testing and data processing. 4.6.2. DISPLACEMENT MEASUREMENT Four d i f f e r e n t sets of displacements (Figure 4.3.) were measured by linear variable d i f f e r e n t i a l transformers (LVDT's). These displacements were : a) The up and down movement of the connection r e l a t i v e to the concrete specimen (LVDT #1 and LVDT #2). 69 LVDT # 4 Figure 4.3. Locations of the LVDTs. 70 b) The sideways movement of the connection r e l a t i v e to the concrete specimen (LVDT #3). c) The displacement and rotation of the specimen r e l a t i v e to the test floor (LVDT #4 and LVDT #5). d) The movement of the hydraulic jack. At the beginning of each test, the LVDTs were calibrated and the c a l i b r a t i o n factors were stored in a f i l e which was used during further data processing. The LVDTs were calibr a t e d by measuring the voltages for several known displacements. From these voltages an average c a l i b r a t i o n factor was calculated for each LVDT. 4.6.2.1. CONNECTION DISPLACEMENT The movements of the connections were measured by two LVDTs. The maximum travel distance of these LVDTs was -12.5 mm to +12.5 mm. One LVDT was mounted on each side of the specimens (Figure 4.4. and Photograph 4.4.). The LVDTs were supported from a reusable seat situated at the back of the specimens because no cracking was expected in this region. The reusable LVDT seats were bolted to the specimen through two holes. These holes were d r i l l e d through the specimens after they had cured. The movements of the connections r e l a t i v e to the concrete panels were measured from an aluminum angle which was bolted to the top of the loading plate. The voltage readings from each LVDT was recorded Top view LVDTs Side view LVDTs Figure 4.4. LVDT mounting for connection displacement. Photograph 4.4. LVDTs measuring connection displacement. 73 separately in the data f i l e but the load-deflection curves were based on the average of the two readings. 4.6.2.2. SIDEWAYS MOVEMENT The sideways movement of the connections was measured with one LVDT which was supported from the cross beam of the inverted T-shape (Figure 4.5.(a) and Photograph 4.5.). This part of the test equipment was bolted to the plate which was welded to the embedded steel connection. The displacement probe was l i g h t l y spring loaded against an aluminum plate that was glued to the concrete. The maximum tra v e l distance of the LVDT was -25.4 mm to +25.4 mm. This measurement was taken in order to see the sideways movements of the connections rebar to the surrounding concrete during the various steps of the c y c l i c t e sts. 4.6.2.3. SPECIMEN ROTATION Two LVDTs were used to record the rotation and displacements of the specimen r e l a t i v e to the test floor (Figure 4.5.(b) and Photograph 4.6.). The movement of the support column was also checked. These measurements were used to check the strength of the support plate which was cast into the back of the specimen. One LVDT was placed at the front edge of the specimen and the other one was placed at the back edge. The maximum travel distance of these LVDTs was -12.5 mm to +12.5 mm. 74 Figure 4.5. LVDT mounting for sideways movement and specimen rotation. Photograph 4.5. LVDT measuring sideways movement. 76 Photograph 4.6. LVDTs measuring specimen ro t a t i o n . 77 4.6.3.4. MOVEMENT OF HYDRAULIC JACK The movement of the hydraulic jack was recorded by an LVDT b u i l t into the jack i t s e l f . The maximum tr a v e l of t h i s LVDT was -75 mm to +75 mm. 4.7. LOADING PROCEDURE The specimens were loaded by a servo-controlled hydraulic jack. The v e r t i c a l shear force applied by the jack was either monotonic or reversed c y c l i c . The monotonic tests loaded the connections in one di r e c t i o n u n t i l they f a i l e d . These monotonic tests were performed in order to compare their maximum strengths with the strengths of similar connections which were tested under c y c l i c loading. The c y c l i c tests usually began with several cycles of loading to about 70 % of the estimated monotonic capacity of the connections. The load was then increased with an increment of about 5 % of the capacity. At t h i s new load l e v e l the load was again cycled u n t i l no further deflection was noticed. This sequence was then continued u n t i l the connections f a i l e d completely or u n t i l the strength f e l l s u b stantially. The testing procedure was quasi-static with each cycle taking several minutes. The load applied to the connection could be held constant several times during each cycle in order to allow the specimen to be examined. The amount of deflection and load placed on the connection was v i s u a l l y controlled by watching an X-Y recorder. The load reversal points were also manually selected by watching the curve on the X-Y recorder. In addition to the curve recorded on the X-Y recorder, selected values were also printed by the computer during the te s t . These readings were selected manually at c r i t i c a l points by depressing the Return button on the computer keyboard. The values were at the same time stored in a separate f i l e on the PDP-11/10 disk. This f i l e could l a t e r be used to reproduce the load-deflection curve, the sideways movement and the specimen rotation for each test. 79 CHAPTER 5  EXPERIMENTAL RESULTS 5.1. INTRODUCTION This chapter discusses in d e t a i l the results of the tests performed in the laboratory. The behaviour during the loading of each connection i s described and the measured connection strength i s compared to the strength calculated by formulae which are discussed in section 6.2.3.3. Two curves are given for each connection. One shows the load applied to the connection versus the movement of the connection r e l a t i v e to the surrounding specimen. The other curve shows the sideways movement of the connection r e l a t i v e to the specimen. The strength values calculated by the formulae are indicated with l i n e s on the graphs. Two separate strength values were calculated and plotted on the load-deflection curves for the connections which were fabricated from separate reinforcing bars (see section 6.2.3.3.1.). These li n e s were denoted : a) Fy - Where the measured y i e l d stress for the rebar, as delivered, was used to calculate V n. 80 b) Fyw - Where the y i e l d stress for the rebar that was cold bent to 30 degrees, straightened and then welded was used to calculate V n. Two strength values were calculated and also plotted on the load-deflection curves for the connections which were welded straight to the reinforcing mesh of the specimens. a) Ca - Type A, perpendicular cracking (see section 6. 2. 