<%BANNER%>

FRP/Steel Strengthening of Unreinforced Concrete Masonry Piers


PAGE 1

FRP/STEEL STRENGTHENING OF UNREINFORCED CONCRETE MASONRY PIERS By VANESSA E. GRILLO A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2003

PAGE 2

Copyright 2003 by Vanessa E. Grillo

PAGE 3

ACKNOWLEDGMENTS Acknowledgments go to the National Science Foundation and the Marketing Development Alliance for the Composites Industry for their financial support of this research. I would like to thank my graduate advisor, Dr. H. R. Hamilton, III, for all his guidance and support during my graduate studies. I would also like to thank the rest of my committee, Dr. Gary R. Consolazio and Dr. Ronald A. Cook, for their support. Most importantly, I would like to thank my parents for their love and guidance throughout my academic career. iii

PAGE 4

TABLE OF CONTENTS Page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES.............................................................................................................x ABSTRACT.....................................................................................................................xix INTRODUCTION...............................................................................................................1 Masonry and Earthquakes.............................................................................................1 FRP Strengthening........................................................................................................2 FRP Composites with a Ductile Connection................................................................3 EXPERIMENTAL PROGRAM..........................................................................................5 Test Specimens.............................................................................................................5 Material Properties........................................................................................................6 FRP Composite Configuration.....................................................................................6 Reinforcing Steel Placement.......................................................................................11 Test Setup...................................................................................................................13 Test Procedures...........................................................................................................15 EXPERIMENTAL RESULTS...........................................................................................17 Observed Behavior.....................................................................................................17 Load-Displacement Envelopes...................................................................................29 System Ductility.........................................................................................................30 Computing Predicted Capacities.................................................................................35 CONCLUSIONS................................................................................................................39 APPENDIX A EXTENDED LITERATURE REVIEW.....................................................................41 A. E. Schultz and R. S. Hutchinson (2001)................................................................41 O. S. Marshall and S. C. Sweeney (2002)..................................................................41 F.L. Moon, T. Yi, R.T. Leon and L.F. Kahn (2002)...................................................43 iv

PAGE 5

J. G. Tumialan, A. San Bartolome and A. Nanni (2003)............................................44 M. J. Chajes, W. W. Finch, Jr., T. F. Januszka and T. A. Thompson, Jr. (1996).......45 K. Bajpai and D. Duthinh (2003)................................................................................46 C. Sittipunt, S.L. Wood, P. Lukkunaprasit and P. Pattarattankul (2001)...................47 T.C. Triantafillou (1998)............................................................................................48 M.J.N. Priestley and F. Seible (1995).........................................................................50 A.M. Holberg and H.R. Hamilton III (2001)..............................................................52 G.M. Calvi, G.R. Kingsley and G. Magenes (1996)..................................................53 M.R.Ehsani, H. Saadatmanesh and J.I. Velazques-Dimas (1999)..............................54 D.P. Abrams (2001)....................................................................................................55 Additional Literature Review.....................................................................................56 B EXPERIMENTAL PROGRAM.................................................................................65 Test Setup...................................................................................................................65 Data Acquisition.........................................................................................................73 C SPECIMEN CONSTRUCTION.................................................................................80 Material Properties......................................................................................................82 FRP Application.........................................................................................................85 Computing Predicted Capacities.................................................................................89 Individual Specimen Details.......................................................................................91 D SPECIMEN RESULTS............................................................................................118 CMU 1......................................................................................................................118 CMU 2......................................................................................................................133 CMU 3......................................................................................................................143 CMU 4......................................................................................................................154 CMU 5......................................................................................................................164 CMU 6......................................................................................................................176 CMU 7......................................................................................................................185 CMU 8......................................................................................................................194 E REVIEW OF FEMA 273..........................................................................................205 F DATA TABLES.......................................................................................................214 Calculated and Measured Lateral Capacities............................................................214 Calculating m-values................................................................................................218 LIST OF REFERENCES.................................................................................................220 BIOGRAPHICAL SKETCH...........................................................................................222 v

PAGE 6

LIST OF TABLES Table page 1 Material Properties.....................................................................................................7 2 Details of FRP Composite Configuration..................................................................9 3 Steel Reinforcement Details.....................................................................................11 4 Summary of Results.................................................................................................20 5 Calculated m-factors................................................................................................33 6 Calculated Displacement Ductility...........................................................................34 7 Measured and Calculated Lateral Capacities...........................................................37 8 Full Prism Compression Test Results......................................................................82 9 Half Prism Compression Test Results......................................................................82 10 Full Stretcher Unit Compression Test Results.........................................................83 11 #3 Reinforcing Bar Tension Test Results................................................................83 12 #4 Reinforcing Bar Tension Test Results................................................................84 13 Masonry Joint Reinforcement Tension Test Results................................................84 14 Masonry Joint Reinforcement Weld Shear Strength Test Results...........................84 15 FRP Composite Coupon Tension Test Results........................................................85 16 CMU 1 FRP Quantities............................................................................................93 17 Critical Dates for CMU 1.........................................................................................94 18 Conditions for CMU 1..............................................................................................94 19 CMU 2 FRP Quantities............................................................................................97 20 Critical Dates for CMU 2.........................................................................................97 vi

PAGE 7

21 Conditions for CMU 2..............................................................................................97 22 CMU 3 FRP Quantities..........................................................................................101 23 Critical Dates for CMU 3.......................................................................................102 24 Conditions for CMU 3............................................................................................102 25 CMU 4 FRP Quantities..........................................................................................106 26 Critical Dates for CMU 4.......................................................................................107 27 Conditions for CMU 4............................................................................................107 28 CMU 5 FRP Quantities..........................................................................................109 29 Critical Dates for CMU 5.......................................................................................110 30 Conditions for CMU 5............................................................................................110 31 CMU 6 FRP Quantities..........................................................................................111 32 Critical Dates for CMU 6.......................................................................................112 33 Conditions for CMU 6............................................................................................112 34 CMU 7 FRP Quantities..........................................................................................114 35 Critical Dates for CMU 7.......................................................................................115 36 Conditions for CMU 7............................................................................................115 37 CMU 8 FRP Quantities..........................................................................................117 38 Critical Dates for CMU 8.......................................................................................118 39 Conditions for CMU 8............................................................................................118 40 Calculated Values for CMU 1................................................................................119 41 Testing Observations for CMU 1...........................................................................119 42 Summary of Results for CMU 1............................................................................132 43 Calculated Values for CMU 2................................................................................133 44 Testing Observations for CMU 2...........................................................................134 45 Summary of Results for CMU 2............................................................................142 vii

PAGE 8

46 Calculated Values for CMU 3................................................................................143 47 Testing Observations for CMU 3...........................................................................144 48 Summary of Results for CMU 3............................................................................153 49 Calculated Values for CMU 4................................................................................154 50 Testing Observations for CMU 4...........................................................................155 51 Summary of Results for CMU 4............................................................................164 52 Calculated Values for CMU 5................................................................................165 53 Testing Observations for CMU 5...........................................................................165 54 Summary of Results for CMU 5............................................................................175 55 Calculated Values for CMU 6................................................................................176 56 Testing Observations for CMU 6...........................................................................177 57 Summary of Results for CMU 6............................................................................184 58 Calculated Values for CMU 7................................................................................185 59 Testing Observations for CMU 7...........................................................................186 60 Summary of Results for CMU 7............................................................................193 61 Calculated Values for CMU 8................................................................................195 62 Testing Observations for CMU 8...........................................................................195 63 Summary of Results for CMU 8............................................................................203 64 Calculated m-Factors for Positive Load Values.....................................................209 65 Calculated m-Factors for Negative Load Values...................................................209 66 Calculated m-Factors Specimens with Non-ductile Reinforcement......................213 67 Measured and Calculated Lateral Capacities.........................................................214 68 Reinforcing Bar Stresses at First Yielding.............................................................215 69 Reinforcing Bar Stresses at = 1...........................................................................217 70 Calculated m-Factors for Positive Load Values using FEMA 273........................218 viii

PAGE 9

71 Calculated m-Factors for Negative Load Values using FEMA 273......................218 72 Calculated m-Factors for Positive Load Values.....................................................218 73 Calculated m-factors for negative load values.......................................................219 74 Calculated m-factors using Moment Curvature at ` y for positive loading...........219 75 Calculated m-factors using Moment Curvature at ` y for negative loading..........219 ix

PAGE 10

LIST OF FIGURES Figure page 1 Pier area outlined on a structure.................................................................................4 2 Pier strengthened for shear and flexure with FRP composites. Rocking and sliding restrained with a ductile connection at the base of the pier............................4 3 Specimen dimensions and location of joint reinforcement........................................6 4 Key to FRP composite configuration on specimen....................................................8 5 Free body diagram of half the base used to determine the quantity and placement of bonded FRP composite..........................................................................................8 6 Typical partially reinforced specimen containing existing reinforcing bars in jambs and sills..........................................................................................................12 7 Test set up with specimen ready for testing (looking at North face of specimen)...14 8 Schematic of specimen in the test set up..................................................................14 9 ICBO test sequence of imposed displacement.........................................................15 10 Drift for CMU 1.......................................................................................................17 11 Drift for CMU 2.......................................................................................................18 12 Drift for CMU 3.......................................................................................................18 13 Drift for CMU 4.......................................................................................................18 14 Drift for CMU 5.......................................................................................................19 15 Drift for CMU 6.......................................................................................................19 16 Drift for CMU 7.......................................................................................................19 17 Drift for CMU 8.......................................................................................................20 18 Compression failure on the west end of the pier and buckled FRP.........................21 x

PAGE 11

19 Cracking into the sill on the east end of the specimen caused by tension in the grouted cells.............................................................................................................22 20 Base splitting of CMU 2...........................................................................................23 21 Cone-shaped section of masonry was pulled out as steel reinforcement was loaded in tension.......................................................................................................24 22 Dowel and jamb steel in CMU 4 pulled out in a V-shape........................................25 23 Splice failure. A) Outline of the section. B) Free body diagram with forces on the section.................................................................................................................26 24 Specimen rocked about the area outlined by the white line.....................................27 25 The FRP composite debonded from the east end of the pier on CMU 8. A) Front view of the FRP composite debonding. B) Side view shows that the debonding started at the joint above the dowels and continued down the pier..........................28 26 Backbone curves for specimens with jamb steel......................................................29 27 Backbone curves for specimens without jamb steel.................................................29 28 Force-displacement curve for URM strengthened with fully bonded FRP composite and debonding FRP composite...............................................................30 29 Schematic of the pier connected to the base with dowels........................................36 30 Pier area outlined on a structure...............................................................................57 31 Failure modes for unreinforced masonry.................................................................58 32 Pier strengthened for shear and flexure with FRP composites. Rocking restrained with the ductile connection at the base of the pier..................................61 33 Idealized load-displacement curve for a URM pier.................................................62 34 URM pier with rocking load of P OT. .........................................................................62 35 Idealized load-displacement curve for a pier strengthened with a ductile connection to the base..............................................................................................63 36 Pier strengthened with a ductile connection with an overturning load of P.............63 37 Test set up with specimen ready for testing (looking at North face of specimen)...65 38 Plan view of test set up. The North direction in the laboratory is defined...............66 39 Angles used to prevent out-of-plane movement.......................................................66 xi

PAGE 12

40 Schematic of specimen in the test set up..................................................................67 41 Dial gauge to monitor slipping between the concrete cap and the specimen's top lintel..........................................................................................................................68 42 Lifting Frame............................................................................................................69 43 Axial load spring system. A) Schematic. B) Photo..................................................71 44 MTS actuator in place..............................................................................................72 45 ICBO test sequence of imposed displacement.........................................................72 46 Locations of instruments for linear displacement measurement..............................74 47 String pots between steel frame and specimen. A) Overall photo of two of the string pots. B) Closer view of a string pot that measure horizontally. C) Closer view of a string pot that measures vertically............................................................74 48 Linear potentiometers on the specimen....................................................................75 49 Rubber tube installation on dowel over the strain gauge.........................................76 50 FRP Strain Gauge.....................................................................................................77 51 FRP strain gauge locations.......................................................................................78 52 LabVIEW program main screen..............................................................................79 53 MTS Signal Generation program screen..................................................................79 54 Specimen dimensions and location of joint reinforcement......................................81 55 Typical partially reinforced specimen containing existing reinforcing bars in jambs and sills..........................................................................................................82 56 Mixing resin.............................................................................................................87 57 FRP application. A) Precoating masonry with resin. B) Pressing FRP cloth into resin. C) Rolling resin onto placed FRP. D) Troweling placed FRP.......................87 58 Grid FRP application. A) Mixing resin. B) Initial coat of resin on masonry. C) Grid FRP pressed into the resin. D) Troweling of additional resin..........................89 59 Schematic of the pier connected to the base with reinforcing bars..........................91 60 FRP composite placement for CMU 1 North face...................................................92 61 FRP composite placement for CMU 1 South face...................................................93 xii

PAGE 13

62 CMU 1 two cycles into testing. A) North face. B) South face.................................93 63 Steel reinforcement location for CMU 2..................................................................94 64 FRP composite placement for CMU 2 North face...................................................96 65 FRP placement for CMU 2 South face.....................................................................96 66 CMU 2 prior to testing. A) North face. B) South face.............................................97 67 Steel reinforcement locations for CMU 3................................................................98 68 Location of transverse GFRP rebar in CMU 3.........................................................99 69 Transverse GFRP block in the faceshell of CMU 3. A) Cross-section of block containing a GFRP bar. B) GFRP bar embedded in faceshell.................................99 70 FRP composite placement for CMU 3 North face.................................................100 71 FRP composite placement for CMU 3 South face.................................................101 72 CMU 3 prior to testing. A) North face. B) South face five cycles into testing......102 73 Repointing of CMU 4. A) Repointed dowel with smoothed epoxy. B) Grooves cut into the mortar joints. C) Passing epoxy over the embedded dowel................104 74 Steel reinforcement locations for CMU 4..............................................................104 75 Details of repointed #3 dowel................................................................................104 76 FRP placement for CMU 4 North face...................................................................105 77 FRP placement for CMU 4 South face...................................................................106 78 CMU 4 prior to testing. A) North face. B) South face two cycles into testing......107 79 Steel reinforcement locations for CMU 5..............................................................108 80 FRP composite placement for CMU 5 North face. (No FRP composite was placed on the South face).......................................................................................109 81 CMU 5 prior to testing. A) North face. B) South face...........................................110 82 FRP composite placement for CMU 6 North face. (No FRP composite was placed on the South face).......................................................................................111 83 CMU 6 prior to testing. A) North face. B) South face...........................................112 xiii

PAGE 14

84 FRP composite placement for CMU 7 North face. (No FRP composite placed on the South face)...................................................................................................113 85 Steel reinforcement locations for CMU 7..............................................................114 86 CMU 7 prior to testing. A) North face. B) South face...........................................114 87 Steel reinforcement locations for CMU 8..............................................................115 88 Installing dowels in CMU 8. A) Grooves cut into the specimen. B) Bars placed onto plywood. C) Dowels placed on plastic spacers with wire. D) Grout was shoveled into the openings in the masonry............................................................116 89 FRP composite placement for CMU 8 North face. (No FRP composite was placed on the South face).......................................................................................117 90 CMU 8 prior to testing. A) North face. B) South face...........................................118 91 Crack Patterns for CMU 1. A) North face. B) South face.....................................120 92 Before and after pictures of CMU 1. A) North face two cycles into testing. B) North face after testing. C) South face two cycles into testing. D) South face after testing.............................................................................................................121 93 Debonding of FRP composite on CMU 1..............................................................122 94 Pier displaced = 1. A) Unbonded length equal to 1. B) Unbonded length equal to 10........................................................................................................................122 95 Compression failure on the west end of the pier and buckled FRP.......................124 96 Drift for CMU 1.....................................................................................................125 97 Backbone curve for CMU 1...................................................................................127 98 Generalized force-displacement relationship for URM and URM retrofit with FRP overlays (Moon, Leon etal 2002)...................................................................128 99 Force-displacement curve for URM strengthened with fully bonded FRP composite and debonding FRP composite.............................................................129 100 Step cracking opened up and allowed the pier of CMU 1 to slide.........................130 101 Sliding of CMU 1...................................................................................................130 102 FRP strain gauge readings for CMU 1. A) Vertical pier strip. B) Vertical base strip. C) Horizontal base strip................................................................................131 103 Out-of-plane movement for CMU 1. A) West. B) East.........................................132 xiv

PAGE 15

104 Crack Patterns for CMU 2. A) North face. B) South face.....................................134 105 Before and after pictures of CMU 2. A) North face taken before testing. B) North face after testing. C) South face taken before testing. D) South face after testing.....................................................................................................................135 106 Base Splitting of CMU 2........................................................................................136 107 Splitting bond failure caused cracks to propagate into the sill at the east end of CMU 2....................................................................................................................137 108 Cracks move radially from the bar to the edges of the pier in CMU 2..................138 109 Drift for CMU 2.....................................................................................................139 110 Backbone Curve for CMU 2..................................................................................140 111 FRP Strains for CMU 2. A) Vertical pier strip. B) Horizontal base strip..............141 112 Steel Strains for CMU 2. A) West dowel. B) East dowel......................................141 113 Out-of-plane Movement for CMU 2. A) East. B) West.........................................142 114 Crack Patterns for CMU 3. A) North face. B) South face.....................................144 115 Debonding pattern on the base of CMU 3..............................................................145 116 Before and after pictures of CMU 3. A) North face taken prior to testing. B) North face after testing. C) South face five cycles into testing. D) South face after testing.............................................................................................................146 117 Rocking occurred about the white line for CMU 3................................................147 118 Cone failure of the grouted jamb containing the jamb steel on the west end of CMU 3....................................................................................................................148 119 Drift for CMU 3.....................................................................................................149 120 Backbone curve for CMU 3...................................................................................150 121 Sliding of CMU 3...................................................................................................150 122 Steel strain gauge readings for CMU 3. A) West jamb steel. B) West dowel. C) East jamb steel. D) East dowel...............................................................................151 123 FRP strain gauge readings for CMU 3. A) Vertical pier strip. B) Horizontal base strip.................................................................................................................152 124 Out-of-plane movement for CMU 3. A) East. B) West.........................................153 xv

PAGE 16

125 Crack Patterns for CMU 4. A) North face. B) South face.....................................155 126 Before and after pictures of CMU 4. A) North face prior to testing. B) North face after testing. C) South face two cycles into testing. D) South face after testing.....................................................................................................................156 127 Crack move from the dowel toward the sill and the lintel. This cracking caused the FRP to delaminate............................................................................................157 128 Crack through the epoxy containing the repointed dowel......................................157 129 Cracking and delamination into the lintel..............................................................158 130 Cracked formed from dowel to pier.......................................................................158 131 Drift for CMU 4.....................................................................................................159 132 Sliding of CMU 4...................................................................................................159 133 Backbone curve for CMU 4...................................................................................160 134 Steel strain gauge readings for CMU 4. A) North face east dowel. B) South face east dowel. C) East jamb steel. D) North face west dowel.............................162 135 FRP strain gauge readings for CMU 4. A) Vertical pier strip. B) Horizontal base strip.................................................................................................................163 136 Out-of-plane movement for CMU 4. A) East. B) West.........................................163 137 Crack Patterns for CMU 5. A) North face. B) South face.....................................166 138 Before and after pictures of CMU 5. A) North face prior to testing. B) North face after testing. C) South face prior to testing. D) South face after testing.........167 139 First cracks in CMU 5 appeared in the base of the south face...............................168 140 Cracks propagating from the grouted core of the east end.....................................169 141 Tensile failure in the masonry exposed the grouted core containing the reinforcing bars......................................................................................................169 142 Splice failure. A) Outline of the section. B) Free body diagram with forces on the section...............................................................................................................170 143 Drift for CMU 5.....................................................................................................171 144 Backbone curve for CMU 5...................................................................................172 xvi

PAGE 17

145 Steel strain gauge readings for CMU 5. A) North face west dowel. B) South face west dowel. C) North face east dowel. D) South face east dowel..........................173 146 FRP strain gauge readings for CMU 5. A) Vertical pier strip. B) Horizontal base strip.........................................................................................................................174 147 Out-of-plane movement for CMU 5. A) East. B) West.........................................174 148 Crack Patterns for CMU 6. A) North face crack pattern. B) South face crack pattern.....................................................................................................................177 149 Before and after pictures of CMU 6. A) North face prior to testing. B) North face after testing. C) South face prior to testing. D) South face after testing................178 150 Toe crushing on the south face east end of CMU 6...............................................179 151 Toe crushing side views for CMU 6. A) East side. B) West side..........................180 152: FRP rupture for CMU 6. A) East end. B) West end................................................180 153 Drift for CMU 6.....................................................................................................181 154 Backbone curve for CMU 6...................................................................................182 155 FRP strain gauge readings for CMU 6. A) Vertical pier strip. B) Vertical base strip. C) Horizontal base strip................................................................................183 156 Out-of-plane movement for CMU 6. A) East. B) West.........................................184 157 Crack Patterns for CMU 7. A) North face. B) South face.....................................186 158 Before and after pictures of CMU 7. A) North face prior to testing. B) North face after testing. C) South face two cycles into testing. D) South face after testing.....................................................................................................................187 159 Specimen rocked about the area outlined by the white line...................................188 160 Hinge formed in the base along the bottom of the grouted column.......................188 161 Drift for CMU 7.....................................................................................................189 162 Backbone curve for CMU 7...................................................................................190 163 Steel strain gauge readings for CMU 7. A) West jamb steel. B) East jamb steel..191 164 FRP strain gauge readings for vertical pier strip on CMU 7..................................191 165 Out-of-plane movement for CMU 7. A) East. B) West.........................................192 xvii

PAGE 18

166 Simple structure with openings..............................................................................194 167 Series of piers under cyclic loading.......................................................................194 168 Crack Patterns for CMU 8. A) North face. B) South face.....................................196 169 Before and after pictures of CMU 8. A) North face prior to testing. B) North face after testing. C) South face two cycles into testing. D) South face after testing....197 170 First crack to form is in the horizonatal bed joint running over the grouted cell...198 171 Debonding of the FRP composite begins over the grouted cell.............................199 172 The FRP composite debonded from the east end of the pier on CMU 8. A) Front view of the FRP composite debonding. B) Side view shows that the debonding started at the joint above the dowels and continued down the pier........................199 173 Debonding of the FRP composite on the west end began at the joint were the grout started and towards the bottom of the pier. The debonding did not reach the bottom of the pier.............................................................................................200 174 Drift for CMU 8.....................................................................................................200 175 Backbone curve for CMU 8...................................................................................201 176 Steel strain gauge readings for CMU 8. A) East dowel (nearest to edge). B) East dowel..............................................................................................................202 177 FRP composite strain gauge readings for CMU 8. A) Vertical pier strip. B) Horizontal base strip...............................................................................................202 178 Out-of-plane movement for CMU 8. A) East. B) West.........................................203 179 Generalized Component Behavior Curves. A) Type 1. B) Type 2. C) Type 3......207 xviii

PAGE 19

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering FRP/STEEL STRENGTHENING OF UNREINFORCED CONCRETE MASONRY PIERS By Vanessa E. Grillo December 2003 Chair: H.R. Hamilton Major Department: Civil and Coastal Engineering This thesis presents research on the strengthening of unreinforced masonry (URM) piers strengthened with fiber-reinforced polymers (FRP) in conjunction with ductile reinforcement. Eight concrete masonry pier specimens were strengthened with a combination of FRP composite strips and reinforcing steel. Specimen construction included the pier and a portion of the masonry just below the pier. FRP composite strips were strategically placed to improve flexure and shear strength in the in-plane direction. Steel dowels were added to the specimen by grouting into cells or by repointing in vertical head joints and were designed to yield prior to the FRP composite rupture, resulting in a ductile response. Improvement in lateral capacity of three to nine times the capacity of URM rocking mode was achieved. Drift capacities ranged from 0.85% to 3.3%. Confinement of the masonry in the base of the specimen was shown to be a controlling factor in the extent of yielding attained in the steel dowels. xix

PAGE 20

INTRODUCTION Masonry and Earthquakes Earthquakes can cause extensive damage to unreinforced masonry (URM) structures. Many older masonry structures currently in use were designed and constructed with little or no consideration of earthquake resistance. In addition, recent changes in seismic requirements have left many URM buildings in need of strengthening. This is particularly true in Midwestern states where seismic design has not been considered in the past. Since the advent of modern reinforced masonry construction, URM structures have been viewed as a significant liability when considering strengthening. Traditional methods of URM strengthening include shotcrete or the addition of steel frames or reinforced concrete walls. In general, these options ignore the contribution of the URM components to the lateral capacity. Furthermore, they are quite expensive and pose significant inconvenience for the building occupants during installation. Significant progress has been made in identifying URM behavior under extreme loads and recognizing the contribution of URM components to both strength and ductility of the building system (Calvi et al. 1996, Marshall and Sweeney 2002). URM structures are usually analyzed as a system of shear walls and/or piers that carry proportional levels of the story shears. Previous research (Calvi et al. 1996, Marshall and Sweeney 2002, Triantafillou 1998) as well as recent design guidelines (FEMA 273) recognize rocking, sliding, toe-crushing and diagonal-tension as the four primary failure modes for URM piers. Relative values of aspect ratio, compressive strength, and axial stress determine the 1

PAGE 21

2 failure mode most likely to occur. It is further recognized that sliding and rocking are relatively stable failure modes and provide some energy dissipation (Calvi et al. 1996, Marshall and Sweeney 2002). This approach allows the engineer to determine the probable level of building performance with no strengthening. If additional capacity is needed, then strengthening is necessary. One option available that can provide added capacity is to apply fiber reinforced polymer (FRP) composites. FRP Strengthening Fiber reinforced polymer (FRP) composites are made of continuous glass, carbon or aramid fibers bonded to the substrate with a resin polymer matrix and are typically unobtrusive to the building occupants, require relatively little surface preparation and are economical. Recent research has shown that FRP composites can be applied to increase strength and change the failure modes of masonry walls (Abrams 2001, Ehsani et al. 1999, Marshall and Sweeney 2002, Triantafillou 1998). The majority of previous testing of FRP composites and masonry has focused on piers with medium to high aspect ratios (ratio of height to width). Higher aspect ratios generally result in an overturning or rocking behavior in the unstrengthened state. Application of bonded FRP composites can cause a shift in the failure mode to one of sliding or shear. The ability of a surface bonded FRP composite strengthening system to prevent masonry from falling off the structure during an extreme event also makes it a favorable alternative to traditional methods (Marshall and Sweeney 2002). Strengthening masonry walls with FRP composites requires that the composite be bonded to the wall surface either in sheets covering the entire surface of the pier or in strips placed at strategic locations on the pier. Unidirectional strips of FRP are thought to be preferable in terms of economy and behavioral response to the two-dimensional

PAGE 22

3 fabrics that cover the entire surface of the masonry wall (Triantafillou 1998). Full coverage may also cause problems by limiting moisture movement through the wall. FRP Composites with a Ductile Connection Although FRP composites increase lateral load capacity, they do not significantly improve ductility and may actually decrease ductility if an undesirable failure mode is precipitated. This is due to the brittle nature of the composite material. Holberg and Hamilton (2002) proposed a hybrid system, consisting of bonded FRP composites in conjunction with steel. The FRP composite adds sufficient strength to the masonry allowing the steel to reach yield, thus incorporating ductility into the system. Research was conducted by Holberg and Hamilton (2002) on URM walls retrofitted with a hybrid strengthening system consisting of FRP composites and steel. Two different types of steel connections were tested, an internal and an external. The internal connection was a steel reinforcing bar placed in the outermost cells of the wall and fully grouted into a concrete foundation. The external connection was a steel angle-plate assembly attached to the foundation. The drift capacities of the reinforced specimens reached up to 1.7%. The lateral capacities of the strengthened specimens were nearly doubled when compared to the lateral capacity of an unstrengthened specimen. Holberg and Hamilton (2002) investigated using a ductile connection between a masonry pier and a concrete foundation. While the results of this research appear promising, the behavior of the connection between the pier and surrounding masonry in multistory buildings (Figure 1) needs to be investigated. One of the problems associated with grouting dowels into masonry is providing confinement to enable the dowels to yield. The focus of the research presented in this paper is on the connection between the

PAGE 23

4 pier and its supporting masonry, and its effect on the ductility and stability of the in-plane pier behavior. Figure 1: Pier area outlined on a structure. In Figure 2, the vertical FRP composite strips are designed to provide enough additional strength to resist the shear and flexural stresses experienced during an earthquake. The ductile connection is designed to yield prior to failure of the FRP composite. Adequate strength must be provided in the masonry surrounding the dowels to ensure yielding at the pier/base interface and prevent a pull out failure. In addition to confinement of the dowels, the masonry below the pier requires strengthening against flexure and shear induced by the tensile forces in the dowels. Sliding and rocking restrained with ductile connection Shear and flexural strength improved Figure 2: Pier strengthened for shear and flexure with FRP composites. Rocking and sliding restrained with a ductile connection at the base of the pier.

PAGE 24

EXPERIMENTAL PROGRAM Eight reinforced concrete pier specimens were constructed and tested to investigate the behavior of the pier, ductile connection and base subjected to in-plane cyclic loading. The experimental parameters included the amount of FRP composite and reinforcing steel placed onto each specimen. Test Specimens Eight concrete masonry specimens were constructed in running bond of medium weight 8-in. (200-mm) concrete masonry units by Painter Masonry, Inc. in Gainesville, FL. Type N mortar was used in face shell bedding for each of the specimens. The ASTM C90 units were purchased locally from Florida Rock Industries, Inc. in Gainesville, FL. All test specimens were single wythe and consisted of a pier 48-in. (1200-mm) tall by 48-in. (1200-mm) wide and a base 9-ft. 4-in. (2800-mm) long and 16-in. (400-mm) tall (Figure 3). Since the test specimen consisted of full scale CMU, no scaling effects were required. The masons installed ASTM A Class B2 Hot Dipped Galvanized After Fabrication standard sized (9 Gauge) wire joint reinforcement every other course in the specimen. This was done in accordance to common practice in the field. The location of the joint reinforcement is indicated in Figure 3. The specimen was constructed on a precast concrete lintel 9-ft. 4-in. (2800-mm) long. A second precast concrete 5-ft. 4-in. (1600-mm) lintel was placed on the top course of the pier. Both lintels were placed with the same mortar used for constructing the specimen. The outer cells of the specimens sills were filled with grout (approximately 5

PAGE 25

6 one month after construction) to prepare them for post tensioning to the concrete base at a force of 42 kips (187 kN). PrecastConcrete Lintel 4'-0" (1220 mm)Specimen 5'-4" (1630 mm) Grouted Cells PrecastConcrete Lintel Lifting Rings 9'-4" (2845 mm) 4'-0" (1220 mm) 1'-4" (406 mm) Joint ReinforcementPierBase Figure 3: Specimen dimensions and location of joint reinforcement. Material Properties Masonry, steel reinforcement, joint reinforcement and FRP composite material properties were determined using ASTM standard test methods (Table 1). Five coupons of the cured FRP composite material were tested in tension. The average width of the tensile specimens was 1.04-in. (26.42-mm) and the average thickness was 0.91-in (23.11-mm). The specimens were cut from a section of cured FRP composite and then milled to a 1-in. width. Preparation of the coupons and testing followed ASTM D3039. FRP Composite Configuration The specimens, except for CMU 6, were strengthened with varying widths and lengths of unidirectional fiberglass fabric (27 ounces/yard 2 .0859 kg/ft 2 ) bonded to the surface of the specimen with a two-part epoxy. The manufacturer-specified tensile strength of the composite was 330 ksi (2275 MPa). Specimen CMU 6 was reinforced

PAGE 26

7 using unidirectional grid reinforcement. This grid system is a high strength, unidirectional reinforcement made by bonding E glass fiber rovings with epoxy resin in a controlled factory environment. The specified strength of the grid system given by the manufacturer was 14,400 lbs/ft (210 kN/m) and was adhered to the specimen using a low modulus epoxy. Table 1: Material Properties Material Test Average Strength No. of specimens CMU Stretcher Unit Unit Strength 2224 psi (15.3 MPa) 5 CMU Full Prism Prism Strength 2167 psi (14.9 MPa) 4 CMU Half Prism Prism Strength 3880 psi (26.7 MPa) 4 #4 Reinforcing Bar Yield Strength 81 ksi (558 MPa) 4 #3 Reinforcing Bar Yield Strength 62 ksi (429 MPa) 4 Joint Reinforcement Yield Strength 107 ksi (738 MPa) 3 FRP Composite Coupon Tensile Strength 2.86 kips/inch (500 N/mm) 5 Quantity and placement of the FRP composites for the pier and the base were determined using a strut and tie analysis and basic mechanics principles. Figure 4 shows the locations for the FRP composite placement. Details of FRP composite configuration are given in Table 2. The basic configuration included strips oriented vertically along the pier jambs as well as diagonally across the pier. The vertical strips (V) increased the in-plane flexural strength and the diagonal strips (X) increased the diagonal tension strength. Bi-directional FRP composite fabric with a fiber orientation was applied along the top of the pier at the pier/lintel interface to prevent separation of the lintel and pier (SP). The base was modeled as a deep beam and it was expected that flexure and shear reinforcement would be required to maintain stability (Figure 5). A strip of unidirectional

PAGE 27

8 FRP composite was placed at the top of the base along its length (HB) to provide reinforcement against bending caused by the tension from the pier reinforcement and compression from the end restraints. Shear reinforcement was added by covering the base with bi-directional FRP composite (SB) to inhibit diagonal cracking in the base and to provide confinement to the grouted dowels. VHBSBXSP Figure 4: Key to FRP composite configuration on specimen. TFRP TDowel/FRPCHold Down Force C M Figure 5: Free body diagram of half the base used to determine the quantity and placement of bonded FRP composite.

PAGE 28

Table 2: Details of FRP Composite Configuration Specimen North Face South Face FRP Composite Configuration* V Unidirectional 4 x 80 (102 x 2032) HB Unidirectional 6 x 112 (152 x 2844) SB 18 x 112 (457 x 2844) X Unidirectional 3 x 68 (76 x 1727) CMU 1 SP 12 x 8 (305 x 1219) V Unidirectional 4 x 80 (102 x 2032) HB Unidirectional 3 x 112 (76 x 2844) SB 18 x 112 (457 x 2844) X Unidirectional 3 x 68 (76 x 1727) CMU 2 SP 12 x 48 (305 x 1219) V Unidirectional 6 x 80 (152 x 2032) HB Unidirectional 3 x 112 (76 x 844) SB 18 x 112 (457x 2844) X Unidirectional 3 x 68 (76 x 1727) CMU 3 SP 12 x 48 (305 x 1219) V Unidirectional 7 x 80 (178 x 2032) HB Unidirectional 3 x 112 (76 x 844) SB 18 x 112 (457x 2844) X Unidirectional 3 x 68 (76 x 1727) CMU 4 SP 12 x 48 (305 x 1219) 9 *Note: Dimensions: in x in (mm x mm)

PAGE 29

Table 2. Continued Specimen North Face South Face FRP Composite Configuration* V Unidirectional 8 x 80 (203 x 2032) HB Unidirectional 3 x 112 (76 x 844) SB 18 x 112 (457x 2844) X Unidirectional 3 x 68 (76 x 1727) CMU 5 SP 12 x 48 (305 x 1219) V Grid 9 x 80 (229 x 2032) HB None SB Grid 18 x 112 (457 x 2844) X Grid 7 x 68 (178 x 1727) CMU 6** SP Grid 12 x 48 (305 x 1219) V Unidirectional 8 x 80 (203 x 2032) HB None SB None X Unidirectional 3 x 68 (76 x 1727) CMU 7 SP 12 x 48 (305 x 1219) V Unidirectional 8 x 80 (203 x 2032) HB Unidirectional 3 x 112 (76 x 844) SB 18 x 112 (457x 2844) X Unidirectional 3 x 68 (76 x 1727) CMU 8 SP 12 x 48 (305 x 1219) 10 *Note: Dimensions: in x in (mm x mm) **Note: CMU 6 was strengthened using a FRP grid system.

PAGE 30

11 Reinforcing Steel Placement In an attempt to force ductile failure, steel dowels were installed into the specimens as indicated in Table 3. The dowels were sized to yield prior to the failure of the FRP composite. Specimens CMU 2, CMU 3, CMU 5 and CMU 8 contained dowels grouted into the cells after the specimens were constructed. CMU 3, CMU 4 and CMU 7 contained jamb and sill reinforcement that was placed as part of the specimen construction. These specimens were intended to represent conditions in which prescriptive amounts of steel have been added during construction. The remainder of the specimens were intended to represent unreinforced conditions. Table 3: Steel Reinforcement Details. Specimen Jamb Steel Dowels Pier Section CMU 1 No None CMU 2 No 2 #4 x 40 in grouted jamb CMU 3* Yes 2 #4 x 40 in grouted cell CMU 4 Yes 4 #3 x 32 repointed in vertical head joint CMU 5 No 4 #4 x 40 grouted in jamb CMU 6 No None CMU 7 Yes None CMU 8 No 4 #3 grouted in jamb *Note: CMU 3 contains transverse GFRP bars in every cell containing the dowel steel. Dowel installation required face shell removal at the fourth course up from the bottom of the pier. The dowels were held in place by hand in the center of the cell and

PAGE 31

12 grout was scooped in through the opening. A tamping rod was used to consolidate the grout around the dowels because there was no room to insert a vibrator into the face shell opening. The face shells were then replaced with mortar. Specimen CMU 4 called for structural repointing with a #3 dowel that was 32inches (812 mm) long. Vertical grooves were cut into the head joint starting at the top of the bottom lintel. The width of the groove matched the width of the mortar joint, 3/8-inch (9.5-mm), so the entire mortar joint was removed. The depth of the groove was 3/4-inch (19-mm) deep. Dimensions for repointing were taken from similar testing by Bajpai and Duthinh (2003). A gel epoxy adhesive was used to install the dowel into the groove. A layer of the epoxy was squeezed into the groove. After coating the bar with the epoxy, it was pressed into the groove. Epoxy was then passed over the bar filling in any gaps and then smoothed by hand, giving the appearance of a repointed joint. Specimens CMU 3, CMU 4 and CMU 7 were constructed with #4 reinforcing bars in the outer jambs and sills (Figure 6). The jamb and sill steel were grouted into the specimen one month after specimen construction. The grout for the jamb and sill steel was shoveled in and consolidated with a concrete vibrator. #4 Reinforcing Bar (o.c.)#4 Reinforcing Bar (o.c.) Figure 6: Typical partially reinforced specimen containing existing reinforcing bars in jambs and sills.

PAGE 32

13 Transverse glass fiber reinforced polymer (GFRP) composite bars were placed in the cells containing the dowels of CMU 3 to avoid the splitting failure that occurred in CMU 2. The purpose of the GFRP bar was to confine the grout containing the dowels in order to develop the strain necessary to yield them. The #2 (3.2 mm) GFRP bars had a specified tensile strength of 120 ksi (825 MPa) and a modulus of elasticity of 5.92 psi (40.8 GPa). One quarter-inch diameter hole was drilled through the unit face shell before dowels were grouted into the cells. The length of the transverse bars matched the thickness of a CMU block, 7.625 inches (194 mm). The GFRP bars were pushed through the block via the drilled holes. The holes in the face shells were filled with silicone caulking and covered with duct tape to prevent the GFRP bars from moving during grouting of the dowels. Test Setup The test setup was designed to apply in-plane cyclic displacements under displacement or load control. A hydraulic actuator was located on the reaction frame and displaces the concrete cap on the specimen (Figure 7). Positive load and displacement correspond to actuator tension. The reaction frame that supports the actuator is constructed of steel and is prevented from overturning by its connection to the laboratory strong floor. A concrete cap and base were constructed for the test fixture and designed to handle loads of 50 kips (220 kN). The concrete base was post-tensioned to the laboratory strong floor. The concrete cap was braced to the laboratory wall by steel angles to prevent out-of-plane twisting. A 55-kip (245 kN) hydraulic actuator was used for testing. The hydraulic actuator was part of a closed-loop hydraulic loading system. The controller used a sinusoidal voltage output from the data acquisition system to impose the displacements. The

PAGE 33

14 actuator load cell was used to monitor the lateral load applied to the specimen during testing. Gravity load was applied to the specimen using rail car springs and threaded rods. This translated into an axial stress of 75 psi (0.5 MPa) on the net section of the base of the pier. The gravity load included load from two rail car springs compressed to develop 6 kips (27 kN), the weight of the concrete cap (2 kips, 9 kN) and the self-weight of the pier and lintel (1 kip, 4 kN). Spring System 55 kip Actuator Reaction Frame Concrete Base Concrete Cap DAQ System Spring System 55 kip Actuator Reaction Frame Concrete Base Concrete Cap DAQ System Figure 7: Test set up with specimen ready for testing (looking at North face of specimen). 8 inches(200 mm)68 inches(1727 mm)24 inches(610 mm)18 inches(457 mm) MTS Load Downward Force = 42 kips (187 kN) Gravity Load SimulatorConcrete BaseLaboratory FloorSpecimen Figure 8: Schematic of specimen in the test set up.