3.3.2.1.). b) Cb - Type B, diagonal cracking (see section 6*2»3*3»2«2«)» 5.2. DETAILS OF CONNECTIONS TESTED Six d i f f e r e n t types of connections were tested in the laboratory. A comprehensive description, of the connections can be found in section 3.5. Section 6.2. summarizes and discusses the results of the t e s t s . The types of connections were : 5.2.1. REBAR #1 AT 45 DEGREES This type of connection consisted of a piece of rebar with two legs bent so as to enter the specimen at 45 degrees to the edge. The portion p a r a l l e l to the edge of the specimen was 200 mm long and each of the legs was 300 mm long. For more d e t a i l s on t h i s connection see section 3.5.1. A t o t a l of three specimens were fabricated and l a t e r tested with t h i s type of connection. The results were : 81 5.2.1.1. TEST #1 The load-deflection curve for t h i s test can be seen in Figure 5.1. Photograph 5.3. shows the connection after f a i l u r e . The following observations were noted during the testing ( Cy stands for Cycle ) : Cy 01 + 79 kN - a crack opened up along the top side of top reinforcing leg. - a small crack could also be noticed along the bottom leg (Photograph 5.1.). - small concrete pieces were sp a l l i n g off along the v e r t i c a l portion of the bar (Photograph 5.1.). Cy 01 -17 kN - no change.in the load but i t was observed that the panel moved s l i g h t l y . Cy 01 -94 kN - the crack around the top side of the connection leg opened up a b i t more. - a crack opened up at the top at the l e v e l of the support plate. Cy 02 +91 kN - a large piece of concrete f e l l off at the bottom corner of the connection (Photograph 5.2.) . - a longer crack opened up along the bottom leg of the rebar. Cy 02 -78 kN - large pieces of concrete were f a l l i n g o f f . - a dialgauge attached to the bottom corner of the panel showed a sideways movement of 82 around 0.5 inches towards the back side of the panel. - no cracking was v i s i b l e on the bottom side of the panel. Cy 03 +75 kN - the rebar broke suddenly at the bottom corner of the connection (Photograph 5.3.). - the break occurred at the end of the weld between the rebar and the loading plate. 83 Figure 5.1. Load-deflection curve for Test #1. Photograph 5.1. Small crack along bottom leg (Cycle #1 + 79 kN). Photograph 5.2. Bottom corner f a l l i n g off (Cycle #2 +91 kN). 86 Photograph 5.3. Connection at f a i l u r e (Cycle #3 + 74 kN). 87 5.2.1.2. TEST #2 The load-deflection curve for t h i s test can be seen in Figure 5.2. Photograph 5.4. shows the connection after f a i l u r e . The following observations were noted during testing : 0 Cy 01 +89 kN - a crack along the rebar became v i s i b l e (Similar to Photograph 5.1.). - the top corner of the specimen moved 2.5 mm re l a t i v e to the support post (measured with a dialgauge). Cy 01 -44 kN - the connection f a i l e d suddenly. - the f a i l u r e of the connection took place at the end of the weld between the rebar and the loading plate (Photograph 5.4.). - the crack along the top leg opened up a b i t more. - a concrete piece (100 mm x 100 mm) f e l l of along the bottom rebar leg (Similar to Photograph 5.2.). 8 8 (0 u in o o o o o o o o o o o o o o o o o o o o o -< I I I I I I I I I 1 I Figure 5.2. Load-deflection curve for Test #2. 89 Photograph 5.4. Connection at f a i l u r e (Cycle #1 -44 kN). 90 5.2.1.3. TEST #6 The load-deflection curve for th i s test can be seen in Figure 5.3. The sideways movement of the connection i s shown in Figure 5.4. Photograph 5.5. shows the connection af t e r f a i l u r e . The following observations were noted during the testing : Cy 03 +42 kN - no v i s i b l e cracking at this load. Cy 04 +60 kN - some concrete at the bottom corner of the connection f e l l off (Similar to Photograph 5.2). Cy 04 -56 kN - the top leg of the connection broke suddenly at end of the weld between the rebar and the loading plate. - at th i s point the dir e c t i o n of the load was reversed. Cy 05 +59 kN - the bottom leg f a i l e d at the end of the weld (Photograph 5.4.). - only one leg contributed to the strength during t h i s cycle since the other leg broke during the previous cycle. 91 Figure 5.3. Load-deflection curve for Test #6. o CO o o o o o o o a» oo ^ CD m o o o o 4 n w H •*-» CM co o o o o 5 0 o o o o o m CD 00 a> o I I I I I I I I Figure 5.4. Sideways movement of connection (Test #6). 93 Photograph 5.5. Connection at f a i l u r e (Cycle #6 +59 kN). 5.2.2. REBAR #1 AT 45 DEGREES WELDED TO AN ANGLE This connection was of the same type as the previous one. The only difference was that a 50 mm x 50 mm x 200 mm angle was welded to the reinforcing bar. For more d e t a i l s see section 3.5.2. In t o t a l three specimens of t h i s type were fabricated. Two of these were tested under c y c l i c loading while one of the specimens was tested monotonically. The results were : The load-deflection curve for t h i s test can be seen in Figure 5.5. The sideways movement of the connection i s shown in Figure 5.6. Photograph 5.7. shows the connection aft e r f a i l u r e . The following observations were noted during the testing : Cy 01 - cracking noises could be heard from the 5.2.2.1. TEST #3 specimen. Cy 02 e l a s t i c behavior, no change in cracking. Cy 04 when the connection was loaded upwards, a small crack along the top end of the angle became s l i g h t l y larger (See Figure 5.7.(a)). when the connection was loaded downwards, a crack at the bottom end of the angle opened up a l i t t l e b i t (See Figure 5.7.(b)). 95 Cy 07 -66 kN - the crack at the top opened up more. - some concrete f e l l of at bottom corner of the connection (See Figure 5.7.(c)). Cy 08 +67 kN - the concrete corner above the angle f e l l off (See Figure 5.7.(d)). - some concrete also f e l l off at the bottom corner of the angle. Cy 09 +69 kN - large pieces of the concrete f e l l off at the bottom corner of the angle (See Figure 5.7.(e)). Cy 09 -65 kN - more concrete f e l l off at the corners. - the panel moved sideways during maximum load. Cy 10 -42 kN - the weld between the angle and the connection rebar suddenly began to f a i l at the top end of the angle. - additional concrete was s p a l l i n g off around the connection rebar. Cy 11 +40 kN - the f a i l u r e of the weld between the rebar and the angle progressed. - the top bar was bending a l o t during upward loading (Photograph 5.6.). - the bottom bar was p u l l i n g out due to the applied tension. - additional concrete s p a l l i n g off around the connection rebar. 96 Cy 13 +32 kN - f a i l u r e of the connection due to pullout of tension leg (Photograph 5.7.). 98 o cn o o o o o o o o o o o o o o o o o o o o o " I I I I I I I I I ~ I Figure 5 . 6 . Sideways movement of connection (Test #3) . 1 (a) (b) (c) 1 J (d) (e) Figure 5.7. Cracking pattern for Test #3. 100 Photograph 5.6. Bending of top bar (Cycle #11 +40 kN). Photograph 5.7. Connection at f a i l u r e (Cycle #13 +32 kN). 102 5.2.2.2. TEST #4 The load-deflection curve for t h i s monotonic test can be seen in Figure 5.8. The following observations were noted during testing : Cy 01 -85 kN - the upper corner rotated out. - the concrete at the bottom corner of the connection was crushing. - a crack opened up under the bottom corner. Cy 01 -87 kN - bond f a i l u r e around the top leg of the connection rebar. - part of the weld between the connection rebar and the angle broke at the top corner due to the sideways movement of the connection. - the bottom rebar leg bent sideways and down without much crushing of the concrete 1 0 3 Figure 5.8. Load -deflection curve for Test #4. 1 04 5.2.2.3. TEST #5 In th i s test the loading plate was welded to the back side of the angle (Figure 5.9.). This means that the load was applied at a s l i g h t l y d i f f e r e n t location from where i t was applied to the two previous tests. The load-deflection curve for thi s test can be seen in Figure 5.10. The sideways movement of the connection i s shown in Figure 5.11. Photographs 5.8. and 5.9. show the connection after f a i l u r e . The following observations were noted during testing : Cy 02 +45 kN - cracks appeared at the top corner of the connection (Similar to Figure 5.7.(a)). - no v i s i b l e cracking at the bottom corner of the connection. Cy 02 -43 kN - cracks appeared at the bottom corner of the connection (Similar to Figure 5.7.(b)). - the concrete was pushed away from around the edges of the angle. Cy 03 +44 kN - more cracking at the corners of the connection. Cy 04 +46 kN - parts of the concrete had f a l l e n off around the angle (Similar to Figures 5.7.(b) and (c) ) . - the connection could move up and down quite fr e e l y . 105 - could see large twisting of the angle as the connection was loaded up and down. - each cycle caused 10-20 mm long concrete pieces to s p a l l off along the legs of the reinforcing bar which was used for the connection (Photograph 5.8.). Cy 05 +50 kN - the connection f a i l e d by complete exposure of the tension leg of the connection (Photograph 5.9.). 106 Connection rebar 1 Angle b) Test #5 Figure 5.9. Load application points for Rebar #1 with angle. 10 Figure 5.10. Load-deflection curve for Test #5. 1 o S S S P S S 2 S 2 c 2 0 0 0 0 0 0 ° c 3 o o o o i i i i i i i i i Y igure 5 . 1 1 . Sideways movement of the connection (Test #5) . Photograph 5.8. Connection at f a i l u r e (Cycle #5 +50 kN). Photograph 5.9. C o n n e c t i o n at f a i l u r e ( C y c l e #5 +50 kN). 5.2.3. REBAR #2 AT 45 DEGREES WITH SHORT RECESS This connection was similar to the connection in section 5.2.1. The only difference was that a 20 mm x 20 mm x 200 mm recess was cast beside the reinforcing bar along the edge of the specimen. For more d e t a i l s see section 3 • 5« 3 • In t o t a l three specimens were fabricated. Two of these were tested under c y c l i c loading while one of the specimens was tested monotonically. The results were : 5.2.3.1. TEST #7 The load-deflection curve for t h i s test can be seen in Figure 5.12. The sideways movement of the connection i s shown in Figure 5.13. Photographs 5.10. to 5.14. show the connection during testing and at f a i l u r e . The following observations were noted during testing : Cy 03 -78 kN - a small part of the top corner above the recess f e l l o f f . Cy 04 -50 kN - small pieces of concrete at both corners was f a l l i n g off (Photograph 5.10.). Cy 04 -87 kN - a loud cracking sound could be heard from the specimen. - a l o t of concrete was f a l l i n g off at both corners of the connection (Photographs 5.11. and 5.12.). 1 12 Cy 05 +56 kN - the concrete was f a l l i n g off around the tension leg of the connection. - the top rebar leg was bending freely at the corner (Photograph 5.13.). Cy 05 +32 kN - the weld between the connection bar and the loading plate was beginning to f a i l from the corners. - the tension leg of the connection was p u l l i n g out without any increase in the load (Photograph 5.14.). The connection f i n a l l y f a i l e d when the bottom tension leg was t o t a l l y exposed. Figure 5.12. Load-deflection curve for Test #7. Displacement (mm) Photograph 5.10. Cycle #4 -50 kN (Test #7). 116 Photograph 5.11. Cycle #4 -87 kN (Test #7). 1 17 Photograph 5.12. Cycle #4 -87 kN (Test #7). Photograph 5.13. Cycle #5 +56 kN (Test #7). 119 Photograph 5.14. Cycle #5 +32 kN (Test #7). 120 5.2.3.2. TEST #8 The load-deflection curve for thi s test can be seen in Figure 5.14. The sideways movement of the connection i s shown in Figure 5.15. Photographs 5.15. and 5.16. show the connection after f a i l u r e . The following observations were noted during testing : Cy 02 - only a small crack was v i s i b l e at the corners of the connection (Similar to Photograph 5.1.). Cy 05 - the cracks at the corners propagated an additional 5-10 mm. Cy 10 - the corners were cracking further. Cy 11 +70 kN - part of the concrete around the top corner of the connection f e l l off (Similar to Photograph 5.11.). Cy 12 -52 kN - part of the concrete around the bottom corner f e l l o f f . - the bar bent outwards at the end of the weld between the rebar and the loading plate. This made i t possible for the bar at the compression corner to bend further without crushing the concrete (Similar to Photograph 5.13.). Cy 15 +31 kN - part of the concrete spalled off around the bottom leg of the connection. 121 Cy 15 -53 kN - the corner of concrete around the top leg spalled o f f . Cy 16 -38 kN - the top leg of the connection rebar broke (Photograph 5.15.). - at this point the load d i r e c t i o n was reversed. Cy 17 +22 kN - the bottom leg of the connection pulled out (Similar to Photograph 5.7.). Figure 5.14. Load-deflection curve for Test #8. Photograph 5.15. Connection at f a i l u r e (Cycle #16 -38 kN). 125 Photograph 5.16. Connection at f a i l u r e (Cycle #16 -38 kN). 5.2.3.3. TEST #13 The load-deflection curve for th i s monotonic test can be seen in Figure 5.16. The sideways movement of the connection i s shown in Figure 5.17. Photograph 5.17. shows the connection after f a i l u r e . The following observations were noted during testing : Cy 01 +93 kN - a small piece of concrete f e l l off at the top corner of the connection (Similar to Photograph 5.10.). Cy 01 +68 kN - the bottom tension corner f e l l o f f . - more concrete f e l l off at the top corner (Similar to Photograph 5.11.). Cy 01 +42 kN - the connection f a i l e d by pullout of the tension leg of the rebar (Photograph 5.17.) 12 (N:>0 P^OT Figure 5.16. Load-deflection curve for Test #13. 128 Figure 5.17. Sideways movement of connection (Test #13). 129 Photograph 5.17. Connection at f a i l u r e (Cycle #1 +42 kN). 130 5.2.4. REBAR #2 AT 45 DEGREES IN A 75 MM THICK SLAB This connection was similar to the connection in section 5.2.1. The only differences were that the slab was 75 mm thick and i t had a 25 mm x 25 mm recess which was cast along the entire-edge of the specimen. For more d e t a i l s see section 3.5.4. In t o t a l three specimens were fabricated. Two of these were tested under c y c l i c loading while one of the specimens was tested monotonically. The results were : 5.2.4.1. TEST #11 The load-deflection curve for th i s test can be seen in Figure 5.18. Figure 5.19. displays an enlargement of the f i r s t cycles of the load-deflection curve. The sideways movement of the connection i s shown in Figure 5.20. Photograph 5.18. to 5.22. show the connection during the testing and at f a i l u r e . The following observations were noted during testing : Cy 08 -81 kN - small pieces of concrete were f a l l i n g off at the corners of the connection. - could see the connection rebar bending at the 45 degree corners. Cy 10 +77 kN - a concrete piece cracked off at the top corner of the connection. 131 - a large gap could be seen at the corner under the bottom leg of the connection. This was caused by the bending of the rebar. Cy 10 -79 kN - the bottom corner of the connection f e l l off (Photograph 5.18.). Cy 11 +74 kN - the concrete was s p a l l i n g off around the bottom tension leg. - the top bar was bending a l o t at the 45 degree bend (Similar to Photograph 5.13.). Cy 12 +65 kN - a large piece of concrete spalled off along the middle of the connection rebar (Photograph 5.19.). Cy 12 +77 kN - more bending of the leg at the compression side of the bar (Photograph 5..20.). Cy 12 -59 kN - the tension leg of the connection broke at end of weld between the loading plate and the reinforcing bar (Photograph 5.21.). - at this point the d i r e c t i o n of the load was reversed. Cy 13 +67 kN - the tension leg of the connection pulled out (Photograph 5.22.). - only one of the legs was contributing to the load since the other leg broke during the previous cycle. Displacement (mm) 13 Figure 5.19. Enlargement of the load-deflection curve for Test #11. Displacement (mm) OO 135 136 Photograph 5.19. Cycle #12 +65 kN (Test #11). Photograph 5.20. Cycle #12 +77 kN (Test #11). 138 Photograph 5.21. Cycle #12 -59 kN (Test #11). 139 Photograph 5.22. Connection at f a i l u r e (Cycle #13 + 67 kN). 