PAGE 34

15 A personal computer, National Instruments LabVIEW software and a 16-bit data acquisition card were used for data acquisition. Linear and string potentiometers were used for measuring displacements. Foil strain gauges were placed on the steel reinforcing bars and FRP composite strips. Test Procedures The test displacement sequence from ICBO Acceptance Criteria for Concrete and Reinforced and Unreinforced Masonry Strengthening Using Fiber-Reinforced Composite Systems (AC125) (ICBO 1997) was followed (Figure 9). The displacement at which the reinforcement was calculated to yield is marked as the first yield point, = 1. For specimens that did not contain any reinforcing steel (CMU 1 and CMU 6), the yield point was taken as the displacement that was expected to cause a rocking failure of an unreinforced specimen. The specimens were loaded in displacement control with three complete cycles for each displacement level, The displacements were = , 1, 2, 3, 4, 6, 8, 10, 12, 16, 18 and 20 and were increased until the specimen experienced a loss in lateral load capacity. The loading rate was approximately 20 seconds per cycle. Figure 9: ICBO test sequence of imposed displacement (ICBO 1997).

PAGE 35

16 It was from CMU 1 that the stiffness of all the walls was determined. At the beginning of the test, the specimen was cycled several times in the elastic range to determine the uncracked stiffness (k) of the specimens, 167 kips/inch (29.2 kN/mm). The stiffness was found using the displacement and load cell reading of the hydraulic actuator. This value of k along with the individual specimens calculated capacity was used to find y the yield point for each specimen. For those specimens containing steel reinforcement the yield point is the displacement associated with measured yield in the steel reinforcement. The cracking load was used as the yield point for those specimens without steel reinforcement.

PAGE 36

EXPERIMENTAL RESULTS Observed Behavior Figure 10 through Figure 17 show the lateral load versus drift ratio plots for all the tests. Drift ratio is the in-plane displacement of the pier divided by the height. The horizontal solid line in the plots represents the calculated lateral load capacity where all the steel reinforcement in the specimen is assumed to yield. Details for these calculated values are discussed in a subsequent section. For CMU 1, which contained no steel reinforcement, the horizontal solid line represents the calculated rocking load for an unstrengthened specimen. For the specimen strengthened with only the grid FRP composite, CMU 6, the horizontal solid line represents the calculated load at which the grid FRP composite was calculated to rupture. -30-20-100102030-2-1012Drift (%)Load (kips)-133.5-89-44.5044.589133.5Load (kN) Figure 10: Drift for CMU 1. 17

PAGE 37

18 -30-20-100102030-2-1012Drift (%)Load (kips)-133.5-89-44.5044.589133.5Load (kN) Figure 11: Drift for CMU 2. -30-20-100102030-2-1012Drift (%)Load (kips)-133.5-89-44.5044.589133.5Load (kN) Figure 12: Drift for CMU 3. -30-20-100102030-2-1012Drift (%)Load (kips)-133.5-89-44.5044.589133.5Load (kN) Figure 13: Drift for CMU 4.

PAGE 38

19 -30-20-100102030-2-1012Drift (%)Load (kips)-133.5-89-44.5044.589133.5Load (kN) Figure 14: Drift for CMU 5. -30-20-100102030-2-1012Drift (%)Load (kips)-133.5-89-44.5044.589133.5Load (kN) Figure 15: Drift for CMU 6. -30-20-100102030-4-3-2-101234Drift (%)Load (kips)-133.5-89-44.5044.589133.5Load (kN) Figure 16: Drift for CMU 7.

PAGE 39

20 -30-20-100102030-2-1012Drift (%)Load (kips)-133.5-89-44.5044.589133.5Load (kN) Figure 17: Drift for CMU 8. Table 4 summarizes the key results of the testing. The maximum lateral load capacity and maximum drift values were observed in the cycle immediately before the specimen experienced a loss in lateral load capacity. Table 4: Summary of Results Specimen Maximum Drift (%) Maximum Lateral Load (kips) [kN] Limiting Mode 1.4 17.7 [79] CMU 1 -1.4 -14.9 [-66] Toe Crushing 1.2 14.7 [65] CMU 2 -1.2 -16.6 [-74] Bond Failure 1.7 28.0 [126] CMU 3 -1.7 -26.5 [-118] Rocking 1.4 27.2 [121] CMU 4 -1.8 -26.8 [-119] Sliding 0.85 12.1 [54] CMU 5 -0.85 -12.0 [-54] Splice Failure 1.6 11.3 [50] CMU 6 -1.6 -10.3 [-46] FRP Rupture 3.2 10.8 [48] CMU 7 -3.3 -12.3 [-55] Flexure 1.0 15.0 [67] CMU 8 -0.99 -16.1 [-72] FRP Bond Failure CMU 1. The FRP composite strip on the east end of the pier fully delaminated on the bottom course at a drift ratio of -1.1%, which is evident in the sharp drop of lateral

PAGE 40

21 capacity in Figure 10. The east end of the pier retained approximately 90% of its maximum capacity in the negative direction. The delamination softened the specimen resulting in larger drift at the same load. It can be concluded that CMU 1 reached its limiting capacity when toe crushing occurred at the west end of the pier, as shown in Figure 18. This is evident in the reduction in load seen at a drift ratio of 1.4% (Figure 10). Even following toe crushing, the specimen retained approximately 40% of its peak capacity. The behavior had changed from that of flexure to sliding, which has favorable energy dissipation characteristics as evidenced by the open loops in Figure 10. Sliding occurred during = 20 across a step crack originating at the bottom west corner of the pier and running diagonally to the top east corner of the pier. Figure 18: Compression failure on the west end of the pier and buckled FRP. CMU 2. Debonding of the vertical FRP composite strip (V) in the pier began early in the cycling at = 1. The debonding initiated at the bed joint just above the

PAGE 41

22 dowel. At = 4, the FRP composite sheet on the base (SB) began debonding around the grout containing the dowel. As the pier was loaded at = 4, the grouted cells containing the dowel were lifted in tension causing cracking into the sill shown in Figure 19. This was observed at both the east and west ends of the pier. As shown in Figure 20, the grout containing the dowel split the specimen at the base. The cause of the bar failure is due to insufficient development length for the dowels. About 80% of the stress to yield was developed in the dowels prior to the splitting failure. Figure 19: Cracking into the sill on the east end of the specimen caused by tension in the grouted cells. Figure 11 shows large open loops during the last set of cycles indicating energy dissipation. These open loops were thought to be caused by the dowel movement after the splitting failure. Inspection after testing showed that the lug marks in the grout had been abraded smooth by the movement of the reinforcing bars.

PAGE 42

23 Figure 20: Base splitting of CMU 2. It can be concluded that for CMU 2, the behavior was limited by a bond failure in splitting mode between the dowels and the grout. The dowels did not develop the amount of strain necessary to fully yield because the bond between the dowel and the surrounding grout split before this could happen. It was decided that for future grouted #4 dowels, a transverse GFRP bars would be placed in every other cell to provide confinement to the grouted core. CMU 3. At = 2, debonding of the FRP composite (SB) was observed around the area containing the dowels and the jamb steel in the base. The tension developed in the dowel and jamb steel caused a cone-shaped section of masonry around the dowel and jamb steel to pull out of the base (Figure 21). A sheet of FRP composite had been applied to both sides of the base to prevent this failure from occurring. This pullout occurred at a drift ratio of approximately 1.15%, resulting in a 35% drop in lateral capacity. Behavior following the pullout was one of rocking. Both the dowel steel and jamb steel on both the east and west ends yielded at = 6, which correspond to a drift ratio of approximately 0.9%. The hysteresis loops tend to open significantly beyond this drift limit indicating good energy dissipation. No movement in the sill was observed, nor was there any cracking in the grout containing the sill steel. It can be concluded that the confinement provided by the FRP composite

PAGE 43

24 around the sill prevented large movements and cracking in the sill area. This allows for a greater lateral capacity and minimal damage to the base of the pier. Figure 21: Cone-shaped section of masonry was pulled out as steel reinforcement was loaded in tension. CMU 4. At = 3, delamination of FRP composite (SB) on the base around the dowels was observed. Cracking propagated from the dowels toward the sills, downward toward the lintel and continuing into the hydrostone laid on the foundation (Figure 22). A V-shaped section of masonry was pulled out as the jamb steel and dowels were loaded in tension. The lintel was damaged by the debonding FRP composite on the base and the cracking caused by tension in the dowels. Open loops in Figure 13 at a drift ratio of .4% correspond to yielding of the east and west jamb steel and the instrumented dowel located on the north face of the pier. The instrumented dowel located on the south face had only reached 60% of its yield stress by the end of testing.

PAGE 44

25 Figure 22: Dowel and jamb steel in CMU 4 pulled out in a V-shape. CMU 5. A loss in lateral load capacity occurred when the units at the east end jamb split through the web, exposing the grouted core of the specimen (Figure 23A) during the last set of displacements ( = 8). The length over which the FRP composite strip overlaps the grouted jamb is known as a splice. The failure at this splice was sudden and resulted in a 54% reduction of peak lateral load. The moment from having FRP composite on only one face of the pier caused a tensile failure in the masonry block. The force transfer between the pier and the base occurred primarily on the north face, which was stiffer due to the added FRP composite. This caused bending toward the FRP composite strip, which in turn caused the masonry to split in tension. Figure 23B shows a free body diagram (FBD) of the face shell section that pulled off the grouted core. The only force transfer to the base of the specimen is though the dowel. The tension in the FRP composite strip causes a moment that is restrained by the

PAGE 45

26 tensile stress in the masonry. The tensile stress in the masonry increases as the force in the tensile strip gets larger. When the tensile stress in the masonry reaches the ultimate tensile strength of the masonry, the face shells crack. A Tfrp Tdowel Tensile strength of masonry B Figure 23: Splice failure. A) Outline of the section. B) Free body diagram with forces on the section. CMU 6. The limiting behavior of CMU 6 was flexure, causing the FRP composite to rupture at the pier/base interface and toe crushing at = 16 on both the east and west ends of the pier. Rupture of the FRP composite strips caused a drop in the lateral load capacity and the test was terminated. This is evident in Figure 15, when there is a reduction in capacity at a drift ratio of .7%. Even after rupture, the pier retained approximately 50% of its peak capacity. The low modulus of elasticity (low stiffness) of the epoxy allowed the FRP composite to stretch during loading and return to its original state during unloading and kept the north face of the pier free from experiencing extensive cracking, which occurred thorough every bed joint in the south face of the pier.

PAGE 46

27 CMU 7. The limiting behavior for CMU 7 was flexure, evident in the pinched loops in Figure 16. No sudden drop in lateral capacity was found for CMU 7 and the drift capacity continued to increase. The steel reinforcement did not reach yield stress. At a = 4, the specimen rocked about the area outlined by the white line in Figure 24. The FRP composite remained unaffected through rocking. Tension in the jamb steel caused cracking in the masonry surrounding the grouted jamb steel in the base, creating an arched crack around the bottom of the grouted column. During loading, the steel reinforcement in the sill flexed, which was observed by the vertical movement of the sills. From Figure 16, the maximum drift ratios were -3.3% and 3.2%. The maximum drift ratio was taken from the last cycle that was tested. The drift ratios achieved from this test were three times larger than achieved for the others. Very little research has been performed on masonry piers with jamb and sill reinforcement. Therefore, it is difficult to conclude if the drift ratios found are reasonable for this situation. Figure 24: Specimen rocked about the area outlined by the white line. CMU 8. The first cracks noticed during testing occurred at = 1 in the horizontal joint right above where the dowels were grouted in the pier. From this location, the FRP

PAGE 47

28 composite (V) started to debond. As displacements increased, the debonding continued toward the bottom of the pier. A drop in the lateral capacity occurred when the FRP composite strip on the east end of the pier debonded along the length of the grouted cell (Figure 25). The FRP composite strip on the west end debonded but was not able to reach the bottom of the pier. It is expected that the debonding on the east end would have continued down the pier if displacements had increased. Yielding of the dowels occurred at a drift ratio of 0.7% ( = 6). The failure of CMU 8 can be attributed to a failure in the bond between the FRP composite and the grout used for installing the masonry. A 60% loss in capacity was observed in Figure 17 as the FRP composite strip on the east end completely debonded. Recall that CMU 8 is the only specimen in which the flexural FRP composite was directly bonded to the grout enclosing the dowels. All other specimens bonded the flexural FRP composite onto the masonry. A B Figure 25: The FRP composite debonded from the east end of the pier on CMU 8. A) Front view of the FRP composite debonding. B) Side view shows that the debonding started at the joint above the dowels and continued down the pier.

PAGE 48

29 Load-Displacement Envelopes Backbone curves for specimens with and without jamb steel are shown in Figure 26 and Figure 27, respectively. These curves were developed in accordance with the acceptance criteria presented in FEMA 273 for new materials. The backbone curves generally start with a higher uncracked stiffness. As the drift approaches 0.1% drift ratio, a dramatic reduction in stiffness is apparent. In Figure 26, the different backbone curve of CMU 7 is attributed to its behavior during testing. CMU 7 did not experience a drop in lateral capacity but instead increased drift ratio without increasing in lateral capacity. -30-20-100102030-2-1012Drift (%)Load (kips)-133.5-89-44.5044.589133.5Load (kN) CMU 3 CMU 4 CMU 7 Figure 26: Backbone curves for specimens with jamb steel. -30-20-100102030-2-1012Drift (%)Load (kips)-133.5-89-44.5044.589133.5Load (kN) CMU 1 CMU 2 CMU 5 CMU 6 CMU 8 Figure 27: Backbone curves for specimens without jamb steel.

PAGE 49

30 The behavior observed can be represented by the force-displacement curve shown in Figure 28. Moon, Leon et al. (2002) suggest that the initial stiffness, k, is that of an uncracked specimen. A reduced stiffness after cracking, k, is caused by a combination of cracking, yielding and damage to the specimen. The specimen stiffness can be further reduced if the FRP composite bond is inadequate. This softening effect is represented by the second branch of the curve and is caused by bond failure. The dotted line represents the sharp reduction in strength caused by the failure of the FRP composite. After this point, the pier returns to the strength found in an unreinforced specimen. % Drift k k'k' Adequate FRP Bond Unstrengthened CapacityInadequate FRP BondLoad Figure 28: Force-displacement curve for URM strengthened with fully bonded FRP composite and debonding FRP composite. System Ductility One of the goals of this research was to assess the ductility of several systems that include FRP composites and steel. One method is to use the FEMA 273 requirements for assessing the ductility of new material. System or component ductility is quantified by the following equation found in FEMA 273: UDCEQQm (1)

PAGE 50

31 where Q CE is the resistance capacity, Q UD is the seismic demand, is the knowledge factor used to account for material property uncertainty and m is the component demand modifier or ductility capacity coefficient. This equation is used to determine component and system capacity using equivalent lateral force procedures. As system or component ductility increase, the value of m increases proportionally, resulting in a reduction in lateral load demand through inelastic response. The procedures outlined in FEMA 273 for determining m for reinforced masonry assume that the reinforcement is a ductile material and, therefore, deformation controlled. Brittle materials (such as masonry and FRP composites) are referred to as force controlled and, at the component level, are assumed to exhibit a linear stiffness to failure. However, systems strengthened with brittle materials show behavior similar to that of a deformation controlled system, with displacement capacity and energy dissipation developed from the cracking and damage sustained by the specimen. This is seen in the bilinear relationship shown in Figure 28. Because CMU 1 and CMU 6 were reinforced with FRP composites only, the traditional definition of displacement ductility, (ultimate displacement divided by the yield displacement), is not applicable. For masonry components and systems with traditional ductility the FEMA 273 m-factors are a function of the displacement curvature and ductility. In an effort to quantify the ductile nature of the specimens response, a moment-curvature analysis was conducted using the test results from specimens CMU 1 and CMU 6. The fundamental difference is that cracking was used instead of yielding when conducting the analysis.

PAGE 51

32 The curvature ductility related to cracking, ` was determined by dividing the ultimate curvature, u by the curvature at first cracking, ` cr : ucrcruMM ` (2) where M y is the yield moment and M u is the ultimate moment. The curvatures for the moment curvature analysis were obtained by using the recorded strains in the FRP composite and basic mechanics principles. The relationship between the displacement ductility and curvature ductility is given by Paulay and Priestly (1992) and is also used in FEMA 273 to determine the displacement ductility. Modifying the equation to give displacement ductility as a function of cracking gives: )5.01()1`(31`LlLlpp (3) The plastic hinge length, l p is defined as the length over which the plastic curvature is assumed to equal the maximum plastic curvature. It is recognized that without ductile material (as in the cases of CMU 1 and CMU 6), plasticity can never be achieved. For the purposes of evaluation, an estimation of the plastic hinge length is made using the equation suggested by FEMA 273 where the plastic rotations at the base of the component have been limited to a plastic hinge zone length equal to: effphLl04.02.0 (4) where L is the length of the wall or the pier and h eff is the height from the base to the lateral force. FEMA 273 uses the moment curvature approach to find m-factors for the Collapse Prevention Performance Level where m is equal to the calculated displacement ductility.

PAGE 52

33 For the Life Safety Performance level, the m-factor is one-third the value of the Collapse Prevention Level. The procedures outlined in FEMA 273 are used to calculate the m-factor for the Life Safety Performance Level. Using the method described, m-factors for components with non-ductile reinforcement were calculated (Table 5). The first column are the ` calculated using the moment curvature analysis and the second column shows the m-factors calculated using the procedures outlined in FEMA 273 for new materials. The m-factor shown in Table 5 using the moment curvature analysis is one-third the value calculated for This allows for a direct comparison between the methods for the Life Safety Performance Level. Table 5: Calculated m-factors Test Moment Curvature Analysis FEMA 273 Procedures CMU 1 2.8 2.7 CMU 6 4.1 1.5 The remainder of the specimens contained steel reinforcement that was instrumented with strain gauges to determine when yielding occurred. Consequently, the displacement ductility, can be calculated directly from: yu (5) where u and y are the measured displacements at ultimate and first yield, respectively. The results of this analysis are shown in Table 6 for positive and negative load direction. For comparison, displacement ductility at first cracking, ` was calculated for each of the specimens with steel reinforcement and are also shown in Table 6. For each specimen, the ductility ratio calculated is higher using ` y than the ratio calculated using y This is reasonable because the change in specimen stiffness occurs at

PAGE 53

34 a lower displacement than the displacement at yielding because of cracking sustained by the specimen prior to yielding. Table 6: Calculated Displacement Ductility Test y (in) [mm] cr (in) [mm] u (in) [mm] ` n/a 0.071 [1.8] 0.33 [6.6] 1.0* 4.6 CMU 2 n/a -0.062 [-1.6] -0.26 [-6.6] 1.0* 4.2 0.32 [8.1] 0.082 [2.1] 0.33 [8.4] 1.03 4.0 CMU 3 -0.37 [-9.4] -0.077 [-1.9] -0.42 [-10.7] 1.1 5.5 0.14 [3.6] 0.077 [1.9] 0.41 [10.4] 2.9 5.3 CMU 4 -0.20 [-5.1] -0.093 [-2.4] -0.58 [-14.7] 2.9 6.2 0.13 [3.3] 0.034 [0.9] 0.29 [7.4] 2.2 8.5 CMU 5 -0.17 [-4.3] -0.034 [-0.9] -0.30 [-7.6] 1.8 8.8 n/a 0.067 [1.7] 0.84 [21.4] 1.0* 12.5 CMU 7 n/a -0.069 [-1.8] -0.98 [-24.8] 1.0* 14.2 n/a 0.047 [1.2] 0.30 [7.6] n/a 6.4 CMU 8 -0.26 [-6.6] -0.028 [-0.7] -0.27 [-6.9] 1.04 9.6 *Note: No yielding in steel, is 1.0. Recall that the displacement ductility is equivalent to the value of m in FEMA 273. Table 7-4 in FEMA 273 provides acceptable m-factors for reinforced masonry. Values given in this table for the cases of CMU 2, CMU 5 and CMU 8 are 4.1, 4.8 and 4.8, respectively. The table in FEMA 273 does not apply to the cases with jamb and sill steel; so acceptable m-factors for CMU 3, CMU 4 and CMU 7 could not be determined. Comparing the ductility ratios in Table 6 to the acceptable values, the ductility ratios found using ` y are closer to the accepted values than the ductility ratios found using y

PAGE 54

35 Further research should be conducted using similar configurations to support the use of the ductility ratios presented in this paper. Computing Predicted Capacities For each of the specimens, the flexural capacity was calculated for two critical sections. Holberg and Hamilton (2001) describe the methodology for the calculations. The first critical section checked is where the FRP composite ends and flexural capacity is provided solely by the reinforcing steel bars. The second critical section checked is where the FRP composite strips placed on the pier jambs provide the flexural capacity. At this section, the FRP composite strip is providing all the flexural capacity. In Figure 29, P s is the axial forces provided by the spring, P w is the self-weight of the pier and concrete cap and Q is the lateral load carrying capacity of the section. The flexural capacity of both sections can be determined using basic mechanics principles and the traditional rectangular stress block assumption. Assuming an under reinforced condition, the depth of the stress block, a, is: emwsysbfPPfAa`85.0 (6) where b e is the effective thickness of the masonry, A s is the cross sectional area of the steel, f` m is the compressive strength of the masonry and f y is the yield strength of the steel reinforcement. Applying equilibrium to the section gives the moment capacity of the section where the flexural strength is provided by the steel reinforcement: )22()()2(alPPadfAMwsysnbar (7) The lateral force, Q, required to yield the bar is: effnbarhMQ (8)

PAGE 55

36 For the critical section reinforced by composite only the depth of the stress block, a, is: emwsbfPPwTa`85.0 (9) where T is the tensile strength of the FRP composite and w is the width of the strip used. Applying equilibrium as before, the moment capacity of the section where the composite is providing the flexural strength is: )22()()2(alPPadwTMwsnbar (10) The lateral force, Q, required to rupture the FRP composite at this critical section is: seffnbarlhMQ (11) where l s is the length of the lap splice between the composite strip and the steel reinforcement. Ps Q h eff ls Pw Critical SectionMnbar d lCritical SectionMnfrp Figure 29: Schematic of the pier connected to the base with dowels.

PAGE 56

37 The measured lateral force at first yield, Q m is compared to the calculated capacity, Q nbar in Table 7 Calculated values for both the steel reinforcement and FRP composite are shown in Table 7. Table 7: Measured and Calculated Lateral Capacities Specimen Measured Capacity Q m (kips) [kN] Q nbar (kips) [kN] Q nfrp (kips) [kN] Ratio (Q m /Q nbar ) 17.7 [79] 17.7 [79] 1.0* CMU 1 -14.9 [-66] n/a -17.7 [-79] 0.84* 14.7 [65] 13.6 [61] 23.1 [103] 1.08 CMU 2 -16.6 [-74] -13.6 [-61] -23.1 [-103] 1.22 24.9 [111] 23.4 [104] 31.9 [142] 1.03 CMU 3 -27.4 [-122] -23.4 [-104] -31.9 [-142] 1.17 n/a 22.1 [98] 36.5 [162] n/a CMU 4 -21.7 [-97] -22.1 [-98] -36.5 [-162] 0.98 9.7 [43] 12 [53] 23.2 [103] 0.81 CMU 5 -11.2 [-50] -12 [-53] -23.2 [-103] 0.93 11.3 [50] 10.6 [47] 1.1 CMU 6 -10.3 [-46] n/a -10.6 [-47] 0.97 9.9 [44] 13.6 [61] 66.7 [297] 0.73 CMU 7 -8.9 [-40] -13.6 [-61] -66.7 [297] 0.65 n/a 12.2 [54] 23.2 [103] n/a CMU 8 -14.1 [63] -12.2 [-54] -23.2 [-103] 1.16 *Note: CMU 1 and CMU 6 did not contain steel. The ratio is calculated as Q m /Q nfrp The ratio of measured capacity at first yield to calculated capacity, Q nbar is also given. CMU 1 and CMU 6 did not contain any steel reinforcement, so there is no calculated value for Q nbar For CMU 1 and CMU 6, the ratio of the maximum measured

PAGE 57

38 capacity to calculated capacity of the FRP composite, Q nfrp is shown. CMU 4 and CMU 8 did not have strain gauges on the west dowels, so it is not known when first yield occurred for the positive loading direction. For most cases, agreement between the measured and calculated capacities falls within 10%. In the case of CMU 5 and CMU 7, there is a more than 20% overestimation by the analytical model. CMU 5 failed prematurely because of the splice failure, which could account for the difference. CMU 7 was limited by rocking about the base of the specimen, below the reinforcing steel bars. Proper confinement and placement of the FRP composite would help in increasing the lateral load capacities of these two specimens.

PAGE 58

CONCLUSIONS Eight concrete masonry piers were strengthened with a combination of FRP composite strips and reinforcing steel. The steel reinforcement was designed to yield prior to rupturing the FRP composite strips. Confinement using FRP composite sheets was provided in the base of the pier to strengthen the masonry against dowel pull out. For most cases, the confinement provided the masonry with the additional strength needed to allow yielding of the steel dowels. In the two cases where the steel reinforcement did not yield, insufficient development length (CMU 2) and a lack of confinement in the base (CMU 7) were to blame. The use of FRP composites in conjunction with steel reinforcement shows potential for improving the behavior of URM structures during seismic events. More research should be conducted on similar configurations in order to quantify the improvement in ductility and appropriate m-factors for reducing seismic demand. Key findings and conclusions are as follows: 1. Improvement in the ductility, lateral capacity and energy dissipation were achieved by adding a FRP/steel strengthening system to the specimens. A drift ratio of 1.8% and a lateral load capacity of 28 kips (125 kN) was achieved. Open loops in load-displacement plots indicate energy dissipation as reinforcing steel yielded. 2. Changes in specimen stiffness (from 10% to 20%) were observed at a drift ratio of 0.1% during testing as the specimens sustained damage through cracking, yielding and debonding of the FRP composite. 3. Yielding was achieved for the specimens that had FRP composite sheets confining the steel reinforcement in the masonry base against bar pull out. 39

PAGE 59

40 4. Ductility factors (m-factors) were developed for the specimens containing steel reinforcement and compared to acceptable values found in FEMA 273. 5. Measured load capacities of the specimens were compared to values calculated using the analytical model presented by Holberg and Hamilton (2002). Agreement between measured and calculated capacities fell within 10% for the majority of the specimens tested.

PAGE 60

APPENDIX A EXTENDED LITERATURE REVIEW A. E. Schultz and R. S. Hutchinson (2001) Schultz and Hutchinson present a completed project on partially-grouted masonry shear walls. The project program consisted of simulated seismic load experiments of partially-grouted masonry walls and both empirical and finite element modeling of shear wall behavior. In this project, all horizontal reinforcement is provided by welded wire grids that are placed in the bed joints of the masonry. The principal variables were height-to-length aspect ratio and horizontal reinforcement ratio. Results presented include modes of response, global force-displacement characteristics and level deformation response. Throughout testing, all the walls showed elastic force-displacement behavior for forces up to at least half of the yield strength of specimen. Inelastic behavior began with vertical cracking of the top course of masonry near the section where ungrouted meets grouted masonry. Increasing the horizontal reinforcement ratio had a small positive influence of deformation capacity and shear strength. Horizontal reinforcement ratio was found to have a modest effect on ultimate shear stress and deformation capacity, but it did not help effect energy dissipation. O. S. Marshall and S. C. Sweeney (2002) In-plane shear tests were conducted on 4 ft. by 4 ft. unreinforced double-wythe brick wall specimens and lightly reinforced single-wythe CMU wall specimens. Tests 41

PAGE 61

42 were conducted at the U.S. Army Engineer Research and Development Center Construction Engineering Research Laboratory (ERDC/CERL). Specimens were constructed to simulate the I shaped piers. Top and bottom of I was wrapped in E-glass or carbon fabric composites to provide confinement, to provide anchorage for the FRP covering the central portion of the specimen and to force failures into the pier section of the specimen. Two gravity load conditions were tested: a low gravity load of 75 psi and a higher gravity load of 150 psi. The researchers expected that the failure modes for the low axial loaded specimens would be bed joint sliding. For low gravity loads, it was discovered that the strength of the specimen was increased in relation to the amount of material that crossed the failure plane (the horizontal bed joint). For the higher gravity loads, the authors claim that the expected failure mode was rocking or X-cracking. In the research, it was discovered that the full-coverage and X-pattern FRP configurations worked best for the higher axial loads. Also, it was found that multiple plies of FRP do not always increase the in-plane capacity. The authors conclude that the multiple plies become too stiff, due to their thickness, and fail due to delamination of the material from the wall. A similar behavior for carbon composites was observed by the researchers. Authors recommend that testing be conducted to develop configurations of FRP that prevent X-cracking while transferring the failure to a more ductile mode, like bed joint sliding or rocking prior to toe crushing.

PAGE 62

43 F.L. Moon, T. Yi, R.T. Leon and L.F. Kahn (2002) This paper is the analysis part of an ongoing project involving a full-scale two-story URM test structure. Both force-controlled and displacement-controlled analyses were conducted for an unreinforced state, retrofit with FRP composites and retrofit through post tensioning of the test structure. Results from the analyses will be used in testing a full-scale URM building at Georgia Tech. The authors believe that the use of FRP composite strips versus full coverage sheets is superior from both an economic and behavioral standpoint. They also state that strips should be placed according to the type of URM failure mode expected for that particular pier and to minimize tensile rupture of FRP composite, compressive failure of masonry (toe crushing) and shear failure of bed joint FRP. For the force-controlled pushover analysis, a lateral force was applied in a triangle distribution to the model test structure. The force was distributed to the individual piers according to their elastic stiffness until yielding of the pier occurred. Yielding of the pier was defined as the point at which the load exceeded the pier capacity (considering all the possible URM failure modes). The displacement-controlled pushover analysis involved modeling each individual pier as a series of force-displacement curves for each of the considered failure modes. This resulted in the analysis considering the difference between brittle and ductile failure modes and providing a displacement capacity for the entire structure. The displacement capacity is the ultimate displacement of a pier before collapse. Results from the analysis showed that the use of FRP strips is effective in increasing the strength while not decreasing the displacement capacity of the structure.

PAGE 63

44 J. G. Tumialan, A. San Bartolome and A. Nanni (2003) The authors present preliminary results of research on the in-plane behavior of unreinforced infill concrete masonry walls strengthened with FRP structural repointing. FRP structural repointing consists of placing FRP bars in the bed joints of the masonry walls in order to improve shear capacity. The authors indicate that the advantages of structural repointing include simplicity, less surface preparation (sandblasting, puttying, etc.) and the masonry structure aesthetics are partially conserved. In the case where infill walls are in contact with a concrete frame, the interaction between the wall and the frame should be considered. Ignoring this will lead to an unconservative design because the infill walls can stiffen the frames and cause redistribution of lateral loads to potentially weaker elements. Walls were tested in both in and out-of-plane loading. The experimental program consisted of three types of test performed on the same walls: Part 1 was testing of specimens subjected to in-plane loading. Part 2 was placing specimens on a shake table where they were subjected to out-of-plane accelerations. Part 3 consisted of retesting the specimens under in-plane loading by loading it until it reached its maximum capacity. Only Part 1 of this experimental program is discussed in this report. Specimens were tested under in-plane cyclic lateral load by following a displacement-controlled method. It was concluded that specimens strengthened with FRP can reach a lateral drift of up to 0.7% without losing lateral load carrying capacity. There were also more, but finer, cracks in the FRP reinforced specimens than in the unreinforced ones. The researchers calculated the absorbed energy as the area under the loading portion of the load vs. displacement curve for each phase. They found that the absorbed

PAGE 64

45 energy increased for the strengthened specimens for a drift greater than 0.5%. For drift greater than 0.7%, the reinforced specimens had 40% more energy absorption than did the unreinforced specimens. M. J. Chajes, W. W. Finch, Jr., T. F. Januszka and T. A. Thompson, Jr. (1996) This paper presents the results of direct bond tests performed on joints consisting of composite material plates bonded to concrete. The focus of the tests was on bond strength and force transfer. To evaluate the effects of surface preparation, type of adhesive and concrete strength on average bond strength, tests were performed using the single-tap shear test specimen and constant bond length. Other tests were also performed to study the force transfer from the composite material plates into the concrete. Researchers concluded that surface preparation of the concrete can influence ultimate bond strength. They also found that the use of ductile adhesives (i.e., those having a low stiffness and large strain to failure) leads to a less effective bond that fails before there is a shear failure of the concrete. If the failure mode of the joint is governed by shearing of the concrete directly beneath the bond, the values of the ultimate bond strength will be proportional to f c They found that the force transfer is largely uniform along the bonded length. Y. C. Kurama (2002) Kurama investigated a precast hybrid wall system that uses mild steel reinforcement as well as post-tensioning steel for flexural strength and inelastic energy dissipation. First, an analytical parametric investigation was performed on the wall specimens. The effect of varying the relative quantities of mild steel and post-tensioning steel under combined gravity and lateral loads were examined. In addition, a series of

PAGE 65

46 nonlinear dynamic time history analyses were conducted to investigate the effect of the mild steel on the behavior of the walls under seismic loading. The paper goes into detail on how to evaluate the amount of energy dissipation and absorption using drift vs. base shear graphs. The investigation concluded that the inelastic-energy dissipations of unbonded post-tensioned precast walls can be increased by using bonded mild steel reinforcement crossing the horizontal joints. It was also found that under high seismic loading, the use of mild steel decreases the maximum lateral displacement of the wall and decreases the number of large displacement peaks because the response of the wall decays faster. For regions of moderate seismisity, it was found in this study that there was little difference in the dynamic response of the unbonded post-tensioned, hybrid and mild steel only precast walls. Also, the use of mild steel reinforcement does not have a significant effect on the self-centering capability of the walls. K. Bajpai and D. Duthinh (2003) Concrete masonry beams were reinforced with surface mounted FRP rebar and tested in four-point bending. Close to full strength development of inch glass FRP bars was achieved in 7.3 inches of concrete masonry (with is less than one half a concrete masonry unit). The bars were embedded in epoxy in grooves cut into the face of the masonry and mortar joints. Bond tests were also performed on concrete masonry prisms in order to investigate the development length of FRP bars. Results showed that smaller bars achieved higher bond strength relative to their tensile strength. It was stated that this was expected because smaller bars have a higher perimeter to cross sectional area ratio than do the

PAGE 66

47 larger bars. Bond tests also concluded that sand-coated bars with a helical fiber achieved higher bond strength than those with circular ribs on a smooth surface finish. It was also concluded that half a masonry unit length is sufficient to develop a inch bar close to its full strength. Two series of bending tests were conducted. Series 1 had FRP bars running parallel to the mortar joints and Series 2 had the bars running perpendicular. Each series consisted of two narrow and two wide beams. The bars were installed in inch square grooves along the entire length of the beam. Beams were fully grouted to ensure shear resistance. The beams were then tested under four-point bending. They all failed in flexure by rupture of the FRP bars. It was concluded that the ACI 530-02 equations for the flexural strength of masonry bars under-reinforced with steel provide a conservative estimate of the flexural strength of concrete masonry beams and walls reinforced with near-surface mounted FRP bars. C. Sittipunt, S.L. Wood, P. Lukkunaprasit and P. Pattarattankul (2001) Four reinforced concrete wall specimens were tested to investigate the influence of diagonal web steel reinforcement on the hysteretic response of structural walls. The walls all had barbell-shaped cross sections. The longitudinal and transverse reinforcement in the boundary elements were the same in all four walls. The primary variables were the amount and orientation of the web reinforcement. The specimens were loaded laterally through the top beam. No axial load was applied to any of the specimens, although they were braced for out-of-plane displacements. The loading was applied in two stages. The first set of cycles was force-controlled, where the walls were pushed until cracking. The second stage of testing

PAGE 67

48 consisted of displacement-controlled cycles. The top of the specimen was pushed so that the deflection at the top was a multiple of the yield displacement in both directions. The yield displacement was determined by monitoring the strains in the longitudinal steel in the boundary elements for each specimen. It was found that all four specimens reached maximum loads that exceeded the calculated nominal capacities. The walls with the diagonal web reinforcement resisted higher loads than those with the conventional vertical and horizontal web reinforcement. There was no significant increase in strength for those walls with higher web reinforcement ratios. The walls with diagonal reinforcement dissipated more energy than those with conventional reinforcement. When loaded laterally, walls with conventional reinforcement transferred shear through compression struts and aggregate interlock within the concrete and dowel action of the web reinforcement. When subjected to cyclic deformations, these mechanisms degraded. Walls with diagonal web reinforcement transferred shear through tensile forces in the reinforcement. This form of energy dissipation is stable and did not degrade during testing. T.C. Triantafillou (1998) Triantafillou proposed an analytical model for the short-term strength of masonry walls reinforced with externally bonded FRP laminates under monotonic out-of-plane bending, in-plane bending and in-plane shear in combination with axial load. The analysis consists of both equations and normalized interaction diagrams. Testing was conducted to verify this model.