5.2.4.2. TEST #12 The load-deflection curve for t h i s monotonic test can be seen in Figure 5.21. The sideways movement of the connection is shown in Figure 5.22. Photograph 5.25. shows the connection after f a i l u r e . The following observations were noted during testing : Cy 01 +95 kN - a crack developed at the compression corner. Cy 01 +97 kN - a concrete piece (100x150 mm) f e l l off from around the tension corner (Photograph 5.23.). Cy 01 +70 kN - no change in the cracking of the concrete. - the top leg of the connection was bending a l o t . - the bottom leg was straightening out to less than 45 degrees. Cy 01 +75 kN - the top corner of concrete was f a l l i n g off (Photograph 5.24.). - more of the concrete around bottom corner f e l l off . - the weld was f a i l i n g at the bottom corner between the load plate and the connection rebar. - the tension leg broke at the end of the weld (Photograph 5.25.). 1 to o o o o o o o o o o o o o o o o o o o o o I I I I I I I I I ~ Figure 5.21. Load-deflection curve for Test #12. 142 o CO Figure 5.22. Sideways movement of connection (Test #12). Photograph 5.23. Tension corner f a l l i n g off (Cycle #1 +97 kN). 144 S i m \i CVCLE J Photograph 5.24. Top corner f a l l i n g off (Cycle #1 +75 kN). Photograph 5.25. Connection at f a i l u r e (Cycle #1 +75 kN). 5.2.4.3. TEST #14 The load-deflection curve for th i s test can be seen in Figure 5.23. An enlargement of the f i r s t cycles of the load-defl e c t i o n curve i s shown in Figure 5.24. The sideways movement of the connection is displayed in Figure 5.25. Photograph 5.26. shows the connection after f a i l u r e . The following observations were noted during testing : Cy 12 +91 kN - small cracks were v i s i b l e around the top corner of the connection. - a small gap could be seen between the concrete and the legs of the connection at the connection corners. Cy 1 3 + 94 kN - the top corner of the concrete f e l l o f f . cy 1 3 -93 kN - the bottom corner f e l l o f f . Cy 1 4 + 86 kN - movement of the legs at the corners, but no additional cracking. Cy 1 4 + 78 kN - a large piece of concrete f e l l off at the tension corner. Cy 1 4 -69 kN - another piece of concrete f e l l off at the tension side. Cy 1 5 + 62 kN - the top tension leg broke at the end of the weld (Photograph 5.26.). 147 Figure 5.23. Load-deflection curve for Test #14. o o o o o o o o o o o o o o o o o o o o o *-* I I I I I I I I I -< Figure 5.24. Enlargement of the load-deflection curve for Test #14. Figure 5.25. Sideways movement of connection (Test #14). 150 Photograph 5.26. C o n n e c t i o n at f a i l u r e ( C y c l e #15 + 151 5.2.5. ANGLE WELDED TO REINFORCING MESH This connection was made up out of a 50 mm x 50 mm x 200 mm angle which was welded straight to the reinforcing mesh in the specimen. For more d e t a i l s see section 3.5.5. In t o t a l two specimens were fabricated. Both of these were tested under c y c l i c loading. The results were : 5.2.5.1. TEST #9 The load-deflection curve for t h i s test can be seen in Figure 5.26. An enlargement of the f i r s t cycles i s shown in Figure 5.27. The sideways movement of the connection i s displayed in Figure 5.28. Photographs 5.27. to 5.31. show the connection during testing and at f a i l u r e . The following observations were noted during testing : Cy 04 -41 kN - diagonal cracks opened up on both sides of the slab (Photograph 5.27.). - no concrete was f a l l i n g off at load l e v e l s of -40 kN to +40 kN. Cy 07 +51 kN - the diagonal cracks was enlarged and they appeared to go straight through the specimen (Photographs 5.28. and 5.29. the arrows in the photographs indicate the d i r e c t i o n of loading which opened up the cracks). - the angle l i f t e d up part of the concrete above the angle when the load was applied upwards. 1 52 - a gap was opened at the bottom edge of the angle during upward loading. - small pieces of concrete were f a l l i n g off at both ends of the angle. Cy 08 +38 kN - a large piece of concrete f e l l off from the back side of the specimen. Cy 09 -29 kN - the concrete at both ends of the angle had f a l l e n off (Photograph 5.30.). - at thi s point the strength of the connection was generated by bending the reinforc i n g bars which were perpendicular to the connection (Photograph 5.31.). - the connection f i n a l l y f a i l e d when the bars broke away from the angle. 153 Figure 5.26. Load-deflection curve for Test #9. 1 54 igure 5.27. ( K 3 l ) P'Ol Enlargement of the load-deflection Test #9. curve for 1 55 i i i i i i—r o CO - S3 - 8 i—i—r i—i—i—r oo - a a a a O «S O. m o —« - CO - CJ cu I o o o o o o O CO CD CO IO Q O O O O O O O O O O O O O O I I I t I I I I I —> Figure 5.28. Sideways movement of connection (Test #9). 156 Photograph 5.28. Cycle #7 +51 kN (Test 158 Photograph 5.29. Cycle # 7 +51 kN (Test #9). 1 5 9 Photograph 5.30. C y c l e #9 -29 kN (Test #9). Photograph 5.31. Cycle #9 -29 kN (Test #9). 161 5.2.5.2. TEST #15 The load-deflection curve for t h i s test can be seen in Figure 5.29. An enlargement of the f i r s t cycles i s shown in Figure 5.30. The sideways movement of the connection i s displayed in Figure 5.31. The following observations were noted during testing : Cy 05 +52 kN - a diagonal crack opened up straight through the slab from the bottom of the angle (Similar to Photograph 5.27.). Cy 05 -52 kN - a diagonal crack opened up straight through the slab from the top of the angle. - the diagonal cracks opened up one aft e r the other during the next four cycles when the connection was cycled. The cracks appeared s l i g h t l y larger with each cycle (Similar to Photographs 5.28. and 5.29.). Cy 09 +47 kN - a piece of concrete f e l l off at the top corner. Cy 09 -47 kN - a piece of concrete f e l l off at the bottom corner. Cy 10 -17 kN - pieces of concrete f e l l off on back side of the slab (Similar to Photographs 5.30. and 5.31.). - the connection f a i l e d when the reinfo r c i n g bars broke off from the angle. 162 (K3i) p-Bo-r. Figure 5.29. Load-deflection curve for Test #15. 16 o o o o o o o o o o o o o o o o o o o o o •** i i i i i i i i i Y igure 5.30. Enlargement of the load-deflection curve for Test #15. 164 -\ . 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 o CO - 8 -S i a a, *» a m a « o m o. •» Q h CM CM I 8 cp o o o> oo i i i i i i i i i Figure 5.31. Sideways movement of connection (Test #15). 165 5.2.6. REVERSED ANGLE WELDED TO REINFORCING MESH This connection was made up out of a 50 mm x 50 mm x 200 mm reversed angle which was welded straight to the reinforcing mesh in the specimen. For more d e t a i l s see section 3.5.6. In t o t a l two specimens were fabricated. Both of these were tested under c y c l i c loading. The results were : 5.2.6.1. TEST #10 The load-deflection curve for t h i s test can be seen in Figure 5.32. An enlargement of the f i r s t cycles i s shown in Figure 5.33. The sideways movement of the connection i s displayed in Figure 5.34. Photographs 5.32. to 5.37. show the connection during testing and at f a i l u r e . The following observations were noted during testing : Cy 02 +42 kN - a diagonal crack on the front side of the specimen opened up from the top of the angle when the connection was loaded upward. A similar crack developed from the bottom of the angle when the specimen was loaded downwards (Photograph 5.32.). - on the back side of the specimen, the cracks developed from the opposite ends of the angle (Photograph 5.33.). - the corners above and below the angle were also cracking. 166 Cy 04 +60 kN - the cracks opened up more at the higher load (Photograph 5.34.). Cy 06 +56 kN - the corner above the angle cracked off (Photograph 5.35.). - loading down, the bottom corner cracked off (Photograph 5.36.). - the connection twisted quite a b i t when the load was applied. This was caused by the thin metal in the angle. Cy 07 -17 kN - the connection f a i l e d when the bar which was p a r a l l e l to the angle yielded at both ends of the angle (Photograph 5.37.). 167 Figure 5.32. Load-deflection curve for Test #10. 168 KO I i i i i i i i i i 1 i i i i i—i—i—i—i—r T o o o o o o o o o o o o o o o o o o o o o -* I I I I I I I I I V (.M3T) P ^ o - i igure 5.33. Enlargement of the load-deflection curve for Test #10. 169 CM a a a a o o o o o o o co IN e fi O O O O O •* n w - i O O O Q CM CO ^ I I I I O O O o o o 10 co t» co 5 o I I I I I *p (K>t) P-BOT; Figure 5.34. Sideways movement of connection (Test #10). Photograph 5.32. Cycle #2 +42 kN (Test #10). Photograph 5.33. Cycle #2 +42 kN (Test #10). Photograph 5.34. Cycle #4 +60 kN (Test #10). Photograph 5.35. Cycle #6 +56 kN (Test #10). 174 Photograph 5.36. Cycle #6 +56 kN (Test #10). Photograph 5.37. Cycle #7 -14 kN (Test #10). 176 5.2.6.2. TEST #16 The load-deflection curve for t h i s test can be seen in Figure 5.35. An enlargement of the f i r s t cycles i s shown in Figure 5.36. The sideways movement of the connection i s displayed in Figure 5.37. Photograph 5.38. shows the connection at f a i l u r e . The following observations were noted during testing : Cy 06 +60 kN - some small pieces of. concrete f e l l off at both ends of the angle. - small diagonal cracks showing up on both sides of the specimen. Cy 07 +72 kN - larger cracks opened up on both sides of the slab. The cracks behaved s i m i l a r l y to the cracks in the previous specimen (Similar to Photographs 5.32. and 5.33.). Cy 08 +72 kN - more large cracks opened up on both sides of the specimen. - the side of the angle that was welded to the loading plate twisted in and out when the load was cycled. Cy 10 +62 kN - the corners above and below the angle cracked off (Similar to Photographs 5.35. and 5.36.). Cy 11 +42 kN - the bar p a r a l l e l to the angle yielded at both ends of the angle. a large piece of concrete f e l l off on back side of the specimen around the reinforcing bars (Photograph 5.38.). 178 Figure 5.35. Load-deflection curve for Test #16. 180 CM ( K 3 1 ) P W O T C Figure 5.37. Sideways movement of connection (Test #16). 181 Photograph 5.38. Cycle #11 + 42 kN (Test #16). 182 CHAPTER 6 DISCUSSION, CONCLUSIONS AND FUTURE SCOPE 6.1. INTRODUCTION This investigation was i n i t i a t e d to develop improved d e t a i l s and ra t i o n a l methods of analysis for a connection for use in a thin flange of an element in an earthquake resistant b u i l d i n g . It was assumed during the testing of the connections that the flange remained e l a s t i c while the non-line a r behaviour was r e s t r i c t e d to the connection. The results of the investigation are summarized and discussed in thi s chapter. The conclusions presented are based on the experimental observations of the behaviour of the connections during testing and the measured forces and deformations. The recommendations for further study are also based on the results discussed here. 183 6.2. DISCUSSION 6.2.1. BEHAVIOUR OF THE CONNECTIONS DURING CYCLIC  LOADING The c y c l i c testing of each connection began with several cycles of loading within the e l a s t i c range. Loading in this range, up to about 80 % of the predicted strength of the connection, appeared to cause only minor cracking close to the corners of the connections. In each case, these i n i t i a l cycles produced stable and narrow load-deflection curves and did not appear to reduce the strength of any of the connections (Figures 5.18. and 5.23.). The connections f a i l e d either by bar f a i l u r e or by sp a l l i n g when the connections were loaded into the i n e l a s t i c range. The bar f a i l u r e s generally occurred at very small deflections (Figures 5.1. and 5.2.). For the connections that f a i l e d by sp a l l i n g , the s t i f f n e s s f e l l gradually while the width of the load-deflection curves increased (Figures 5.5. and 5.14.). Successive cycles of i n e l a s t i c displacement for these connections caused increased cracking which was followed by a progressive loss of concrete around the embedded bar. The bar carrying compression tended to buckle at t h i s point and became i n e f f e c t i v e . This gave some of the connections a reduced but r e l a t i v e l y stable strength with increasing i n e l a s t i c displacements (Figures 5.5., 5.14. and 5.29..). However, other connections showed r e l a t i v e l y l i t t l e loss of strength with increased displacement (Figure 5.18.). 184 The connections that were welded straight to the specimen's reinforcing mesh lost most of their load carrying capacity af t e r a few cycles in the i n e l a s t i c range (Figures 5.26. and 5.32.). At t h i s point the specimens were also severely cracked. In general, the connections fabricated from embedded rebar with the lower y i e l d stress (Rebar #2) behaved better than similar connections fabricated from rebar with higher y i e l d stress (Rebar #1). Both of the above connections also behaved better than those in which the reinforci n g mesh was used as part of the connection. 6.2.2. BEHAVIOUR OF CONNECTIONS DURING MONOTONIC  LOADING Three of the connections were tested monotonically (see Tabie 6.1.). The maximum load for these connections was reached just before the concrete around the compression corner of the connection cracked off (see Photograph 5.2.). The load-deflection curves for these connections had a f a l l i n g branch which was formed during increased i n e l a s t i c displacement. The strength of the connections remained at a le v e l of about 60 % of the maximum strength u n t i l a def l e c t i o n of about 10 to 20 times the deflection at the maximum strength was reached (Figure 5.21.). None of the connections that were welded straight to the reinforcing mesh were tested monotonically. Of the other connections, the connections that were fabricated from the rebar with the lower y i e l d stress (Rebar #2) again behaved Table 6.1. MAXIMUM CONNECTION STRENGTHS POSITIVE NEGATIVE CYCLES # OF CONNECTION TEST STRENGTH STRENGTH AT CYCLES TYPE # y 1 (kN) (kN) i _ 1 MAXIMUM (.__ 1 IN TEST iREBAR # 1 1 90.64 -93.94 2 3 2 88.60 -50.32 1 1 6 60.85 -56.41 4 5 REBAR # 1 3 69.31 -67.33 10 1 3 WITH ANGLE 4 0.00 -89.92 1 MONOTONIC 5 50.36 -46.26 6 6 REBAR # 2 7 88.76 -77.19 4 5 8 70.52 -85.31 1 1 1 6 1 3 93.93 0.00 1 MONOTONIC REBAR # 2 1 1 81 .62 -81.34 8 1 3 7 5 mm SLAB 1 2 97.57 0.00 1 MONOTONIC 1 4 93.60 -92.99 13 1 5 ANGLE TO 9 50.99 -73.66 8 9 MESH 1 5 59.82 -65.26 9 1 1 REV. ANGLE 10 60.66 -63.36 4 7 TO MESH 16 76.45 -73.49 9 1 1 better than the connections fabricated from the rebar with the higher y i e l d stress (Rebar #1). 6.2.3. STRENGTH OF CONNECTIONS 6.2.3.1. MAXIMUM MEASURED STRENGTHS The maximum measured posit i v e and negative strengths for the d i f f e r e n t connections are given in Table 6.1. A comparison of these strength values shows the the maximum and minimum strengths are usually similar in magnitude. The table does not show a large v a r i a t i o n between the maximum strengths of the connections that were fabricated from the two di f f e r e n t strength rebars. The table also shows the cycle during which the maximum strengths occurred and the number of cycles to f a i l u r e . A comparison of these cycle numbers and the maximum strengths for each connection shows that while the y i e l d stress does not seem to have a large eff e c t on the connection strength i t does appear to affe c t the number of cycles to f a i l u r e . The connections made from Rebar #2, which had the lower y i e l d stress appeared to be able to sustain more loading cycles at lower load l e v e l s , after the maximum strengths had been reached, than those made from Rebar #1. 