PAGE 68

49 For the analysis, the author assumed that the laminates act as truss like elements. The author states that this is a fair assumption because FRP is usually applied as narrow straps and work in tension, much like truss elements. In the case of full coverage FRP, he would have had to consider a more complicated laminate theory. The author states that the typical failure for out-of-plane bending with axial force is compressive crushing and that the typical failure for in-plane bending can be compressive crushing or shearing of FPR in the tension zone directly beneath the bond (in the cases of short development length). Triantafillou proposed that the analysis and design of reinforced masonry is based on the assumption that the total shear capacity is the sum of two terms, V RD1 and V RD2 V RD1 is the contribution from the uncracked masonry and V RD2 is from the effect of shear reinforcement (modeled from truss analogy). The shear capacity of masonry, V RD is given from Equation A-1 (in SI units) adopted from Eurocode 6, where f vk is the characteristic shear strength of the masonry. VRDfvktdm VRD2 x.3fktdm (A-1) The author assumes that the contribution of vertical FRP reinforcement, which provides mainly a dowel effect, is negligible. Therefore, the only shear resistance mechanism left is the action of the horizontal laminates. This can be modeled similar to stirrups in reinforced concrete beams. Triantafillou reported Equation A-2, where frpe is the effective FRP strain from Equation A-3. Both Equation A-2 and A-3 are in SI units. VRD2.7frp hEfrpfrpelt (A-2)

PAGE 69

50 frpe.0119.0205hEfrp .0104hEfrp 2 (A-3) To test his analytical model, twelve identical small wall specimens were constructed of perforated clay units and statically tested in four-point bending to study the effect of FRP reinforcement on the failure mechanism and load capacity of the wall. Six were tested out-of-plane and six in-plane. Walls tested in out-of-plane bending had FRP laminates bonded to the tension face. Those walls tested in in-plane bending were reinforced symmetrically on both faces. The FRP covered walls tested in out-of-plane bending failed by crushing of the masonry in the compression zone, indicating flexural failure. There was good agreement between analysis and experiments, with a maximum error equal to approximately 15%. Those walls tested in in-plane bending failed prematurely due to peeling off of the FRP laminates in the tension zone directly beneath the bond. It was found that this action was due to the short bond development length because of the small size of the specimens. Consequently, it was concluded that the achievement of full in-plane flexural strength depends on proper anchorage. M.J.N. Priestley and F. Seible (1995) This experimental program focuses on implementing techniques aimed at increasing flexural ductility by limiting shear deformations and stabilizing critical compression toe regions on masonry walls. The program consisted of testing a full-scale five-story reinforced hollow concrete masonry building. The test building was tested under simulated seismic actions to failure. The first series of tests were run on the building with no FRP retrofits. Afterwards, the test structure was repaired by means of

PAGE 70

51 structural carbon overlays on the first two stories of the structural walls. A single layer was applied on each side of the walls with two layers in the toe regions. Results showed that repairing with composites doubled the displacement capacity of the top story. Measured shear displacements in the repaired walls were reduced to half the measured shear displacements of the original five-story building. It was stated that the forces that are transferred in the composite overlays are limited by the laminar shear or principal tensile strength of the original structural wall material. This is because the polymer resin usually has a higher tensile capacity than the masonry it lays on. Equation A-4 for shear capacity of the composite overlay was developed using a conservative diagonal tension crack angle assumption of 45 where f o is the allowable overlay stress level based on a maximum allowable strain of .004, d is the effective structural wall length and t is the thickness of the composite overlay. Vofotid (A-4) Stiffness criteria, rather than strength criteria, was employed in the wall overlay design. Shear deformations were limited to deformation levels which can be expected in concrete walls with conventional horizontal reinforcement, A shreq by scaling the amount of horizontal overlay fabric, A oh from the required horizontal reinforcement in Equation A-5. E o is the modulus of elasticity of the overlay material and E s is the modulus of elasticity of the substrate (in this case it is masonry). AohAshreqEsEo (A-5)

PAGE 71

52 The researchers compared the bond between the FRP composite and masonry to that of traditional reinforcement and unconfined concrete. They borrowed the upper limits to the total gain on shear capacity from conventional concrete design (ACI-318) in Equation A-6, where f`c is the nominal concrete strength in MPa, b w is the wall width and d is the effective wall length. max Vo = max Vs x V s.66f'c bwd (A-6) A.M. Holberg and H.R. Hamilton III (2001) URM walls strengthened with a hybrid system containing both FRP composites and steel was investigated. The ductile connection was designed to yield before the composite failed. This system was to increase lateral capacity and ductility to the URM. Four unreinforced hollow concrete masonry shear wall specimens were strengthened with the hybrid system and then tested. Vertical strips of the FRP were placed at each end of the wall to increase the in-plane flexural capacity and diagonal strips were applied to improve shear capacity. The walls were sized to ensure a rocking failure mode for an unreinforced specimen. Two types of steel connections were used: internal and external. The internal connection was a steel reinforcing bar placed in the outermost cells of the wall and fully grouted. The external connection was a steel angle-plate assembly attached to the foundation of the test setup. The GFRP strips were then folded into two layers under the plate. The drift capacities of the reinforced specimens reached up to 1.7%. The lateral capacity of a strengthened specimen was nearly doubled. The external connection did not prevent sliding, which provides good energy dissipation. For the specimens with an

PAGE 72

53 internal connection, an eccentricity was caused between the reinforcing bar and the GFRP composite which caused an out-of-plane displacement at the critical section. The eccentricity was improved for other specimens which caused an improvement in the displacement capacity of the specimen. G.M. Calvi, G.R. Kingsley and G. Magenes (1996) The paper addresses problems associated with the experimental evaluation of the URM buildings subjected to seismic loading. It focuses on the merits and roles of several experimental techniques, which include quasi-static, dynamic and pseudo dynamic loading at full and reduced scale. The authors indicate that masonry is highly sensitive to loading rates, boundary conditions and effects of axial loading. They also noted that the final collapse of URM buildings is often associated with in-plane shear failure of the piers of a critical story (usually the ground story most axial load). Paper discusses test procedures of past testing of masonry piers. The large majority of previous testing focused on in-plane shear. The boundary conditions at the top and bottom of the wall are either fixed-fixed or fixed-free. The authors suggest that a real seismic excitation is better simulated in dynamic tests, but acknowledge that quasi-static tests have several advantages over dynamic shaking table tests. Quasi-static tests allow for an easier application of large forces, for easier observance of cracking and for the more accurate measurement of forces and displacements. Quasi-static tests tend to show more extensive damage and lower strength than dynamic tests because masonry exhibits rate-dependant behavior, which allows

PAGE 73

54 propagation of cracking at constant load or at constant imposed displacement. It was concluded that, in general, static cyclic testing is a conservative approach. From past experiments, given material properties, the main parameters determining the pier failure mechanism are the axial load, the aspect ratio and the boundary conditions at the ends. Once boundary conditions are determined, increasing aspect ratios tend to lead to a rocking failure and decreasing aspect ratios lead to shear failures, like diagonal cracking or shear sliding. Maintaining constant geometry and boundary conditions, an increase in axial load leads to increased chances for diagonal shear failure and low axial loads are usually associated with sliding and rocking. The ability to predict the failure mechanism is important. Both static and dynamic experiments have shown how rocking and sliding tend to be associated with good seismic behavior and diagonal cracking to brittle behavior. Rocking provides a limit to the maximum shear and dissipated energy through impact, keeping the integrity of the wall except for the localized damage in the corners. Sliding on horizontal bedjoints is another good dissipative mechanism also maintaining the overall integrity of the wall, although it can also result in larger displacements. Diagonal shear cracking is associated with brittle failure under dynamic loading, can be catastrophic. M.R.Ehsani, H. Saadatmanesh and J.I. Velazques-Dimas (1999) This paper presents the experimental results of three half-scale URM walls retrofitted with E-glass FRP strips and tested under cyclic out-of-plane bending. The unstable and explosive out-of-plane failure endangers the gravity load carrying capacity of the wall. For this series of tests, no axial load was applied out-of-plane failures are characterized by cracks along the mortar bedjoints while in-plane failures have a diagonal

PAGE 74

55 crack pattern. The study presented was conducted to demonstrate the feasibility of retrofitting with GFRP on the tension face for structures subjected to reverse cyclic loading. The specimens were intended to represent a typical load-bearing wall in a low-rise building away from corners. Specimens were retrofitted with vertical strips of FRP applied using the wet lay-up procedure. Two different glass fabric densities and five reinforcement ratios were investigated. It was concluded from the observed behavior of this set of URM masonry walls retrofitted with FRP that the ultimate flexural strength was significantly increased. The walls deflected almost 14 times the maximum allowable deflection found in masonry specifications. These deflections were as much as 2.5% of the wall height. D.P. Abrams (2001) Author states that because of rocking and sliding along mortar bed joints, URM wall and pier elements can be considered to behave as displacement controlled components with significant capacity to accept large nonlinear deflections. One common explanation for damage to numerous URM buildings during earthquakes is that the building has not been engineered to resist lateral seismic forces either at the time of the original design or at a later time when rehabilitation should have been considered. The US is researching and developing rehabilitation focused on the implementation of performance-based guidelines (FEMA 273 and FEMA 356). This will result in various performance levels for varying degrees of structural intervention. These new guidelines will not just be based on strengthening but also on enhancing deformation

PAGE 75

56 capacity. This results in different ductilities, or m factors. The ductility factor, m, expresses the lateral drift at a specific performance state versus yield displacement. This paper describes a recent testing program at the Mid-America Earthquake Center which investigated the basic mechanisms for unreinforced brick masonry wall. The piers used had a constant height-to-length ratio equal to 1.8, so that flexural mechanisms (rocking and toe compression) could be studied. Piers strengthened by FRP, reinforced and prestressed cores, reinforced shotcrete overlays and surface coating were tested. The pier strengthened with the FRP behaved with a significant increase in lateral strength (over three times that of the unreinforced pier). As the FRP delaminated from the surface, the lateral stiffness of the pier gradually reduced. First signs of delamination were observed at 0.04% drift and continued until the lateral pier strength was completely lost at 1.9% drift. Diagonal stair-stepped cracks developed at 0.6% drift. This was due to the larger shear forces brought on by the enhanced flexural strength provided by the FRP. Additional Literature Review Masonry and Earthquakes Recent earthquakes have revealed extensive damage to existing unreinforced masonry structures. Many older buildings currently in use were designed and constructed with little or no consideration to earthquake resistance. In addition, changed usage and more stringent seismic requirements have left many buildings in need of additional strengthening. In deciding to retrofit a structure, it is important to know the failure mode of the components in order to restrain that mode or change it to a more favorable failure mode.

PAGE 76

57 Failure modes for masonry piers depend on the combination of applied loads, pier geometry and properties of the materials. Typical failure modes for URM are illustrated in Figure 31. These failure modes are characterized by brittle behavior with rapid decreases in capacity and very limited deformations after reaching the ultimate load. The most energy dissipating failure mode is bed-joint sliding. This is a result of a low, vertical compressive force. Rocking is the second best failure mode in terms of energy dissipation. It is deformation controlled, which allows large lateral displacements. Diagonal tension, or X-cracking, failure is characterized by diagonal cracks formed through the mortar joints or the masonry units themselves. Toe crushing is the least energy dissipating of the four failure modes. In this failure mode, high gravity loads or large overturning forces crush the masonry units. Figure 30: Pier area outlined on a structure.

PAGE 77

58 Figure 31: Failure modes for unreinforced masonry. The goal of retrofit design is to change the failure mode of a pier from a brittle mode, X-cracking or toe crushing, to a more favorable failure mode, rocking or bed joint sliding. If a URM pier is determined to have a rocking or bed joint sliding failure mode, not adding any strengthening may be the best option. The URM piers that have been determined to fail in X-cracking or toe crushing would need to be strengthened to change the failure mode to bed joint sliding or rocking. The ideal retrofit design would allow the wall to dissipate energy while not losing its ability to carry axial load, give the wall an in-plane strength close to or greater than the strength of the wall without any reinforcement and, during a seismic event, not allow crushed masonry to fall on those in the structure or those trying to exit. Conventional Strengthening and Rehabilitation Options There are several methods available for retrofitting existing structures that increase the strength and ductility of masonry buildings. Conventional strengthening methods can

PAGE 78

59 be costly, add significant mass to the structure, reduce the amount of useable space in the structure and negatively affect aesthetics. One of the most common methods of retrofitting existing masonry structures involves removing one or more wythes of masonry and filling the void with pneumatically applied concrete, commonly known as shotcrete (Moon et al. 2002). This method has showed to be effective in increasing both the strength and ductility of the URM walls. Shotcrete is costly because of the amount of surface preparation needed and formwork required. This method also adds considerable weight to the structure, resulting in larger inertia forces during an earthquake. A second method involves applying a thin surface coating to one or both sides of the URM wall. Typical coatings include glass-reinforced cement, ferrocement and a wire mesh reinforced cement. Other methods of strengthening include adding reinforcing steel to the structure. Post-tensioning or prestressing has been used to enhance the tensile and flexural capacity of concrete. After core drilling from the top of masonry walls, steel is post-tensioned to the foundation. This method is costly, but does not alter the exterior of the structure. It can also be installed without disturbing the occupants. FRP Strengthening Fiber reinforced polymer (FRP) laminates are made of continuous glass, carbon or Kevlar fibers bonded to the substrate with a resin polymer matrix. FRP composite systems have been used extensively in recent years to retrofit concrete structures in high seismic zones (Marshall and Sweeney 2002). The FRP composite system provides additional reinforcement that enables the structure to have better resistance against earthquake damage. The composite gives the masonry the added strength it needs to resist the shear and flexural stresses experienced during an earthquake.

PAGE 79

60 Some of the advantages of FRP composite systems are that it is lightweight, has high stiffness, excellent fatigue properties and has a high resistance to corrosion. This technique is typically unobtrusive to the building occupants, requires little surface preparation and is very economical. The ability of a FRP composite strengthening system to keep crushed masonry from falling off the structure also makes it a favorable alternative to traditional methods. Strengthening masonry walls with FRP composites requires that the composite be bonded to one or both sides of the wall. Currently, there are two approached for this. The first approach bonds FRP composite sheets over the entire surface of the pier. The other uses FRP composite strips bonded to specific locations on the pier. The use of unidirectional strips of FRP is preferable in terms of economy and behavioral response to the two-dimensional fabrics that cover the entire surface of the masonry wall (Triantafillou 1998). Past research FRP composites have been tested as a reinforcing overlay for seismic retrofit of reinforced concrete and masonry. Results of testing done on masonry with FRP composites have shown that both in-plane and out-of-plane strength were significantly increased without decreasing the displacement capacity of the structure. Tests have concluded that the FRP composites can be applied to increase strength and change the failure modes of masonry walls (Marshall and Sweeney 2002, Abrams 2001). FRP linked with a ductile connection Although FRP composites increases lateral load capacity, it does not improve ductility. This is because of the brittle nature of the composite material. This has lead to the idea of a hybrid system, consisting of FRP composites in conjunction with steel. The

PAGE 80

61 FRP composite adds sufficient strength to the masonry allowing the steel to reach yield, incorporating ductility into the system. Figure 32 illustrates a masonry pier strengthened against flexure and shear using the hybrid system. The added FRP composite improves the flexural behavior. Terminating the FRP composite at the pier allows for sliding at the top and bottom of the pier. The ductile connection provides restraint against rocking. Sliding and rocking restrained with ductile connection Shear and flexural strength improved Sliding at top allowed Figure 32: Pier strengthened for shear and flexure with FRP composites. Rocking restrained with the ductile connection at the base of the pier. The idealized load-displacement curve for statically loaded URM cantilever pier is shown in Figure 33 (Holberg and Hamilton 2002). The lateral load capacity for the rocking mode, P OT is a function of pier geometry and axial loading (Figure 34). The slope to the rocking load is the stiffness of the URM pier. After reaching this point, the pier will continue displace without gaining additional capacity until it becomes unstable.

PAGE 81

62 Figure 33: Idealized load-displacement curve for a URM pier (Holberg and Hamilton 2002). POT Figure 34: URM pier with rocking load of P OT. Adding a ductile connection to the base will increase the lateral capacity to P = P OT + P c (Figure 35). The additional capacity, P c is a function of the tensile force T in the ductile connection and pier geometry (Figure 36). Energy dissipation will also increase, as shown by the shaded area in Figure 35. The shaded area is a result of the energy dissipation from the yielding ductile reinforcement.

PAGE 82

63 Figure 35: Idealized load-displacement curve for a pier strengthened with a ductile connection to the base (Holberg and Hamilton 2002). P T Figure 36: Pier strengthened with a ductile connection with an overturning load of P. Research was conducted by Holberg and Hamilton (2001) on URM walls retrofitted with a hybrid strengthening system consisting of FRP composites and steel. Two different types of steel connections were tested, an internal and an external. The internal connection was a steel reinforcing bar placed in the outermost cells of the wall and fully grouted. The external connection was a steel angle-plate assembly attached to the foundation. The drift capacities of the reinforced specimens reached up to 1.7%.

PAGE 83

64 There was an almost doubling of the reinforced specimens lateral capacity when compared to an unreinforced specimens capacity.

PAGE 84

EXPERIMENTAL PROGRAM Test Setup Chapter B The test set up is shown in Figure 37. A 55 kip (240 kN) hydraulic actuator is located on the reaction frame and displaces the concrete cap on the specimen. Positive load and displacement correspond to actuator tension (Figure 38). Negative direction is actuator compression. The reaction frame that supports the actuator is constructed of steel and is prevented from overturning by its connection to the laboratory strong floor. A concrete cap and base were constructed for the test fixture and designed to handle loads of 50 kips (220 kN). The concrete base was tied to the laboratory strong floor. The concrete cap was braced to the laboratory wall by steel angles to prevent out-of-plane twisting (Figure 39). Spring System 55 kip Actuator Reaction Frame Concrete Base Concrete Cap DAQ System Spring System 55 kip Actuator Reaction Frame Concrete Base Concrete Cap DAQ System Figure 37: Test set up with specimen ready for testing (looking at North face of specimen). 65

PAGE 85

66 NorthConcrete CapLateral BracingSteel Frame48 inchesActuator Displacement Sign Convention+_ Figure 38: Plan view of test set up. The North direction in the laboratory is defined. Figure 39: Angles used to prevent out-of-plane movement. The specimens were constructed separately and placed into the test fixture at the time of testing (Figure 40). The specimens were placed on the concrete base on a bed of hydrostone to allow for an even bearing surface. A layer of fresh mortar was spread on the top lintel of the specimen and the concrete cap was then immediately placed onto the specimen. The actuator was then connected to the concrete cap by four threaded rods running through the concrete cap. A steel bearing plate was placed on plywood over a vertical head joint in the base of the specimen where it would be tied to the concrete base. Steel channels were placed on the bearing plate and along the side of the base. The threaded rod connecting the sill to the base was post-tensioned using a hydraulic jack at a force of 42 kips (187 kN). The use

PAGE 86

67 of bearing plates allowed for the force of the post-tensioned threaded rod to be directed over any vertical head joint desired. The concrete cap and the top lintel were connected using threaded rod and steel channels. The nuts on these threaded rods were hand tightened. Figure 40 shows the confinement details for the specimen. Concrete BaseSpecimen 8 inches(200 mm)68 inches(1727 mm)24 inches(610 mm)18 inches(457 mm) MTS Load Downward Force = 42 kips (187 kN)Laboratory Floor Figure 40: Schematic of specimen in the test set up. A dial gauge was placed between the cap and the specimens top lintel to monitor slip between the lintel and the concrete cap (Figure 41). Throughout testing, no slipping occurred between the concrete cap and the top lintel of any of the specimens.

PAGE 87

68 Figure 41: Dial gauge to monitor slipping between the concrete cap and the specimen's top lintel. Lifting Frame A lifting frame was constructed to move the specimens (Figure 42). It consisted of a channel placed on the top of the specimen and a series of threaded rod that were tied to angles. Prior to specimen construction, 5/8 inch holes were drilled two inches from the top of the lintels. Steel pipes were then inserted in the holes and cast into place when the concrete was poured. The specimen was constructed on what was the bottom of the lintel. Threaded rod was run through the pipes and attached to steel angles. Hooks at the top of the lifting frame were attached to the laboratory crane.

PAGE 88

69 Figure 42: Lifting Frame. Axial Loading A gravity load of 9 kips (40 kN) was applied to the specimen. This translated to an axial stress of 75 psi (0.5 MPa) at the base of the pier. This gravity load included load from two rail car springs compressed to develop 6 kips (27 kN), the weight of the concrete cap (2 kips, 9 kN) and the self-weight of the pier and lintel (1 kip, 4 kN). The two springs were compressed by a hydraulic jack pushing upward on the top steel plate and held in compression by tightening the nuts on the top of two threaded rods tied between the steel angle system on the concrete cap and the steel connection post

PAGE 89

70 tensioned to the concrete base (Figure 43). The hydraulic jack was pressurized to 500 psi. With a 12 inch 2 cross-sectional area, 500 psi in the jack translated to 6 kips. This procedure was repeated for each specimen. The compressed springs allowed the specimen to move laterally without large changes in the axial load. If the specimen lost the ability to carry axial load, the threaded rod would lose tension as the spring loses compression. No instrument was placed to monitor the amount of axial load sustained during testing. The only way to tell if any axial load remained is by checking the tension in the rod and monitoring the amount of pressure in the hydraulic jack needed to release the spring. The springs were purchased from Barber Spring, Co. in Pittsburgh, PA. The specified spring stiffness was 4,288 lbs per inch of deflection (751 kN/m). A force of 6 kips was reached at a 1.4 inches (35 mm) deflection. The spring compression was monitored visually with measured tick marks that were attached to the spring system. Also, wooden blocks cut to allow only a 1.4 inch deflection were attached to the middle steel plate (flush with the top of the springs) to maintain a consistent spring compression among all the specimens.

PAGE 90

71 SpecimenSpringHydraulic A B Figure 43: Axial load spring system. A) Schematic. B) Photo. Lateral Loading A 55 kip (245 kN) MTS hydraulic actuator was used for testing (Figure 44). The hydraulic actuator was part of a closed-loop hydraulic loading system. The system was controlled by an MTS 407 controller. The controller used a sinusoidal voltage output from the data acquisition system to impose the displacements. The actuator also contained a load cell that was used to monitor the lateral load capacity of the specimen during testing.

PAGE 91

72 Figure 44: MTS actuator in place. The ICBO Acceptance Criteria for Concrete and Reinforced and Unreinforced Masonry Strengthening Using Fiber-Reinforced Composite Systems (AC125) (ICBO 1997) was followed for cycling in this testing program ( Figure 45). The displacement at which the reinforcement was expected to yield is marked as the first yield point, = 1. For specimens that did not contain any steel (CMU 1 and CMU 6), the yield point was taken as the displacement that was expected to caused failure of an unreinforced specimen. The specimens were loaded in displacement control with three complete cycles for each displacement level, The displacements were = , 1, 2, 3, 4, 6, 8 and 10. The displacements were increased until the specimen lost its lateral load carrying capacity. The loading rate was approximately 20 seconds per cycle. Figure 45: ICBO test sequence of imposed displacement.

PAGE 92

73 Data Acquisition A personal computer, National Instruments LabVIEW software and a 16-bit data acquisition card were used for data acquisition. Linear potentiometers and string pots were used for measuring displacements. Strain gauges were placed on the steel rebar and FRP strips. Linear Displacement Measurement Instrumentation was applied to the specimens and the actuator to monitor linear displacements and applied loads (Figure 46). Instruments labeled Carol, Diane, Flo and Gina were string potentiometers (SP) attached between a solid steel frame and hooks attached to the specimen (Figure 47). The SPs were BEI Duncan miniature spring return linear motion sensors, models 9610 and 9615. The string was tied to a wire and pulled out to the half way point before being attached to the specimen, allowing positive and negative measurements. The calibration factor used for the SP was taken from the value imprinted on each instrument. The data acquisition program measured the SP supply voltage as well as the output of the SP so that an accurate linear measurement could be computed.

PAGE 93

74 Out of PlaneCarolGinaHenryIan DianeVanessaFloKarlJimWillEmily Figure 46: Locations of instruments for linear displacement measurement. B A C Figure 47: String pots between steel frame and specimen. A) Overall photo of two of the string pots. B) Closer view of a string pot that measure horizontally. C) Closer view of a string pot that measures vertically. The rest of the displacement instruments were linear potentiometers. The linear potentiometers were calibrated using a calibration tool and a LabVIEW program developed by the researchers. The tool held the instrument in place while the stroke

PAGE 94

75 length was measured and recorded by the LabVIEW program. Values recorded were used to calculate the calibration factor. Each instrument was glued to a small metal box. This metal box protected the instrument during application and removal. The metal box was adhered to the specimen using hot glue and measured off the steel frames located on either side of the pier (Figure 48). These steel frames were constructed by the researchers and stood solidly on the laboratory floor. The stroke was compressed to the half way point when it was placed onto the specimen. This enabled the instrument to read in the positive and negative directions, depending on the movement of the wall. Figure 48: Linear potentiometers on the specimen Strain Gauges The strain gauges used on the steel were purchased from Texas Measurements, Inc. and were type FLA-5-11-1L. The dowel was ground, polished flat and cleaned where the strain gauge was to be placed. The strain gauge was glued in place using ethyl cyanoacrylate (as recommended by the strain gauge manufacturer). After the glue dried, a thin layer of silicone sealant was placed over the strain gauge to protect it from the masonry grout. Electrical tape was bound two-inches above and below the strain gauge.

PAGE 95

76 The tape was then covered with a shrink rubber tube to provide a four inch (101.6 mm) unbonded length of bar, which was centered on the pier/base interface during installation (Figure 49). Having a four inch gauge length allows for the strain at that location to be distributed over the unbonded length. The advantage of this is that the elongation of the bar occurs over the four inch length and not over a smaller length. Figure 49: Rubber tube installation on dowel over the strain gauge. Strain gauges were placed into the wet resin during FRP composite application (Figure 50). The strain gauges used in the FRP composite were also purchased from Texas Measurements, Inc. and were type PFL-30-11. See Figure 51 for strain gauge locations. All strain gauges were placed in the center of the FRP strap that they were measuring. One of the strain gauges on the FRP (Linda) was initially placed immediately above the pier/base intersection and then moved to the FRP/steel transition point (Linda (2)) after it was noticed that there was very little strain read above the pier/base intersection point. On those specimens that had the vertical FRP strip running continuously across the pier/base interface, a strain gauge (Marie) was placed immediately below the interface. On those that had vertical FRP strips terminating at the

PAGE 96

77 bottom of the pier no strain gauge was placed below the pier/base intersection. For all specimens, a strain gauge (Nancy) was placed on the horizontal strip that ran along the top of the base at the corner where the pier met the base. Figure 50: FRP Strain Gauge.

PAGE 97

78 Strain Measurement Direction Rebar LindaMarieNancy Linda (2) Figure 51: FRP strain gauge locations. LabVIEW Programs A LabVIEW program was developed to obtain the measurement readings. On the main screen (Figure 52), the user could specify file names for recording the data, identify cycle number, monitor the strain gauges and view a load vs. displacement graph for the test. The load and displacement shown on the screen are from the actuator load cell and LVDT. At the beginning of every test, a Zero Scan was run before any specimen displacements occurred. The Zero Scan measured the initial positions of each of the instruments. A Regular Scan was run while the specimen was displacing. The Regular Scan recorded the value of the current position minus the value of the zero scan. The value from the Regular Scan was also shown on the screen.

PAGE 98

79 Figure 52: LabVIEW program main screen. A LabVIEW program was also written to control the MTS actuator (Figure 53). The user inputs the displacement value, frequency and the number of cycles for each displacement value. The program sent a voltage output corresponding to the particular displacement to the MTS controller. The MTS controller controlled the amount of hydraulic fluid entering the actuator. A calibration factor of 0.5 inches per volt output remained constant throughout the entire testing program. Figure 53: MTS Signal Generation program screen.

PAGE 99

APPENDIX C SPECIMEN CONSTRUCTION Eight concrete masonry specimens were constructed in running bond of medium weight 8-in. (200-mm.) concrete masonry units by Painter Masonry, Inc. in Gainesville, FL. Type N mortar was used in faceshell bedding for each of the specimens. The ASTM C90 units were purchased locally from Florida Rock Industries, Inc. in Gainesville, FL. All test specimens were single wythe and consisted of a pier 48-in. (1200-mm.) tall by 48-in. (1200-mm.) wide and a base 9-ft. 4-in. (2800-mm.) long and 16-in. (400-mm.) tall (Figure 54). Since the test specimen consisted of full scale CMU, no scaling effects were required. The masons installed the joint reinforcement every other course in the specimen. This was done in accordance to common practice in the field. The location of the joint reinforcement is located in Figure 54. The specimens were constructed to simulate typical piers between windows of a building. Tests at the Georgia Institute of Technology (Moon et al. 2002) and at the U.S. Army Engineer Research and Development Center Construction Engineering Research Laboratory (Marshall and Sweeney 2002) used a similar pier configuration. The specimen was constructed on a precast concrete lintel 9-ft. 4-in. (2800-mm.) long. A second precast concrete 5-ft. 4-in. (1600-mm) lintel was placed on the top course of the pier. Both lintels were placed with the same mortar used for constructing the specimen. Lintels were filled with concrete and the bottom lintel also contained two #6 reinforcing bars embedded into the concrete along its length. The precast lintels contained reinforcement in the bottom. To take advantage of this reinforcement for lifting 80

PAGE 100

81 the specimen, the lintels were inverted before specimen construction began. The bottom lintel facilitated moving the specimen and the top lintel was used to attach the specimen to the concrete cap for loading. The outer cells of the base (at the tie down locations) were filled with grout (approximately one month after construction) to prevent overstressing the masonry during post-tensioning of the tie down. PrecastConcrete Lintel 4'-0" (1220 mm)Specimen 5'-4" (1630 mm) Grouted Cells PrecastConcrete Lintel Lifting Rings 9'-4" (2845 mm) 4'-0" (1220 mm) Joint ReinforcementPierBase PrecastConcrete Lintel PrecastConcrete Lintel Concrete fill placed after installation. #6 Rebar Figure 54: Specimen dimensions and location of joint reinforcement. Three specimens were constructed with #4 reinforcing bars in the jambs and sills (Figure 55). The existing steel was placed and grouted approximately one month after specimen construction. The grout for the existing steel was deposited with a shovel and consolidated with a vibrator. Many concrete masonry structures contain this type of prescriptive reinforcement around openings. These specimens allowed evaluation of the performance of partially reinforced masonry with composites added.

PAGE 101

82 #4 Reinforcing Bar (o.c.)#4 Reinforcing Bar (o.c.) Figure 55: Typical partially reinforced specimen containing existing reinforcing bars in jambs and sills. Material Properties Material properties are based on five full stretcher unit tests, four full prism tests and four half prism tests. The average full prism compressive strength was 2167 psi (14.9 MPa) (Table 8), the average half prism compressive strength was 3880 psi (26.7 MPa) (Table 9) and the average full stretcher unit compressive strength was 2224 psi (15.3 MPa) (Table 10). Table 8: Full Prism Compression Test Results Specimen Compressive Stress (psi) [MPa] I 2230 [15.4] III 2108 [14.5] V 2310 [15.9] VII 2021 [13.9] Average 2167 [14.9] Table 9: Half Prism Compression Test Results Specimen Compressive Stress (psi) [MPa] II 4088 [28.2] IV 3651 [25.2] VI 3593 [24.8] VIII 4189 [28.9] Average 3880 [26.7]

PAGE 102

83 Table 10: Full Stretcher Unit Compression Test Results Trial # Compressive Strength (psi) [MPa] 1 2234 [15.4] 2 2228 [15.4] 3 2331 [16.1] 4 2105 [14.5] 5 2035 [14.0] Average 2224 [15.3] It was expected that the compressive strength of the full prisms and half prisms would be similar. This was not the case. The half prisms were 21% stronger that the full prisms. It is possible that the half prisms were cast from a stronger batch of concrete than were the full prisms. The half prisms were not cut from the full prisms but were cast separately. It was observed that the full units absorbed more resin than the half units during the FRP application of CMU 1, leading to the conclusion that the full blocks were cast from a more porous concrete than the half blocks. Steel reinforcement was purchased from Rinker Materials in Gainesville, FL. Samples were taken from the inventory for tension testing. The average yield stress for the #3 bars was 62 ksi (429 MPa) and an average elongation of 16.0% (Table 11). The average yield stress for the #4 bars was 81 ksi (557 MPa) and an average elongation of 10.3% (Table 12). Procedures given in ASTM A370 were followed for testing. Table 11: #3 Reinforcing Bar Tension Test Results Trial # Ultimate Stress (ksi) [MPa] Yield Stress (ksi) [MPa] Elongation 1 93 [641] 61 [423] 16.4% 2 92 [634] 61 [423] 16.4% 3 94 [648] 63 [435] 17.3% 4 94 [648] 63 [435] 14.0% Average 93 [641] 62 [429] 16.0%

PAGE 103

84 Table 12: #4 Reinforcing Bar Tension Test Results Trial # Ultimate Stress (ksi) [MPa] Yield Stress (ksi) [MPa] Elongation 1 90 [621] 80 [554] 10.7% 2 92 [634] 81 [558] 12.2% 3 91 [627] 81 [558] 6.4% 4 91 [627] 81 [558] 12.0% Average 91 [627] 81 [558] 10.3% Masonry joint reinforcement was tested according to ASTM A951. The masonry joint reinforcement was provided by the masons. Specification sheets acquired from the mason report that the joint reinforcement is ASTM A Class B2 Hot Dipped Galvanized After Fabrication Standard Sized (9 Gauge) wire. Masonry Reinforcing Corporation of America supplied the joint reinforcement to the masons. Both the tension test and the weld shear strength tests from ASTM A951 were performed. The average yield stress for the joint reinforcement was found to be 107 ksi (738 MPa) (Table 13) and an average weld shear stress of 53 ksi (364 MPa) (Table 14). Table 13: Masonry Joint Reinforcement Tension Test Results Trial # Yield Stress (ksi) [MPa] 1 107 [738] 2 107 [738] Average 107 [738] Table 14: Masonry Joint Reinforcement Weld Shear Strength Test Results Trial # Shear Stress (ksi) [MPa] 1 51 [352] 2 55 [376] Average 53 [364] Five coupons of the cured composite material were tested in tension. The average width of the tensile specimens was 1.04-in. (26.42-mm) and the average thickness was 0.91-in (23.11-mm). The specimens were cut from a large piece of cured FRP composite and then milled to a 1-in. width. Two inch tabs for each specimen were made of hard

PAGE 104

85 circuit board and adhered to the ends of the coupons at a development length of 10-in. (250-mm.). Preparation of the coupons and testing followed ASTM D3039. The average tensile strength was 2.87 kips/inch (Table 15). Table 15: FRP Composite Coupon Tension Test Results Specimen Strength/Width (kips/inch) [N/mm] Stress (ksi) [MPa] 1 2.94 [512] 3.42 [23.56] 2 2.99 [521] 3.25 [22.38] 3 2.66 [463] 2.70 [18.57] 4 2.93 [510] 3.27 [22.53] 5 2.83 [493] 2.67 [21.09] Average 2.87 [500] 3.06 [21.09] FRP Application Before application, the specimen was wire brushed with a grinder and then swept clean of loose particles. A wet lay-up process was used to apply the FRP composite fabric to the surface of the masonry. For CMU 1, the initial coating of resin was applied immediately before lay-up. It was then noticed that the units were absorbing much of the resin and the FRP composite fabric was not as impregnated as it should be. On subsequent specimens a coat of resin was applied to the masonry an hour before lay-up to allow for the block to soak in some of the resin. This delay also allowed time for the resin to become tacky, improving the hold on the wetted FRP composite fabric. For CMU 7 and CMU 8, the specimens were sandblasted prior to FRP application. This was done to test if there was a significant difference in surface adhesion. There proved not to be any difference between the methods for this type of testing. No difference in delamination pattern or strength was observed between the sandblasted and wire brushed specimens. The specimens were strengthened with varying widths and lengths of unidirectional SikaWrap Hex 100G fiberglass fabric (27 ounces/yard 2 [.0859 kg/ft 2 ]) placed on the

PAGE 105

86 surface of the specimen with Sikadur Hex 300 epoxy. The manufacturers specified tensile strength of the material is 330 ksi (2275 MPa). The configuration included strips oriented vertically along the pier jambs as well as diagonally across the pier. The vertical strips increase the in-plane flexural strength. The X-pattern increases the diagonal tension strength. A strip of unidirectional FRP composite was also placed at the top of the base along its length. This was provided to reinforce the base in flexure. Bi-directional FRP composite fabric with a fiber orientation of was also used on each specimen. It was placed along the top of the pier at the pier/lintel intersection to prevent separation of the lintel and pier. The north and south face of the base was completely covered with bi-directional FRP composite to inhibit diagonal cracking in the base. The strips of cloth fiber were cut from large rolls prior to application. The Sikadur 300 epoxy comes in two parts, a hardener and the resin. The two parts were mechanically mixed using a drill with a mixing attachment (Figure 56). After the initial coating of resin was allowed to tack for approximately one hour (Figure 57A), a second coat was applied immediately prior to FRP lay-up. The cut strips were then saturated resin using a resin saturated roller and pressed onto the specimen (Figure 57B). A resin saturated roller was rolled over the wet fabric to ensure complete wet-out (Figure 57C). After the FRP turned clear (a sign of complete absorption of resin), a plastic trowel was passed over the composite to remove any air bubbles (Figure 57D). The troweling also ensured that the fabric laid flat on the specimen. For CMU 7 and CMU 8, the cloth was soaked in a tub of resin before lay-up. Excess resin was then screened off. It was then pressed onto the specimen using the plastic trowel.

PAGE 106

87 Figure 56: Mixing resin. A B C D Figure 57: FRP application. A) Precoating masonry with resin. B) Pressing FRP cloth into resin. C) Rolling resin onto placed FRP. D) Troweling placed FRP. Specimen CMU 6 was reinforced using Clark Schwebel Tech-Fab Companys G15000 unidirectional grid reinforcement. This grid system is a high strength,

PAGE 107

88 unidirectional reinforcement made by bonding E glass fiber rovings with epoxy resin in a controlled factory environment. The specified strength given by the manufacturer is 14,400 lbs/ft (210 kN/m). The strips used were cut from large rolls of the grid FRP. The manufacturer suggests laying the grid flat for 15 minutes before application to remove the waves formed from being stored in a roll. This did not work for this particular grid. Instead, the grid was laid as flat as possible and heat was spread evenly over the surface of the grid using a heat gun. This process removed the kinks and waves in the grid which allowed for easier application to the specimen. The resin for this system was thicker in consistency than Sikadur resin. Both parts of the resin were poured into a clean, plastic bin and mixed using a large, clean paddle (Figure 58A). It required that a layer of resin be applied with a trowel to the masonry (Figure 58B). The grid system strips were then pressed into the thick resin and additional resin was troweled on (Figure 58C and D). The grid slipped slightly in the resin under its own weight. The grid was then held in place using a series of two-by-fours and clamps.

PAGE 108

89 A B C D Figure 58: Grid FRP application. A) Mixing resin. B) Initial coat of resin on masonry. C) Grid FRP pressed into the resin. D) Troweling of additional resin. Computing Predicted Capacities For each of the specimens, the flexural capacity was calculated for two critical sections. The methodology for calculations is described by Holberg and Hamilton (2001). The first critical section checked is where the composite ends and flexural capacity is provided by the reinforcing steel bars. The second critical section checked is when the flexural capacity is provided by the composite strips placed on the pier jambs. At this section, the composite is carrying all the flexural tensile force. In Figure 59, P s is the axial forces provided by the spring, P w is the self weight of the pier and concrete cap and Q is the lateral load carrying capacity of the section.

PAGE 109

90 The flexural capacity of both sections can be determined using basic mechanics principles and the traditional rectangular stress block assumption. Assuming an under reinforced condition, the depth of the stress block, a, is: aAsfyPsPw0.85f'mbe (C.1) where b e is the effective thickness of the masonry, A s is the cross sectional area of the steel, f` m is the compressive strength of the masonry and f y is the yield strength of the steel reinforcement. Applying equilibrium to the section gives the moment capacity of the section where the flexural strength is provided by the steel reinforcement: MnbarAsfyda2 PsPwl2 a2 (C.2) The lateral force, Q, required to yield the bar is: QMnbarheff (C.3) For the critical section reinforced by composite only the depth of the stress block, a, is: aTwPsPw0.85f'mbe (C.4) where T is the tensile strength of the FRP and w is the width of the strip used. Applying equilibrium as before, the moment capacity of the section where the composite is providing the flexural strength is: MnfrpTwda2 PsPwl2 a2 (C.5) The lateral force, Q, required to rupture the FRP at this critical section is:

PAGE 110

91 QMnfrpheffls (C.6) where l s is the length of the lap splice between the composite strip and the steel reinforcement. Ps Q h eff ls Pw Critical SectionMnbar d lCritical SectionMnfrp Figure 59: Schematic of the pier connected to the base with reinforcing bars. Individual Specimen Details CMU Test 1 To evaluate the change in behavior caused by adding a steel component to the FRP system a specimen containing only FRP was tested (Figure 60, Figure 61 and Table 16). No steel reinforcement was placed in the cells of the specimen. The amount of FRP placed on CMU 1 is identical to that placed on CMU 2, which contained two #4 dowel bars. It was decided that the quantity of FRP to be installed would be enough to reach twice the lateral load carrying capacity than needed to yield the dowels in CMU 2.

PAGE 111

92 Photos of the specimen prior to testing are shown in Figure 62. Initial photos for CMU 1 were taken two cycles into testing. Table 17 contains critical dates for CMU 1. Table 18 contains the outside weather conditions on the date of the task. Temperature and relative humidity were obtained from a weather station located on the University of Florida campus. ABCDE Figure 60: FRP composite placement for CMU 1 North face.

PAGE 112

93 ABC Figure 61: FRP composite placement for CMU 1 South face. Table 16: CMU 1 FRP Quantities Mark Direction Width (in) [mm] Length (in) [mm] A Unidirectional 4 [101.6] 80 [2032] B Unidirectional 6 [152.4] 112 [2844.8] C 18 [457.2] 112 [2844.8] D Unidirectional 3 [76.2] 68 [1727.2] E 12 [304.8] 48 [1219.2] A B Figure 62: CMU 1 two cycles into testing. A) North face. B) South face.

PAGE 113

94 Table 17: Critical Dates for CMU 1 Construction Dowel Installation FRP Application Testing 5/24/02 N/A 8/23/02 9/5-6/02 Table 18: Conditions for CMU 1 Task Type Temperature (F) Relative Humidity (%) FRP Application 90 60 Testing 80 78 CMU Test 2 CMU 2 was the first specimen tested with the hybrid system. A #4 dowel bar was centered in the cells of the jambs (Figure 63). The bars were 40-inches (1016-mm) in length. The faceshells were removed from the fourth course up from the bottom of the pier to allow for bar installation. They were cut out using a circular saw with a masonry blade. The bars were held in place by hand in the center of the cell and grout was scooped in through the opening. A tamping rod was used to consolidate the grout around the dowels. A tamping rod was used instead of a vibrator because there was no room to insert the vibrator in the removed faceshell opening. The faceshells were then replaced into the pier with mortar. Each dowel was named to identify the strain gauge placed on it for data acquisition (Figure 63). 4.4" (1122 mm)Typ. 3.8" (97 mm)Typ. (Oscar) (Pete) N#4 x 40" dowel#4 x 40" dowel Figure 63: Steel reinforcement location for CMU 2. The same width of vertical FRP composite strips was placed on CMU 2 as was placed on CMU 1, the control specimen (Figure 64, Figure 65 and Table 19). This

PAGE 114

95 allowed for a direct comparison between the FRP only system (CMU 1) and the hybrid system (CMU 2). The 6-inch horizontal strip on the base was replaced with a 3-inch strip in order to minimize the quantity of FRP composite needed to retrofit the specimens. No change in the behavior of the base was observed. The lateral load capacity of the specimen with dowels only was calculated for CMU 2. The width of the vertical FRP composite strips was adjusted to achieve a lateral load capacity twice that of the capacity of dowels alone. This ensured that the dowels would yield prior to FRP composite rupture. To force yielding at the pier/base intersection, the FRP strip was terminated at the base of the pier. Photos of the specimen prior to testing are shown in Figure 66. Table 20 contains critical dates for CMU 2. Table 21 contains the outside weather conditions on the date of the task. Temperature and relative humidity were obtained from a weather station located on the University of Florida campus.

PAGE 115

96 ABCDE Figure 64: FRP composite placement for CMU 2 North face. ABC Figure 65: FRP placement for CMU 2 South face.

PAGE 116

97 Table 19: CMU 2 FRP Quantities Mark Direction Width (in) [mm] Length (in) [mm] A Unidirectional 4 [101.6] 56 [1422.4] B Unidirectional 3 [76.2] 112 [2844.8] C 18 [457.2] 112 [2844.8] D Unidirectional 3 [76.2] 68 [1727.2] E 12 [304.8] 48 [1219.2] A B Figure 66: CMU 2 prior to testing. A) North face. B) South face. Table 20: Critical Dates for CMU 2 Construction Dowel Installation FRP Application Testing 5/23/02 7/19/02 9/11/02 11/23/02 Table 21: Conditions for CMU 2 Task Type Temperature (F) Relative Humidity (%) FRP Application 85 65 Testing 76 70 CMU Test 3 CMU 3 was one of the three specimens containing reinforcement in the jambs and sills. This reinforcement was intended to represent the existing condition prior to installation of the strengthening system. Dowel reinforcement was installed as in CMU 2

PAGE 117

98 (Figure 67). The steel bars were named to identify the strain gauge placed on it for data acquisition. 4.4" (1122 mm)Typ. 3.8" (97 mm)Typ. (Pete) (Oscar) N 12.8 (325 mm)Typ. (Quinn)(Rick)#4 jamb steel#4 jamb steel#4 x 40" dowel#4 x 40" dowel #2 GFRP Transverse bar Figure 67: Steel reinforcement locations for CMU 3. Transverse GFRP composite bars were placed in the cells containing the dowel bars (Figure 68 and Figure 69). The purpose of the GFRP bar was to confine the grout containing the dowels in order to develop the strain necessary to yield the dowels. The GFRP bars are Aslan 100 GFRP Rebar with a tensile strength of 120 ksi (825 MPa) and a Modulus of Elasticity of 5.92 psi (40.8 GPa).The GFRP bars are #2 bars. The same size standards apply for GFRP bars as for steel reinforcing bars. Quarter inch holes were drilled through the unit faceshell using a hammer drill before dowel bars were installed. The length of the transverse bars matched the thickness of a CMU block, 7.625 inches (194 mm). The GFRP bars were pushed through the block via the drilled holes. The holes in the faceshells were filled with silicone caulking and covered with duct tape to prevent the GFRP bars from moving during grouting of the dowels.