187 6.2.3.2. STRENGTHS AT FAILURE The usable connection strengths at f a i l u r e can be seen in Table 6.2. The usable connection strength was estimated as follows. If the connection f a i l e d suddenly, the strength at f a i l u r e for loading in one di r e c t i o n was taken as the load measured immediately before f a i l u r e and the strength for loading in the other di r e c t i o n was taken as the load at the end of the previous half cycle. If the connection f a i l e d gradually, one value of strength at f a i l u r e was taken as the strength at the end of the la s t cycle in which the capacity was s t i l l increasing when the deflection went through zero and the other was taken as the strength at the end of the previous half cycle. The monotonic tests were considered to have f a i l e d when the connection rebar broke or when the strength had f a l l e n to 50 % of the maximum strength. Table 6.2. also shows the number of cycles to f a i l u r e for the connections and the number of cycles to maximum strength. 6.2.3.3. CALCULATED STRENGTHS The calculated shear strengths for the two d i f f e r e n t types of connections were obtained by the following two procedures. 6.2.3.3.1. SEPARATE REBAR AT 45 DEGREES. A procedure for c a l c u l a t i n g the shear strength of t h i s type of connection i s given in [1] and [2]. As shown in 188 Table 6.2. MEASURED CONNECTION STRENGTH AT FAILURE STRENGTH PREVIOUS CONNECTION TEST CYCLES FAILURE CYCLES AT MAXIMUM TYPE # AT TYPE TO FAILURE STRENGTH MAXIMUM FAILURE (kN) (kN) REBAR # 1 1 2 BAR 3 74.35 -77.70 2 1 BAR 1 -50.32 88.60 6 4 BAR 5 59.90 -56.41 REBAR # 1 3 10 SPALLING 13 32.26 -43.86 WITH ANGLE 4 MONO. SPALLING 1 -51.00 5 6 SPALLING 6 50.36 -46.26 REBAR # 2 7 4 SPALLING 5 -56.14 88.75 8 1 1 BAR 16 22.26 -37.40 1 3 MONO. SPALLING 1 41 .72 REBAR # 2 1 1 8 SPALLING 12 77.00 -69.93 75 mm SLAB 1 2 MONO. BAR 1 73.09 1 4 13 BAR 15 61 .77 -69.30 ANGLE TO 9 8 BAR 8 32.60 -73.66 MESH 15 9 BAR 10 42.34 -65.26 REV. ANGLE 10 4 BAR 6 -55.79 55.56 TO MESH 1 6 9 BAR 1 1 44.54 -63.53 189 Figure 6.1., each leg contributes to the strength of the connection. If each leg i s assumed to y i e l d when the connection f a i l s , then C = T = A b * F y [1] where A^ i s the cross sectional area of the rebar and F v i s the y i e l d strength. When the two legs are at 45 degrees, the strength of the connection V n i s : V n = Sqrt(2) * Afa * F y [2] However, t h i s approach for ca l c u l a t i n g the strength must be modified when the concrete i s lost around the legs by s p a l l i n g . At t h i s point the compression leg becomes in e f f e c t i v e and the expected shear strength i s 0.5 * V n. Two separate values were calculated and plotted on the load-deflection curves for t h i s type of connection. These were denoted Fy and Fyw (see section 5.1. for d e f i n i t i o n of the terms). It has been suggested by Spencer [4] that the above approach must be modified for a connection with an angle welded to the connection rebar. Before the connection can f a i l the concrete bearing against the end of the angle must be pushed o f f . The force Fp required to f a i l t his concrete was found to be given by F p = A p * f c [3] Figure 6.1. Forces in rebar connection. 191 where Ap i s the area of the recess formed by the angle and f c i s the concrete cylinder strength. The shear strength under monotonic loading w i l l therefore be given by V n = Sqrt(2) * A b * Fy + A p * f c [4] However, t h i s approach does not seem to be v a l i d for thin panels. For the panels studied in thi s investigation, the strength of the connections that had an angle welded to the rebar was predicted with formula [2], This can be explained by the fact that after only a few cycles the concrete at the ends of the angle had cracked off and the connection worked as a normal rebar connection. 6.2.3.3.2. CONNECTIONS WELDED TO REINFORCING  MESH. During the testing of the specimens which used these connections, i t was observed that cracks developed straight through the specimens. These cracks started at the ends of the connections and extended from the edge of the flange towards the thickened portion of the specimen, which represents the web. The cracks appear to be consistent with a f l e x u r a l mode of f a i l u r e , with the maximum t e n s i l e s t r a i n at the edge of the flange and the neutral axis at flange-web junction. Some of the cracks that developed during the testing extended from the end of the connections straight across the flange to the web. These cracks were c a l l e d type A. Other cracks were seen to extend from the end of the connections across the web at approximately 45 degrees. These cracks were c a l l e d type B. Pauley, P r i e s t l e y and Synge [13] reported similar diagonal crack development during testing of squat shearwalls. Types A and B represent the two extreme cracking cases. 6.2.3.3.2.1. TYPE A The procedure for cal c u l a t i n g the connection shear strength for type A cracking i s given in [5] and shown in Figure 6.2. as Ca = Pi*Fy*(Db/2)*2*((W-A1)+(W-A2)+...)/W) [5] where Fy i s the y i e l d strength, i s the diameter of the bar, W i s the panel width and A-| , A 2, ... are the distances to the reinforcing bars from the panel edge. The bars p a r a l l e l to the connection which were part of the re i n f o r c i n g mesh extended across these types of cracks. The above formula i s derived by assuming that a l l the bars crossing the crack y i e l d and that the y i e l d strength of the steel i s F v . 6.2.3.3.2.2. TYPE B The procedure for ca l c u l a t i n g the shear strength of thi s type of connection i s given in [6] and shown in Figure 6.3. as 193 Ca = Pi *F y * (D b/2)~2 * ((A, + A 2 + A3 + A 4)/W) Ca = Connection capacity Fy = Steel y i e l d stress D b = Diameter of rebar Figure 6.2. Forces in connection welded to reinforcing mesh (perpendicular cracking). 194 Cb = Pi *F y * (D b/2)~2 * ((B, + B 2 + B 3 + B 4)/W) Cb = Connection capacity F y = Steel y i e l d stress D b = Diameter of rebar Figure 6 . 3 . Forces in connection welded to reinforcing mesh (diagonal cracking). 195 Cb = Pi*F y*(D b/2r2*( (L-B 1 ) + (L-B 2) + . . . )/W) [6] where F y is the y i e l d strength, D b i s the diameter of the bar, W i s the distance from the edge of the flange to the point where the diagonal compression strut is assumed to meet the web and B 1, B2, ... are the distances from the reinforcing bars from th i s point. The above formula i s derived by assuming that a l l the bars crossing the crack y i e l d and that the y i e l d strength of the steel i s F y . However, i t i s assumed that bars p a r a l l e l to a connection do not contribute to i t s strength. This i s a conservative assumption which i s l i k e l y to be true when there are a number of connections along a flange. 6.2.3.3.3. CALCULATED CONNECTION STRENGTHS The calculated strengths and the measured maximum strengths are shown in Table 6.3. and 6.4. Table 6.3. displays the strength values obtained from using formula [2] for both Fy and Fyw. Both calculated values are also plotted on the load-deflection curves for each test (section 5.2.). Table 6.4. shows the values obtained by formulas [5] and [6]. Figure 6.4. shows the rela t i o n s h i p between the measured strength and the calculated strength for the connections (using Fy and Ca). Figure 6.5. shows the relationship between the measured strength and the calculated strength for the connections (using Fyw and Cb). Figure 6.6. shows Table 6.3. CALCULATED CONNECTION STRENGTHS CONNECTION TYPE TEST # POSITIVE STRENGTH (kN) NEGATIVE STRENGTH (kN) STRENGTH USING Fy (kN) STRENGTH USING Fyw (kN) REBAR # 1 1 90.64 -93.94 84.34 98.50 2 88.60 -50.32 84. 34 98.50 6 60.85 -56.41 84.34 98.50 REBAR # 1 WITH ANGLE 3 4 69.31 0.00 -67.33 -89.92 84.34 84.34 98.50 98.50 5 50.36 -46.26 84.34 98.50 REBAR # 2 7 88.76 -77. 1 9 57.70 69.20 8 70.52 -85.31 57.70 69.20 1 3 93.93 0.00 57.70 69.20 REBAR # 2 7 5 mm SLAB 1 1 1 2 81 .62 97.57 -81.34 0.00 57.70 57.70 69.20 69.20 1 4 93.60 -92.99 57.70 69.20 Table 6.4.  CALCULATED CONNECTION STRENGTHS POSITIVE NEGATIVE STRENGTH STRENGTH CONNECTION TEST STRENGTH STRENGTH USING Ca USING Cb TYPE # (kN) (kN) (kN) (kN) ANGLE TO 9 50.99 -73.66 39.20 62.02 MESH 1 5 59.82 -65.26 39.20 62.02 REV. ANGLE 10 60.66 -63.36 59.00 53.00 TO MESH 16 76.45 -73.49 59.00 53.00 Measured versus Calculated strength Calculated strength (kN) • Rebar #1 + Rebar #1 with Angle o Rebar #2 vr. Recess A Rebar #2 75 mm X Angle to mesh V Rev. angle to mesh M e a s u r e d v e r s u s C a l c u l a t e d s t r e n g t h Calculated strength (kN) • Rebar #1 + Rebar #1 -with Angle o Rebar #2 yr. Recess A Rebar #2 75 m m X Angle to mesh V Rev. angle to mesh Failure versus Calculated strength Using Fy or Ca 0 20 40 60 80 100 Calculated strength (kN) + Rebar #1 -with Angle o Rebar #2 YT. Recess X Angle to mesh v" Rev. angle to mesh 201 the re l a t i o n s h i p between the f a i l u r e strength and the calculated strength for the connections (using Fy and Ca). Figure 6.7. shows the rel a t i o n s h i p between the f a i l u r e strength and the calculated strength for the connections (using Fyw and Cb). By comparing the values in the table and examining the graphs and the load-deflection curves for a l l the tests, i t can be seen that the models used for the calculations work reasonably well. However, as pointed out in section 6.2.3.1., the agreement was not good enough to confirm that the maximum strengths depended on the value of F y. The models described above are considered to be adequate to predict the actual maximum strength of the connections when used with the actual F y. 6.2.3.4. INFLUENCE OF CONCRETE STRENGTH ON  CONNECTION STRENGTHS Figure 6.8. shows a graph of the cylinder strengths and the connection strengths. An examination of the graph reveals no direct relationship between the two strengths for the specimens which had an embedded reinforcing for the connection in addition to the mesh in the panel. However, the graph seems to indicate some relationship between the plotted strength values for the two types of connections that were welded straight to the reinforcing mesh. This r e l a t i o n s h i p can be inferred since the te n s i l e resistance of the concrete could contribute to the strength, as a Fai lure versus Calculated s trength Using F J T V or Cb • A Calculated strength (kN) Rebar #1 + Rebar #1 with Angle o Rebar #2 nr. Recess Rebar #2 75 mm X Angle to mesh V Rev. angle to mesh M O lO. C fD cn co O O 3 3 fD O r t O 3 W r r n fD 3 i Q r r 3* cn < fD i-l cn c cn O 3 & fD r-t in r t fD 3 lO r r 3* cn 100 90 60 50 30 C o n n e c t i o n s t r e n g t h v s . C y l i n d e r s t r . 80 H 70 H 40 H -Metve lorvtc-Monotonic Monotonic ijebar #1 Rebar #2 75mm A n g l e Rebar #1 TV. Angle Rebar #2 ReY. Anglle —I i 1 1 1 1 1 I 1 1 1 l I I "I 1 1 1 1 1 1 2 6 3 4 5 7 8 9 11 15 16 10 12 13 14 • Cylinder str. (MPa) Mix g A Connection st. (kN) to o CA) 204 secondary e f f e c t , during the tests of these connections. 6.2.4. TYPES OF CONNECTION FAILURES The connection f a i l u r e s can be divided into two general types. The f i r s t and most common type was the bar f a i l u r e , while the second was f a i l u r e caused by s p a l l i n g . The f a i l u r e types for each of the tested connections are l i s t e d in Table 6.2. The bar f a i l u r e s for Rebar #1 occurred suddenly, without warning, at r e l a t i v e l y high loads and with a very small connection displacement. This sudden f a i l u r e was most l i k e l y caused by the high y i e l d stress of the rebar which made the rebar very b r i t t l e . The deflections at f a i l u r e for the tests can be found in Table 6.5. The bar f a i l u r e s for Rebar #2 took place aft e r the connection rebar had been bent back and forth several times. This indicates that this rebar was much more d u c t i l e which means that the bar f a i l u r e s would occur at much higher deflections and usually not before some concrete had spalled off from around the corners of the connections. Spalling f a i l u r e s occurred only in the connections which were made from a separate rebar. The f a i l u r e s were a result of one of the connection legs p u l l i n g out from the specimen (see Photograph 5.7.). This type of f a i l u r e happened gradually with more and more of the connection legs exposed with each cycle. The connections f a i l e d when the las t piece of the reinforcing leg pulled out. Table 6.5. MOVEMENTS OF CONNECTIONS CONNECTION TYPE TEST '# FAILURE TYPE DEFLECTION AT FAILURE (mm) MAXIMUM DEFLECTION (mm) SIDEWAYS MOVEMENT FAILURE (mm) MAXIMUM SIDEWAYS MOVEMENT (mm) REBAR # 1 1 BAR 1.45 1.45 2 BAR -0.42 -0.92 6 BAR -0.03 -3.09 1.25 5.83 REBAR # 1 WITH ANGLE 3 4 SPALLING SPALLING 9.43 -9.02 10.22 -9.02 26.44 26.44 5 SPALLING 5.86 10.95 11.35 25.18 REBAR # 2 7 BAR -0.06 3.44 4.11 18.39 8 BAR 5.64 -8.43 16.21 23.38 13 SPALLING 15.97 17.52 19.64 19.64 REBAR # 2 75 mm SLAB 11 12 SPALLING BAR 11.88 14.35 12.28 14.35 9.04 17.68 16.22 17.68 14 BAR 12.20 14.48 6.42 11.12 ANGLE TO MESH 9 15 BAR BAR 5.38 6.13 11.23 15.82 1.96 2.06 12.00 16.24 REV. ANGLE TO MESH 10 16 BAR BAR -0.18 -0.01 5.53 -6.49 -0.03 -6.75 -1.30 -7.55 206 The connections that were welded straight to the rein f o r c i n g mesh f a i l e d when the steel in the reinforcing mesh yielded in the region around the connections. The f a i l u r e did not however take place before there had been extensive cracking around the connections and considerable loss of concrete. Diagonal cracks also developed from the connection corners straight through the specimen (see Photographs 5.28. and 5.29.). These cracks extended a l l the way towards the back of the specimens. 6.2.5. DISPLACEMENT OF THE CONNECTIONS The displacement capacity of the connections should be compared to the required displacement capacity in an earthquake which w i l l depend on the strength of the building and the intensity of the earthquake. The calculated response of a one story panel structure with a shear wall connection model subjected to a moderately severe earthquake i s discussed by Spencer and Tong [5]. They found that the maximum displacement at roof l e v e l varied from about 15 mm for a structure with strong shear walls in which the connections remained e l a s t i c to 40 mm for a building with weaker walls in which the connections were loaded well into the i n e l a s t i c range. It was also assumed that strength of the connections could f a l l to about 50 % of the o r i g i n a l value at maximum displacement. Spencer and Tong [5] showed that the shear displacement v between the wall panels was v = x * (b/h) [7] where b was the panel width, h was the height to roof l e v e l and x was the displacement at roof l e v e l . The displacement of an individual connection w i l l then be 0.5 * v. For the building they analysed, the r a t i o of b/h was 0.44 which meant that the largest displacement capacity required for a connection would be approximately 0.44*0.50*40 mm or about 9.0 mm. The above displacement can be compared to the connection displacement at f a i l u r e and the maximum connection displacement which can be found in Table 6.5. The values show that the connections in the 75 mm specimens could be used in buildings in which the connections would be loaded into the i n e l a s t i c range in an earthquake. The rebar with the lower y i e l d stress (Rebar #2) behaved again also better than the rebar with the higher y i e l d stress(Rebar #1 ). 