PAGE 118

99 FRP Rebar Extent of grouting for dowels Figure 68: Location of transverse GFRP rebar in CMU 3. GFRP barGrout Faceshell Tape A B Figure 69: Transverse GFRP block in the faceshell of CMU 3. A) Cross-section of block containing a GFRP bar. B) GFRP bar embedded in faceshell. The width of the vertical FRP composite strips was adjusted to achieve a lateral load capacity twice that of the capacity of steel reinforcement alone. The capacity calculated included both the existing steel and the dowel steel. FRP was placed on both sides of the specimen in the same configuration as described for the previous specimens (Figure 70, Figure 71 and Table 22). This ensured that the steel reinforcement (existing

PAGE 119

100 and dowel) would yield prior to FRP rupture. In order to force yielding at the pier/base intersection, the FRP strip was terminated at the base of the pier. Photos of the specimen are shown in Figure 72. Table 23 contains critical dates for CMU 3. Table 24 contains the outside weather conditions on the date of the task. Temperature and relative humidity were obtained from a weather station located on the University of Florida campus. ABCDE Figure 70: FRP composite placement for CMU 3 North face.

PAGE 120

101 ABC Figure 71: FRP composite placement for CMU 3 South face. Table 22: CMU 3 FRP Quantities Mark Direction Width (in) [mm] Length (in) [mm] A Unidirectional 6 [152.4] 56 [1422.4] B Unidirectional 3 [76.2] 112 [2844.8] C 18 [457.2] 112 [2844.8] D Unidirectional 3 [76.2] 68 [1727.2] E 12 [304.8] 48 [1219.2]

PAGE 121

102 A B Figure 72: CMU 3 prior to testing. A) North face. B) South face five cycles into testing. Table 23: Critical Dates for CMU 3 Construction Dowel Installation FRP Application Testing 5/23/02 11/7/02 11/14/02 1/22/03 Table 24: Conditions for CMU 3 Task Type Temperature (F) Relative Humidity (%) FRP Application 67 52 Testing 77 82 CMU Test 4 CMU 4 also contained reinforcement to represent existing conditions. Four #3 dowel bars were installed by repointing into the outer head joints (Figure 74). For details on dowel locations see Figure 74. The steel bars were named to identify the strain gauge placed on it for data acquisition. This configuration was chosen to see if there is a difference in performance between specimens with dowels grouted into the cell versus dowels placed using the repointing method. Knowing the differences in behavior and capacities among the two methods allows for the option of installing dowels using more than just the grouting method. In some cases, it may be easier to install dowels using a

PAGE 122

103 repointing method than using grouting. Repointing does not add the extra concrete material that is needed for grouting dowels. Repointing required that the mortar joints be removed. Since 32 inch (812-mm) long dowel was used, 32 inch long grooves were cut into the specimen (Figure 73B) starting at the top of the bottom lintel. The width of the groove matched the width of the mortar joint, 3/8-inch (9.5-mm), so the entire mortar joint was removed. The depth of the groove was 3/4-inch (19-mm) deep. See Figure 75 for details of the repointed dowel. Dimensions for repointing were taken from similar testing by Bajpai and Duthinh (2003). Concresive 1420 General Purpose Gel Epoxy Adhesive from Master Builders Technologies was used to install the dowel. A layer of the epoxy was squeezed into the groove. After coating the bar with the epoxy, it was pressed into the groove. The epoxy was then passed over the bar filling in any gaps (Figure 73C). The epoxy was then smoothed by hand. Removing the mortar head joints for the purpose of repointing was difficult. The repointing tool used by masons did not cut deep or wide enough to insert the #3 dowel. Instead, a grinder with a masonry cutting wheel was used. It required several passes of the cutting wheel to remove enough mortar to insert the dowel.

PAGE 123

104 A B C Figure 73: Repointing of CMU 4. A) Repointed dowel with smoothed epoxy. B) Grooves cut into the mortar joints. C) Passing epoxy over the embedded dowel. 4.4" (1122 mm)Typ. 3.8" (97 mm)Typ. (Pete) (Oscar) N (Quinn) (Rick)#4 jamb steel#3 x 32" dowel#3 x 32" dowel#3 x 32" dowel Figure 74: Steel reinforcement locations for CMU 4. 3/8" (19 mm) 3/4" (19 mm) Figure 75: Details of repointed #3 dowel. The width of the vertical FRP composite strips was adjusted to achieve a lateral load capacity twice that of the capacity of steel reinforcement alone (both existing and dowel). Details of FRP composite configuration for CMU 4 are provided in Figure 76, Figure 71 and Table 25). This ensured that the steel reinforcement would yield prior to

PAGE 124

105 FRP rupture. In order to force yielding at the pier/base intersection, the FRP strip was terminated at the base of the pier. Photos of specimen prior to testing are shown in Figure 78. The photo of the south face was taken two cycles into testing. Table 26 contains critical dates for CMU 4. Table 27 contains the outside weather conditions on the date of the task. Temperature and relative humidity were obtained from a weather station located on the University of Florida campus. ABCDE Figure 76: FRP placement for CMU 4 North face.

PAGE 125

106 ABC Figure 77: FRP placement for CMU 4 South face. Table 25: CMU 4 FRP Quantities Mark Direction Width (in) [mm] Length (in) [mm] A Unidirectional 7 [177.8] 56 [1422.4] B Unidirectional 3 [76.2] 112 [2844.8] C 18 [457.2] 112 [2844.8] D Unidirectional 3 [76.2] 68 [1727.2] E 12 [304.8] 48 [1219.2]

PAGE 126

107 A B Figure 78: CMU 4 prior to testing. A) North face. B) South face two cycles into testing. Table 26: Critical Dates for CMU 4 Construction Dowel Installation FRP Application Testing 5/23/02 1/10/03 1/13/03 1/22/03 Table 27: Conditions for CMU 4 Task Type Temperature (F) Relative Humidity (%) FRP Application 68 85 Testing 75 96 CMU Test 5 One of the goals of this testing program was to minimize the amount of FRP needed. This not only reduces the cost of the system, but also reduces the amount of labor needed to retrofit a building. It was necessary to investigate the behavioral changes of placing FRP composite on one face of a wall versus placing the FRP composite on both faces of a wall. This presents the problem of out-of-plane bending and movement. The steel for the specimens with one-sided FRP was placed to minimize any out-of-plane displacements.

PAGE 127

108 CMU 5 contained two #3 dowels in the jambs of the pier and FRP composite on only one face of the pier (Figure 79). The dowels were placed so that each bar was close to either the north or south face of the pier. This was done to mitigate the out-of-plane bending due to having the vertical FRP composite strips on only the north face of the specimen. Two #3 dowel bars were placed in the jambs of the pier (Figure 79). The dowels were 32-inches (812-mm) in length. The installation of the dowels is similar to that of CMU 2. The dowels were named to identify the strain gauge placed on it for data acquisition. 4.4" (1122 mm)Typ. 2.4" (61 mm)Typ. (Pete) (Oscar) N (Quinn) (Rick) #3 x 32" dowel#3 x 32" dowel#3 x 32" dowel#3 x 32" dowel Figure 79: Steel reinforcement locations for CMU 5. The width of the vertical FRP composite strip was adjusted to achieve a lateral load capacity twice that of the capacity of dowels alone. This ensured that the dowels would yield prior to FRP rupture. In order to force yielding at the pier/base intersection, the FRP composite strip was terminated at the base of the pier. Details of FRP composite configuration for CMU 4 are provided in Figure 80 and Table 28. Photos of the specimen prior to testing are shown in Figure 81. Table 29 contains critical dates for CMU 5. Table 30 contains the outside weather conditions on the date of the task. Temperature and relative humidity were obtained from a weather station located on the University of Florida campus.

PAGE 128

109 ABCDE Figure 80: FRP composite placement for CMU 5 North face. (No FRP composite was placed on the South face) Table 28: CMU 5 FRP Quantities Mark Direction Width (in) [mm] Length (in) [mm] A Unidirectional 8 [203.2] 56 [1422.4] B Unidirectional 3 [76.2] 112 [2844.8] C 18 [457.2] 112 [2844.8] D Unidirectional 3 [76.2] 68 [1727.2] E 12 [304.8] 48 [1219.2]

PAGE 129

110 A B Figure 81: CMU 5 prior to testing. A) North face. B) South face. Table 29: Critical Dates for CMU 5 Construction Dowel Installation FRP Application Testing 5/23/02 1/10/03 1/13/03 2/19/03 Table 30: Conditions for CMU 5 Task Type Temperature (F) Relative Humidity (%) FRP Application 68 85 Testing 77 54 CMU Test 6 The FRP composite system used has been the wet lay-up of unidirectional FRP composite fabric with Sikadur 330 High Modulus Epoxy. CMU 6 was strengthened using unidirectional grid reinforcement. Quantities of the grid system were chosen to achieve the calculated lateral load capacity of CMU 1. Since only the north face of CMU 6 was strengthened, the capacity of CMU 6 became half of the calculated capacity of CMU 1. Details of FRP reinforcement are shown in Figure 82 and Table 31. No steel reinforcement was installed in CMU 6.

PAGE 130

111 Photos of the specimen prior to testing are shown in Figure 83. Table 32 contains critical dates for CMU 6. Table 33 contains the outside weather conditions on the date of the task. Temperature and relative humidity were obtained from a weather station located on the University of Florida campus. ABCD Figure 82: FRP composite placement for CMU 6 North face. (No FRP composite was placed on the South face) Table 31: CMU 6 FRP Quantities Mark Direction Width (in) [mm] Length (in) [mm] A Grid 9 [228.6] 80 [2032] B Grid 18 [457.2] 112 [2844.8] C Grid 7 [177.8] 68 [1727.2] D Grid 12 [304.8] 48 [1219.2]

PAGE 131

112 A B Figure 83: CMU 6 prior to testing. A) North face. B) South face. Table 32: Critical Dates for CMU 6 Construction Dowel Installation FRP Application Testing 5/23/02 N/A 1/29/03 2/21/03 & 2/26/03 Table 33: Conditions for CMU 6 Task Type Temperature (F) Relative Humidity (%) FRP Application 77 52 Testing 75 72 CMU Test 7 CMU 7 was strengthened using the unidirectional FRP composite fabric system on only the pier portion of the specimen. This specimen contained jamb and sill steel reinforcement to represent existing conditions (Figure 85). The purpose of this test was to observe the behavior of the specimen with FRP on the pier alone. This would be an attractive option for retrofitting because the least amount of material and labor is required. It would have little impact on the occupants during installation and would minimally alter the exterior of a structure. Details of FRP composite reinforcement are shown in Figure 84 and Table 34.

PAGE 132

113 The width of the vertical FRP composite strip was adjusted to achieve a lateral load capacity twice that of the capacity of jamb reinforcement alone. This ensured that the jamb reinforcement would yield prior to FRP rupture. Photos of the specimen prior to testing are shown in Figure 86. Table 35 contains critical dates for CMU 7. Table 36 contains the outside weather conditions on the date of the task. Temperature and relative humidity were obtained from a weather station located on the University of Florida campus. ABC Figure 84: FRP composite placement for CMU 7 North face. (No FRP composite placed on the South face)

PAGE 133

114 Table 34: CMU 7 FRP Quantities Mark Direction Width (in) [mm] Length (in) [mm] A Unidirectional 8 [203.2] 56 [1422.4] B Unidirectional 3 [76.2] 68 [1727.2] C 12 [457.2] 48 [2844.8] 4.4" (1122 mm)Typ. 3.8" (97 mm)Typ. (Oscar) (Pete) N#4 jamb steel#4 jamb steel Figure 85: Steel reinforcement locations for CMU 7. A B Figure 86: CMU 7 prior to testing. A) North face. B) South face.

PAGE 134

115 Table 35: Critical Dates for CMU 7 Construction Dowel Installation FRP Application Testing 5/23/02 N/A 3/19/03 3/26/03 Table 36: Conditions for CMU 7 Task Type Temperature (F) Relative Humidity (%) FRP Application 84 79 Testing 77 47 CMU Test 8 From the results of CMU 5 (see Appendix D), there was a splice failure in the masonry from the eccentricity caused by the placement of the dowels. It was decided that for CMU 8, dowels would be placed near the face of the masonry where the FRP was applied. See Figure 87 for details. 1.0" (25 mm) ClearTyp. (Oscar) (Pete) N 1.0" (25 mm) ClearTyp. #3 x 32" dowel#3 x 32" dowel Figure 87: Steel reinforcement locations for CMU 8. The faceshells were removed over the full length of the dowel in the jambs of the pier and base. Two #3 dowels 32-inches (812-mm) long were mounted to plywood using wire and plastic spacers. The masonry joint reinforcement was exposed after the faceshells were removed. The dowels were placed by working around the joint reinforcement. The plywood held the dowel in place and retained the grout in the cells. Grout was deposited through an opening in the top and a tamping rod and mallet used to consolidate the grout around the dowel. After removing the plywood (after 4 days of

PAGE 135

116 curing), the grout was ground smooth using a mechanical grinder. This was done to improve the contact between the grout and the FRP composite. A B C D Figure 88: Installing dowels in CMU 8. A) Grooves cut into the specimen. B) Bars placed onto plywood. C) Dowels placed on plastic spacers with wire. D) Grout was shoveled into the openings in the masonry. The width of the vertical FRP composite strips was adjusted to achieve a lateral load capacity twice that of the capacity of dowels alone. Details of FRP composite reinforcement are shown in Figure 89 and Table 37. This ensured that the dowels would

PAGE 136

117 yield prior to FRP composite rupture. In order to force yielding at the pier/base intersection, the FRP strip was terminated at the base of the pier. Photos of the specimen prior to testing are shown in Figure 90. Table 38 contains critical dates for CMU 8. Table 39 contains the outside weather conditions on the date of the task. Temperature and relative humidity were obtained from a weather station located on the University of Florida campus. ABCDE Figure 89: FRP composite placement for CMU 8 North face. (No FRP composite was placed on the South face) Table 37: CMU 8 FRP Quantities Mark Direction Width (in) [mm] Length (in) [mm] A Unidirectional 8 [203.2] 56 [1422.4] B Unidirectional 3 [76.2] 112 [2844.8] C 18 [457.2] 112 [2844.8] D Unidirectional 3 [76.2] 68 [1727.2] E 12 [304.8] 48 [1219.2]

PAGE 137

118 A B Figure 90: CMU 8 prior to testing. A) North face. B) South face. Table 38: Critical Dates for CMU 8 Construction Dowel Installation FRP Application Testing 5/23/02 2/25/03 3/19/03 4/10/03 Table 39: Conditions for CMU 8 Task Type Temperature (F) Relative Humidity (%) FRP Application 84 79 Testing 77 93 APPENDIX D SPECIMEN RESULTS CMU 1 CMU 1 was the control specimen for the testing program. The only reinforcement applied was FRP (See Appendix C for details). No steel dowels were installed.

PAGE 138

119 Testing The values calculated for determining the deflections that would be used in the displacement controlled testing are contained in Table 40. It was from CMU 1 that the stiffness for the walls was determined. At the beginning of the test, the specimen cycled several times in the elastic range to determine the uncracked stiffness (k) of the specimens, 167 kips/inch (29.2 kN/mm). The stiffness was found using the displacement and load cell reading of the hydraulic actuator. This value of k along with the calculated capacities in Table 40 was used to find y or the yield point. The yield point is where the section goes from being an uncracked to being cracked for specimens not containing any steel reinforcement. For those specimens containing steel reinforcement the yield point is the displacement for the calculated lateral capacity that will yield the steel reinforcement. Observations were taken after each cycle during testing for each displacement. Values of found in Table 41 are the increments of increasing displacement after yielding. The frequency for testing was .05 Hz. Table 40: Calculated Values for CMU 1. Q n unreinforced Q nfrp y 3.4 kips [15.1 kN] 17.7 kips [78.7 kN] 0.02 in. [0.5 mm] Table 41: Testing Observations for CMU 1. Observations < 1 Flexural cracking along pier/base intersection. 1 Cracks move up the pier. 2-4 Hysteresis plot begins to have wider loops. 5-8 Continued cracking. >8 Toe crushing observed in the bottom corner unit in the west end. Pier faceshells were pulled apart from the tension in the FRP strips. FRP strips delaminated from the base of the pier up toward the top of the pier.

PAGE 139

120 Results Results of testing CMU 1 are reported in this section. Figure 91 shows the final crack pattern for the specimen. Crack patterns are indicative of step cracking. Figure 92 shows pictures taken of the specimen before and after testing. The before photos were actually taken two cycles into testing. A B Figure 91: Crack Patterns for CMU 1. A) North face. B) South face. A B

PAGE 140

121 C D Figure 92: Before and after pictures of CMU 1. A) North face two cycles into testing. B) North face after testing. C) South face two cycles into testing. D) South face after testing. Debonding of FRP composite It evident from Figure 93 that FRP debonding occurred. The debonding allowed for the FRP straps to hold tensile forces without rupturing. The debonding acted similarly to the unbonded length on a reinforcing bar. The unbonded length allows for the strain concentrated at a crack to be distributed over a larger length of FRP composite without rupture. Had there been a perfect bond between the masonry and the FRP, the length over which the tensile strain would have been distributed would have been the width of the crack formed in the masonry behind the FRP. Since the crack would be very small, the strain would be distributed over a small length causing greater tensile forces in the strap at that location.

PAGE 141

122 Figure 93: Debonding of FRP composite on CMU 1. The softening brought to the system by the debonded FRP length can be explained using a simple model and following basic mechanics. Figure 94 illustrates two different pier sections that have been displaced the same amount. One has an unbonded length, g, equal to 1 and the other has an unbonded length equal to 10. D = 1g = 1 D = 1P1 A D = 1g = 10 D = 1P2 B Figure 94: Pier displaced = 1. A) Unbonded length equal to 1. B) Unbonded length equal to 10. Using basic mechanics of materials and assuming that the Elastic Modulus for the FRP is linear, the lateral force, P1, for g = 1 is: 1 = /g = 1/1 = 1

PAGE 142

123 1 = E 1 = 1 1 = P1/A P1 = E*A where A is the cross sectional area of the FRP composite strip. Similarly, the lateral force, P2, for g =10 is: 2 = /g = 1/10 = 0.1 2 = E 2 = 0.1*E 2 = P2/A P2 = 0.1* E*A The relationship between P1 and P2 becomes: P1/A = E = 10*P2/A P1 =10* P2 The lateral force needed to displace a pier is directly related to the unbonded length of the FRP composite. For this example, ten times the force is needed to displace the pier with the unbonded length of 1 than for the unbonded length of 10. As g becomes larger, the stress in the FRP composite becomes smaller. As g becomes smaller, the stress in the FRP composite becomes larger. Buckling of tensile reinforcement Figure 95 shows a compression failure (toe crushing) at the base of the pier on the west end and that the FRP strip has buckled away from the block. The apparent buckling of the FRP composite under cyclic loading can be related to the buckling of steel reinforcement in concrete and masonry.

PAGE 143

124 Figure 95: Compression failure on the west end of the pier and buckled FRP. Buckling of tensile reinforcement during cyclic loading can occur in the plastic hinge region of reinforced concrete and masonry sections. If the linear rotations are large enough the tensile reinforcement will strain well beyond yield causing permanent elongation of the reinforcement. At this stage, wide horizontal cracks form across the width of the section. During reverse cycling the tensile stresses in the reinforcement reduce to zero and the cracks remain open. A reversal of the lateral force produces a compression force in the reinforcement. Until the crack closes, the compression force must be resisted by the reinforcement only. If confinement is not adequate, reinforcement buckling may cause a rotation of the blocks bound by the horizontal cracks. This often leads to the specimen sliding or moving out-of-plane. In the case of FRP composite on masonry, as the specimen was loaded the FRP composite strip stretch from the tensile stress. During reverse loading as the pier returned

PAGE 144

125 to its original position, the stretch FRP buckled out-of-plane because of the additional gain in length from the stretching. The toe crushing caused the specimen to lose its lateral load carrying capacity and was a result of the large the tensile strain in the FRP composite strip during the last displacement cycle. To balance the larger tensile force, the toe of the pier must sustain a higher compressive force. Lateral load capacity is limited by either the compressive strength of the masonry at the toe or the tensile capacity of the FRP composite. Drift capacity From Figure 96 the specimen reached a maximum drift ratio of .4% before losing lateral load carrying capacity. Testing by Abrams (2001) on hollow concrete masonry piers reinforced with FRP composites found a drift ratio of 1.9%. Tumialan, Bartolome and Nanni (2003) reported a maximum drift ratio of 0.5% of an unstrengthened URM wall specimen. The CMU 1 shows a 29% improvement in drift capacity over an unstrengthened specimen. -30-20-100102030-4-3-2-101234Drift (%)Load (kips) Figure 96: Drift for CMU 1.

PAGE 145

126 Load capacity and backbone curve The maximum load reached was 17.7 kips [78.7 kN] in the negative direction and 14.9 kips [66.3 kN] in the positive direction (Figure 97). The horizontal solid line in the plots represents the lateral strength of an unreinforced, unstrengthened specimen, 3.4 kips [15.1 kN], computed by taking moments about the pier toe of the vertical and horizontal forces. The strengthened specimen showed a considerable improvement over the unstrengthened specimen. It was able to achieve a lateral load capacity of almost five times greater. Because of the sudden toe crushing failure on the east end of the specimen, the lateral load carrying capacity was not able to reach the same 17.7 kips [78.7 kN] in the positive direction. If there had been no to crushing, there would have been an increase capacity in the positive direction. Load versus displacement plots of the test data were prepared and the backbone curve was developed in accordance with the acceptance criteria presented in FEMA-273 for new materials (Figure 97). The flexure cracking produced pinched loops in the load versus displacement plot. As cracks developed, there was a change in the slope of the curves, indicative of a decreasing stiffness. The open loops are a result of sliding that occurred before toe crushing.

PAGE 146

127 -30-20-100102030-2-1012Displacement (in)Load (kips) Figure 97: Backbone curve for CMU 1. A typical force-displacement curve for unreinforced masonry strengthened with FRP composites is shown in Figure 98 (Moon, Leon etal 2002). The initial positive stiffness, k, is the stiffness of an uncracked specimen. The decreased positive stiffness after yield, k, is due to the masonry cracking and the FRP strips becoming stressed. The sharp drop in strength represents the shear force that causes the failure of the FRP. After this point, the pier returns to the strength found in an unreinforced specimen. The backbone curve developed for CMU 1 (Figure 97) shows a similar behavior as described in the generalized curve.

PAGE 147

128 % Drift URM with FRPURM kk' Figure 98: Generalized force-displacement relationship for URM and URM retrofit with FRP overlays (Moon, Leon etal 2002). A similar force-displacement relationship can be added to Figure 98 for fully-bonded FRP composites and debonding FRP composites. The curve derived from the testing program is shown in Figure 99. Instead of having one curve representing URM strengthened with FRP, two separate curves can be substituted to represent two conditions of FRP composites, fully-bonded and debonding. Each of these cases will have a different k because of the softening effect of the debonding FRP composite. The debonding FRP composite will rupture at a lower load and have a lower stiffness than the fully bonded FRP composite.

PAGE 148

129 % Drift URM with Fully Bonded FRPURM with Debonding FRPURM k k'k' Figure 99: Force-displacement curve for URM strengthened with fully bonded FRP composite and debonding FRP composite. Sliding Sliding occurred across a step crack near the base of the pier. The white line in Figure 100 illustrates the location of the step cracking where the pier opened up, allowing it to slide. Sliding was observed during the last set of displacements, = 20. Sliding was also monitored and shown in Figure 101. For the first couple set of displacements, sliding occurred during cycling but returned to its original position after the three cycles. For the last set of displacements, the pier did not regain its original position. Looking at the sliding at a displacement of zero, one can see that there is a displaced value for sliding. This means that the wall did not go back to where it had been prior to the cycle.

PAGE 149

130 Figure 100: Step cracking opened up and allowed the pier of CMU 1 to slide. -0.5-0.2500.250.5-2-1012Displacement (in)In-plane Sliding(in) Figure 101: Sliding of CMU 1. Strains in FRP composite FRP composite strains are shown in Figure 102. The FRP composite ruptures at a strain of 24,500 The maximum strain in the FRP composite strip was 10,544 The FRP reached 43% of its capacity.

PAGE 150

131 03000600090001200015000-2-1012Displacement (in)FRP Pier Strain (uE) A 03000600090001200015000-2-1012Displacement (in)FRP Base Strain (uE) B 03000600090001200015000-2-1012Displacement (in)FRP Horizontal Base Strain (uE) C Figure 102: FRP strain gauge readings for CMU 1. A) Vertical pier strip. B) Vertical base strip. C) Horizontal base strip. Out-of-plane movement Out-of-plane displacements of the bottom corners of the pier are shown in Figure 103. There was no significant out-of-plane movement noted for this specimen, even at higher lateral displacements.

PAGE 151

132 00.250.50.751-2-1012Displacement (in)Out of Plane Movement (in) A 00.250.50.751-2-1012Displacement (in)Out of Plane Movement (in) B Figure 103: Out-of-plane movement for CMU 1. A) West. B) East. Table 42: Summary of Results for CMU 1. Value At a disp. Of (in) [mm] At a load of (kips) [kN] Location Max Load (kips) [kN] -17.7, 14.9 [78.7, 66.3] -0.53, 0.44 [-13.5, 11.2] Max Drift (%) -1.4, 1.4 -0.69, 0.67 [-17.5, 17.0] 14.3, 8.1 [63.6, 36.0 ] Max displacement (in) [mm] -0.69, 0.67 [-17.5, 17.0] -13.7, 8.1 [-60.9, 36.0] Max FRP Strain () 10544 0.66 [16.8] -13.7 [-60.9] Vertical base strip Max Steel Strain () n/a n/a n/a Max Out-of-plane (in) [mm] -.003, 0.06 [1.5, 1.52] 0.67, 0.41 [17.0, 10.4] 8.1, 4.4 [36.0, 19.6] East Overall performance It can be concluded that for CMU 1, the lateral load capacity was controlled by a combination of flexure and sliding. A flexure failure is evident in the toe crushing at the base of the pier. As discussed earlier, the larger the lateral displacements the larger the tensile strain in the FRP. To balance the larger tensile force, the toe of the pier must sustain a higher compressive force. Lateral load capacity is limited by either the

PAGE 152

133 compressive strength of the masonry at the toe or the tensile capacity of the FRP composite. The pier continued to hold axial load even after the toe crushing. If the specimen had lost its ability to carry the axial load provided by the spring, the threaded rod held in tension by the compressed spring would have become loose as the spring expanded. As the specimen was taken out of the test fixture, there was still tension in the rod which needed to be released by the hydraulic jack. CMU 2 CMU 2 was reinforced with the same FRP composite configuration as CMU 1. It also contained a #4 dowel bar grouted for a length of 40 inches (1016 mm) in the jambs of the pier and into the base. See Appendix C for details. Testing The values calculated for determining the deflections that would be used in the displacement controlled testing are contained in Table 43. The predicted lateral strengths given are the values given by taking the moments about the pier toe of the vertical and horizontal forces. Table 43: Calculated Values for CMU 2. Q nbar Q nfrp y 13.6 kips [60.5 kN] 23.1 kips [102.7 kN] 0.082 in. [2.1 mm] Observations and notes were taken during testing at different values of The observations can be found in Table 44. During testing, the most prevalent action observed was debonding of the FRP in the base. This is due to the cracking formed behind the fiberglass.

PAGE 153

134 Table 44: Testing Observations for CMU 2. Observations < 1 Flexure cracking on back side of wall at pier/base intersection. The hysteresis loop is linear and symmetric. 1 The hysteresis loop is linear and symmetric. Flexure cracks move up the pier. Signs of step cracking. 2-4 Yielding of the bar (as indicated by strain gauges and graph). Shear cracking through some units. Series of cracks formed around window used for dowel placement. Debonding of FRP along the bottom of the base. 5-6 Faceshell spreading. Results Results of testing CMU 2 are reported in this section. Figure 104 shows the final crack pattern for the specimen. Figure 105 shows pictures taken of the specimen before and after testing. A B Figure 104: Crack Patterns for CMU 2. A) North face. B) South face.

PAGE 154

135 A B C D Figure 105: Before and after pictures of CMU 2. A) North face taken before testing. B) North face after testing. C) South face taken before testing. D) South face after testing.

PAGE 155

136 Bond failure As shown in Figure 106, the grout containing the dowel split the specimen at the base. When removing the bar from the grout, it was noted that the bar had slipped within the grouted core as well, indicating a bond failure. The cause of the bar failure is due to insufficient development length for the dowels. Figure 106: Base Splitting of CMU 2. A reinforcement bar embedded in concrete develops bond by adhesion between the bar and mechanical friction between the deformations (ribs) of the bar and the concrete. When the bar is loaded in tension, the effect of the adhesion is quickly lost. This is because the diameter of the bar decreases slightly due to Poissons ratio. This leaves the bond to be transferred by bearing on the deformations of the bar. Equal and opposite bearing stresses act on the surrounding concrete. The forces on the concrete have both longitudinal and radial components. The radial component causes circumferential tensile stresses in the concrete around the bar. As these forces get larger, the concrete splits parallel to the bar and the resulting crack propagates to the surface. The cracks follow the bar along the bottom or side surfaces. The bond transfer drops after these cracks have developed unless reinforcement is provided to restrain them. Cracks tend to develop along the shortest distance between a bar and the surface or between two bars.

PAGE 156

137 In the case of CMU 2, the cracks propagated to the surface. The FRP composite in the base served to restrain the cracking in the base but leaving the cracks to move towards the bottom lintel and the sills (Figure 107). Cracking was also formed in the masonry surrounding the end of the dowels in the pier (Figure 108). Figure 107: Splitting bond failure caused cracks to propagate into the sill at the east end of CMU 2.

PAGE 157

138 Figure 108: Cracks move radially from the bar to the edges of the pier in CMU 2. Drift capacity From Figure 109 the maximum drift ratio was .2% before a loss in the lateral load carrying capacity. Tests by Holberg and Hamilton (2001) concluded that for a concrete masonry wall with reinforcing steel and strengthened with FRP, the drift capacity was 1.7%. The lower drift capacity in CMU 2 is probably due to premature bond failure in the specimen. For the last set of cycles there open loops in Figure 109 for positive loading. Open loops in load versus drift plots are indication of energy dissipation. The energy dissipation for CMU 2 can be attributed to the abrasion caused by the reinforcing bars against the grout. When the grout containing the reinforcing bars was inspected after testing, the grout surrounding the reinforcing bars had smooth rib marks. The smoothening of the rib marks was caused when the bar was pulled during loading.

PAGE 158

139 -30-20-100102030-4-3-2-101234Drift (%)Load (kips) Figure 109: Drift for CMU 2. Load capacity and backbone curve Load versus displacement plots of the test data were prepared and the backbone curve was developed in accordance with the acceptance criteria presented in FEMA-273 for new materials (Figure 110). The flexure cracking produced pinched loops in the load versus displacement plot. As more cracking developed, there was a change in the slope of the curves, indicative of a decreasing stiffness, much like CMU 1. There was a change in stiffness as the pier went from an uncracked section to a cracked section relying on the tensile capacity of the dowel for stability. The maximum load reached was 16.6 kips [73.8 kN] in the negative direction and 14.7 kips [65.4 kN] in the positive direction (Figure 110). The horizontal solid line in the plots represents the lateral load capacity, Q nbar of 13.6 kips (60.5 kN) from Table 43, where it was calculated that the dowels would yield.

PAGE 159

140 -30-20-100102030-2-1012Displacement (in)Load (kips) Figure 110: Backbone Curve for CMU 2. Strains in FRP composite and steel reinforcement Strains recorded from the strain gauges embedded in the resin of the FRP composite are located in Figure 111. There was no significant strain in the vertical FRP composite pier strip. This could be because strain gauge on the vertical FRP composite strip ended was placed at the bottom of the pier, where the strip was terminated. There was insufficient development length to develop any considerable strain in the strip at the gauge location. There is more considerable strain recorded in the horizontal FRP composite strip on the base. The maximum strain in this strip was 4607 18% of the strain to rupture. This indicated that the FRP in the base was being stressed. As the base cracked, the FRP composites stretched to keep the masonry together.

PAGE 160

141 03000600090001200015000-2-1012Displacement (in)FRP Pier Strain (uE) A 03000600090001200015000-2-1012Displacement (in)FRP Horizontal Base Strain (uE) B Figure 111: FRP Strains for CMU 2. A) Vertical pier strip. B) Horizontal base strip. Strains recorded from the strain gauges placed on the dowels in the pier/base interface area are located in Figure 112. From this figure, the dowels approached yielding. This can be seen as the curve approaches the horizontal line marking the yield strain of a #4 reinforcing bar with a yield stress of 81 ksi [558.5 MPa] or 2793 Because of the bond failure, the masonry could not resist higher displacements which would have eventually caused the dowels to yield. 010002000300040005000-2-1012Displacement (in)Dowel Bar Strain (uE) A 010002000300040005000-2-1012Displacement (in)Dowel Bar Strain (uE) B Figure 112: Steel Strains for CMU 2. A) West dowel. B) East dowel. Out-of-plane movement Out-of-plane displacements at the base of the pier are shown in Figure 113. There is some movement in the instruments negative direction which means that the out-of

PAGE 161

142 plane movement was toward the south. Figure 113 also shows permanent out-of-plane deformation at the east end of the pier. -0.5-0.2500.250.5-2-1012Displacement (in)Out of Plane Movement (in) A -0.5-0.2500.250.5-2-1012Displacement (in)Out of Plane Movement (in) B Figure 113: Out-of-plane Movement for CMU 2. A) East. B) West. Table 45: Summary of Results for CMU 2. Value At a disp. Of (in) [mm] At a load of (kips) [kN] Location Max Load (kips) [kN] -16.6, 14.7 [-73.8, 65.4] -0.55, 0.30 [-14.0, 7.6] Max Drift (%) -1.2, 1.2 -0.56, 0.56 [-14.2, 14.2] -14.5, 9.3 [-64.5,41.4] Max displacement (in) [mm] -0.56, 0.56 [-14.2, 14.2] -14.8, 9.3 [-65.8, 41.4] Max FRP Strain () 4607 0.35 [8.9] 13.3 [59.2] Horizontal base strip Max Steel Strain () 2333 -0.55 [14.0] -16.6 kips [-73.8] West dowel Max Out-of-plane (in) [mm] -0.15, 0.016 [-3.8, 0.40] 0.40, -0.011 [10.2, -0.28] 5.3, -2.2 [23.6, -9.8] East Overall performance It was expected that CMU 2 perform better than CMU 1 because of the added dowels. This did not happen. Instead, the two specimens behaved similarly. CMU 2 reached a maximum load of about 1 kips (4.4 kN) less than that of CMU 1. The splitting failure due to the bond failure of the dowels caused an earlier loss in lateral capacity than

PAGE 162

143 expected. The force required for the bond failure in the concrete was less than the force required to fail the units in compression. It can be concluded that for CMU 2, the loss in lateral capacity was a result of bond failure in splitting mode between the dowels and the grout. The dowels did not develop the amount of strain necessary to fully yield because the bond between the dowel and the surrounding grout split before this could happen. It was decided that for future grouted #4 dowels, a transverse GFRP bars will be placed every other cell to provide confinement to the grouted core. The specimen continued to hold axial load after testing. CMU 3 For CMU 3, the specimen contained FRP, jamb steel and #4 dowels. Transverse GFRP bars were placed in locations shown in schematic to confine the grouted column in the retrofitted steel dowel in order to achieve sufficient development. This was decided after evaluation of the results of CMU 2. See Appendix C for details. Testing The values calculated for determining the deflections that would be used in the displacement controlled testing are contained in Table 46. The predicted lateral capacities were computed by taking moments about the pier toe of the vertical and horizontal forces. The value of k from CMU 1 along with the calculated capacities in Table 46 was used to find y or the yield point. Observations were taken during testing for each displacement. Values of found in Table 47 are the increments of increasing displacement after yielding. The frequency for testing was .05 Hz. Table 46: Calculated Values for CMU 3 Q nbar Q nfrp y 23.4 kips [104 kN] 31.9 kips [141.9 kN] 0.14 in. [3.6 mm]

PAGE 163

144 Table 47: Testing Observations for CMU 3 Observations < 1 Load vs. graph slightly flattens on the negative loading side. No visible cracks. 1 Loops begin to open on the plot, but still relatively linear. Very few cracks visible. 2-4 Loops get wider. Step cracking visible. During testing, the FRP wrinkles in the base where the jamb steel is located. 5-8 Loss of capacity noticed. Piece in masonry base holding the dowel cracks under the FRP, causing wrinkling in the FRP as the pier was loaded. >8 Base FRP piece delaminating. Pier seems to be moving out-of-plane, but at closer inspection, the faceshells were spreading. Results Figure 114 shows the final crack pattern for the specimen. The cracking in the base was difficult to access due to the FRP placed on it. It can be safe to assume that the debonding pattern seen on the FRP is related to the cracking of the masonry underneath it (Figure 115). A B Figure 114: Crack Patterns for CMU 3. A) North face. B) South face.

PAGE 164

145 Figure 115: Debonding pattern on the base of CMU 3. Figure 116 shows pictures taken of the specimen before and after testing. No photo is available of the south side prior to testing; instead, a photo taken five cycles into testing is provided.

PAGE 165

146 A B C D Figure 116: Before and after pictures of CMU 3. A) North face taken prior to testing. B) North face after testing. C) South face five cycles into testing. D) South face after testing. Observed behavior During testing, debonding of the FRP composite was observed around the area containing the dowels in the base. As the pier lifted off the base, the area of grout

PAGE 166

147 surrounding the steel reinforcement rocked. The white line shown in Figure 117 outlines the area of the specimen that rocked with increasing displacements. The FRP composite in the base between the dowels wrinkled as the pier rocked. The specimen lost its lateral load capacity when it freely rocked about the white line at a = 8. Figure 117: Rocking occurred about the white line for CMU 3. After inspection, the grouted jamb containing the jamb steel on the west end looks as if it failed in a cone-like manner at the pier/base intersection (Figure 118). Signs of reinforcing bar slipping were found after removing the grout. At the pier/base intersection, the rib marks from the bar showed signs of abrasion. This means the bar was slipping through the grout as the ends of the pier were displaced from the base.

PAGE 167

148 Figure 118: Cone failure of the grouted jamb containing the jamb steel on the west end of CMU 3. Drift capacity From Figure 119 the maximum drift ratio was -1.6% and 1.7% before a loss in lateral load carrying capacity. Marshall and Sweeney (2002) discovered from their testing that for a CMU pier of the same dimensions with fiberglass reinforcement and #4 reinforcing bars in the jambs had a maximum drift ratio of 1.7%.

PAGE 168

149 -30-20-100102030-4-3-2-101234Drift (%)Load (kips) Figure 119: Drift for CMU 3. Load capacity and backbone curve Load versus displacement plots of the test data were prepared and the backbone curve was developed in accordance with the acceptance criteria presented in FEMA-273 for new materials (Figure 120). The flexure cracking produced pinched loops in the load versus displacement plot. As more cracking developed, there was a change in the slope of the curves, indicative of a decreasing stiffness, much like CMU 1. This decrease in stiffness is due to the change from an uncracked section to a cracked section relying on the tensile reinforcement to remain stable. CMU 3 was able to reach a load of 17 kips [75.6 kN] before there was a change in stiffness. CMU 2 reached 12 kips [53.4] and CMU 1 reached 8 kips [35.6] before their change in stiffness. A larger load needs to be developed in order to crack a section with four grouted cores. The maximum load reached was 28.0 kips [124.5 kN] in the negative direction and 26.5 kips [117.9 kN] in the positive direction (Figure 120). The horizontal solid line in the plot represents the lateral load capacity Q nbar 23.4 kips [104 kN] from Table 46, where it was calculated that the dowels and jamb steel will yield. The measured load reached the calculated load.

PAGE 169

150 -30-20-100102030-2-1012Displacement (in)Load (kips) Figure 120: Backbone curve for CMU 3. Sliding Evidence of sliding in was recorded in Figure 121. The largest sliding displacements occurred in the negative (south) direction. No permanent sliding deformation is seen in the figure because only zero sliding is recorded at a displacement of zero. This means that the pier slid back to its original position after sliding. -0.5-0.2500.250.5-2-1012Displacement (in)In-plane Sliding(in) Figure 121: Sliding of CMU 3.

PAGE 170

151 Strains in steel reinforcement and FRP composite The strain readings (Figure 122) indicate that both the dowel and the jamb steel reached and surpassed yield strain. Both jamb steel bars achieved their maximum strain at approximately 23 kips (102 kN), positive loading for the jamb steel located on the east end and negative loading for the jamb steel located on the west end. There is also a permanent strain in the dowel on the east end for the last cycle as shown by the positive strain value obtained when the displacement is zero. The dowel yielded and permanently deformed. 010002000300040005000-2-1012Displacement (in)Jamb Steel Strain (uE) A 010002000300040005000-2-1012Displacement (in)Dowel Bar Strain (uE) B 010002000300040005000-2-1012Displacement (in)Jamb Steel Strain (uE) C 010002000300040005000-2-1012Displacement (in)Dowel Bar Strain (uE) D Figure 122: Steel strain gauge readings for CMU 3. A) West jamb steel. B) West dowel. C) East jamb steel. D) East dowel.

PAGE 171

152 Strains recorded from the strain gauges embedded in the resin of the FRP composite are located in Figure 122. Strains in the horizontal strip on the base are indicative of bending in the masonry base. Strain readings in the vertical strip on the pier means that the FRP strip held the masonry together despite the stretching of the grouted core containing the reinforcing steel. Without the FRP composite, it is possible that the masonry would have been damaged at a lower load. The maximum strain read in the vertical strip was 3000 12% of the strain to rupture. 03000600090001200015000-2-1012Displacement (in)FRP Pier Strain (uE) A 03000600090001200015000-2-1012Displacement (in)FRP Horizontal Base Strain (uE) B Figure 123: FRP strain gauge readings for CMU 3. A) Vertical pier strip. B) Horizontal base strip. Out-of-plane movement Out-of-plane displacement of the bottom corners of the pier are shown in Figure 124. For the later cycles, the readings do not start from zero but instead have a value. This indicates that a permanent out-of-plane deformation was recorded. After close inspection after testing, it was concluded the pier did not move, just the faceshells around the grouted reinforcement. As the load was applied in the negative displacement direction, the tension in the FRP composite strip caused the units containing the steel

PAGE 172

153 reinforcement to split and move outward. This was movement was recorded as out-of-plane displacement. -0.5-0.2500.250.5-2-1012Displacement (in)Out of Plane Movement (in) A -0.5-0.2500.250.5-2-1012Displacement (in)Out of Plane Movement (in) B Figure 124: Out-of-plane movement for CMU 3. A) East. B) West. Table 48: Summary of Results for CMU 3. Value At a disp. Of (in) [mm] At a load of (kips) [kN] Location Max Load (kips) [kN] -28.0, 26.5 [-124.5, 117.9] -0.43, 0.45 [-10.9, 11.4] Max Drift (%) -1.6, 1.7 0.70, -0.87 [ 17.8, -22.1] -17.7 ,9.8 [-78.7, 43.6] Max displacement (in) [mm] -0.70, 0.87 [-17.8, 22.1] -17.7, 9.8 [-78.7, 43.6] Max FRP Strain () 5247 -0.68 [-17.3] -22.1 [-98.3] Horizontal base strip Max Steel Strain () 4694 0.41 [10.7] 22.4 [99.6] West jamb steel Max Out-of-plane (in) [mm] -0.003, 0.16 [-0.08, 4.1] -0.07, 0.84 [-1.8, 21.3] -15.2, 9.9 [-67.6, 44.0] West & East Overall performance The limiting behavior for CMU 3 was rocking. As the pier rocked, the grout containing the reinforcing steel rocked with it. The FRP on the base began delaminating from the cracking formed beneath it. There were no sudden compression failures.

PAGE 173

154 CMU 3 reached a much higher load capacity (28 kips/124.5 kN) than the previous two tests. It had an increase lateral load capacity of 37% versus CMU 1 (17.7 kips/ 78.7 kN). A reason that this specimen performed well is that the dowels were confined to allow development. The pier continued to hold axial load. If the specimen had lost its ability to carry the axial load provided by the spring, the threaded rod held in tension by the compressed spring would have become loose as the spring expanded. As the specimen was taken out of the test fixture, there was still tension in the rod which needed to be released by the hydraulic jack. CMU 4 For CMU 4, the method of repointing dowels into the vertical masonry joints around the jamb steel was implemented. FRP composite reinforcement was also applied to both sides of the specimen. See Appendix C for more details. Testing The values calculated for determining the deflections that would be used in the displacement controlled testing are contained in Table 49. These values are computed by taking moments about the pier toe of the vertical and horizontal forces. The value of k from CMU 1 along with the calculated capacities in Table 49 was used to find y or the yield point. The frequency for testing was .05 Hz. Observations were taken during testing for each displacement. Values of found in Table 50 are the increments of increasing displacement after yielding. Table 49: Calculated Values for CMU 4. Q nbar Q nfrp y 22.1 kips [98.3 kN] 36.6 kips [162.8 kN] 0.13 in. [3.3 mm]

PAGE 174

155 Table 50: Testing Observations for CMU 4. Observations < 1 No visible damage. Load vs. displacement graph linear and symmetric. 1 Open loops in load vs. displacement graph for the first cycle. Next two cycles reach the same point, but are not open. Small cracks around the repointed bars and the vertical strips at pier-base joint. 2-4 Open first loops. East side of the specimen reaches a lower load than the west side. Cracks form through the dowel epoxy. Cracking visible along base. 5-8 Wide loops observed. Many more visible cracks. FRP is damaged along the base in the area of the reinforcing steel. >8 More cracking. Mortar is knocked loose. Crack forms on the right sill. FRP in base becomes unbonded and pulls some of the lintel concrete off along the bottom. Crack visible through lintel and hydrostone. Results Figure 125 shows the final crack pattern for the specimen. CMU 4 experienced cracking into the lintel, which is made of solid concrete. Cracks propagated into the hydrostone laid between the specimen and the concrete base. This cracking was due to the cracks formed from the stressing of the reinforcing steel. Figure 126 shows pictures taken of the specimen before and after testing. No photo is available of the south side prior to testing; instead, a photo taken two cycles into testing is provided. A B Figure 125: Crack Patterns for CMU 4. A) North face. B) South face.

PAGE 175

156 A B C D Figure 126: Before and after pictures of CMU 4. A) North face prior to testing. B) North face after testing. C) South face two cycles into testing. D) South face after testing. Observed behavior The first thing observed during testing was delamination of FRP composite on the base around the dowels. The delamination was formed as the cracks propagating from the dowels moved upward toward the sill and downward toward the lintel (Figure 127).

PAGE 176

157 Cracks also formed in the epoxy containing the dowels as the pier separated from the base. The lintel was damaged by the debonding FRP composite on the base and the cracking caused by the dowels (Figure 129). Figure 130 shows the crack formed from the dowels to the pier. Figure 127: Crack move from the dowel toward the sill and the lintel. This cracking caused the FRP to delaminate. Figure 128: Crack through the epoxy containing the repointed dowel.

PAGE 177

158 Figure 129: Cracking and delamination into the lintel. Figure 130: Cracked formed from dowel to pier. Drift capacity From Figure 131 the maximum drift ratio was -1.8% and 1.4% before the specimen experienced a loss in lateral load carrying capacity. At the end of testing, the open loops in Figure 131 indicate that the pier was sliding.

PAGE 178

159 -30-20-100102030-4-3-2-101234Drift (%)Load (kips) Figure 131: Drift for CMU 4. Sliding Evidence of sliding in the positive direction is recorded in Figure 132. From this figure, permanent sliding deformation is shown. The climbing sliding position at a displacement of zero means that the pier did not return to its original position after sliding. -0.5-0.2500.250.5-2-1012Displacement (in)In-plane Sliding(in) Figure 132: Sliding of CMU 4. Load capacity and backbone curve The maximum load reached was 27.2 kips [121 kN] in the negative direction and 26.8 kips [119.2] kN in the positive direction (Figure 133). The horizontal solid line in

PAGE 179

160 the plots represents the calculated the lateral load capacity, Q nbar of 22.1kips (98.3 kN) from Table 43, where it was calculated that the dowels and jamb steel will yield. The measured load reached the calculated load. Comparing to CMU 3, there was very little difference in values of drift and maximum lateral load capacity. This can be expected because the same amount of steel was used in both specimens. Recall, CMU 3 contained jamb steel and #4 dowel grouted in the cells next to the jamb steel. CMU 4 also contained jamb steel as well as two #3 dowels on each side of the specimen. Two #3 bars have the same cross sectional area as one #4 bar. Load versus displacement plots of the test data were prepared and the backbone curve was developed in accordance with the acceptance criteria presented in FEMA-273 for new materials (Figure 133). From these figures, it can be noted that the reinforcing steel should have yielded under the loading. Notice that for positive loading, the backbone curve did not surpass yield but only just reached it. -30-20-100102030-2-1012Displacement (in)Load (kips) Figure 133: Backbone curve for CMU 4.

PAGE 180

161 Strain in the steel reinforcement and FRP composite From the strain readings (Figure 134), it can be noted that two of the three dowels that were instrumented reached and surpassed yield strain. The strain gauge on the dowel located on the south face west end of the specimen was not recorded. The flat line from the recorded data marks the limit constraint from the data acquisition system. It can not be known for sure what happened in terms of strain for the dowel beyond this point. For the strain in the dowel on the north face east end in Figure 134, a permanent strain is seen. As the specimen unloaded, the strain in the dowel remained. Recall from Figure 128, a crack was formed in the horizontal bed joint between the pier and the base through the epoxy containing this dowel. It was recorded that the two of the three instrumented dowels yielded (Figure 134). The yielding was recorded for the two dowels located on the north face. The dowel on the south face did not yield. The diagonal FRP composite strips on the north face of the specimen could be the reason for the uneven stress transfer across the joint. From the recorded data, more strain (therefore, more force) transferred through the north face of the specimen. Because no gauge was recorded for the bar on the south side west end, it cannot be concluded that this is true.

PAGE 181

162 010002000300040005000-2-1012Displacement (in)Dowel Bar Strain (uE) A 010002000300040005000-2-1012Displacement (in)Dowel Bar Strain (uE) B 010002000300040005000-2-1012Displacement (in)Jamb Steel Strain (uE) C 010002000300040005000-2-1012Displacement (in)Dowel Bar Strain (uE) D Figure 134: Steel strain gauge readings for CMU 4. A) North face east dowel. B) South face east dowel. C) East jamb steel. D) North face west dowel. Strains recorded from the strain gauges embedded in the resin of the FRP composite are shown in Figure 135. No strain was developed in the gauge placed on the vertical strap in the pier. No strain should be read at the bottom of the pier because the force should be transferred through the dowels at the pier/base interface. Recall that the FRP composite strip was terminated at the bottom of the pier. Strain was recorded for the gauge placed on the horizontal strap in the base. Strain in this location is indicative of bending of the masonry base. The strain in this strip reached 32% of its strain to rupture. Cracks in the grout filling the sills show that bending was occurring. The FRP composite on the base held the masonry together as bending

PAGE 182

163 occurred. It is possible that without the FRP composite, the base would have come apart at a lower lateral load capacity. 03000600090001200015000-2-1012Displacement (in)FRP Pier Strain (uE) A 03000600090001200015000-2-1012Displacement (in)FRP Horizontal Base Strain (uE) B Figure 135: FRP strain gauge readings for CMU 4. A) Vertical pier strip. B) Horizontal base strip. Out-of-plane movement Out-of-plane displacements are shown in Figure 136. The positive readings indicate that the wall is moving northward. The magnitude of these displacements is not significant (when compared to the displacements from the other specimens). -0.5-0.2500.250.5-2-1012Displacement (in)Out of Plane Movement (in) A -0.5-0.2500.250.5-2-1012Displacement (in)Out of Plane Movement (in) B Figure 136: Out-of-plane movement for CMU 4. A) East. B) West.

PAGE 183

164 Table 51: Summary of Results for CMU 4. Value At a disp. Of (in) [mm] At a load of (kips) [kN] Location Max Load (kips) [kN] -27.2, 26.8 [-121.0, 119.2] -0.56, 0.64 [-14.2, 16.3] Max Drift (%) -1.8, 1.4 -0.88, 0.67 [-22.4, 17.0] -18.3, 22.8 [-81.4, 101.4] Max displacement (in) [mm] -0.88, 0.67 [-22.4, 17.0] -18.3, 22.8 [-81.4, 101.4] Max FRP Strain () 7864 -0.88 [-22.4] -18.3 [-81.4] Horizontal base strip Max Steel Strain () 4694 0.42 [10.7] 25.7 [114.3] West end South face dowel Max Out-of-plane (in) [mm] -0.05, 0.07 [1.8, 1.3] -0.57, -0.26 [14.5, 6.6] 27.0, -2.0 [120.1, 8.9] West Overall performance The limiting behavior for CMU 4 was sliding. From Figure 126, there is damage in the masonry and FRP in the base portion surrounding the reinforcing steel. The FRP held the masonry together and allowed the pier to reach a larger capacity. The pier continued to hold axial load. If the specimen had lost its ability to carry the axial load provided by the spring, the threaded rod held in tension by the compressed spring would have become loose as the spring expanded. As the specimen was taken out of the test fixture, there was still tension in the rod which needed to be released by the hydraulic jack. CMU 5 CMU 5 contained two #3 dowels in each jamb. The bars were positioned near opposing faceshells in each jamb. FRP was placed only on the north face of the specimen. See Appendix C for details.

PAGE 184

165 Testing The values calculated for determining the deflections that would be used in the displacement controlled testing are contained in Table 52. The predicted capacities were computed by taking moments about the pier toe of the vertical and horizontal forces. The value of k from CMU 1 along with the calculated capacities in Table 52 was used to find y Observations were taken during testing for each displacement. Values of found in Table 53 are the increments of increasing displacement after yielding. The frequency for testing was .05 Hz. Table 52: Calculated Values for CMU 5. Q nbar Q nfrp y 12.1 kips [53.8 kN] 23.2 kips [103.2 kN] 0.07 in. [1.8 mm] Table 53: Testing Observations for CMU 5. Observations < 1 Load vs. graph linear and symmetric during first 2 cycles. At a = the first open loop was observed. Crack pattern in the base on the back side developed. This pattern is symmetric on both left and right sides. 1 First cycle on hysterisis plot is open. The slope begins to flatten out. 2-4 Change in stiffness observed in plot. Step cracking pattern develops in the area around the dowels. Vertical straps begin to debond near the area of the window. 5-8 Debonding continues down the vertical straps. Gaining little load for large displacements. Debonding in base FRP piece near the cleanouts. >8 Test ended at = 8. Results Results of testing CMU 5 are reported in this section. Figure 137 shows the final crack pattern for the specimen. Figure 138 shows pictures taken of the specimen before and after testing.

PAGE 185

166 A B Figure 137: Crack Patterns for CMU 5. A) North face. B) South face.

PAGE 186

167 A B C D Figure 138: Before and after pictures of CMU 5. A) North face prior to testing. B) North face after testing. C) South face prior to testing. D) South face after testing.

PAGE 187

168 Observed behavior The first cracks appeared during the third set of displacements ( = ) on the south face of the specimen in the base (Figure 139). These cracks were not observed in previous specimens because FRP composite was placed on both the north and south faces. The cause of this cracking can be a result of the restraint at the end of the base. As the pier is loaded, it lifts the grout containing the dowels. The restraint at the end of the base contains this rotation causing the crack. It is possible that this crack pattern occurred on the other specimens containing steel reinforcement but was covered by the FRP composite covering the south face of the base. Figure 139: First cracks in CMU 5 appeared in the base of the south face. Cracks were also observed propagating from the dowels, specifically where the grouted core ended in the pier (Figure 140). There is no cracking evident in the grouted core of the specimen. This is evidence that the FRP composite caused the cracking as it pulled away from the face of the specimen. A loss in lateral load capacity occurred when the units at the jamb split through the web, exposing the grouted core on the east end of the specimen (Figure 141). The failure was sudden and resulted in a sharp reduction in lateral load as seen in Figure 143.

PAGE 188

169 Figure 140: Cracks propagating from the grouted core of the east end. Figure 141: Tensile failure in the masonry exposed the grouted core containing the reinforcing bars. Figure 142 shows a free body diagram (FBD) of the faceshell section that pulled off the grouted core. There is no force at the bottom of the FBD because after the pier

PAGE 189

170 separates from the base there is no longer a force transfer at this interface. The axial force and the tension in the FRP composite strip causes a moment that is restrained by the bearing and tensile stress in the masonry. The tensile stress in the masonry grows as the force in the tensile strip gets larger. When the tensile stress in the masonry reaches the ultimate tensile strength of the masonry, the faceshells crack. A Tfrp Axial Tensile strength of masonryBearing B Figure 142: Splice failure. A) Outline of the section. B) Free body diagram with forces on the section. In strengthening an existing building, it is best to avoid placing FRP composites directly unto one side of the units containing grouted cores. A more effective approach would be to remove the faceshells prior to grouting and place the FRP composite directly unto the grout. If the existing situation already contains grouted reinforcement, GFRP bars can be inserted through the cores to contain the faceshells. This could prove to be very difficult, but it should be considered to avoid the splice failure.

PAGE 190

171 Drift capacity From Figure 143, the maximum drift ratio was 0.85% before the specimen experienced as loss in lateral load carrying capacity. This is the lowest drift ratio of all the specimens tested. -30-20-100102030-4-3-2-101234Drift (%)Load (kips) Figure 143: Drift for CMU 5. Load capacity and backbone curve Load versus displacement plots of the test data were prepared and a backbone curve developed in accordance with the acceptance criteria prescribed in FEMA 273 for new materials (Figure 144). The horizontal solid line in the plots represents the calculated lateral load capacity, Q nbar from Table 52, where the dowels will yield. From this figure, it is evident that the specimen just reached its predicted capacity, but was not able to surpass it.

PAGE 191

172 -30-20-100102030-2-1012Displacement (in)Load (kips) Figure 144: Backbone curve for CMU 5. The maximum load reached was 12.1 kips [53.8 kN] in the positive direction and 12.0 kips [53.4 kN] in the negative direction (Figure 144), which compares well to the calculated yield capacity of Q nbar 12.1 kips [53.8 kN]. CMU 5 had the lowest lateral load capacity of the specimens tested. There was a 32% drop in lateral load capacity from CMU 5 to CMU 1, the control specimen with FRP only. Strains in the steel reinforcement and FRP composite From the strain readings (Figure 145), all four dowels reached yield strain. The horizontal line in the figure marks the strain at which the bar would be expected to yield. The flat line from the recorded data marks where the strain gauge strained beyond the maximum value allowed by the data acquisition system.

PAGE 192

173 010002000300040005000-2-1012Displacement (in)Dowel Bar Strain (uE) A 010002000300040005000-2-1012Displacement (in)Dowel Bar Strain (uE) B 010002000300040005000-2-1012Displacement (in)Dowel Bar Strain (uE) C 010002000300040005000-2-1012Displacement (in)Dowel Bar Strain (uE) D Figure 145: Steel strain gauge readings for CMU 5. A) North face west dowel. B) South face west dowel. C) North face east dowel. D) South face east dowel. FRP composite strains are located in Figure 146. Very little strain was recorded for either strain gauge. The low reading observed in horizontal base strip is a result of very little bending in the base. Unlike CMU 4, which experienced high strain values in horizontal base strip, no deformations in the FRP composite on the base were observed.

PAGE 193

174 03000600090001200015000-2-1012Displacement (in)FRP Pier Strain (uE) A 03000600090001200015000-2-1012Displacement (in)FRP Horizontal Base Strain (uE) B Figure 146: FRP strain gauge readings for CMU 5. A) Vertical pier strip. B) Horizontal base strip. Out-of-plane movement There was no significant out-of-plane movement for this specimen (Figure 147). -0.5-0.2500.250.5-2-1012Displacement (in)Out of Plane Movement (in) A -0.5-0.2500.250.5-2-1012Displacement (in)Out of Plane Movement (in) B Figure 147: Out-of-plane movement for CMU 5. A) East. B) West.

PAGE 194

175 Table 54: Summary of Results for CMU 5. Value At a disp. Of (in) [mm] At a load of (kips) [kN] Location Max Load (kips) [kN] -12.0, 12.1 [-15.4, 53.8] -0.39, 0.28 [-9.9, 7.1] Max Drift (%) -0.85, 0.85 -0.40, 0.41 [-10.1, 10.4] -10.2, 5.7 [-45.4, 25.4] Max displacement (in) [mm] -0.41, 0.41 [-10.4, 10.4] -10.0, 5.7 [-44.5, 25.4] Max FRP Strain () 864 -0.28 [-7.1] -12.0 [-53.4] Horizontal base strip Max Steel Strain () 4693 -0.17 [-4.3] -6.6 [-29.4] East end South face dowel Max Out-of-plane (in) [mm] Negligible, 0.30 [7.6] 0.24 [6.1] -5.1 [22.7] East Overall performance It can be concluded that for CMU 5, the specimen failed because splice failure in the masonry. The moment from having FRP composite on only one side of the pier caused a tensile failure in the masonry block. The force transfer between the pier and the base was occurring on mostly the north face, which was stiffer due to the added FRP composite. This caused bending toward the FRP composite strip, causing the masonry to split in tension. The tensile strength in the masonry is much less than the compression strength of masonry. Recall, CMU 1 experienced toe crushing which is a compressive failure in the masonry. It can be expected that the masonry will fail in tension before it fails in compression. A way of minimizing the eccentricity would be to place more of the ductile material, the steel reinforcement, closer to the face containing the FRP composite reducing the offset in the splice and placing the FRP composite directly unto the grout containing the dowels. This will be tested in a subsequent test.

PAGE 195

176 Even after failure, the pier continued to hold axial load. If the specimen had lost its ability to carry the axial load provided by the spring, the threaded rod held in tension by the compressed spring would have become loose as the spring expanded. As the specimen was taken out of the test fixture, there was still tension in the rod which needed to be released by the hydraulic jack. CMU 6 CMU 6 was strengthened using the grid system of FRP. No steel was placed in the specimen. See Appendix C for details. Testing The values calculated for determining the deflections that would be used in the displacement controlled testing are contained in Table 55. These lateral strength capacities were computed by taking moments about the pier toe of the vertical and horizontal forces. The value of k from CMU 1 along with the calculated capacities in Table 55 was used to find y or the displacement increment. Observations were taken during testing for each displacement. Values of found in Table 53 are the increments of increasing displacement after yielding. The frequency for testing was .05 Hz. Table 55: Calculated Values for CMU 6. Q n unreinforced Q nfrp y 3.4 kips [15.1 kN] 10.6 kips [47.1 kN] 0.06 in. [1.5 mm]

PAGE 196

177 Table 56: Testing Observations for CMU 6. Observations < 1 Load vs. displacement graph is linear and symmetric. Base on the south face of specimen begins to crack. 1 First cycle on the plot begins to lose stiffness. Smaller load increase for a larger displacement. 2-4 As hysterisis loops open, cracks form along the bottom of the pier on the north face. It marks the first crack formed on the north face. 5-8 Slight increase in capacity for larger displacement. Cracks increase in length horizontally and propagate vertically up the pier. >8 Cracks form into the bottom and top lintel. Toe crushing on the south face east side of the pier. This causes specimen to lose capacity. FRP ruptures on both the east and west sides. Results Results of testing CMU 6 are reported in this section. Figure 137 shows the final crack pattern for the specimen. On the south face, each bed joint in the pier experienced cracking. There was toe crushing on both the west and east ends of the pier. Figure 138 shows pictures taken of the specimen before and after testing. A B Figure 148: Crack Patterns for CMU 6. A) North face crack pattern. B) South face crack pattern.

PAGE 197

178 A B C D Figure 149: Before and after pictures of CMU 6. A) North face prior to testing. B) North face after testing. C) South face prior to testing. D) South face after testing.

PAGE 198

179 Observed behavior As loading occurred, the FRP stretched and returned to its original state, behaving in an elastic manner. The low Modulus of Elasticity (low stiffness) of the epoxy is what allowed the FRP to stretch during loading and return to its original state during unloading. The first incident of toe crushing occurred on the east end of the pier (Figure 150). Toe crushing then occurred on the west end, as well as on the sides of the pier (Figure 151). The FRP ruptured along the bottom of the pier connecting to the base on both the west and east ends (Figure 152). The FRP rupture caused a drop in the lateral load capacity and the test was terminated. Figure 150: Toe crushing on the south face east end of CMU 6.

PAGE 199

180 A B Figure 151: Toe crushing side views for CMU 6. A) East side. B) West side. A B Figure 152: FRP rupture for CMU 6. A) East end. B) West end. Drift capacity From Figure 153, the maximum drift ratio was -1.7% and 1.6% before there was a reduction of lateral load carrying capacity. The specimens lateral capacity dropped to -4.0 kips (-17.8 kN) and 5.0 kips (22.2 kN), which is close to the calculated rocking capacity, 3.4 kips (15.0 kN).

PAGE 200

181 -30-20-100102030-4-3-2-101234Drift (%)Load (kips) Figure 153: Drift for CMU 6. Lateral load capacity and backbone curve A load versus displacement plot of the test data was prepared and a backbone curve developed for each specimen in accordance with the acceptance criteria prescribed in FEMA 273 for new materials (Figure 154). The horizontal solid line in the plots represents the predicted lateral load capacity of the specimen. Because the FRP ruptured when it reached the predicted capacity shows that the equation and methods used to calculate these values is accurate. The maximum load reached was 10.3 kips [45.8 kN] in the negative direction and 11.3 kips [50.3kN] in the positive direction (Figure 154). The difference in load for the positive and negative sides is due to the differences in the toe crushing observed at the different ends.

PAGE 201

182 -30-20-100102030-2-1012Displacement (in)Load (kips) Figure 154: Backbone curve for CMU 6. Strains in FRP composite Strains recorded from the strain gauges embedded in the resin of the FRP composite are shown in Figure 155. The largest strains occurred in the vertical strip of the configuration. Little strain was observed in the horizontal strip on the base. This means there was minimal bending of the base.

PAGE 202

183 03000600090001200015000-2-1012Displacement (in)FRP Pier Strain (uE) A 03000600090001200015000-2-1012Displacement (in)FRP Base Strain (uE) B 03000600090001200015000-2-1012Displacement (in)FRP Horizontal Base Strain (uE) C Figure 155: FRP strain gauge readings for CMU 6. A) Vertical pier strip. B) Vertical base strip. C) Horizontal base strip. Out-of-plane movement The pier displaced out-of-plane toward the north (Figure 156). This is due to the crushing of the bed joints. As the pier unloaded and returned to its original position, the crushed masonry did not allow this and displaced the wall. This is similar to what was observed in CMU 1.

PAGE 203

184 -0.5-0.2500.250.5-2-1012Displacement (in)Out of Plane Movement (in) A -0.5-0.2500.250.5-2-1012Displacement (in)Out of Plane Movement (in) B Figure 156: Out-of-plane movement for CMU 6. A) East. B) West. Table 57: Summary of Results for CMU 6. Value At a disp. Of (in) [mm] At a load of (kips) [kN] Location Max Load (kips) [kN] -10.3, 11.3 [-45.8, 50.3] -0.54, 0.54 [-13.7, 13.7] Max Drift (%) -1.7, 1.6 -0.80, 0.78 [-20.3, 19.8] -4.0, 5.4 [-17.8, 24.0] Max displacement (in) [mm] -0.80, 0.78 [-20.3, 19.8] -4.0, 5.4 [-17.8, 24.0] Max FRP Strain () 12005 -0.61 [-15.5] -9.0 [-40.0] Vertical pier strip Max Steel Strain () Max Out-of-plane (in) [mm] -0.36, 0.12 [-9.1, 3.0] -0.10, -0.47 [-2.5, -11.9] .75, 4.1 [3.1, 18.23] West Overall performance The limiting behavior of CMU 6 was rocking. The rocking caused the FRP composite rupture and the toe crushing. As the pier rocked, it compressed the toes and pulled on the FRP composite strip. An advantage of this system was that the low modulus of the grid FRP composite kept the face of the masonry it was placed on free from many cracks. The majority of the cracks occurred on the face with no FRP composite. The low

PAGE 204

185 modulus of the grid also allowed for greater displacements under loading. It would be interesting to investigate the behavior of this system placed on both sides of the wall. Even after failure, the pier continued to hold axial load. If the specimen had lost its ability to carry the axial load provided by the spring, the threaded rod held in tension by the compressed spring would have become loose as the spring expanded. As the specimen was taken out of the test fixture, there was still tension in the rod which needed to be released by the hydraulic jack. CMU 7 CMU 7 contained FRP composite on one side of the pier only. It also contained jamb steel in the outermost jambs of the pier and sills. See Appendix C for details. Testing The values calculated for determining the deflections that would be used in the displacement controlled testing are contained in Table 58. These predicted values of lateral load capacity were computed by taking the moment about the pier toe of the vertical and horizontal forces. The value of k from CMU 1 along with the calculated capacities in Table 58 was used to find y or the yield point. Observations were taken during testing for each displacement. Values of found in Table 59 are the increments of increasing displacement after yielding. The frequency for testing was .05 Hz. Table 58: Calculated Values for CMU 7. Q nbar Q nfrp y 13.6 kips [60.5 kN] 66.7 kips [296.7 kN] 0.08 in. [2.0 mm]

PAGE 205

186 Table 59: Testing Observations for CMU 7. Observations < 1 No visible damage. Load vs. displacement graph linear and symmetric. 1 Crack heard as first open loop formed in plot. 2-4 The specimen gained little lateral load capacity for larger displacements. Crack forms in the base around the grout containing the steel reinforcement. 5-8 Pier rocks about the arch formed from the crack around the bottom of the grouted column. Cracking in the bottom lintel as the base flexes. >8 Entire pier forms a hinge extending from the west side of the west grouted column in the base around to the east side of the east grouted column in the base. FRP remains unaffected by this rocking mode. Rocks about this hinge until a crack forms along the pier base intersection. Results Results of testing CMU 7 are reported in this section. Figure 157 shows the final crack pattern for the specimen. CMU 7 experienced cracking into the lintel, which is made of solid concrete. Figure 158 shows pictures taken of the specimen before and after testing. A B Figure 157: Crack Patterns for CMU 7. A) North face. B) South face.

PAGE 206

187 A B C D Figure 158: Before and after pictures of CMU 7. A) North face prior to testing. B) North face after testing. C) South face two cycles into testing. D) South face after testing.

PAGE 207

188 Observed behavior The pier began to rock at a = 4. The white line in Figure 159 outlines the area of the base that rocked as the pier was loaded. The FRP remained unaffected through rocking. Cracking occurred in the base as the grouted columns were lifted as the pier rocked. As the pier lifted, it brought the masonry in the base around the jamb steel up with it. This created an arched crack around the bottom of the grouted column (Figure 160). Testing ended when the pier separated from the base. Figure 159: Specimen rocked about the area outlined by the white line. Figure 160: Hinge formed in the base along the bottom of the grouted column.

PAGE 208

189 Drift capacity From Figure 161, the maximum drift ratios were -3.3% and 3.2%. The maximum drift ratio was taken from the last cycle that was tested. The testing ended when the pier completely severed from the base (i.e. a crack formed between the pier and the base below it). Unlike the other specimens tested thus far, CMU 7 did not experience any catastrophic or sudden loss of lateral capacity. If testing had continued, it is likely that the drift capacity could have reached much higher values. The drift ratios achieved from this test were three times larger than achieved for the others. Very little research has been performed on masonry piers with jamb and sill reinforcement. Therefore, it is difficult to conclude if the drift ratios found are reasonable for this situation. These large drift ratios are a result of the steel in the sill confining the rocking pier in the base. -30-20-100102030-4-3-2-101234Drift (%)Load (kips) Figure 161: Drift for CMU 7. Lateral capacity and backbone curve The backbone curve in Figure 162 was determined using the standards found in FEMA-273. The horizontal solid line in the plots represents the lateral load capacity,

PAGE 209

190 Q nbar from Table 60, where the jamb steel will yield. The curves did not reach the predicted yielding capacity. The maximum load reached was 12.3 kips [54.7 kN] in the negative direction and 10.8 kips in the positive direction [48 kN] (Figure 162). The capacity predicted for this specimen was 13.6 kips [60.5 kN]. In this case, the method used for calculating the capacity of the specimen is not accurate. -30-20-100102030-2-1012Displacement (in)Load (kips) Figure 162: Backbone curve for CMU 7. Strains in steel reinforcement and FRP composite Strain readings from the jamb steel bar located in the east jamb of the pier, are not conclusive (Figure 163). The pattern of the strain recorded does not make any sense and indicate that something went wrong with the strain gauge. For this reason, readings from this gauge will be ignored. From the strain gauge on the west jamb steel, it can be seen that this bar never yielded. This is consistent with Figure 162 which showed that the specimen never reached its predicted yielding capacity. Readings indicate that there was a force transfer through this bar.

PAGE 210

191 010002000300040005000-2-1012Displacement (in)Jamb Steel Strain (uE) A 010002000300040005000-2-1012Displacement (in)Jamb Steel Strain (uE) B Figure 163: Steel strain gauge readings for CMU 7. A) West jamb steel. B) East jamb steel. There was no significant strain in the strain gauge embedded in the FRP composite on the pier (Figure 164). This is consistent with the fact that no cracking was formed in the pier. This leads to the question of whether or not the FRP improved performance or if the steel in the jambs and sill was the source of the added strength. As a recommendation for future testing of this type of configuration, strain gauges should be placed in the diagonal straps as well as along the length of the vertical straps. It can be that the diagonal straps kept the pier from coming apart during the cycling. Without the FRP composite it is possible that the pier would have come apart (diagonal step cracking, sliding) at an earlier drift ratio. 03000600090001200015000-2-1012Displacement (in)FRP Pier Strain (uE) Figure 164: FRP strain gauge readings for vertical pier strip on CMU 7.

PAGE 211

192 Out-of-plane movement During the later cycles of testing, the bending in the sills was so great that the steel frames holding the instruments had to be removed because they were lifting off the ground. For this reason, there is no concluded maximum values of out-of-plane movement were recorded in Table 60. Even so, instruments were able to record that the wall rotated as it was displacing. The east end of the pier moved in (south) and the west end moved out (north) during cycling. This is evident in the instrument reading of the out-of-plane movement (Figure 165). The change in starting point between cycles is evidence that the pier underwent permanent out-of-plane deformations. It did not return to its original position after cycling. Also, because cracking did not form between the pier and the base until the last cycle it can be assumed that the base was also moving out-of-plane. -0.5-0.2500.250.5-2-1012Displacement (in)Out of Plane Movement (in) A -0.5-0.2500.250.5-2-1012Displacement (in)Out of Plane Movement (in) B Figure 165: Out-of-plane movement for CMU 7. A) East. B) West.

PAGE 212

193 Table 60: Summary of Results for CMU 7. Value At a disp. Of (in) [mm] At a load of (kips) [kN] Location Max Load (kips) [kN] -12.3, 10.8 [-54.7, 48.0] -0.10, 0.10 [-2.54, 2.54] Max Drift (%) -3.3, 3.2 -1.60, 1.52 [-40.6, 38.6] -9.2, 7.7 [-40.9, 34.2] Max displacement (in) [mm] -1.60, 1.52 [-40.6, 38.6] -9.2, 7.7 [-40.9, 34.2] Max FRP Strain () Max Steel Strain () 1540 0.34 [8.64] 9.9 [44.0] West jamb steel Max Out-of-plane (in) [mm] -Overall performance The limiting behavior for CMU 7 was rocking. No sudden lost in lateral capacity was found for CMU 7. Boundary effects The simple structure shown in Figure 166 contains a series of piers and openings. The movement of the center pier is restrained by the movements of the two outer piers. This relationship is shown in Figure 167. Under cyclic loading, the piers displace and rock in the direction of the force. For the case of grout filled sills, like in CMU 7, the grouted sill runs the entire length of the structure under each opening and through the sills. These grouted cells connect from one pier to another. As one pier rocks and lifts at it the bottom corner, the adjacent pier provides a vertical restraint to the sill. The restraint provided by the adjacent pier is represented by the force in the post-tensioned threaded rods in the test set up.

PAGE 213

194 The sills in CMU 7 showed this behavior. The displacement of the sill was observed to resemble the shape of a cantilevered fixed end. It resembled the displaced sill shape shown in Figure 167. From this observation it can be concluded that the post-tensioned rods provided adequate restraint to model the situation found in a real structure. OpeningPierPierOpeningPier Figure 166: Simple structure with openings. OpeningPier Pier Figure 167: Series of piers under cyclic loading. CMU 8 CMU 8 contained FRP on one side. It also contained two #3 dowels near the face of the glassed masonry. See Appendix C for details.

PAGE 214

195 Testing The values calculated for determining the deflections that would be used in the displacement controlled testing are contained in Table 61. The predicted capacities were computed by taking moments about the pier toe of the vertical and horizontal forces. The value of k from CMU 1 along with the calculated capacities in Table 61 was used to find y or the yield point. Observations were taken during testing for each displacement. Values of found in Table 62 are the increments of increasing displacement after yielding. The frequency for testing was .05 Hz. Table 61: Calculated Values for CMU 8. Q nbar Q nfrp y 12.0 kips [53.4 kN] 23.2 kips [103.2 kN] 0.07 in. [1.8 mm] Table 62: Testing Observations for CMU 8. Observations < 1 No visible damage. Load vs. displacement graph linear and symmetric. 1 Cracking heard. Loops in load vs. displacement graph opened up a bit, indicating damage. Slop changes from k to k`. Cracks form along bedjoint where the grouted dowels end in the pier. 2-4 Cracks form on both the north and south sides of the specimen. Debonding is heard and progresses from the cell where the dowel starts towards the bottom of the pier. 5-8 FRP completely debonds on the west side. Loses lateral load carrying capacity. Results Results of testing CMU 8 are reported in this section. Figure 168 shows the final crack pattern for the specimen. Figure 169 shows pictures taken of the specimen before and after testing.

PAGE 215

196 A B Figure 168: Crack Patterns for CMU 8. A) North face. B) South face.

PAGE 216

197 A B C D Figure 169: Before and after pictures of CMU 8. A) North face prior to testing. B) North face after testing. C) South face two cycles into testing. D) South face after testing.

PAGE 217

198 Observed behavior The first cracks noticed during testing occurred in the horizontal joint right above where the dowels were grouted in the pier at a = 1 (Figure 170). From this location, the FRP composite started to debond (Figure 171). As displacements increased, the debonding continued toward the bottom of the pier. A drop in the lateral capacity occurred when the FRP composite strip on the east end of the pier debonded along the length of the grouted cell (Figure 172). The strip on the west end was debonding but was not able to reach the bottom of the pier (Figure 173). It is expected that the debonding on the east end would have continued down the pier if displacements had increased. Figure 170: First crack to form is in the horizonatal bed joint running over the grouted cell.

PAGE 218

199 Figure 171: Debonding of the FRP composite begins over the grouted cell. 1. 2. A 3. 4. B Figure 172: The FRP composite debonded from the east end of the pier on CMU 8. A) Front view of the FRP composite debonding. B) Side view shows that the debonding started at the joint above the dowels and continued down the pier.

PAGE 219

200 Figure 173: Debonding of the FRP composite on the west end began at the joint were the grout started and towards the bottom of the pier. The debonding did not reach the bottom of the pier. Drift capacity From Figure 174 the maximum drift ratio was -0.99% and 1.0% before there was a drop in the lateral load capacity. There was a steady increase in load until the FRP composite strip debonded. The loss in lateral capacity occurred during positive loading which indicates that the specimen lost strength in the east end. -30-20-100102030-4-3-2-101234Drift (%)Load (kips) Figure 174: Drift for CMU 8.

PAGE 220

201 Lateral load capacity and backbone curve A load versus displacement plot of the test data was prepared and a backbone curve was developed in accordance with the acceptance criteria prescribed in FEMA 273 (Figure 175). The horizontal solid line in the plot represents the lateral load capacity, Q nbar from Table 61, where the dowels will yield. From these figures, it can be noted that the dowels should have yielded under the loading. The maximum load reached was 16.1 kips [72 kN] in the negative direction and 15.0 kips [67 kN] in the positive direction (Figure 175). Notice that there was a positive increase in k until the FRP composite strip debonded. -30-20-100102030-2-1012Displacement (in)Load (kips) Figure 175: Backbone curve for CMU 8. Strain in steel reinforcement and FRP composite All four dowels were instrumented with strain gauges. During the wire brushing of the specimen, the strain gauges on the west end of the specimen were damaged. The only steel strain gauges that were recorded were for the dowels in the east of the specimen. From the strain readings (Figure 176), it can be noted that the dowels that were instrumented reached and surpassed yield strain.