6.2.6. SIDEWAYS MOVEMENT OF THE CONNECTIONS The sideways movement of the connections at f a i l u r e and at the maximum movement can be seen in Table 6.5. The sideways movement i s also plotted versus the load for each test and can be found in section 5.2. The graphs show very small sideways movements during the i n i t i a l cycles of the tes t s . However, with increasing number of cycles the sideways movement was growing. This is mainly the result of 208 the thin cover around the connection reinforcing bars which offers minimal restraint against sideways movement of the connection under c y c l i c loading. Another reason for the sideways movement could be the fact that the connection rebar was not placed exactly in the middle of the specimens. The bars in the reinforcing mesh and the connection rebar were together centered in the specimen (see Figure 3.9.). This meant that the connection rebar was placed s l i g h t l y off center. 6.2.7. CONNECTION DETAILS 6.2.7.1. PANEL THICKNESS A comparison of the results obtained for the 75 mm thick specimens with the results for the 50 mm thick specimens indicates that the thicker specimens behave better during loading cycles. The tests show that the thicker specimens lose less concrete around the connection re i n f o r c i n g . This is mainly due to the thicker cover over the reinfo r c i n g bars. The connections in the thicker specimens can also maintain a higher load l e v e l without f a i l u r e and the connections experience much lower sideways movement. 209 6.2.7.2. PANEL RECESS The type of connection that had a short recess beside the connection reinforcing appeared to perform better than the other types of connections during the t e s t i n g . The short recess also works better than the longer one since t h i s preserves the concrete at the connection corners. Another function of the recess is to expose the side of the rebar so the specimen can e a s i l y be welded to another specimen. 6.2.7.3. ANGLE WELDED TO CONNECTION The main function of the angle welded to the connection reinforcing i s to make the attachment to an adjacent connection easier. The angle does not however add any strength to the connection since the concrete around the angle w i l l be spalled off after the f i r s t few cycles. 6.3. CONCLUSION a) Good quality control of the steel that goes into the connections i s e s s e n t i a l . B r i t t l e f a i l u r e s of the connections, which are more l i k e l y with high y i e l d strength bars, are not desirable. b) A thicker flange w i l l behave better during earthquake loading since there i s more cover around the reinf o r c i n g . A thicker flange is also less f r a g i l e during transport. 210 c) The design strength of a connection should be taken as 50 % of the expected maximum capacity of the connection. The spacing between connections can be reduced in order to reduce the load on each connection. d) Loading cycles in the e l a s t i c range do not seem to reduce the capacity of the connections. e) The connections with a recess in the panel edge and embedded bars at 45 degrees appear to perform best under simulated earthquake loading. These connections can also be ea s i l y attached to an adjacent connection. f) Connections in which an angle i s welded d i r e c t l y to the reinforcing mesh are not recommended since they damage the reinforcing s t e e l and cause severe cracking in the flange i t s e l f . g) The models used to predict the connection strength are adequate for predicting the lower l i m i t of the connection strengths. 6.4. FUTURE SCOPE The research described in thi s d i s s e r t a t i o n covers only a small part of the overall objective of predicting the behaviour of precast concrete buildings under earthquake loading. Based on the results of t h i s study, the following suggestions are made for further research on the behaviour of embedded rebar connections in thin flanges. 21 1 a) Tests should be performed where the connections were cycled at lower loads for a longer period of time. This would show the effect of fatigue on the connections. b) A large number of similar connections should be tested in order to establish a strength d i s t r i b u t i o n . c) Connections should be tested under c y c l i c loading while a perpendicular force i s simultaneously applied to the connection. This perpendicular force should be applied perpendicular to the plane of the flange, and in the plane of the flange, perpendicular to the edge (see forces and P 2, Figure 6 . 9 . ) . The effect of the combined application of these forces should also be investigated. Both the a b i l i t y of the connection to r e s i s t these forces and their effect on the shear capacity are important and need to be better understood. 2 1 2 Figure 6.9. Forces perpendicular to connection. 213 REFERENCES 1. S.V. Polyakov et a l . . Investigation into Eartquake Resistance of Large Panel Buildings. Proc. Fourth World Conference on Earthquake Engineering, Santiago, Chile 1969, Vol. 1. 2. PCI Design Handbook. 3rd Ed., Prestressed Concrete Ins t i t u t e , Chicago, 1985. 3. Metric Design Handbook. 1st Ed., Canadian Prestressed Concrete In s t i t u t e , Ottawa, 1982. 4. R.A. Spencer. Earthquake Resistant Connections for Low Rise Precast Concrete Buildings. Seminar on Precast Concrete Construction in Seismic Zones, Tokyo, October 27-31, 1986, pp. 61-81. 5. R.A. Spencer and W.K.T. Tong. Design of a One-Story Precast Concrete Building for Earthquake Loading. Proc. Eight World Conference on Earthquake Engineering, San Francisco 1984, pp. Vol V, pp. 653-660. 6. J.M. Becker and C. Lorente. Seismic Design of Precast Concrete Panel Buildings. Proc. Workshop on Earthquake Resistant Reinforced Concrete Building Construction, University of C a l i f o r n i a at Berkeley, July, 1977. 7. L.D. Martin and W.J. Korkosz. Connections for Precast Prestressed Concrete Buildings. Technical Report No. 2, Prestressed Concrete I n s t i t u t e , Chicago, March, 1982. 8. A. Aswad. Selected Precast Connections: Low-Cycle Behaviour and Strength, Proc. 2nd U.S. National Conference on Earthquake Engineering, Stanford University, C a l i f o r n i a , August 22-24, 1979. 9. W.F. Dawson and M. Shemie. Bolted Connections as a Substitute for on Site Welding and Wet Joints in Precast Concrete. Proc. Canadian Structural Concrete Conference, Ottawa 1977, pp. 269-289. 10. R.A. Spencer and D.S. N e i l l e . C y c l i c Tests of Welded HeadedStud Connections. PCI Journal, May-June, 1976, Vol. 15, No. 1, pp. 67-78. 214 11. N.M. Hawkins. State-of-the-Art Report on Seismic Resistance of Prestressed and Precast Concrete Structures: Part 2 - Precast Concrete, PCI Journal, January-February, 1978, Vol. 23, No. 1, pp. 40-58. 12. R.P. Saxena. An Experimental Investigation of a Precast Concrete Connection. M.Ap.Sc. Thesis, Department of C i v i l Engineering, U.B.C., September 1983. 13. T. Paulay, M.J.N. Pr i e s t l e y and A.J. Synge. D u c t i l i t y in Earthquake Resisting Squat Shearwalls. ACI Journal, July-August, 1982, Vol. 79, No. 4, pp. 257-269. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0062538/manifest

Comment

Related Items