PAGE 221

202 010002000300040005000-2-1012Displacement (in)Dowel Bar Strain (uE) A 010002000300040005000-2-1012Displacement (in)Dowel Bar Strain (uE) B Figure 176: Steel strain gauge readings for CMU 8. A) East dowel (nearest to edge). B) East dowel. Strains recorded from the strain gauge embedded in the resin of the FRP are located in Figure 177. The strain gauge in the FRP composite on the pier was placed above the dowels. See Appendix B for details. Strain in this gauge shows that the FRP composite above the dowels was in tension. The maximum strain achieved was in the vertical pier strip. It reached 27% of its strain to rupture. 03000600090001200015000-2-1012Displacement (in)FRP Base Strain (uE) A 03000600090001200015000-2-1012Displacement (in)FRP Horizontal Base Strain (uE) B Figure 177: FRP composite strain gauge readings for CMU 8. A) Vertical pier strip. B) Horizontal base strip. Out-of-plane movement There was no significant out-of-plane movement for this specimen until a jump on the east end of the specimen (Figure 178). The instrument was removed when the FRP

PAGE 222

203 composite strip peeled off. Even though it was removed, the instrument was not disconnected from the data acquisition system and continued to record voltage. This sharp increase is attributed to the full range value of the instrument and should be ignored. Notice there was no significant out-of-plane movement for the instrument on the west end. It can be concluded that placing the dowels against the face closest to the FRP composite strip limited the out-of-plane movement. -0.5-0.2500.250.5-2-1012Displacement (in)Out of Plane Movement (in) A -0.5-0.2500.250.5-2-1012Displacement (in)Out of Plane Movement (in) B Figure 178: Out-of-plane movement for CMU 8. A) East. B) West. Table 63: Summary of Results for CMU 8. Value At a disp. of At a load of Location Max Load (kips) [kN] -16.1, 15.0 [-71.6, 66.7] -0.39, 0.35 [-9.9, 8.9] Max Drift (%) -0.99, 1.0 -0.40, 0.42 [-10.2, 10.7] -15.3, 5.7 [-68.1, 25.4] Max displacement (in) [mm] -0.40, 0.42 [-10.2, 10.7] -15.3, 10.1 [-15.3, 44.9] Max FRP Strain () 6538 0.40 [10.2] 14.9 [66.3] Vertical pier strip Max Steel Strain () 2715 -0.39 [-9.9] -16.1 [71.6] East dowel Max Out-of-plane (in) [mm] -0.06, 0.30 [-1.5, 7.6] 0.38, -0.07 [9.7, 1.8] 14.8, -2.2 [65.8, 9.8] East

PAGE 223

204 Overall performance The failure of CMU 8 can be attributed to a failure in the bond between the FRP composite and the grout used for installing the masonry. The loss in capacity did not occur until the FRP composite strip on the east end debonded. An improvement to the bond between the grout and the FRP composite could result in higher drift values and higher lateral capacity. Recall that CMU 8 is the only specimen in which the flexural FRP composite was directly bonded to the grout enclosing the dowels. All other specimens bonded the flexural FRP composite onto the masonry. A primer is probably needed when bonding FRP composite to a concrete grout. Another way to avoid debonding is to provide a restraint along the FRP composite strip at the bottom of the pier. A horizontal FRP composite strip can be placed over the vertical strips at the bottom of the pier. The horizontal strip would confine debonding at the bottom of the pier.

PAGE 224

APPENDIX E REVIEW OF FEMA 273 Systematic Rehabilitation Rehabilitation Objectives In order to apply the rehabilitation guidelines in FEMA 273, a Rehabilitation Objective needs to be selected as a basis for design. The Rehabilitation Objectives are the level of the desired building performance when a building is subjected to a specified earthquake severity. The guidelines describe four basic Building Performance levels. Each of the four levels describes the allowable levels of damage to a building after an earthquake. Two levels of earthquake severity are described in the guidelines, Basic Safety Earthquake-1 (BSE 1) and Basic Safety Earthquake-2 (BSE 2). Depending upon the location of the structure to be rehabilitated, on of these two hazards are chosen. The goal for seismic rehabilitation in FEMA 273 is to obtain the Basic Safety Objective (BSO). In order to achieve this objective, both the Life Safety Performance Level for BSE 1 and the Collapse Prevention Level for BSE 2 must be met. Under the Life Safety Performance Level, extensive damage has occurred to the structure but it has not collapsed. For URM, there is extensive cracking, noticeable in-plane offsets and minor out-of-plane offsets. For reinforced masonry, extensive cracking is distributed throughout the wall and there is some isolated crushing. The typical drift value for URM and reinforced masonry is 0.6%. These drift values are values associated with the Life Safety Performance Level and are not provided as drift limit requirements. 205

PAGE 225

206 Acceptance Criteria The purpose of analysis using the guidelines is to predict the forces and deformation demands on the structure for a given design earthquake. The acceptance criteria for linear analysis procedure are given as: mKQCEQUD (E.1) QCE is the resistance capacity of the component being analyzed and QUD is the seismic demand placed on the pier. The knowledge factor, K, is used to account for the uncertainty regarding the validity of the computed component strength. The value of K is based on the reliability of the available knowledge of the strength of the existing component, its material properties and engineering judgement. When a minimal level of knowledge is available, a K value of 0.75 is used. A K value of 1.0 may be used where comprehensive knowledge and understanding of the component is known. The component demand modifier or ductility capacity coefficient, m, depends on the structural performance level desired and to which elements they will be applied. This factor can only be applied to deformation controlled components. One of the goals of the research program was to develop m factors for the piers tested. Calculating the m-factor FEMA 273 outlines the method needed to obtain m-factors for new materials or new methods of construction. The required parameters and acceptance criteria are based on load-deflection curves developed through cyclic testing. FEMA states that the objective of the experiment should be to develop the stiffness for the subassembly at different loading increments. It requires that three identical or similar tests be performed

PAGE 226

207 for a unique design configuration. The test set up should incorporate the effects of axial load, moment and shear. A backbone curve, or load deflection envelope, should be developed for each set of data. FEMA 273 outlines the method used to develop the backbone curve. For each specimen, the backbone curve is compared to the Generalized Component Behavior Curves found in FEMA 273 and classified as being either force-controlled or deformation-controlled (Figure 179). 123QyQDgde ab 12QyQDge ab 1QyQDg A B C Figure 179: Generalized Component Behavior Curves. A) Type 1. B) Type 2. C) Type 3. The Type 1 curve is representative of typical ductile behavior. The elastic range, from 0 to 1, is followed by a plastic range, 1 to 2. This plastic range may include strain-hardening or softening. This period of plasticity is then followed by a strength-degraded range, 2 to 3. In this range, the component is resisting a residual force. The component is considered deformation-controlled if the strain-hardening or softening is large, when e is greater than 2g; otherwise, it is classified as force-controlled. The Type 2 curve represents a different type of ductile behavior. Just like the Type 1 curve, Type 2 is characterized by an elastic range followed by a plastic range. The plastic range is then followed by a rapid and complete loss in strength. The component is considered to be deformation-controlled if e is greater than 2g. If not, it is force-controlled.

PAGE 227

208 The Type 3 curve represents brittle behavior. The elastic range is followed by a rapid and complete loss in strength. Components displaying this behavior are always classified as force-controlled. If the component is classified as force-controlled, no m-factor is calculated. The value for m is used as acceptance criteria for deformation-controlled components in linear procedures. It should be taken as 0.75 times the ratio of the deformation acceptance criteria to the deformation at yield. The deformation at yield is given as the deformation at g in Figure 179 for Type 1 or Type 2 curves. To decide on which deformation acceptance criteria to use, the component must first be classified as either a Primary Element or a Secondary Element. A Primary Element is defined as those components that are essential to the ability of the structure to resist earthquake-induced deformations. Secondary Elements are elements that are not essential to the structures ability to resist earthquake-induced deformations. Masonry piers, like those found in this research program, are classified as Primary Elements. After a component is classified as Primary or Secondary, a deformation acceptance criterion is chosen for the Building Performance Level previously chosen. The Basic Safety Objective requires that both the Life Safety Level and the Collapse Prevention Level be considered. For Primary Elements, the acceptance criteria shall be the deformation corresponding to: 0.75 times the deformation at point 2 on the curves for Life Safety and 0.75 times the deformation at point 3 on the Type 1 curve, but this value can not be greater than point 2.

PAGE 228

209 Table 64: Calculated m-Factors for Positive Load Values. Test y (in) [mm] 2 (in) [mm] .75*2 (in) [mm] m 0.071 0.261 0.196 CMU 2 [1.8] [6.6] [5.0] 2.1 CMU 3 0.082 [2.1] 0.285 [7.2] 0.214 [5.4] 2.0 CMU 4 0.077 [1.9] 0.219 [5.6] 0.164 [4.2] 1.6 CMU 5 0.034 [0.9] 0.294 [7.5] 0.221 [5.6] 4.9 CMU 7 0.067 [1.7] 0.843 [21.4] 0.632 [16.1] 7.1 CMU 8 0.047 [1.2] 0.299 [7.6] 0.224 [5.7] 3.6 Table 65: Calculated m-Factors for Negative Load Values. Test y (in) [mm] 2 (in) [mm] .75*2 (in) [mm] m CMU 2 0.062 [1.6] 0.258 [6.6] 0.194 [4.9] 2.3 CMU 3 0.077 [1.9] 0.270 [6.9] 0.203 [5.2] 2.0 CMU 4 0.093 [2.4] 0.389 [9.9] 0.292 [7.4] 2.4 CMU 5 0.034 [0.9] 0.189 [4.8] 0.142 [3.6] 3.1 CMU 7 0.069 [1.8] 0.977 [24.8] 0.733 [18.6] 8.0 CMU 8 0.028 [0.7] 0.271 [6.9] 0.203 [5.2] 5.4 FEMA 273 provides a table of computed m-factors for different performance levels. Values found in the table are based on the ratio of vertical compressive stress to expected compressive strength of the masonry and the ratio of wall/pier height to distance from base to load. It also takes into account the ratio of reinforcement to the cross sectional area of the pier at the critical section. Since the FRP reinforcement does not cross the critical section where the pier meets the base, it will not be included in the reinforcement ratio.

PAGE 229

210 For CMU 2, containing one #4 rebar in its outermost jamb, the m-value from the table is 4.1. For CMU 5 and CMU 8, both containing two #3 rebar in the outermost jambs, the m-value from the table is 4.8. Both these m-values are for the Life Safety Level. Looking at the positive loading direction in Table 64, the m-values for CMU 2, CMU 5 and CMU 8 are 2.1, 4.9 and 3.6, respectively. For the negative loading direction, the m-values are 2.3, 3.1 and 5.4. The m-values calculated for CMU 2 were only half of what is found in the FEMA table. Recall that for CMU 2 the test was terminated when there was a bond strength failure between the rebar and the grout, causing the masonry to split. This premature failure could be the reason for the low m-value. The table from FEMA 273 was not applied to CMU 3, CMU 4 and CMU 7, all containing what was termed as existing steel. The table developed by FEMA is to be applied to those components in which steel is added after construction. The steel in the jambs and sills of this CMU 3, CMU 4 and CMU 7 cause them to behave in a different way then they would have if they only had retrofit reinforcing steel. The table is not applicable to CMU 1 and CMU 6. These two specimens contained no steel and were reinforced by FRP only. A procedure for developing an m-value for these specimens is outlined in the following section. Finding m-factors for non-ductile components The procedures outlined in FEMA 273 for finding m-factors for reinforced masonry make the assumption that the reinforcement is a ductile material. For CMU 1 and CMU 6, which have no steel embedded, the reinforcement is FRP, a brittle material. It was necessary to derive m-factors for non-ductile reinforcement from the procedure outlined in the Commentary that was used to derive the m-factors for ductile

PAGE 230

211 reinforcement. Analytical procedures in the Commentary are based on those presented in Paulay and Priestley (1992). The m-factors were determined from the analysis of lateral deflections for reinforced walls or piers. A moment-curvature diagram needed to be developed in order to find the curvature at yield, y, and at ultimate loading, u. Paulay and Priestley (1992) explain that yield curvature may not necessarily coincide with the yielding of the reinforcement, which generally occurs at a lower curvature. They prescribe a method for approximating linear behavior before yielding. Yielding for brittle components is defined as the point where cracking of the section has occurred and the tensile capacity is provided by the reinforcement. From the moment curvature diagrams for CMU 1 and CMU 6, the curve before yielding is already linear. This simplifies defining a yield point in the curve. For FRP, the curvature at first cracking was defined as its yield curvature, or `y. Curvature ductilites, were determined by dividing the ultimate curvature, u, by the curvature at yield, y: uMyyMu (E.2) where My is the moment capacity at yield and Mu is the ultimate moment capacity. Substituting y with `y gives the following equation for curvature ductility for a non-ductile component, `: 'uMy'yMu (E.3)

PAGE 231

212 Displacement ductilities, were determined from curvature ductilities and from the plastic hinge zone length, lp. Displacement ductilites are important because they relate interstory deflections to each other. The relationship between displacement and curvature ductilities can be expressed by this simple equation found in Paulay and Priestley (1992): K (E.4) where K is a constant. K is derived from the integration of the curvatures along the height of the wall or pier. Approximations for K have been made to simplify the iterations needed to find it. A relationship between the displacement and curvature ductilities is finally concluded in Paulay and Priestly and is used in FEMA 273 to find the displacement ductility: 131lpL 10.5lpL (E.5) Substituting ` for this equation is developed for the displacement ductility for non-ductile reinforced components: '13'1lpL 10.5lpL (E.6) The plastic hinge length, lp, is defined as the length over which the plastic curvature is assumed to equal the maximum plastic curvature. Plastic rotations at the base of the component have been limited to a plastic hinge zone length given in FEMA 273 and this length is equal to: lp0.2L0.04heff (E.6)

PAGE 232

213 where L is the length of the wall or the pier and heff is the height from the base to the lateral force. The m-factors for the Collapse Prevention Level are equal to the displacement ductility from Equation E.6. The m-factor for Life Safety is one-third of the value for the Collapse Prevention Level. The difference in m-factors among the different Performance Levels corresponds to the differences in inelastic deflections associated with the different damage conditions. Using the described methods and developing moment-curvature diagrams for CMU 1 and CMU 6, m-factors for these components with non-ductile reinforcement were developed. These values are located in Table 66. Table 66: Calculated m-Factors Specimens with Non-ductile Reinforcement Test Analytical Procedure FEMA 273 CMU 1 2.8 2.7 CMU 6 4.1 1.5 For CMU 1, the m-values calculated using the different procedures gave approximately the same value. There is quite a large difference between the m-factors calculated for CMU 6. The analytical procedure resulted in a value more than twice that found using FEMA. One reason for this is that FEMA assumes that the reinforcement will not rupture for the given Performance Level. Since the FRP in CMU 6 ruptured, perhaps a lower ultimate value should be used. The value used should depend on the drift capacity the designer wants for the component. The value of u can be chosen for the moment that will give the desired drift capacity. The lower value of u will yield a lower m-factor.

PAGE 233

214 APPENDIX F DATA TABLES Calculated and Measured Lateral Capacities The lateral capacity at first yield measured during testing from the hydraulic actuator, Qm, is compared to the calculated capacity, Qnbar, in Table 67. For most cases, the agreement between the measured and calculated capacities falls within 10%. In the case of CMU 5 and CMU 7, there is a more than 20% overestimation by the analytical model. CMU 5 failed prematurely because of the splice failure, which could account for the difference. CMU 7 was limited by rocking about the base of the specimen, below the reinforcing steel bars. Proper confinement and placement of the FRP composite would help in increasing the lateral load capacities of these two specimens. Table 67: Measured and Calculated Lateral Capacities Specimen Measured Capacity Qm (kips) [kN] Calculated Capacity Qnbar (kips) [kN] Ratio (Qm/Qn) CMU 1* 14.7 13.6 1.08 CMU 2 -16.6 -13.6 1.22 24.9 23.4 1.03 CMU 3 -27.4 -23.4 1.17 N/A N/A N/A CMU 4 -21.7 -22.1 0.98 9.7 12 0.81 CMU 5 -11.2 12 0.93 CMU 6* 9.9 13.6 0.73 CMU 7 -8.9 -13.6 0.65 N/A N/A N/A CMU 8 -14.1 -12.2 1.16 *Note: CMU 1 and CMU 6 did not contain steel.

PAGE 234

215 First yielding is defined as the lateral load at which one of the reinforcing steel bars first reaches yield. Table 68 shows the measured stress, fm, in the reinforcing steel bars at first yielding. The load measured in the actuator load cell is Qm. Reinforcing steel in specimens CMU 2 and CMU 7 never reached yielding. Table 68: Reinforcing Bar Stresses at First Yielding. Specimen Bar fm (ksi) [MPa] Qm (kips) [kN] CMU 1* West dowel 55.6 [383] 14.7 [65] CMU 2 East dowel 67.7 [467] -16.6 [-74] West jamb 79.6 [549] West dowel 46.3 [319] 24.9 [111] East jamb 81 [558] CMU 3 East dowel 74 [510] -27.4 [-122] East jamb 81 [558] East dowel 45.6 [314] CMU 4 East dowel 8.4 [58] -21.7 [-97] West dowel 58.6 [404] West dowel 36.4 [251] 9.7 [43] East dowel 61 [421] CMU 5 East dowel 56.1 [387] -11.2 [-50] CMU 6* West jamb 44.7 [308] 9.9 [44] CMU 7 East jamb 63.3 [436] 8.9 [40] East dowel 61 [421] CMU 8 East dowel 58.9 [406] -14.1 [-63] *Note: CMU 1 and CMU 6 did not contain steel.

PAGE 235

216 For the displacement ductility of = 1, the reinforcing bar stresses are shown in Table 69. The displacement ductility of = 1 in this table corresponds to the = 1 cycles used during testing. The corresponding displacement for = 1 was calculated by dividing the calculated Qnbar by the stiffness of the pier, k. Recall that k was determined during the testing of CMU 1. By definition, the cycles of = 1 should have yielding the reinforcing bars. From Table 69 it can be noted that the reinforcing bars did not yield.

PAGE 236

217 Table 69: Reinforcing Bar Stresses at = 1 Specimen Bar fm (ksi) [MPa] % of yield Qm (kips) [kN] CMU 1* West dowel 38.8 [268] 48% 11.8 [52] CMU 2 East dowel 47.9 [330] 59% -13.2 [-58] West jamb 30.9 [213] 38% West dowel 16.9 [117] 21% 15.3 [68] East jamb 1.8 [12] 2% CMU 3 East dowel 8.3 [57] 10% -14.5 [-64] East jamb 19.5 [134] 24% East dowel 18.8 [130] 31% CMU 4 East dowel 6.6 [46] 11% -11.3 [-50] West dowel 1.6 [11] 3% West dowel 0.87 [6] 1% 8 [36] East dowel 8.0 [55] 13% CMU 5 East dowel 8.2 [57] 13% -8 [-36] CMU 6* West jamb 5.2 [36] 6% 8.6 [38] CMU 7 East jamb 6.9 [48] 9% -9.5 [-42] East dowel 2.6 [18] 4% CMU 8 East dowel 2.8 [19] 5% -8.8 [-39] *Note: CMU 1 and CMU 6 did not contain steel.

PAGE 237

218 Calculating m-values Table 70: Calculated m-Factors for Positive Load Values using FEMA 273 Test y 2 .75*2 m (in) (in) (in) CMU 2 1.0* CMU 3 0.32 0.29 0.22 1.0** CMU 4 0.14 0.41 0.31 1.6 CMU 5 0.13 0.29 0.22 1.3 CMU 7 1.0* CMU 8 N/A N/A N/A N/A *Note: No yielding in steel, m-factor is 1.0. **Note: Value for m-factor was less than 1.0. Table 71: Calculated m-Factors for Negative Load Values using FEMA 273 Test y (in) 2 (in) .75*2 (in) m CMU 2 1.0* CMU 3 -0.37 -0.42 -0.32 1.0** CMU 4 -0.20 -0.58 -0.44 1.6 CMU 5 -0.17 -0.30 -0.22 1.0** CMU 7 1.0* CMU 8 -0.26 -0.27 -0.20 1.0** *Note: No yielding in steel, m-factor is 1.0. **Note: Value for m-factor was less than 1.0. Table 72: Calculated m-Factors for Positive Load Values Test `y (in) [mm] 2 (in) [mm] .75*2 (in) [mm] m CMU 2 0.071 [1.8] 0.261 [6.6] 0.196 [5.0] 2.1 CMU 3 0.082 [2.1] 0.285 [7.2] 0.214 [5.4] 2.0 CMU 4 0.077 [1.9] 0.219 [5.6] 0.164 [4.2] 1.6 CMU 5 0.034 [0.9] 0.294 [7.5] 0.221 [5.6] 4.9 CMU 7 0.067 [1.7] 0.843 [21.4] 0.632 [16.1] 7.1 CMU 8 0.047 [1.2] 0.299 [7.6] 0.224 [5.7] 3.6

PAGE 238

219 Table 73: Calculated m-factors for negative load values Test `y (in) [mm] 2 (in) [mm] .75*2 (in) [mm] m CMU 2 0.062 [1.6] 0.258 [6.6] 0.194 [4.9] 2.3 CMU 3 0.077 [1.9] 0.270 [6.9] 0.203 [5.2] 2.0 CMU 4 0.093 [2.4] 0.389 [9.9] 0.292 [7.4] 2.4 CMU 5 0.034 [0.9] 0.189 [4.8] 0.142 [3.6] 3.1 CMU 7 0.069 [1.8] 0.977 [24.8] 0.733 [18.6] 8.0 CMU 8 0.028 [0.7] 0.271 [6.9] 0.203 [5.2] 5.4 Table 74 and Table 75 show m-factors calculated using the analytical procedure described. The curvature at yield, y, is taken as `y instead of y. Table 74: Calculated m-factors using Moment Curvature at `y for positive loading Test m CMU 2 1.02 CMU 3 2.0 CMU 4 N/A CMU 5 3.5 CMU 7 1.0** CMU 8 N/A Table 75: Calculated m-factors using Moment Curvature at `y for negative loading Test m CMU 2 1.03 CMU 3 1.7 CMU 4 1.4 CMU 5 N/A CMU 7 N/A CMU 8 10.8

PAGE 239

LIST OF REFERENCES Abrams, D. P. (2001) "Seismic Rehabilitation Technologies for Masonry Walls." National Symposium on Comprehensive Force Protection, The Citadel, Charleston, SC. Applied Technology Council (1997a). "Commentary on the NEHRP Guidelines for the Seismic Rehabilitation of Buildings (FEMA Publication 273)." Federal Emergency Management Agency, Washington, D.C. Applied Technology Council (1997b). "NEHRP Guidelines for the Seismic Rehabilitation of Buildings (FEMA Publication 273)." Federal Emergency Management Agency, Washington, D.C. Bajpai, K., and Duthinh, D. (2003) "Bending Performance of Masonry Walls Strengthened with Surface Mounted FRP Bars." Ninth North American Masonry Conference, Clemson, SC. Calvi, G. M., Kingsley, G. R., and Magenes, G. (1996). "Testing of masonry structures for seismic assessment." Earthquake Spectra, 12(1), 145-162. Chajes, M. J., Finch, W. W., Januszka, T. F., and Thomson, T. A. (1996). "Bond and force transfer of composite material plates bonded to concrete." ACI Structural Journal, 93(2), 208-217. Ehsani, M. R., Saadatmanesh, H., and Velazquez-Dimas, J. I. (1999). "Behavior of retrofitted URM walls under simulated earthquake loading." Journal of Composites for Construction, 3(3), 134-142. Holberg, A. M., and Hamilton, H. R. (2002). "Strengthening URM with GFRP composites and ductile connections." Earthquake Spectra, 18(1), 63-84. International Conference of Building Officials (ICBO) (1997). "Acceptance Criteria for Concrete and Reinforced and Unreinforced Masonry Strengthening Using Fiber-Reinforced, Composite Systems (AC125)." Whittier, CA, 8. Kurama, Y. C. (2002). "Hybrid post-tensioned precast concrete walls for use in seismic regions." PCI Journal, 47(5), 36-59. 220

PAGE 240

221 Marshall, O. S., and Sweeney, S. C. (2002). "In-plane shear performance of masonry walls strengthened with FRP." International SAMPE Symposium and Exhibition (Proceedings), 47(II), 929-940. Moon, F. L., Yi, T., Leon, R. T., and Kahn, L. F. (2002). "Seismic Strengthening of Unreinforced Masonry Structures with FRP Overlays and Post-Tensioning." 12th European Conference on Earthquake Engineering, London, UK. Paulay, T., and Priestley, M. J. N. (1992). Seismic Design of Reinforced Concrete and Masonry Buildings, New York, NY: John Wiley & Sons, Inc. Priestley, M. J. N., and Seible, F. (1995). "Design of seismic retrofit measures for concrete and masonry structures." Construction and Building Materials, 9(6), 365-377. Schultz, A. E., and Hutchinson, R. S. (2001). "Seismic Behavior of Partially-Grouted Masonry Shear Walls, Phase 2: Effectiveness of Bed-Joint Reinforcement." ST-98-6, National Institute of Standards and Technology, Minneapolis. Sittipunt, C., Wood, S. L., Lukkunaprasit, P., and Pattararattanakul, P. (2001). "Cyclic behavior of reinforced concrete structural walls with diagonal web reinforcement." ACI Structural Journal, 98(4), 554-562. Triantafillou, T. C. (1998). "Strengthening of masonry structures using epoxy-bonded FRP laminates." Journal of Composites for Construction, 2(2), 96-103. Tumialan, J. G., San Bartolome, A., and Nanni, A. (2003) "Strengthening of URM Infill Walls by FRP Structural Repointing." Ninth North American Masonry Conference, Clemson, SC.

PAGE 241

BIOGRAPHICAL SKETCH Vanessa E. Grillo was born on August 27, 1979, in Ridgewood, NJ. In 1981, she moved to Miami, FL. After graduating high school, she was admitted to the College of Engineering at the University of Florida in the fall of 1997. She graduated with a bachelors degree in civil engineering in August 2001. She immediately began graduate studies in structural engineering at the University of Florida in the Department of Civil and Coastal Engineering. After receiving her Master of Engineering degree in 2003, she plans on working for a structural engineering consulting firm in Miami, FL. 222


Permanent Link: http://ufdc.ufl.edu/UFE0001398/00001

Material Information

Title: FRP/Steel Strengthening of Unreinforced Concrete Masonry Piers
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0001398:00001

Permanent Link: http://ufdc.ufl.edu/UFE0001398/00001

Material Information

Title: FRP/Steel Strengthening of Unreinforced Concrete Masonry Piers
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0001398:00001


This item has the following downloads:


Full Text












FRP/STEEL STRENGTHENING OF UNREINFORCED CONCRETE MASONRY
PIERS















By

VANESSA E. GRILLO


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ENGINEERING

UNIVERSITY OF FLORIDA


2003

































Copyright 2003

by

Vanessa E. Grillo
















ACKNOWLEDGMENTS

Acknowledgments go to the National Science Foundation and the Marketing

Development Alliance for the Composites Industry for their financial support of this

research. I would like to thank my graduate advisor, Dr. H. R. Hamilton, III, for all his

guidance and support during my graduate studies. I would also like to thank the rest of

my committee, Dr. Gary R. Consolazio and Dr. Ronald A. Cook, for their support. Most

importantly, I would like to thank my parents for their love and guidance throughout my

academic career.
















TABLE OF CONTENTS
Page

A C K N O W L E D G M E N T S ......... .................................................................................... iii

LIST OF TABLES ........... ................... .............. ....... ....... vi

LIST OF FIGURES ............................... ... ...... ... ................. .x

ABSTRACT .............. .......................................... xix

INTRODUCTION .................................. .. ... .... ......................

M asonry and E arthquakes..................................................................................1
FR P Strengthening .................. ............... .. ................ .......... ......... ............. 2
FRP Com posites w ith a D uctile Connection ........................................ .....................3

EX PER IM EN TAL PR O G R A M ............................................................... .....................5

T est Sp ecim en s ................................................................... ............................. . 5
M material P rop erties......... ...................................................................... ........ .. .. ....
FRP Composite Configuration .................................. .....................................6
R enforcing Steel P lacem ent................................................................. ........... 11
T e st S e tu p .......................................................................... 1 3
T est P ro cedu res ...................................... ............................................. 15

EXPERIM EN TAL RESULTS... ...................................................................................17

O observed B behavior ............................................................. .............. 17
Load-D isplacem ent Envelopes ........................................................ ............... 29
S y stem D u ctility ................................................................................................... 3 0
Com putting Predicted Capacities........................................... .......................... 35

C O N CLU SIO N S............................................. ............ ............................. .. 39

APPENDIX

A EXTENDED LITERATURE REVIEW ............... .............................................41

A E. Schultz and R. S. Hutchinson (2001) ..................................... ............... 41
O S. M marshall and S. C. Sw eeney (2002) ....................................... ............... 41
F.L. Moon, T. Yi, R.T. Leon and L.F. Kahn (2002)........................................43









J. G. Tumialan, A. San Bartolome and A. Nanni (2003).........................................44
M. J. Chajes, W. W. Finch, Jr., T. F. Januszka and T. A. Thompson, Jr. (1996).......45
K B ajpai and D D uthinh (2003)................. ...... .................. .................... ... 46
C. Sittipunt, S.L. Wood, P. Lukkunaprasit and P. Pattarattankul (2001) ...................47
T .C T riantafillou (1998) ................................................... ............................. 48
M .J.N Priestley and F. Seible (1995)..................................... ......... ............... 50
A.M. Holberg and H.R. Hamilton III (2001)........... ...................................... 52
G.M. Calvi, G.R. Kingsley and G. Magenes (1996) ...............................................53
M.R.Ehsani, H. Saadatmanesh and J.I. Velazques-Dimas (1999).............................54
D .P A bram s (200 1)............. ............................................................ .......... .. 55
Additional Literature Review ................... ................. ..... ............... 56

B EXPERIMENTAL PROGRAM ........................... ................... ................... 65

T e st S e tu p .............................................................................6 5
Data Acquisition ...... ......... .............................73

C SPECIMEN CONSTRUCTION.......................................... 80

M material P ro p erties......................................................................................... .. 82
F R P A p p location ..................................................... ................ 8 5
Com putting Predicted Capacities.......................................... ........................... 89
Individual Specim en D etails............................................................ .....................9 1

D SPE C IM EN R E SU L T S .................................... ........... ................ .......................118

CMU 1 .............. ..... ......................................118
C M U 2 ..............................................................................1 3 3
C M U 3 ..............................................................................14 3
C M U 4 ..............................................................................1 5 4
C M U 5 ..............................................................................1 6 4
C M U 6 ..............................................................................1 7 6
C M U 7 ..............................................................................1 8 5
C M U 8 ..............................................................................1 9 4

E REVIEW OF FEMA 273 ..... ............ ......... ............205

F D A T A T A B L E S ...............................................................2 14

Calculated and Measured Lateral Capacities ................ ....... ..............214
C alcu latin g m -v alu es .......................................................................................... 2 18
LIST OF REFERENCES ......................... ......... .........220

BIOGRAPHICAL SKETCH ................................................ ............... 222
















LIST OF TABLES

Tablege

1 M material P rop erties ................. .................................... ...... ........ .......... .. ....

2 D details ofFRP Com posite Configuration ........................................ .....................9

3 Steel R einforcem ent D details ......... ......... ..................................... ............... 11

4 Summary of Results .................................. .................... ............ 20

5 C calculated m -factors ...................... ................ ................. .... ....... 33

6 Calculated Displacem ent Ductility....... ................... ...................... ............... 34

7 M measured and Calculated Lateral Capacities ................................. ............... 37

8 Full Prism Compression Test Results ........................................... ............... 82

9 Half Prism Compression Test Results.................................... .................... ...............82

10 Full Stretcher Unit Compression Test Results ............... ............................... 83

11 #3 Reinforcing Bar Tension Test Results ..................................... .................83

12 #4 Reinforcing Bar Tension Test Results ..................................... .................84

13 Masonry Joint Reinforcement Tension Test Results............................. .............84

14 Masonry Joint Reinforcement Weld Shear Strength Test Results ...........................84

15 FRP Composite Coupon Tension Test Results ................................ ...............85

16 CM U 1 FR P Q uantities .................................................. .............................. 93

17 C critical D ates for C M U 1 ........................................ ...................... .....................94

18 C conditions for C M U 1.................................................... .............................. 94

19 CM U 2 FR P Q uantities ................................................. ............................... 97

20 C critical D ates for C M U 2 .............................................................. .....................97









2 1 C conditions for C M U 2 ...................................................................... .................. 97

22 CM U 3 FR P Q uantities ................................................. ............................. 101

23 Critical D ates for CM U 3 .............................................. ............................. 102

24 C condition s for C M U 3 ............................................... ....................................... 102

25 CM U 4 FR P Q uantities ................................................ .............................. 106

26 C critical D ates for C M U 4 .......................................................................... .. .... 107

27 C conditions for C M U 4............................................................................ ....... 107

28 CM U 5 FR P Q uantities ................................................ .............................. 109

29 C critical D ates for C M U 5 .......................................................................... ...... 110

30 C conditions for C M U 5............................................................................ ....... 110

31 CM U 6 FRP Quantities .............. .................................. ................................. 111

32 C critical D ates for C M U 6 .......................................................................... ...... 112

33 C conditions for C M U 6............................................................................ ....... 112

34 C M U 7 FR P Q quantities ............ ........................................................................114

35 C critical D ates for C M U 7 .......................................................................... ...... 115

36 C conditions for C M U 7............................................................................ ....... 115

37 C M U 8 FR P Q quantities ............ ........................................................................117

38 C critical D ates for C M U 8 .......................................................................... ...... 118

39 C conditions for C M U 8............................................................................ ....... 118

40 C alculated V alues for CM U 1....................................................... ................... 119

41 Testing Observations for CMU 1..................................................................... 119

42 Summary of Results for CMU 1. ................... ........................................... 132

43 Calculated V alues for CM U 2 ....................................................... .............. 133

44 Testing Observations for CMU 2. ........................................................................134

45 Summary of Results for CMU 2. ................ ........... ..................142









46 C alculated V alues for C M U 3 .................................................................... ..... 143

47 Testing Observations for CM U 3 ........................................ ........ ............... 144

48 Summary of Results for CMU 3 .... ......... .............. .. ...............153

49 Calculated V alues for CM U 4 ....................................................... .............. 154

50 Testing Observations for CM U 4. ......................................................................... 155

51 Summary of Results for CMU 4. ................................. ...................164

52 Calculated V alues for CM U 5 ...................................................... .............. 165

53 Testing Observations for CM U 5.......................................................................... 165

54 Summary of Results for CM U 5. ........................................ ....................... 175

55 Calculated V values for CM U 6. ........................................................... .... ........ ... 176

56 Testing Observations for CM U 6. ......................................................................... 177

57 Summary of Results for CMU 6. ................................. ...................184

58 Calculated V alues for CM U 7 ....................................................... .............. 185

59 Testing Observations for CM U 7. .........................................................................186

60 Summary of Results for CM U 7. ..................................................................... 193

61 Calculated V alues for CM U 8 ....................................................... .............. 195

62 Testing Observations for CM U 8. ......................................................................... 195

63 Summary of Results for CM U 8. .....................................................................203

64 Calculated m-Factors for Positive Load Values..................................................209

65 Calculated m-Factors for Negative Load Values. .............................................209

66 Calculated m-Factors Specimens with Non-ductile Reinforcement ....................213

67 M measured and Calculated Lateral Capacities ............................... ............... 214

68 Reinforcing Bar Stresses at First Yielding. ....................................................... 215

69 R enforcing B ar Stresses at = 1 ................................................ .....................217

70 Calculated m-Factors for Positive Load Values using FEMA 273 ......................218









71 Calculated m-Factors for Negative Load Values using FEMA 273 ....................218

72 Calculated m-Factors for Positive Load Values............................................218

73 Calculated m-factors for negative load values ............................... .......... ..... 219

74 Calculated m-factors using Moment Curvature at A y for positive loading ..........219

75 Calculated m-factors using Moment Curvature at A y for negative loading..........219
















LIST OF FIGURES

Figure page

1 Pier area outlined on a structure. ....................................................... ....................4

2 Pier strengthened for shear and flexure with FRP composites. Rocking and
sliding restrained with a ductile connection at the base of the pier .........................

3 Specimen dimensions and location of joint reinforcement.............. .................6

4 Key to FRP composite configuration on specimen................ .... ...............

5 Free body diagram of half the base used to determine the quantity and placement
of bonded FRP com posite. ............................................... .............................. 8

6 Typical partially reinforced specimen containing existing reinforcing bars in
jam b s an d sills. ....................................................... ................ 12

7 Test set up with specimen ready for testing (looking at North face of specimen)... 14

8 Schematic of specimen in the test set up .... ........... ...................................... 14

9 ICBO test sequence of imposed displacement. ......................................................15

10 Drift for CM U 1. .............. ...... ........ .................. .... .............. 17

11 D rift for C M U 2. .......................... ........................... .... ......... .. .....18

12 D rift for CM U 3 .... ............... ... ... ................ ........... ... ........... ..... ........... 18

13 D rift for C M U 4. .......................... ........................... .... ......... .. .....18

14 D rift for C M U 5. .......................... ........................ .. ............. .. .....19

15 D rift for C M U 6. .......................... ........................... .... ......... .. .....19

16 D rift for C M U 7. .......................... ........................... ............. .. .....19

17 D rift for C M U 8. .......................... ......................................... ........ 20

18 Compression failure on the west end of the pier and buckled FRP ....................21









19 Cracking into the sill on the east end of the specimen caused by tension in the
grouted cells. .........................................................................22

20 B ase splitting of CM U 2................................................. .............................. 23

21 Cone-shaped section of masonry was pulled out as steel reinforcement was
loaded in tension............. .... ............. ......................................24

22 Dowel and jamb steel in CMU 4 pulled out in a V-shape.................. .......... 25

23 Splice failure. A) Outline of the section. B) Free body diagram with forces on
the section ....................... ......... ..... ................. .. .......... 26

24 Specimen rocked about the area outlined by the white line...................................27

25 The FRP composite debonded from the east end of the pier on CMU 8. A) Front
view of the FRP composite debonding. B) Side view shows that the debonding
started at the joint above the dowels and continued down the pier ..........................28

26 Backbone curves for specimens with jamb steel .................................. ...............29

27 Backbone curves for specimens without jamb steel................ .......... .........29

28 Force-displacement curve for URM strengthened with fully bonded FRP
composite and debonding FRP composite. ................................... ............... 30

29 Schematic of the pier connected to the base with dowels............ ...............36

30 Pier area outlined on a structure. ........................................ .......................... 57

31 Failure modes for unreinforced masonry. ..................................... ............... 58

32 Pier strengthened for shear and flexure with FRP composites. Rocking
restrained with the ductile connection at the base of the pier. ...............................61

33 Idealized load-displacement curve for a URM pier. .............................................. 62

34 URM pier with rocking load of POT................. .............................................. 62

35 Idealized load-displacement curve for a pier strengthened with a ductile
connection to the base. ........................................ ........................ 63

36 Pier strengthened with a ductile connection with an overturning load of P.............63

37 Test set up with specimen ready for testing (looking at North face of specimen)...65

38 Plan view of test set up. The North direction in the laboratory is defined ..............66

39 Angles used to prevent out-of-plane movement ......................................66









40 Schematic of specimen in the test set up........ ............ ........................67

41 Dial gauge to monitor slipping between the concrete cap and the specimen's top
lin te l .................................................................................................................... 6 8

42 L lifting Fram e ............. .............................. ............ ..... 69

43 Axial load spring system. A) Schematic. B) Photo...............................................71

44 M T S actuator in place. ..................................................................... ..................72

45 ICBO test sequence of imposed displacement. ........... .................... .................72

46 Locations of instruments for linear displacement measurement.............................74

47 String pots between steel frame and specimen. A) Overall photo of two of the
string pots. B) Closer view of a string pot that measure horizontally. C) Closer
view of a string pot that measures vertically .......................................................74

48 Linear potentiometers on the specimen ................................................... .............. 75

49 Rubber tube installation on dowel over the strain gauge. ......................................76

50 FRP Strain Gauge .................... .................. ...... ....... .............. 77

51 FR P strain gauge locations ............................................... ............................ 78

52 LabVIEW program main screen. ........................................ ......................... 79

53 MTS Signal Generation program screen. ....................................... ............... 79

54 Specimen dimensions and location of joint reinforcement.............. ................ 81

55 Typical partially reinforced specimen containing existing reinforcing bars in
jam bs and sills. .........................................................................82

56 M ixing resin. .........................................................................87

57 FRP application. A) Precoating masonry with resin. B) Pressing FRP cloth into
resin. C) Rolling resin onto placed FRP. D) Troweling placed FRP .....................87

58 Grid FRP application. A) Mixing resin. B) Initial coat of resin on masonry. C)
Grid FRP pressed into the resin. D) Troweling of additional resin........................89

59 Schematic of the pier connected to the base with reinforcing bars..........................91

60 FRP composite placement for CMU 1 North face. .............................................92

61 FRP composite placement for CMU 1 South face. .............................................93









62 CMU 1 two cycles into testing. A) North face. B) South face.............................93

63 Steel reinforcement location for CM U 2.......................................... .................. 94

64 FRP composite placement for CMU 2 North face. .............................................96

65 FRP placement for CM U 2 South face.................................... ...... ............... 96

66 CMU 2 prior to testing. A) North face. B) South face..................... ..............97

67 Steel reinforcement locations for CM U 3. ....................... ....... ....................98

68 Location of transverse GFRP rebar in CM U 3.................................... ............... 99

69 Transverse GFRP block in the faceshell of CMU 3. A) Cross-section of block
containing a GFRP bar. B) GFRP bar embedded in faceshell. ..............................99

70 FRP composite placement for CMU 3 North face. ..............................................100

71 FRP composite placement for CMU 3 South face. ..............................................101

72 CMU 3 prior to testing. A) North face. B) South face five cycles into testing......102

73 Repointing of CMU 4. A) Repointed dowel with smoothed epoxy. B) Grooves
cut into the mortar joints. C) Passing epoxy over the embedded dowel..............104

74 Steel reinforcement locations for CM U 4. .................................. ............... 104

75 D details of repointed #3 dow el. .................................................... ........................104

76 FRP placement for CM U 4 North face........................................ ............... 105

77 FRP placement for CM U 4 South face........................................ ............... 106

78 CMU 4 prior to testing. A) North face. B) South face two cycles into testing......107

79 Steel reinforcement locations for CM U 5. .................................. ............... 108

80 FRP composite placement for CMU 5 North face. (No FRP composite was
placed on the South face) ............................................... ............................ 109

81 CMU 5 prior to testing. A) North face. B) South face ............... .....................110

82 FRP composite placement for CMU 6 North face. (No FRP composite was
placed on the South face) ............. ............ ....................... 11

83 CMU 6 prior to testing. A) North face. B) South face........................................112









84 FRP composite placement for CMU 7 North face. (No FRP composite placed
on the South face) ....... .. .... ........................ ..... ............ ...... .... 113

85 Steel reinforcement locations for CMU 7. .............. ...... ..................114

86 CMU 7 prior to testing. A) North face. B) South face ................. ....................114

87 Steel reinforcement locations for CMU 8. .............. ...... ..................115

88 Installing dowels in CMU 8. A) Grooves cut into the specimen. B) Bars placed
onto plywood. C) Dowels placed on plastic spacers with wire. D) Grout was
shoveled into the openings in the m asonry. ..........................................................116

89 FRP composite placement for CMU 8 North face. (No FRP composite was
placed on the South face) ............. ................................................................. 117

90 CMU 8 prior to testing. A) North face. B) South face ............. ...............118

91 Crack Patterns for CMU 1. A) North face. B) South face. .................................... 120

92 Before and after pictures of CMU 1. A) North face two cycles into testing. B)
North face after testing. C) South face two cycles into testing. D) South face
after testing ..................... .. ....... ................................ ..... ......... 121

93 Debonding of FRP composite on CMU 1. ................................... .....................122

94 Pier displaced A = 1. A) Unbonded length equal to 1. B) Unbonded length equal
to 10........... .......... ........... ................. 122

95 Compression failure on the west end of the pier and buckled FRP. ...................124

96 D rift for CM U ...................................................... ................................... 125

97 B backbone curve for C M U 1. ................................................. ........................... 127

98 Generalized force-displacement relationship for URM and URM retrofit with
FRP overlays (Moon, Leon etal 2002). ........................................ ...............128

99 Force-displacement curve for URM strengthened with fully bonded FRP
composite and debonding FRP composite. ................................. .................129

100 Step cracking opened up and allowed the pier of CMU 1 to slide....................130

101 Sliding of CM U 1 .............. ...... ........ .... ............ .... ............ 130

102 FRP strain gauge readings for CMU 1. A) Vertical pier strip. B) Vertical base
strip. C) H orizontal base strip. ..........................................................................131

103 Out-of-plane movement for CMU 1. A) West. B) East. ........................................ 132









104 Crack Patterns for CMU 2. A) North face. B) South face. ....................................134

105 Before and after pictures of CMU 2. A) North face taken before testing. B)
North face after testing. C) South face taken before testing. D) South face after
te stin g .......................................................................... 13 5

106 B ase Splitting of CM U 2. ........ ................................................... ............... 136

107 Splitting bond failure caused cracks to propagate into the sill at the east end of
C M U 2 ......... ...... .................................................. ............................ 137

108 Cracks move radially from the bar to the edges of the pier in CMU 2..................138

109 D rift for C M U 2 ......................................................................... .....................139

110 Backbone Curve for CM U 2. ............................................................................ 140

111 FRP Strains for CMU 2. A) Vertical pier strip. B) Horizontal base strip. .............141

112 Steel Strains for CMU 2. A) West dowel. B) East dowel ....................................141

113 Out-of-plane Movement for CMU 2. A) East. B) West............... ... .................142

114 Crack Patterns for CMU 3. A) North face. B) South face. ........................ 144

115 Debonding pattern on the base of CMU 3................................... ............... 145

116 Before and after pictures of CMU 3. A) North face taken prior to testing. B)
North face after testing. C) South face five cycles into testing. D) South face
after testing g ........................................................................................... 14 6

117 Rocking occurred about the white line for CMU 3............................147

118 Cone failure of the grouted jamb containing the jamb steel on the west end of
C M U 3 ......................................................................................... 14 8

119 D rift for CM U 3. ............................................................................... ..... ........... 149

120 Backbone curve for CM U 3. .............................................................................150

121 Sliding of CM U 3 .................. .................................... ................ 150

122 Steel strain gauge readings for CMU 3. A) West jamb steel. B) West dowel. C)
E ast jam b steel. D ) E ast dow el ...................................................................... .. 151

123 FRP strain gauge readings for CMU 3. A) Vertical pier strip. B) Horizontal
b ase strip .................................................. .... ................. 152

124 Out-of-plane movement for CMU 3. A) East. B) West. ............. ..................153









125 Crack Patterns for CMU 4. A) North face. B) South face. ....................................155

126 Before and after pictures of CMU 4. A) North face prior to testing. B) North
face after testing. C) South face two cycles into testing. D) South face after
te stin g .......................................................................... 15 6

127 Crack move from the dowel toward the sill and the lintel. This cracking caused
the FR P to delam inate. ............................... .... .......... .............. ........... .. 157

128 Crack through the epoxy containing the repointed dowel.................................. 157

129 Cracking and delamination into the lintel. .................................. ............... 158

130 Cracked formed from dowel to pier. ........................................... ............... 158

13 1 D rift for C M U 4 ......................................................................... .....................159

132 Sliding of CM U 4 .................. ..................................... .. ........ .. .. 159

133 Backbone curve for CM U 4. .............................................................................160

134 Steel strain gauge readings for CMU 4. A) North face east dowel. B) South
face east dowel. C) East jamb steel. D) North face west dowel...........................162

135 FRP strain gauge readings for CMU 4. A) Vertical pier strip. B) Horizontal
b ase strip .................................................. .... ................. 16 3

136 Out-of-plane movement for CMU 4. A) East. B) West. ......................................163

137 Crack Patterns for CMU 5. A) North face. B) South face. ....................................166

138 Before and after pictures of CMU 5. A) North face prior to testing. B) North
face after testing. C) South face prior to testing. D) South face after testing.........167

139 First cracks in CMU 5 appeared in the base of the south face.............................168

140 Cracks propagating from the grouted core of the east end................................... 169

141 Tensile failure in the masonry exposed the grouted core containing the
reinforcing bars. .....................................................................169

142 Splice failure. A) Outline of the section. B) Free body diagram with forces on
th e se ctio n ...................................... ............................................... 1 7 0

143 D rift for C M U 5. ......................................................................... .....................17 1

144 Backbone curve for CM U 5. .............................................................................172









145 Steel strain gauge readings for CMU 5. A) North face west dowel. B) South face
west dowel. C) North face east dowel. D) South face east dowel........................173

146 FRP strain gauge readings for CMU 5. A) Vertical pier strip. B) Horizontal base
strip ...................................... ................................................... . 1 7 4

147 Out-of-plane movement for CMU 5. A) East. B) West. ......................................174

148 Crack Patterns for CMU 6. A) North face crack pattern. B) South face crack
pattern .......................... ......... ...................................... 177

149 Before and after pictures of CMU 6. A) North face prior to testing. B) North face
after testing. C) South face prior to testing. D) South face after testing ...............178

150 Toe crushing on the south face east end of CMU 6. ..........................................179

151 Toe crushing side views for CMU 6. A) East side. B) West side........................180

152: FRP rupture for CMU 6. A) East end. B) West end.........................................180

153 D rift for C M U 6. ......................................................................... .....................18 1

154 Backbone curve for CM U 6. .............................................................................182

155 FRP strain gauge readings for CMU 6. A) Vertical pier strip. B) Vertical base
strip. C) H orizontal base strip. ........................................ ......................... 183

156 Out-of-plane movement for CMU 6. A) East. B) West. ......................................184

157 Crack Patterns for CMU 7. A) North face. B) South face. ....................................186

158 Before and after pictures of CMU 7. A) North face prior to testing. B) North
face after testing. C) South face two cycles into testing. D) South face after
te stin g .......................................................................... 18 7

159 Specimen rocked about the area outlined by the white line ..............................188

160 Hinge formed in the base along the bottom of the grouted column ......................188

16 1 D rift for C M U 7. ......................................................................... .....................189

162 Backbone curve for CM U 7. .............................................................................190

163 Steel strain gauge readings for CMU 7. A) West jamb steel. B) East jamb steel..191

164 FRP strain gauge readings for vertical pier strip on CMU 7 ..............................191

165 Out-of-plane movement for CMU 7. A) East. B) West. ........................................192









166 Simple structure with openings. ........................................................................... 194

167 Series of piers under cyclic loading. ................ ...... .. .... .. ...................... 194

168 Crack Patterns for CMU 8. A) North face. B) South face. ....................................196

169 Before and after pictures of CMU 8. A) North face prior to testing. B) North face
after testing. C) South face two cycles into testing. D) South face after testing....197

170 First crack to form is in the horizontal bed joint running over the grouted cell... 198

171 Debonding of the FRP composite begins over the grouted cell ..........................199

172 The FRP composite debonded from the east end of the pier on CMU 8. A) Front
view of the FRP composite debonding. B) Side view shows that the debonding
started at the joint above the dowels and continued down the pier........................199

173 Debonding of the FRP composite on the west end began at the joint were the
grout started and towards the bottom of the pier. The debonding did not reach
the bottom of the pier. ............................ .......... ............ .... 200

174 D rift for C M U 8. ......................................................................... .....................200

175 Backbone curve for CM U 8. ............................................................................201

176 Steel strain gauge readings for CMU 8. A) East dowel (nearest to edge). B)
E ast dow el. .......................................... ........................... 202

177 FRP composite strain gauge readings for CMU 8. A) Vertical pier strip. B)
H horizontal base strip ............ ... ......................................................... .... .... ... .. 202

178 Out-of-plane movement for CMU 8. A) East. B) West. ......................................203

179 Generalized Component Behavior Curves. A) Type 1. B) Type 2. C) Type 3......207


xviii















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Engineering

FRP/STEEL STRENGTHENING OF UNREINFORCED CONCRETE MASONRY
PIERS

By

Vanessa E. Grillo

December 2003

Chair: H.R. Hamilton
Major Department: Civil and Coastal Engineering

This thesis presents research on the strengthening of unreinforced masonry (URM)

piers strengthened with fiber-reinforced polymers (FRP) in conjunction with ductile

reinforcement. Eight concrete masonry pier specimens were strengthened with a

combination of FRP composite strips and reinforcing steel. Specimen construction

included the pier and a portion of the masonry just below the pier. FRP composite strips

were strategically placed to improve flexure and shear strength in the in-plane direction.

Steel dowels were added to the specimen by grouting into cells or by repointing in

vertical head joints and were designed to yield prior to the FRP composite rupture,

resulting in a ductile response.

Improvement in lateral capacity of three to nine times the capacity of URM rocking

mode was achieved. Drift capacities ranged from 0.85% to 3.3%. Confinement of the

masonry in the base of the specimen was shown to be a controlling factor in the extent of

yielding attained in the steel dowels.















INTRODUCTION

Masonry and Earthquakes

Earthquakes can cause extensive damage to unreinforced masonry (URM)

structures. Many older masonry structures currently in use were designed and constructed

with little or no consideration of earthquake resistance. In addition, recent changes in

seismic requirements have left many URM buildings in need of strengthening. This is

particularly true in Midwestern states where seismic design has not been considered in

the past. Since the advent of modern reinforced masonry construction, URM structures

have been viewed as a significant liability when considering strengthening. Traditional

methods of URM strengthening include shotcrete or the addition of steel frames or

reinforced concrete walls. In general, these options ignore the contribution of the URM

components to the lateral capacity. Furthermore, they are quite expensive and pose

significant inconvenience for the building occupants during installation.

Significant progress has been made in identifying URM behavior under extreme

loads and recognizing the contribution of URM components to both strength and ductility

of the building system (Calvi et al. 1996, Marshall and Sweeney 2002). URM structures

are usually analyzed as a system of shear walls and/or piers that carry proportional levels

of the story shears. Previous research (Calvi et al. 1996, Marshall and Sweeney 2002,

Triantafillou 1998) as well as recent design guidelines (FEMA 273) recognize rocking,

sliding, toe-crushing and diagonal-tension as the four primary failure modes for URM

piers. Relative values of aspect ratio, compressive strength, and axial stress determine the









failure mode most likely to occur. It is further recognized that sliding and rocking are

relatively stable failure modes and provide some energy dissipation (Calvi et al. 1996,

Marshall and Sweeney 2002). This approach allows the engineer to determine the

probable level of building performance with no strengthening. If additional capacity is

needed, then strengthening is necessary. One option available that can provide added

capacity is to apply fiber reinforced polymer (FRP) composites.

FRP Strengthening

Fiber reinforced polymer (FRP) composites are made of continuous glass, carbon

or aramid fibers bonded to the substrate with a resin polymer matrix and are typically

unobtrusive to the building occupants, require relatively little surface preparation and are

economical. Recent research has shown that FRP composites can be applied to increase

strength and change the failure modes of masonry walls (Abrams 2001, Ehsani et al.

1999, Marshall and Sweeney 2002, Triantafillou 1998). The majority of previous testing

of FRP composites and masonry has focused on piers with medium to high aspect ratios

(ratio of height to width). Higher aspect ratios generally result in an overturning or

rocking behavior in the unstrengthened state. Application of bonded FRP composites can

cause a shift in the failure mode to one of sliding or shear. The ability of a surface bonded

FRP composite strengthening system to prevent masonry from falling off the structure

during an extreme event also makes it a favorable alternative to traditional methods

(Marshall and Sweeney 2002).

Strengthening masonry walls with FRP composites requires that the composite be

bonded to the wall surface either in sheets covering the entire surface of the pier or in

strips placed at strategic locations on the pier. Unidirectional strips of FRP are thought to

be preferable in terms of economy and behavioral response to the two-dimensional









fabrics that cover the entire surface of the masonry wall (Triantafillou 1998). Full

coverage may also cause problems by limiting moisture movement through the wall.

FRP Composites with a Ductile Connection

Although FRP composites increase lateral load capacity, they do not significantly

improve ductility and may actually decrease ductility if an undesirable failure mode is

precipitated. This is due to the brittle nature of the composite material. Holberg and

Hamilton (2002) proposed a hybrid system, consisting of bonded FRP composites in

conjunction with steel. The FRP composite adds sufficient strength to the masonry

allowing the steel to reach yield, thus incorporating ductility into the system.

Research was conducted by Holberg and Hamilton (2002) on URM walls

retrofitted with a hybrid strengthening system consisting of FRP composites and steel.

Two different types of steel connections were tested, an internal and an external. The

internal connection was a steel reinforcing bar placed in the outermost cells of the wall

and fully grouted into a concrete foundation. The external connection was a steel angle-

plate assembly attached to the foundation. The drift capacities of the reinforced

specimens reached up to 1.7%. The lateral capacities of the strengthened specimens were

nearly doubled when compared to the lateral capacity of an unstrengthened specimen.

Holberg and Hamilton (2002) investigated using a ductile connection between a

masonry pier and a concrete foundation. While the results of this research appear

promising, the behavior of the connection between the pier and surrounding masonry in

multistory buildings (Figure 1) needs to be investigated. One of the problems associated

with grouting dowels into masonry is providing confinement to enable the dowels to

yield. The focus of the research presented in this paper is on the connection between the










pier and its supporting masonry, and its effect on the ductility and stability of the in-plane

pier behavior.

















Figure 1: Pier area outlined on a structure.

In Figure 2, the vertical FRP composite strips are designed to provide enough

additional strength to resist the shear and flexural stresses experienced during an

earthquake. The ductile connection is designed to yield prior to failure of the FRP

composite. Adequate strength must be provided in the masonry surrounding the dowels to

ensure yielding at the pier/base interface and prevent a pull out failure. In addition to

confinement of the dowels, the masonry below the pier requires strengthening against

flexure and shear induced by the tensile forces in the dowels.






Shear and
flexural strength
,_/ "" improved



Sliding and rocking restrained
with ductile connection

Figure 2: Pier strengthened for shear and flexure with FRP composites. Rocking and
sliding restrained with a ductile connection at the base of the pier.















EXPERIMENTAL PROGRAM

Eight reinforced concrete pier specimens were constructed and tested to investigate

the behavior of the pier, ductile connection and base subjected to in-plane cyclic loading.

The experimental parameters included the amount of FRP composite and reinforcing

steel placed onto each specimen.

Test Specimens

Eight concrete masonry specimens were constructed in running bond of medium

weight 8-in. (200-mm) concrete masonry units by Painter Masonry, Inc. in Gainesville,

FL. Type N mortar was used in face shell bedding for each of the specimens. The ASTM

C90 units were purchased locally from Florida Rock Industries, Inc. in Gainesville, FL.

All test specimens were single wythe and consisted of a pier 48-in. (1200-mm) tall by 48-

in. (1200-mm) wide and a base 9-ft. 4-in. (2800-mm) long and 16-in. (400-mm) tall

(Figure 3). Since the test specimen consisted of full scale CMU, no scaling effects were

required. The masons installed ASTM A Class B2 Hot Dipped Galvanized After

Fabrication standard sized (9 Gauge) wire joint reinforcement every other course in the

specimen. This was done in accordance to common practice in the field. The location of

the joint reinforcement is indicated in Figure 3.

The specimen was constructed on a precast concrete lintel 9-ft. 4-in. (2800-mm)

long. A second precast concrete 5-ft. 4-in. (1600-mm) lintel was placed on the top course

of the pier. Both lintels were placed with the same mortar used for constructing the

specimen. The outer cells of the specimens' sills were filled with grout (approximately










one month after construction) to prepare them for post tensioning to the concrete base at a

force of 42 kips (187 kN).

5'-4" (1630 mm)
4'-0" (1220 mm) Precast
SConcrete Lintel




I Joint
S Reinforcement Grouted Cells






Precast C=
Lifting Rings Precast .
Lifting Rings Concrete Lintel
9'-4" (2845 mm) ,



Figure 3: Specimen dimensions and location of joint reinforcement.


Material Properties

Masonry, steel reinforcement, joint reinforcement and FRP composite material

properties were determined using ASTM standard test methods (Table 1). Five coupons

of the cured FRP composite material were tested in tension. The average width of the

tensile specimens was 1.04-in. (26.42-mm) and the average thickness was 0.91-in (23.11-

mm). The specimens were cut from a section of cured FRP composite and then milled to

a 1-in. width. Preparation of the coupons and testing followed ASTM D3039.

FRP Composite Configuration

The specimens, except for CMU 6, were strengthened with varying widths and

lengths of unidirectional fiberglass fabric (27 ounces/yard2, .0859 kg/ft2) bonded to the

surface of the specimen with a two-part epoxy. The manufacturer-specified tensile

strength of the composite was 330 ksi (2275 MPa). Specimen CMU 6 was reinforced










using unidirectional grid reinforcement. This grid system is a high strength,

unidirectional reinforcement made by bonding E glass fiber rovings with epoxy resin in a

controlled factory environment. The specified strength of the grid system given by the

manufacturer was 14,400 lbs/ft (210 kN/m) and was adhered to the specimen using a low

modulus epoxy.

Table 1: Material Properties
Material Average No. of
Material Test
Strength specimens
CMU Unit 2224 psi 5
Stretcher Unit Strength (15.3 MPa)
CMU Full Prism 2167 psi
Prism Strength (14.9 MPa)
CMU Half Prism 3880 psi
Prism Strength (26.7 MPa)
#4 Reinforcing Yield 81 ksi
Bar Strength (558 MPa)
#3 Reinforcing Yield 62 ksi
Bar Strength (429 MPa)
Joint Yield 107 ksi
Reinforcement Strength (738 MPa)
FRP 2.86
Tensile
Composite n kips/inch 5
Strength
Coupon S (500 N/mm)

Quantity and placement of the FRP composites for the pier and the base were

determined using a strut and tie analysis and basic mechanics principles. Figure 4 shows

the locations for the FRP composite placement. Details of FRP composite configuration

are given in Table 2. The basic configuration included strips oriented vertically along the

pier jambs as well as diagonally across the pier. The vertical strips (V) increased the in-

plane flexural strength and the diagonal strips (X) increased the diagonal tension strength.

Bi-directional FRP composite fabric with a +450 fiber orientation was applied along the

top of the pier at the pier/lintel interface to prevent separation of the lintel and pier (SP).

The base was modeled as a deep beam and it was expected that flexure and shear

reinforcement would be required to maintain stability (Figure 5). A strip of unidirectional








FRP composite was placed at the top of the base along its length (HB) to provide

reinforcement against bending caused by the tension from the pier reinforcement and

compression from the end restraints. Shear reinforcement was added by covering the base

with bi-directional FRP composite (SB) to inhibit diagonal cracking in the base and to

provide confinement to the grouted dowels.


Figure 4: Key to FRP composite configuration on specimen.


CHold Down Force T owel/FRP

I TFRP
OM
C

Figure 5: Free body diagram of half the base used to determine the quantity and
placement of bonded FRP composite.














Table 2: Details of FRP Composite Configuration
Specimen North Face South Face FRP Composite Configuration*


~I '










. I
2..




t I I


V Unidirectional 4 x 80 (102 x 2032)

HB Unidirectional 6 x 112 (152 x 2844)

SB 450 18 x 112 (457 x 2844)

X Unidirectional 3 x 68 (76 x 1727)

SP 450 12 x 8 (305 x 1219)

V Unidirectional 4 x 80 (102 x 2032)
HB Unidirectional 3 x 112 (76 x 2844)
SB 450 18 x 112 (457 x 2844)
X Unidirectional 3 x 68 (76 x 1727)
SP 450 12 x 48 (305 x 1219)

V Unidirectional 6 x 80 (152 x 2032)
HB Unidirectional 3 x 112 (76 x 844)
SB 450 18 x 112 (457x 2844)
X Unidirectional 3 x 68 (76 x 1727)
SP 450 12 x 48 (305 x 1219)

V Unidirectional 7 x 80 (178 x 2032)
HB Unidirectional 3 x 112 (76 x 844)
SB 450 18 x 112 (457x 2844)
X Unidirectional 3 x 68 (76 x 1727)
SP 450 12 x48 (305 x 1219)


*Note: Dimensions: in x in (mm x mm)


CMU 1







CMU 2







CMU 3







CMU 4














North Face





I I


Table 2. Continued
South Face


41ti


Specimen


*Note: Dimensions: in x in (mm x mm)
**Note: CMU 6 was strengthened using a FRP grid system.


FRP Composite Configuration*
Unidirectional 8 x 80 (203 x 2032)
Unidirectional 3 x 112 (76 x 844)
450 18 x 112 (457x 2844)
Unidirectional 3 x 68 (76 x 1727)
450 12 x 48 (305 x 1219)

Grid 9 x 80 (229 x 2032)
None
Grid 18 x 112 (457 x 2844)
Grid 7 x 68 (178 x 1727)
Grid 12 x 48 (305 x 1219)
Unidirectional 8 x 80 (203 x 2032)
None
None
Unidirectional 3 x 68 (76 x 1727)
450 12 x 48 (305 x 1219)
Unidirectional 8 x 80 (203 x 2032)
Unidirectional 3 x 112 (76 x 844)
450 18 x 112 (457x 2844)
Unidirectional 3 x 68 (76 x 1727)
450 12 x 48 (305 x 1219)


CMU 5


CMU 6**






CMU 7






CMU 8










Reinforcing Steel Placement

In an attempt to force ductile failure, steel dowels were installed into the specimens

as indicated in Table 3. The dowels were sized to yield prior to the failure of the FRP

composite. Specimens CMU 2, CMU 3, CMU 5 and CMU 8 contained dowels grouted

into the cells after the specimens were constructed. CMU 3, CMU 4 and CMU 7

contained jamb and sill reinforcement that was placed as part of the specimen

construction. These specimens were intended to represent conditions in which

prescriptive amounts of steel have been added during construction. The remainder of the

specimens were intended to represent unreinforced conditions.

Table 3: Steel Reinforcement Details.
Jamb
Specimen Jamb Dowels Pier Section
Steel

CMU1 No None E

2 #4 x 40"
CMU 2 No in grouted s
jamb
2 #4 x 40"
CMU 3* Yes in grouted
cell
4 #3 x 32"
repointed
in vertical
head joint
4 #4 x 40"
CMU 5 No grouted in ElE E M
jamb

CMU 6 No None


CMU 7 Yes None

4 #3
CMU 8 No grouted in i D
jamb
*Note: CMU 3 contains transverse GFRP bars in every cell containing the dowel steel.

Dowel installation required face shell removal at the fourth course up from the

bottom of the pier. The dowels were held in place by hand in the center of the cell and









grout was scooped in through the opening. A tamping rod was used to consolidate the

grout around the dowels because there was no room to insert a vibrator into the face shell

opening. The face shells were then replaced with mortar.

Specimen CMU 4 called for "structural repointing" with a #3 dowel that was 32-

inches (812 mm) long. Vertical grooves were cut into the head joint starting at the top of

the bottom lintel. The width of the groove matched the width of the mortar joint, 3/8-inch

(9.5-mm), so the entire mortar joint was removed. The depth of the groove was 3/4-inch

(19-mm) deep. Dimensions for repointing were taken from similar testing by Bajpai and

Duthinh (2003). A gel epoxy adhesive was used to install the dowel into the groove. A

layer of the epoxy was squeezed into the groove. After coating the bar with the epoxy, it

was pressed into the groove. Epoxy was then passed over the bar filling in any gaps and

then smoothed by hand, giving the appearance of a repointed joint.

Specimens CMU 3, CMU 4 and CMU 7 were constructed with #4 reinforcing bars

in the outer jambs and sills (Figure 6). The jamb and sill steel were grouted into the

specimen one month after specimen construction. The grout for the jamb and sill steel

was shoveled in and consolidated with a concrete vibrator.


Figure 6: Typical partially reinforced specimen containing existing reinforcing bars in
jambs and sills.









Transverse glass fiber reinforced polymer (GFRP) composite bars were placed in

the cells containing the dowels of CMU 3 to avoid the splitting failure that occurred in

CMU 2. The purpose of the GFRP bar was to confine the grout containing the dowels in

order to develop the strain necessary to yield them. The #2 (3.2 mm) GFRP bars had a

specified tensile strength of 120 ksi (825 MPa) and a modulus of elasticity of 5.92 psi

(40.8 GPa). One quarter-inch diameter hole was drilled through the unit face shell before

dowels were grouted into the cells. The length of the transverse bars matched the

thickness of a CMU block, 7.625 inches (194 mm). The GFRP bars were pushed through

the block via the drilled holes. The holes in the face shells were filled with silicone

caulking and covered with duct tape to prevent the GFRP bars from moving during

grouting of the dowels.

Test Setup

The test setup was designed to apply in-plane cyclic displacements under

displacement or load control. A hydraulic actuator was located on the reaction frame and

displaces the concrete cap on the specimen (Figure 7). Positive load and displacement

correspond to actuator tension. The reaction frame that supports the actuator is

constructed of steel and is prevented from overturning by its connection to the laboratory

strong floor. A concrete cap and base were constructed for the test fixture and designed to

handle loads of 50 kips (220 kN). The concrete base was post-tensioned to the laboratory

strong floor. The concrete cap was braced to the laboratory wall by steel angles to prevent

out-of-plane twisting.

A 55-kip (245 kN) hydraulic actuator was used for testing. The hydraulic actuator

was part of a closed-loop hydraulic loading system. The controller used a sinusoidal

voltage output from the data acquisition system to impose the displacements. The











actuator load cell was used to monitor the lateral load applied to the specimen during

testing.

Gravity load was applied to the specimen using rail car springs and threaded rods.

This translated into an axial stress of 75 psi (0.5 MPa) on the net section of the base of

the pier. The gravity load included load from two rail car springs compressed to develop

6 kips (27 kN), the weight of the concrete cap (2 kips, 9 kN) and the self-weight of the

pier and lintel (1 kip, 4 kN).


Spring Syslem
55 kip Aclualor




Concrete Cap Reacllon Frame







Concrete Base


Figure 7: Test set up with specimen ready for testing (looking at North face of specimen).

Gravity Load Simulator


-- MTS Load


68inches 8 inches
(1727 nm) (200 ) Downward Force 42 laps
(187 kN)


Figure 8: Schematic of specimen in the test set up.










A personal computer, National Instruments LabVIEW software and a 16-bit data

acquisition card were used for data acquisition. Linear and string potentiometers were

used for measuring displacements. Foil strain gauges were placed on the steel reinforcing

bars and FRP composite strips.

Test Procedures

The test displacement sequence from ICBO Acceptance Criteria for Concrete and

Reinforced and Unreinforced Masonry Strengthening Using Fiber-Reinforced Composite

Systems (AC125) (ICBO 1997) was followed (Figure 9). The displacement at which the

reinforcement was calculated to yield is marked as the first yield point, t = 1. For

specimens that did not contain any reinforcing steel (CMU 1 and CMU 6), the yield point

was taken as the displacement that was expected to cause a rocking failure of an

unreinforced specimen. The specimens were loaded in displacement control with three

complete cycles for each displacement level, [t. The displacements were t = 1/4, 1/2, 3/4, 1,

2, 3, 4, 6, 8, 10, 12, 16, 18 and 20 and were increased until the specimen experienced a

loss in lateral load capacity. The loading rate was approximately 20 seconds per cycle.

12-
I0





4


Figure 9: ICBO test sequence of imposed displacement (ICBO 1997).









It was from CMU 1 that the stiffness of all the walls was determined. At the

beginning of the test, the specimen was cycled several times in the elastic range to

determine the uncracked stiffness (k) of the specimens, 167 kips/inch (29.2 kN/mm). The

stiffness was found using the displacement and load cell reading of the hydraulic

actuator. This value of k along with the individual specimen's calculated capacity was

used to find Ay, the yield point for each specimen. For those specimens containing steel

reinforcement the yield point is the displacement associated with measured yield in the

steel reinforcement. The cracking load was used as the yield point for those specimens

without steel reinforcement.
















EXPERIMENTAL RESULTS

Observed Behavior

Figure 10 through Figure 17 show the lateral load versus drift ratio plots for all the

tests. Drift ratio is the in-plane displacement of the pier divided by the height. The

horizontal solid line in the plots represents the calculated lateral load capacity where all

the steel reinforcement in the specimen is assumed to yield. Details for these calculated

values are discussed in a subsequent section. For CMU 1, which contained no steel

reinforcement, the horizontal solid line represents the calculated rocking load for an

unstrengthened specimen. For the specimen strengthened with only the grid FRP

composite, CMU 6, the horizontal solid line represents the calculated load at which the

grid FRP composite was calculated to rupture.

30 133.5
20 89
i10 44.5

o0 0 0 o
-j-10 -44.5 -
-20 -89
-30 -133.5
-2 -1 0 1 2
Drift (%)

Figure 10: Drift for CMU 1.










30

20

. 10


0
o1
S1 0
J-10

-20

-30


-2 -1 0
Drift (%)


Figure 11: Drift for CMU 2.


30

20

,,10

"0
o
J-10

-20

-30


-2 -1 0
Drift (%)


Figure 12: Drift for CMU 3.


30

20

10

"a 0
-10

-20

-30
-2 -1 0
Drift (%)


Figure 13: Drift for CMU 4.


133.5

89

44.5

0 *

-44.5 -

-89

-133.5


1 2


133.5

89

44.5

0 *

-44.5 -

-89

-133.5


1 2


133.5

89

44.5

0

-44.5 -

-89

-133.5


1 2










30

20

.10


0
n5
J-10

-20

-30


-2 -1 0
Drift (%)


133.5

89

44.5

0 *

-44.5 -

-89

-133.5


1 2


Figure 14: Drift for CMU 5.


30

20

&.10

"0
o
J-10

-20

-30


-2 -1 0


-2 -1 0
Drift (%)


Figure 15: Drift for CMU 6.

30






.-10

-20

-30


1 2


-4 -3 -2 -1 0 1 2 3 4
Drift (%)


Figure 16: Drift for CMU 7.


133.5

89

44.5

0 *

-44.5 -

-89

-133.5


133.5

89

44.5

0 r

-44.5 -

-89

-133.5


- K3


a
I ~
c,










30

20

.10

"a 0

.-10

-20

-30


-2 -1 0
Drift (%)


133.5

89

44.5 z

0 "
cc
0
-44.5 --

-89

-133.5


1 2


Figure 17: Drift for CMU 8.

Table 4 summarizes the key results of the testing. The maximum lateral load

capacity and maximum drift values were observed in the cycle immediately before the

specimen experienced a loss in lateral load capacity.

Table 4: Summary of Results
Maxi Maximum
Maximum
ecin Dt Lateral Limiting
Specimen Drift
Load Mode
(kips) [kN]
1.4 17.7 [79]
CMU 1 1.4 17.7 [79] Toe Crushing
-1.4 -14.9 [-66]

CMU 2 12 14.7 [65 Bond Failure
-1.2 -16.6 [-74]
CM 3 1.7 28.0 [126] R
-1.7 -26.5 [-118]

CMU 4 1.4 27.2 [121] Sliding
-1.8 -26.8 [-119]
0.85 12.1 [54]
CMU 5 0.85 12. [54] Splice Failure
-0.85 -12.0 [-54]
1.6 11.3 [50]
CMU 6 1.3 [] FRP Rupture
-1.6 -10.3 [-46]
CMUI T7 3.2 10.8 [48] Flexure
-3.3 -12.3 [-55]
1.0 15.0 [67] FRP Bond
CMU 8
-0.99 -16.1 [-72] Failure

CMU 1. The FRP composite strip on the east end of the pier fully delaminated on

the bottom course at a drift ratio of-1.1%, which is evident in the sharp drop of lateral









capacity in Figure 10. The east end of the pier retained approximately 90% of its

maximum capacity in the negative direction. The delamination softened the specimen

resulting in larger drift at the same load.

It can be concluded that CMU 1 reached its limiting capacity when toe crushing

occurred at the west end of the pier, as shown in Figure 18. This is evident in the

reduction in load seen at a drift ratio of 1.4% (Figure 10). Even following toe crushing,

the specimen retained approximately 40% of its peak capacity. The behavior had changed

from that of flexure to sliding, which has favorable energy dissipation characteristics as

evidenced by the open loops in Figure 10. Sliding occurred during [ = 20 across a step

crack originating at the bottom west corner of the pier and running diagonally to the top

east corer of the pier.






















Figure 18: Compression failure on the west end of the pier and buckled FRP.

CMU 2. Debonding of the vertical FRP composite strip (V) in the pier began

early in the cycling at t = 1. The debonding initiated at the bed joint just above the









dowel. At [ = 4, the FRP composite sheet on the base (SB) began debonding around the

grout containing the dowel. As the pier was loaded at t = 4, the grouted cells containing

the dowel were lifted in tension causing cracking into the sill shown in Figure 19. This

was observed at both the east and west ends of the pier.

As shown in Figure 20, the grout containing the dowel split the specimen at the

base. The cause of the bar failure is due to insufficient development length for the

dowels. About 80% of the stress to yield was developed in the dowels prior to the

splitting failure.

























Figure 11 shows large open loops during the last set of cycles indicating energy

dissipation. These open loops were thought to be caused by the dowel movement after the

splitting failure. Inspection after testing showed that the lug marks in the grout had been

abraded smooth by the movement of the reinforcing bars.


















Figure 20: Base splitting of CMU 2.

It can be concluded that for CMU 2, the behavior was limited by a bond failure in

splitting mode between the dowels and the grout. The dowels did not develop the amount

of strain necessary to fully yield because the bond between the dowel and the surrounding

grout split before this could happen. It was decided that for future grouted #4 dowels, a

transverse GFRP bars would be placed in every other cell to provide confinement to the

grouted core.

CMU 3. At [t = 2, debonding of the FRP composite (SB) was observed around the

area containing the dowels and the jamb steel in the base. The tension developed in the

dowel and jamb steel caused a cone-shaped section of masonry around the dowel and

jamb steel to pull out of the base (Figure 21). A sheet of +450 FRP composite had been

applied to both sides of the base to prevent this failure from occurring. This pullout

occurred at a drift ratio of approximately 1.15%, resulting in a 35% drop in lateral

capacity. Behavior following the pullout was one of rocking.

Both the dowel steel and jamb steel on both the east and west ends yielded at [t = 6,

which correspond to a drift ratio of approximately 0.9%. The hysteresis loops tend to

open significantly beyond this drift limit indicating good energy dissipation. No

movement in the sill was observed, nor was there any cracking in the grout containing the

sill steel. It can be concluded that the confinement provided by the FRP composite









around the sill prevented large movements and cracking in the sill area. This allows for a

greater lateral capacity and minimal damage to the base of the pier.






















Figure 21: Cone-shaped section of masonry was pulled out as steel reinforcement was
loaded in tension.

CMU 4. At t = 3, delamination of FRP composite (SB) on the base around the

dowels was observed. Cracking propagated from the dowels toward the sills, downward

toward the lintel and continuing into the hydrostone laid on the foundation (Figure 22). A

V-shaped section of masonry was pulled out as the jamb steel and dowels were loaded in

tension. The lintel was damaged by the debonding FRP composite on the base and the

cracking caused by tension in the dowels.

Open loops in Figure 13 at a drift ratio of +0.4% correspond to yielding of the east

and west jamb steel and the instrumented dowel located on the north face of the pier. The

instrumented dowel located on the south face had only reached 60% of its yield stress by

the end of testing.






























Figure 22: Dowel and jamb steel in CMU 4 pulled out in a V-shape.

CMU 5. A loss in lateral load capacity occurred when the units at the east end

jamb split through the web, exposing the grouted core of the specimen (Figure 23A)

during the last set of displacements (t = 8). The length over which the FRP composite

strip overlaps the grouted jamb is known as a splice. The failure at this splice was sudden

and resulted in a 54% reduction of peak lateral load.

The moment from having FRP composite on only one face of the pier caused a

tensile failure in the masonry block. The force transfer between the pier and the base

occurred primarily on the north face, which was stiffer due to the added FRP composite.

This caused bending toward the FRP composite strip, which in turn caused the masonry

to split in tension.

Figure 23B shows a free body diagram (FBD) of the face shell section that pulled

off the grouted core. The only force transfer to the base of the specimen is though the

dowel. The tension in the FRP composite strip causes a moment that is restrained by the









tensile stress in the masonry. The tensile stress in the masonry increases as the force in

the tensile strip gets larger. When the tensile stress in the masonry reaches the ultimate

tensile strength of the masonry, the face shells crack.










Tensile strength
of masonry


A B

Figure 23: Splice failure. A) Outline of the section. B) Free body diagram with forces on
the section.

CMU 6. The limiting behavior of CMU 6 was flexure, causing the FRP composite

to rupture at the pier/base interface and toe crushing at t = 16 on both the east and west

ends of the pier. Rupture of the FRP composite strips caused a drop in the lateral load

capacity and the test was terminated. This is evident in Figure 15, when there is a

reduction in capacity at a drift ratio of +1.7%. Even after rupture, the pier retained

approximately 50% of its peak capacity.

The low modulus of elasticity (low stiffness) of the epoxy allowed the FRP

composite to stretch during loading and return to its original state during unloading and

kept the north face of the pier free from experiencing extensive cracking, which occurred

thorough every bed joint in the south face of the pier.









CMU 7. The limiting behavior for CMU 7 was flexure, evident in the pinched

loops in Figure 16. No sudden drop in lateral capacity was found for CMU 7 and the drift

capacity continued to increase. The steel reinforcement did not reach yield stress.

At a t = 4, the specimen rocked about the area outlined by the white line in Figure

24. The FRP composite remained unaffected through rocking. Tension in the jamb steel

caused cracking in the masonry surrounding the grouted jamb steel in the base, creating

an arched crack around the bottom of the grouted column. During loading, the steel

reinforcement in the sill flexed, which was observed by the vertical movement of the

sills.

From Figure 16, the maximum drift ratios were -3.3% and 3.2%. The maximum

drift ratio was taken from the last cycle that was tested. The drift ratios achieved from

this test were three times larger than achieved for the others. Very little research has been

performed on masonry piers with jamb and sill reinforcement. Therefore, it is difficult to

conclude if the drift ratios found are reasonable for this situation.
















Figure 24: Specimen rocked about the area outlined by the white line.

CMU 8. The first cracks noticed during testing occurred at [ = 1 in the horizontal

joint right above where the dowels were grouted in the pier. From this location, the FRP









composite (V) started to debond. As displacements increased, the debonding continued

toward the bottom of the pier. A drop in the lateral capacity occurred when the FRP

composite strip on the east end of the pier debonded along the length of the grouted cell

(Figure 25). The FRP composite strip on the west end debonded but was not able to reach

the bottom of the pier. It is expected that the debonding on the east end would have

continued down the pier if displacements had increased. Yielding of the dowels occurred

at a drift ratio of 0.7% (t = 6).

The failure of CMU 8 can be attributed to a failure in the bond between the FRP

composite and the grout used for installing the masonry. A 60% loss in capacity was

observed in Figure 17 as the FRP composite strip on the east end completely debonded.

Recall that CMU 8 is the only specimen in which the flexural FRP composite was

directly bonded to the grout enclosing the dowels. All other specimens bonded the

flexural FRP composite onto the masonry.
















A B

Figure 25: The FRP composite debonded from the east end of the pier on CMU 8. A)
Front view of the FRP composite debonding. B) Side view shows that the
debonding started at the joint above the dowels and continued down the pier.










Load-Displacement Envelopes

Backbone curves for specimens with and without jamb steel are shown in Figure 26

and Figure 27, respectively. These curves were developed in accordance with the

acceptance criteria presented in FEMA 273 for new materials. The backbone curves

generally start with a higher uncracked stiffness. As the drift approaches 0.1% drift ratio,

a dramatic reduction in stiffness is apparent. In Figure 26, the different backbone curve of

CMU 7 is attributed to its behavior during testing. CMU 7 did not experience a drop in

lateral capacity but instead increased drift ratio without increasing in lateral capacity.

30 133.5
20 \ 89

"- 10 44.5 "
0 I 0

o -10 -CMU3 -445 0
_- -CMU 4
-20 CMU7 -89
-30 -133.5


2 -1 0 1
Drift (%)


Figure 26: Backbone curves for specimens with jamb steel.

30
20
10
0 I
c CMU 1
S-10 --CMU2
S-0 -CMU 5
-20 -- -CMU6
-CMU 8
-30
-2 -1 0 1 2
Drift (%)


133.5
89
44.5 z
0
-0
-44.5 --
-89
-133.5


Figure 27: Backbone curves for specimens without jamb steel.


2









The behavior observed can be represented by the force-displacement curve shown

in Figure 28. Moon, Leon et al. (2002) suggest that the initial stiffness, k, is that of an

uncracked specimen. A reduced stiffness after cracking, k', is caused by a combination of

cracking, yielding and damage to the specimen. The specimen stiffness can be further

reduced if the FRP composite bond is inadequate. This softening effect is represented by

the second branch of the curve and is caused by bond failure. The dotted line represents

the sharp reduction in strength caused by the failure of the FRP composite. After this

point, the pier returns to the strength found in an unreinforced specimen.

S Adequate FRP Bond





Inadequate FRP Bond

Unstrengthened Capacity





% Drift

Figure 28: Force-displacement curve for URM strengthened with fully bonded FRP
composite and debonding FRP composite.

System Ductility

One of the goals of this research was to assess the ductility of several systems that

include FRP composites and steel. One method is to use the FEMA 273 requirements for

assessing the ductility of new material. System or component ductility is quantified by

the following equation found in FEMA 273:

mKQcCE uo (1)









where QCE is the resistance capacity, QUD is the seismic demand, K is the knowledge

factor used to account for material property uncertainty and m is the component demand

modifier or ductility capacity coefficient. This equation is used to determine component

and system capacity using equivalent lateral force procedures. As system or component

ductility increase, the value of m increases proportionally, resulting in a reduction in

lateral load demand through inelastic response.

The procedures outlined in FEMA 273 for determining m for reinforced masonry

assume that the reinforcement is a ductile material and, therefore, "deformation

controlled". Brittle materials (such as masonry and FRP composites) are referred to as

"force controlled" and, at the component level, are assumed to exhibit a linear stiffness to

failure. However, systems strengthened with brittle materials show behavior similar to

that of a deformation controlled system, with displacement capacity and energy

dissipation developed from the cracking and damage sustained by the specimen. This is

seen in the bilinear relationship shown in Figure 28. Because CMU 1 and CMU 6 were

reinforced with FRP composites only, the traditional definition of displacement ductility,

[t (ultimate displacement divided by the yield displacement), is not applicable. For

masonry components and systems with traditional ductility the FEMA 273 m-factors are

a function of the displacement curvature and ductility. In an effort to quantify the ductile

nature of the specimen's response, a moment-curvature analysis was conducted using the

test results from specimens CMU 1 and CMU 6. The fundamental difference is that

cracking was used instead of yielding when conducting the analysis.









The curvature ductility related to cracking, [i',, was determined by dividing the

ultimate curvature, yp,, by the curvature at first cracking, qpcr:

u Mcr
/= -- (2)


where My is the yield moment and M, is the ultimate moment. The curvatures for

the moment curvature analysis were obtained by using the recorded strains in the FRP

composite and basic mechanics principles.

The relationship between the displacement ductility and curvature ductility is given

by Paulay and Priestly (1992) and is also used in FEMA 273 to determine the

displacement ductility. Modifying the equation to give displacement ductility as a

function of cracking gives:


,a= 1+3(,'- 1)(1 0.5 ) (3)
L L

The plastic hinge length, lp, is defined as the length over which the plastic curvature is

assumed to equal the maximum plastic curvature. It is recognized that without ductile

material (as in the cases of CMU 1 and CMU 6), plasticity can never be achieved. For the

purposes of evaluation, an estimation of the plastic hinge length is made using the

equation suggested by FEMA 273 where the plastic rotations at the base of the

component have been limited to a plastic hinge zone length equal to:

p = 0.2L + 0.04heff (4)

where L is the length of the wall or the pier and heffis the height from the base to the

lateral force.

FEMA 273 uses the moment curvature approach to find m-factors for the Collapse

Prevention Performance Level where m is equal to the calculated displacement ductility.









For the Life Safety Performance level, the m-factor is one-third the value of the Collapse

Prevention Level. The procedures outlined in FEMA 273 are used to calculate the m-

factor for the Life Safety Performance Level.

Using the method described, m-factors for components with non-ductile

reinforcement were calculated (Table 5). The first column are the Lt'A calculated using the

moment curvature analysis and the second column shows the m-factors calculated using

the procedures outlined in FEMA 273 for new materials. The m-factor shown in Table 5

using the moment curvature analysis is one-third the value calculated for [LA. This allows

for a direct comparison between the methods for the Life Safety Performance Level.

Table 5: Calculated m-factors
Moment Curvature FEMA 273
Test
Analysis Procedures
CMU 1 2.8 2.7
CMU 6 4.1 1.5

The remainder of the specimens contained steel reinforcement that was

instrumented with strain gauges to determine when yielding occurred. Consequently, the

displacement ductility, [LA, can be calculated directly from:

A.
A= (5)


where Au and Ay are the measured displacements at ultimate and first yield,

respectively. The results of this analysis are shown in Table 6 for positive and negative

load direction. For comparison, displacement ductility at first cracking, L' A, was

calculated for each of the specimens with steel reinforcement and are also shown in Table

6.

For each specimen, the ductility ratio calculated is higher using A'y than the ratio

calculated using Ay. This is reasonable because the change in specimen stiffness occurs at










a lower displacement than the displacement at yielding because of cracking sustained by

the specimen prior to yielding.

Table 6: Calculated Disolacement Ductility


N' ote: INo yielding in steel, [t


is I.U.


Recall that the displacement ductility is equivalent to the value of m in FEMA 273.

Table 7-4 in FEMA 273 provides acceptable m-factors for reinforced masonry. Values

given in this table for the cases of CMU 2, CMU 5 and CMU 8 are 4.1, 4.8 and 4.8,

respectively. The table in FEMA 273 does not apply to the cases with jamb and sill steel;

so acceptable m-factors for CMU 3, CMU 4 and CMU 7 could not be determined.

Comparing the ductility ratios in Table 6 to the acceptable values, the ductility ratios

found using A'y are closer to the accepted values than the ductility ratios found using Ay.


Ay A, A,
Test (in) (in) (in) LA P A
[mmn] [mm] [mm]
0.071 0.33
n/a 1.0* 4.6
CMU 2 [1.8] [6.6]
-0.062 -0.26
n/a 1.0* 4.2
[-1.6] [-6.6]
0.32 0.082 0.33
1.03 4.0
CMU 3 [8.1] [2.1] [8.4]
-0.37 -0.077 -0.42
1.1 5.5
[-9.4] [-1.9] [-10.7]
0.14 0.077 0.41
2.9 5.3
CMU 4 [3.6] [1.9] [10.4]
-0.20 -0.093 -0.58
2.9 6.2
[-5.1] [-2.4] [-14.7]
0.13 0.034 0.29
2.2 8.5
CMU 5 [3.3] [0.9] [7.4]
-0.17 -0.034 -0.30
1.8 8.8
[-4.3] [-0.9] [-7.6]
0.067 0.84
n/a 1.0* 12.5
[1.7] [21.4]
CMU 7
-0.069 -0.98
n/a 1.0* 14.2
[-1.8] [-24.8]
0.047 0.30
n/a n/a 6.4
CM1U 8 [1.2] [7.6]
CMU 8
-0.26 -0.028 -0.27
1.04 9.6
[-6.6] [-0.7] [-6.9]
*T 1 In









Further research should be conducted using similar configurations to support the use of

the ductility ratios presented in this paper.

Computing Predicted Capacities

For each of the specimens, the flexural capacity was calculated for two critical

sections. Holberg and Hamilton (2001) describe the methodology for the calculations.

The first critical section checked is where the FRP composite ends and flexural capacity

is provided solely by the reinforcing steel bars. The second critical section checked is

where the FRP composite strips placed on the pier jambs provide the flexural capacity. At

this section, the FRP composite strip is providing all the flexural capacity. In Figure 29,

P, is the axial forces provided by the spring, Pw is the self-weight of the pier and concrete

cap and Q is the lateral load carrying capacity of the section.

The flexural capacity of both sections can be determined using basic mechanics

principles and the traditional rectangular stress block assumption. Assuming an under

reinforced condition, the depth of the stress block, a, is:

AsA,f+P, +P
a = (6)
0.85f',mbe

where be is the effective thickness of the masonry, A, is the cross sectional area of

the steel, fm is the compressive strength of the masonry andfy is the yield strength of the

steel reinforcement. Applying equilibrium to the section gives the moment capacity of the

section where the flexural strength is provided by the steel reinforcement:

a l a
Jnbar = A /y (d -) + (P P -) ( ) (7)
2 22

The lateral force, Q, required to yield the bar is:

AMnbar
Q=- (8)
heff











For the critical section reinforced by composite only the depth of the stress block,

a, is:

Tw +P, +Pw
a = (9)
0.85 f'mbe

where Tis the tensile strength of the FRP composite and w is the width of the strip used.

Applying equilibrium as before, the moment capacity of the section where the composite

is providing the flexural strength is:

a I a
nbar = Tw (d )+ (P P)(- ) (10)
2 22

The lateral force, Q, required to rupture the FRP composite at this critical section is:

AMnbar
Q-= (11)
heff l

where ls is the length of the lap splice between the composite strip and the steel

reinforcement.

P



Critical Section
Mnf,,
Critical Section I

.............. ., i ..... ... ... U ...................... .. .

,................ ..i



Figure 29: Schematic of the pier connected to the base with dowels.










The measured lateral force at first yield, Qm, is compared to the calculated capacity,

Qnbar, in Table 7. Calculated values for both the steel reinforcement and FRP composite

are shown in Table 7.

Table 7: Measured and Calculated Lateral Capacities
Measured Capacity Qnbar Qnfrp
Specimen Q, (kips) (kips) Ratio (Q,/Qnbar)
(kips) [kN] [kN] [kN]
17.7 17.7
1.0"
CMU 1 [79] n/a [79]
-14.9 -17.7
0.84*
[-66] [-79]
14.7 13.6 23.1
1.08
CMU 2 [65] [61] [103]
-16.6 -13.6 -23.1
1.22
[-74] [-61] [-103]
24.9 23.4 31.9
1.03
CM 3 [111] [104] [142]
-27.4 -23.4 -31.9
1.17
[-122] [-104] [-142]
22.1 36.5
n/a n/a
CMU 4 [98] [162]
-21.7 -22.1 -36.5
0.98
[-97] [-98] [-162]
9.7 12 23.2
0.81
CM 5 [43] [53] [103]
-11.2 -12 -23.2
0.93
[-50] [-53] [-103]
11.3 10.6
1.1
CMU 6 [50] n/a [47]
-10.3 -10.6
0.97
[-46] [-47]
9.9 13.6 66.7
0.73
CMU 7 [44] [61] [297]
-8.9 -13.6 -66.7
0.65
[-40] [-61] [297]
na 12.2 23.2
n/a n/a
CU8 [54] [103]
-14.1 -12.2 -23.2
1.16
[63] [-54] [-103]
*Note: CMU 1 and CMU 6 did not contain steel. The ratio is calculated as Qm/Qnfp.

The ratio of measured capacity at first yield to calculated capacity, Qnbar, is also

given. CMU 1 and CMU 6 did not contain any steel reinforcement, so there is no

calculated value for Qnbar. For CMU 1 and CMU 6, the ratio of the maximum measured









capacity to calculated capacity of the FRP composite, Qfrp, is shown. CMU 4 and CMU

8 did not have strain gauges on the west dowels, so it is not known when first yield

occurred for the positive loading direction.

For most cases, agreement between the measured and calculated capacities falls

within 10%. In the case of CMU 5 and CMU 7, there is a more than 20% overestimation

by the analytical model. CMU 5 failed prematurely because of the splice failure, which

could account for the difference. CMU 7 was limited by rocking about the base of the

specimen, below the reinforcing steel bars. Proper confinement and placement of the FRP

composite would help in increasing the lateral load capacities of these two specimens.















CONCLUSIONS

Eight concrete masonry piers were strengthened with a combination of FRP

composite strips and reinforcing steel. The steel reinforcement was designed to yield

prior to rupturing the FRP composite strips. Confinement using FRP composite sheets

was provided in the base of the pier to strengthen the masonry against dowel pull out. For

most cases, the confinement provided the masonry with the additional strength needed to

allow yielding of the steel dowels. In the two cases where the steel reinforcement did not

yield, insufficient development length (CMU 2) and a lack of confinement in the base

(CMU 7) were to blame.

The use of FRP composites in conjunction with steel reinforcement shows potential

for improving the behavior of URM structures during seismic events. More research

should be conducted on similar configurations in order to quantify the improvement in

ductility and appropriate m-factors for reducing seismic demand.

Key findings and conclusions are as follows:

1. Improvement in the ductility, lateral capacity and energy dissipation were achieved
by adding a FRP/steel strengthening system to the specimens. A drift ratio of 1.8%
and a lateral load capacity of 28 kips (125 kN) was achieved. Open loops in load-
displacement plots indicate energy dissipation as reinforcing steel yielded.

2. Changes in specimen stiffness (from 10% to 20%) were observed at a drift ratio of
0.1% during testing as the specimens sustained damage through cracking, yielding
and debonding of the FRP composite.

3. Yielding was achieved for the specimens that had FRP composite sheets confining
the steel reinforcement in the masonry base against bar pull out.






40


4. Ductility factors (m-factors) were developed for the specimens containing steel
reinforcement and compared to acceptable values found in FEMA 273.

5. Measured load capacities of the specimens were compared to values calculated
using the analytical model presented by Holberg and Hamilton (2002). Agreement
between measured and calculated capacities fell within 10% for the majority of the
specimens tested.














APPENDIX A
EXTENDED LITERATURE REVIEW

A. E. Schultz and R. S. Hutchinson (2001)

Schultz and Hutchinson present a completed project on partially-grouted masonry

shear walls. The project program consisted of simulated seismic load experiments of

partially-grouted masonry walls and both empirical and finite element modeling of shear

wall behavior.

In this project, all horizontal reinforcement is provided by welded wire grids that

are placed in the bed joints of the masonry. The principal variables were height-to-length

aspect ratio and horizontal reinforcement ratio. Results presented include modes of

response, global force-displacement characteristics and level deformation response.

Throughout testing, all the walls showed elastic force-displacement behavior for

forces up to at least half of the yield strength of specimen. Inelastic behavior began with

vertical cracking of the top course of masonry near the section where ungrouted meets

grouted masonry.

Increasing the horizontal reinforcement ratio had a small positive influence of

deformation capacity and shear strength. Horizontal reinforcement ratio was found to

have a modest effect on ultimate shear stress and deformation capacity, but it did not help

effect energy dissipation.

O. S. Marshall and S. C. Sweeney (2002)

In-plane shear tests were conducted on 4 ft. by 4 ft. unreinforced double-wythe

brick wall specimens and lightly reinforced single-wythe CMU wall specimens. Tests









were conducted at the U.S. Army Engineer Research and Development Center

Construction Engineering Research Laboratory (ERDC/CERL). Specimens were

constructed to simulate the "I" shaped piers. Top and bottom of "I" was wrapped in E-

glass or carbon fabric composites to provide confinement, to provide anchorage for the

FRP covering the central portion of the specimen and to force failures into the pier

section of the specimen.

Two gravity load conditions were tested: a low gravity load of 75 psi and a higher

gravity load of 150 psi. The researchers expected that the failure modes for the low axial

loaded specimens would be bed joint sliding. For low gravity loads, it was discovered

that the strength of the specimen was increased in relation to the amount of material that

crossed the failure plane (the horizontal bed joint). For the higher gravity loads, the

authors claim that the expected failure mode was rocking or X-cracking. In the research,

it was discovered that the full-coverage and X-pattern FRP configurations worked best

for the higher axial loads.

Also, it was found that multiple plies of FRP do not always increase the in-plane

capacity. The authors conclude that the multiple plies become too stiff, due to their

thickness, and fail due to delamination of the material from the wall. A similar behavior

for carbon composites was observed by the researchers.

Authors recommend that testing be conducted to develop configurations of FRP

that prevent X-cracking while transferring the failure to a more ductile mode, like bed

joint sliding or rocking prior to toe crushing.









F.L. Moon, T. Yi, R.T. Leon and L.F. Kahn (2002)

This paper is the analysis part of an ongoing project involving a full-scale two-story

URM test structure. Both force-controlled and displacement-controlled analyses were

conducted for an unreinforced state, retrofit with FRP composites and retrofit through

post tensioning of the test structure. Results from the analyses will be used in testing a

full-scale URM building at Georgia Tech.

The authors believe that the use of FRP composite strips versus full coverage sheets

is superior from both an economic and behavioral standpoint. They also state that strips

should be placed according to the type of URM failure mode expected for that particular

pier and to minimize tensile rupture of FRP composite, compressive failure of masonry

(toe crushing) and shear failure of bed joint FRP.

For the force-controlled pushover analysis, a lateral force was applied in a triangle

distribution to the model test structure. The force was distributed to the individual piers

according to their elastic stiffness until yielding of the pier occurred. Yielding of the pier

was defined as the point at which the load exceeded the pier capacity (considering all the

possible URM failure modes).

The displacement-controlled pushover analysis involved modeling each individual

pier as a series of force-displacement curves for each of the considered failure modes.

This resulted in the analysis considering the difference between brittle and ductile failure

modes and providing a displacement capacity for the entire structure. The displacement

capacity is the ultimate displacement of a pier before collapse.

Results from the analysis showed that the use of FRP strips is effective in

increasing the strength while not decreasing the displacement capacity of the structure.









J. G. Tumialan, A. San Bartolome and A. Nanni (2003)

The authors present preliminary results of research on the in-plane behavior of

unreinforced infill concrete masonry walls strengthened with FRP structural repointing.

FRP structural repointing consists of placing FRP bars in the bed joints of the masonry

walls in order to improve shear capacity. The authors indicate that the advantages of

structural repointing include simplicity, less surface preparation (sandblasting, putting,

etc.) and the masonry structure aesthetics are partially conserved. In the case where infill

walls are in contact with a concrete frame, the interaction between the wall and the frame

should be considered. Ignoring this will lead to an unconservative design because the

infill walls can stiffen the frames and cause redistribution of lateral loads to potentially

weaker elements.

Walls were tested in both in and out-of-plane loading. The experimental program

consisted of three types of test performed on the same walls: Part 1 was testing of

specimens subjected to in-plane loading. Part 2 was placing specimens on a shake table

where they were subjected to out-of-plane accelerations. Part 3 consisted of retesting the

specimens under in-plane loading by loading it until it reached its maximum capacity.

Only Part 1 of this experimental program is discussed in this report. Specimens were

tested under in-plane cyclic lateral load by following a displacement-controlled method.

It was concluded that specimens strengthened with FRP can reach a lateral drift of

up to 0.7% without losing lateral load carrying capacity. There were also more, but finer,

cracks in the FRP reinforced specimens than in the unreinforced ones.

The researchers calculated the absorbed energy as the area under the loading

portion of the load vs. displacement curve for each phase. They found that the absorbed









energy increased for the strengthened specimens for a drift greater than 0.5%. For drift

greater than 0.7%, the reinforced specimens had 40% more energy absorption than did

the unreinforced specimens.

M. J. Chajes, W. W. Finch, Jr., T. F. Januszka and T. A. Thompson, Jr. (1996)

This paper presents the results of direct bond tests performed on joints consisting of

composite material plates bonded to concrete. The focus of the tests was on bond strength

and force transfer. To evaluate the effects of surface preparation, type of adhesive and

concrete strength on average bond strength, tests were performed using the single-tap

shear test specimen and constant bond length. Other tests were also performed to study

the force transfer from the composite material plates into the concrete.

Researchers concluded that surface preparation of the concrete can influence

ultimate bond strength. They also found that the use of ductile adhesives (i.e., those

having a low stiffness and large strain to failure) leads to a less effective bond that fails

before there is a shear failure of the concrete. If the failure mode of the joint is governed

by shearing of the concrete directly beneath the bond, the values of the ultimate bond

strength will be proportional to /f'c. They found that the force transfer is largely uniform

along the bonded length.

Y. C. Kurama (2002)

Kurama investigated a precast "hybrid" wall system that uses mild steel

reinforcement as well as post-tensioning steel for flexural strength and inelastic energy

dissipation. First, an analytical parametric investigation was performed on the wall

specimens. The effect of varying the relative quantities of mild steel and post-tensioning

steel under combined gravity and lateral loads were examined. In addition, a series of









nonlinear dynamic time history analyses were conducted to investigate the effect of the

mild steel on the behavior of the walls under seismic loading. The paper goes into detail

on how to evaluate the amount of energy dissipation and absorption using drift vs. base

shear graphs.

The investigation concluded that the inelastic-energy dissipations of unbonded

post-tensioned precast walls can be increased by using bonded mild steel reinforcement

crossing the horizontal joints. It was also found that under high seismic loading, the use

of mild steel decreases the maximum lateral displacement of the wall and decreases the

number of large displacement peaks because the response of the wall decays faster. For

regions of moderate seismisity, it was found in this study that there was little difference

in the dynamic response of the unbonded post-tensioned, hybrid and mild steel only

precast walls. Also, the use of mild steel reinforcement does not have a significant effect

on the self-centering capability of the walls.

K. Bajpai and D. Duthinh (2003)

Concrete masonry beams were reinforced with surface mounted FRP rebar and

tested in four-point bending. Close to full strength development of 14 inch glass FRP bars

was achieved in 7.3 inches of concrete masonry (with is less than one half a concrete

masonry unit). The bars were embedded in epoxy in grooves cut into the face of the

masonry and mortar joints.

Bond tests were also performed on concrete masonry prisms in order to investigate

the development length of FRP bars. Results showed that smaller bars achieved higher

bond strength relative to their tensile strength. It was stated that this was expected

because smaller bars have a higher perimeter to cross sectional area ratio than do the









larger bars. Bond tests also concluded that sand-coated bars with a helical fiber achieved

higher bond strength than those with circular ribs on a smooth surface finish. It was also

concluded that half a masonry unit length is sufficient to develop a /4 inch bar close to its

full strength.

Two series of bending tests were conducted. Series 1 had FRP bars running parallel

to the mortar joints and Series 2 had the bars running perpendicular. Each series consisted

of two narrow and two wide beams. The bars were installed in /2 inch square grooves

along the entire length of the beam. Beams were fully grouted to ensure shear resistance.

The beams were then tested under four-point bending. They all failed in flexure by

rupture of the FRP bars. It was concluded that the ACI 530-02 equations for the flexural

strength of masonry bars under-reinforced with steel provide a conservative estimate of

the flexural strength of concrete masonry beams and walls reinforced with near-surface

mounted FRP bars.

C. Sittipunt, S.L. Wood, P. Lukkunaprasit and P. Pattarattankul (2001)

Four reinforced concrete wall specimens were tested to investigate the influence of

diagonal web steel reinforcement on the hysteretic response of structural walls. The walls

all had barbell-shaped cross sections. The longitudinal and transverse reinforcement in

the boundary elements were the same in all four walls. The primary variables were the

amount and orientation of the web reinforcement.

The specimens were loaded laterally through the top beam. No axial load was

applied to any of the specimens, although they were braced for out-of-plane

displacements. The loading was applied in two stages. The first set of cycles was force-

controlled, where the walls were pushed until cracking. The second stage of testing









consisted of displacement-controlled cycles. The top of the specimen was pushed so that

the deflection at the top was a multiple of the yield displacement in both directions. The

yield displacement was determined by monitoring the strains in the longitudinal steel in

the boundary elements for each specimen.

It was found that all four specimens reached maximum loads that exceeded the

calculated nominal capacities. The walls with the diagonal web reinforcement resisted

higher loads than those with the conventional vertical and horizontal web reinforcement.

There was no significant increase in strength for those walls with higher web

reinforcement ratios.

The walls with diagonal reinforcement dissipated more energy than those with

conventional reinforcement. When loaded laterally, walls with conventional

reinforcement transferred shear through compression struts and aggregate interlock

within the concrete and dowel action of the web reinforcement. When subjected to cyclic

deformations, these mechanisms degraded. Walls with diagonal web reinforcement

transferred shear through tensile forces in the reinforcement. This form of energy

dissipation is stable and did not degrade during testing.

T.C. Triantafillou (1998)

Triantafillou proposed an analytical model for the short-term strength of masonry

walls reinforced with externally bonded FRP laminates under monotonic out-of-plane

bending, in-plane bending and in-plane shear in combination with axial load. The

analysis consists of both equations and normalized interaction diagrams. Testing was

conducted to verify this model.









For the analysis, the author assumed that the laminates act as truss like elements.

The author states that this is a fair assumption because FRP is usually applied as narrow

straps and work in tension, much like truss elements. In the case of full coverage FRP, he

would have had to consider a more complicated laminate theory.

The author states that the typical failure for out-of-plane bending with axial force is

compressive crushing and that the typical failure for in-plane bending can be compressive

crushing or shearing of FPR in the tension zone directly beneath the bond (in the cases of

short development length).

Triantafillou proposed that the analysis and design of reinforced masonry is based

on the assumption that the total shear capacity is the sum of two terms, VRDI and V2,.

VRI is the contribution from the uncracked masonry and VR2 is from the effect of shear

reinforcement (modeled from truss analogy). The shear capacity of masonry, VR, is

given from Equation A-i (in SI units) adopted from Eurocode 6, wherefvk is the

characteristic shear strength of the masonry.

fvk-t-d .3-fk-t.d
VRD= + VRD2 <
Y m Ym (A-1)

The author assumes that the contribution of vertical FRP reinforcement, which

provides mainly a dowel effect, is negligible. Therefore, the only shear resistance

mechanism left is the action of the horizontal laminates. This can be modeled similar to

stirrups in reinforced concrete beams. Triantafillou reported Equation A-2, where Cfrpe is

the effective FRP strain from Equation A-3. Both Equation A-2 and A-3 are in SI units.

.7
VRD2= P h-Efrps frpe--t
Y frp (A-2)










Sfe = .0119- .0205(p hEfrp) + .0104(p hEfrp) (A-3)

To test his analytical model, twelve identical small wall specimens were

constructed of perforated clay units and statically tested in four-point bending to study

the effect of FRP reinforcement on the failure mechanism and load capacity of the wall.

Six were tested out-of-plane and six in-plane. Walls tested in out-of-plane bending had

FRP laminates bonded to the tension face. Those walls tested in in-plane bending were

reinforced symmetrically on both faces.

The FRP covered walls tested in out-of-plane bending failed by crushing of the

masonry in the compression zone, indicating flexural failure. There was good agreement

between analysis and experiments, with a maximum error equal to approximately 15%.

Those walls tested in in-plane bending failed prematurely due to peeling off of the FRP

laminates in the tension zone directly beneath the bond. It was found that this action was

due to the short bond development length because of the small size of the specimens.

Consequently, it was concluded that the achievement of full in-plane flexural strength

depends on proper anchorage.

M.J.N. Priestley and F. Seible (1995)

This experimental program focuses on implementing techniques aimed at

increasing flexural ductility by limiting shear deformations and stabilizing critical

compression toe regions on masonry walls. The program consisted of testing a full-scale

five-story reinforced hollow concrete masonry building. The test building was tested

under simulated seismic actions to failure. The first series of tests were run on the

building with no FRP retrofits. Afterwards, the test structure was repaired by means of









structural carbon overlays on the first two stories of the structural walls. A single layer

was applied on each side of the walls with two layers in the toe regions.

Results showed that repairing with composites doubled the displacement capacity

of the top story. Measured shear displacements in the repaired walls were reduced to half

the measured shear displacements of the original five-story building.

It was stated that the forces that are transferred in the composite overlays are

limited by the laminar shear or principal tensile strength of the original structural wall

material. This is because the polymer resin usually has a higher tensile capacity than the

masonry it lays on.

Equation A-4 for shear capacity of the composite overlay was developed using a

conservative diagonal tension crack angle assumption of 450 wherefo is the allowable

overlay stress level based on a maximum allowable strain of .004, dis the effective

structural wall length and t is the thickness of the composite overlay.

Vo = fo-tid (A-4)

Stiffness criteria, rather than strength criteria, was employed in the wall overlay

design. Shear deformations were limited to deformation levels which can be expected in

concrete walls with conventional horizontal reinforcement, Ashreq, by scaling the amount

of horizontal overlay fabric, Aoh, from the required horizontal reinforcement in Equation

A-5. Eo is the modulus of elasticity of the overlay material and Es is the modulus of

elasticity of the substrate (in this case it is masonry).

Es
A oh = A shreq
0 (A-5)









The researchers compared the bond between the FRP composite and masonry to

that of traditional reinforcement and unconfined concrete. They borrowed the upper

limits to the total gain on shear capacity from conventional concrete design (ACI-318) in

Equation A-6, where fc is the nominal concrete strength in MPa, bw is the wall width and

d is the effective wall length.

max Vo= max Vs Vs = .66.-fc-.b d (A-6)

A.M. Holberg and H.R. Hamilton III (2001)

URM walls strengthened with a hybrid system containing both FRP composites and

steel was investigated. The ductile connection was designed to yield before the composite

failed. This system was to increase lateral capacity and ductility to the URM.

Four unreinforced hollow concrete masonry shear wall specimens were

strengthened with the hybrid system and then tested. Vertical strips of the FRP were

placed at each end of the wall to increase the in-plane flexural capacity and diagonal

strips were applied to improve shear capacity. The walls were sized to ensure a rocking

failure mode for an unreinforced specimen.

Two types of steel connections were used: internal and external. The internal

connection was a steel reinforcing bar placed in the outermost cells of the wall and fully

grouted. The external connection was a steel angle-plate assembly attached to the

foundation of the test setup. The GFRP strips were then folded into two layers under the

plate.

The drift capacities of the reinforced specimens reached up to 1.7%. The lateral

capacity of a strengthened specimen was nearly doubled. The external connection did not

prevent sliding, which provides good energy dissipation. For the specimens with an









internal connection, an eccentricity was caused between the reinforcing bar and the

GFRP composite which caused an out-of-plane displacement at the critical section. The

eccentricity was improved for other specimens which caused an improvement in the

displacement capacity of the specimen.

G.M. Calvi, G.R. Kingsley and G. Magenes (1996)

The paper addresses problems associated with the experimental evaluation of the

URM buildings subjected to seismic loading. It focuses on the merits and roles of several

experimental techniques, which include quasi-static, dynamic and pseudo dynamic

loading at full and reduced scale.

The authors indicate that masonry is highly sensitive to loading rates, boundary

conditions and effects of axial loading. They also noted that the final collapse of URM

buildings is often associated with in-plane shear failure of the piers of a critical story

(usually the ground story most axial load).

Paper discusses test procedures of past testing of masonry piers. The large majority

of previous testing focused on in-plane shear. The boundary conditions at the top and

bottom of the wall are either fixed-fixed or fixed-free.

The authors suggest that a real seismic excitation is better simulated in dynamic

tests, but acknowledge that quasi-static tests have several advantages over dynamic

shaking table tests. Quasi-static tests allow for an easier application of large forces, for

easier observance of cracking and for the more accurate measurement of forces and

displacements. Quasi-static tests tend to show more extensive damage and lower strength

than dynamic tests because masonry exhibits rate-dependant behavior, which allows









propagation of cracking at constant load or at constant imposed displacement. It was

concluded that, in general, static cyclic testing is a conservative approach.

From past experiments, given material properties, the main parameters determining

the pier failure mechanism are the axial load, the aspect ratio and the boundary conditions

at the ends. Once boundary conditions are determined, increasing aspect ratios tend to

lead to a rocking failure and decreasing aspect ratios lead to shear failures, like diagonal

cracking or shear sliding. Maintaining constant geometry and boundary conditions, an

increase in axial load leads to increased chances for diagonal shear failure and low axial

loads are usually associated with sliding and rocking.

The ability to predict the failure mechanism is important. Both static and dynamic

experiments have shown how rocking and sliding tend to be associated with good seismic

behavior and diagonal cracking to brittle behavior. Rocking provides a limit to the

maximum shear and dissipated energy through impact, keeping the integrity of the wall

except for the localized damage in the covers. Sliding on horizontal bedjoints is another

good dissipative mechanism also maintaining the overall integrity of the wall, although it

can also result in larger displacements. Diagonal shear cracking is associated with brittle

failure under dynamic loading, can be catastrophic.

M.R.Ehsani, H. Saadatmanesh and J.I. Velazques-Dimas (1999)

This paper presents the experimental results of three half-scale URM walls

retrofitted with E-glass FRP strips and tested under cyclic out-of-plane bending. The

unstable and explosive out-of-plane failure endangers the gravity load carrying capacity

of the wall. For this series of tests, no axial load was applied out-of-plane failures are

characterized by cracks along the mortar bedjoints while in-plane failures have a diagonal









crack pattern. The study presented was conducted to demonstrate the feasibility of

retrofitting with GFRP on the tension face for structures subjected to reverse cyclic

loading.

The specimens were intended to represent a typical load-bearing wall in a low-rise

building away from corners. Specimens were retrofitted with vertical strips of FRP

applied using the wet lay-up procedure. Two different glass fabric densities and five

reinforcement ratios were investigated.

It was concluded from the observed behavior of this set of URM masonry walls

retrofitted with FRP that the ultimate flexural strength was significantly increased. The

walls deflected almost 14 times the maximum allowable deflection found in masonry

specifications. These deflections were as much as 2.5% of the wall height.

D.P. Abrams (2001)

Author states that because of rocking and sliding along mortar bed joints, URM

wall and pier elements can be considered to behave as displacement controlled

components with significant capacity to accept large nonlinear deflections.

One common explanation for damage to numerous URM buildings during

earthquakes is that the building has not been engineered to resist lateral seismic forces

either at the time of the original design or at a later time when rehabilitation should have

been considered. The US is researching and developing rehabilitation focused on the

implementation of performance-based guidelines (FEMA 273 and FEMA 356). This will

result in various performance levels for varying degrees of structural intervention. These

new guidelines will not just be based on strengthening but also on enhancing deformation









capacity. This results in different ductilities, or m factors. The ductility factor, m,

expresses the lateral drift at a specific performance state versus yield displacement.

This paper describes a recent testing program at the Mid-America Earthquake

Center which investigated the basic mechanisms for unreinforced brick masonry wall.

The piers used had a constant height-to-length ratio equal to 1.8, so that flexural

mechanisms (rocking and toe compression) could be studied. Piers strengthened by FRP,

reinforced and prestressed cores, reinforced shotcrete overlays and surface coating were

tested.

The pier strengthened with the FRP behaved with a significant increase in lateral

strength (over three times that of the unreinforced pier). As the FRP delaminated from the

surface, the lateral stiffness of the pier gradually reduced. First signs of delamination

were observed at 0.04% drift and continued until the lateral pier strength was completely

lost at 1.9% drift. Diagonal stair-stepped cracks developed at 0.6% drift. This was due to

the larger shear forces brought on by the enhanced flexural strength provided by the FRP.

Additional Literature Review

Masonry and Earthquakes

Recent earthquakes have revealed extensive damage to existing unreinforced

masonry structures. Many older buildings currently in use were designed and constructed

with little or no consideration to earthquake resistance. In addition, changed usage and

more stringent seismic requirements have left many buildings in need of additional

strengthening. In deciding to retrofit a structure, it is important to know the failure mode

of the components in order to restrain that mode or change it to a more favorable failure

mode.









Failure modes for masonry piers depend on the combination of applied loads, pier

geometry and properties of the materials. Typical failure modes for URM are illustrated

in Figure 31. These failure modes are characterized by brittle behavior with rapid

decreases in capacity and very limited deformations after reaching the ultimate load. The

most energy dissipating failure mode is bed-joint sliding. This is a result of a low, vertical

compressive force. Rocking is the second best failure mode in terms of energy

dissipation. It is deformation controlled, which allows large lateral displacements.

Diagonal tension, or X-cracking, failure is characterized by diagonal cracks formed

through the mortar joints or the masonry units themselves. Toe crushing is the least

energy dissipating of the four failure modes. In this failure mode, high gravity loads or

large overturning forces crush the masonry units.


Figure 30: Pier area outlined on a structure.






58


Light Axial Load

1 1





Rocking Toe Crushing
SHeavy Axial Load






Bed-Joint Sliding Diagonal-Tension


Figure 31: Failure modes for unreinforced masonry.

The goal of retrofit design is to change the failure mode of a pier from a brittle

mode, X-cracking or toe crushing, to a more favorable failure mode, rocking or bed joint

sliding. If a URM pier is determined to have a rocking or bed joint sliding failure mode,

not adding any strengthening may be the best option. The URM piers that have been

determined to fail in X-cracking or toe crushing would need to be strengthened to change

the failure mode to bed joint sliding or rocking. The ideal retrofit design would allow the

wall to dissipate energy while not losing its ability to carry axial load, give the wall an in-

plane strength close to or greater than the strength of the wall without any reinforcement

and, during a seismic event, not allow crushed masonry to fall on those in the structure or

those trying to exit.

Conventional Strengthening and Rehabilitation Options

There are several methods available for retrofitting existing structures that increase

the strength and ductility of masonry buildings. Conventional strengthening methods can









be costly, add significant mass to the structure, reduce the amount of useable space in the

structure and negatively affect aesthetics.

One of the most common methods of retrofitting existing masonry structures

involves removing one or more wythes of masonry and filling the void with

pneumatically applied concrete, commonly known as shotcrete (Moon et al. 2002). This

method has showed to be effective in increasing both the strength and ductility of the

URM walls. Shotcrete is costly because of the amount of surface preparation needed and

formwork required. This method also adds considerable weight to the structure, resulting

in larger inertia forces during an earthquake. A second method involves applying a thin

surface coating to one or both sides of the URM wall. Typical coatings include glass-

reinforced cement, ferrocement and a wire mesh reinforced cement.

Other methods of strengthening include adding reinforcing steel to the structure.

Post-tensioning or prestressing has been used to enhance the tensile and flexural capacity

of concrete. After core drilling from the top of masonry walls, steel is post-tensioned to

the foundation. This method is costly, but does not alter the exterior of the structure. It

can also be installed without disturbing the occupants.

FRP Strengthening

Fiber reinforced polymer (FRP) laminates are made of continuous glass, carbon or

Kevlar fibers bonded to the substrate with a resin polymer matrix. FRP composite

systems have been used extensively in recent years to retrofit concrete structures in high

seismic zones (Marshall and Sweeney 2002). The FRP composite system provides

additional reinforcement that enables the structure to have better resistance against

earthquake damage. The composite gives the masonry the added strength it needs to resist

the shear and flexural stresses experienced during an earthquake.









Some of the advantages of FRP composite systems are that it is lightweight, has

high stiffness, excellent fatigue properties and has a high resistance to corrosion. This

technique is typically unobtrusive to the building occupants, requires little surface

preparation and is very economical. The ability of a FRP composite strengthening system

to keep crushed masonry from falling off the structure also makes it a favorable

alternative to traditional methods.

Strengthening masonry walls with FRP composites requires that the composite be

bonded to one or both sides of the wall. Currently, there are two approached for this. The

first approach bonds FRP composite sheets over the entire surface of the pier. The other

uses FRP composite strips bonded to specific locations on the pier. The use of

unidirectional strips of FRP is preferable in terms of economy and behavioral response to

the two-dimensional fabrics that cover the entire surface of the masonry wall

(Triantafillou 1998).

Past research

FRP composites have been tested as a reinforcing overlay for seismic retrofit of

reinforced concrete and masonry. Results of testing done on masonry with FRP

composites have shown that both in-plane and out-of-plane strength were significantly

increased without decreasing the displacement capacity of the structure. Tests have

concluded that the FRP composites can be applied to increase strength and change the

failure modes of masonry walls (Marshall and Sweeney 2002, Abrams 2001).

FRP linked with a ductile connection

Although FRP composites increases lateral load capacity, it does not improve

ductility. This is because of the brittle nature of the composite material. This has lead to

the idea of a hybrid system, consisting of FRP composites in conjunction with steel. The










FRP composite adds sufficient strength to the masonry allowing the steel to reach yield,

incorporating ductility into the system.

Figure 32 illustrates a masonry pier strengthened against flexure and shear using

the hybrid system. The added FRP composite improves the flexural behavior.

Terminating the FRP composite at the pier allows for sliding at the top and bottom of the

pier. The ductile connection provides restraint against rocking.






Sliding at top
allowed -
allod Shear and flexural
strength improved





Sliding and rocking restrained
with ductile connection


Figure 32: Pier strengthened for shear and flexure with FRP composites. Rocking
restrained with the ductile connection at the base of the pier.

The idealized load-displacement curve for statically loaded URM cantilever pier is

shown in Figure 33 (Holberg and Hamilton 2002). The lateral load capacity for the

rocking mode, POT, is a function of pier geometry and axial loading (Figure 34). The

slope to the rocking load is the stiffness of the URM pier. After reaching this point, the

pier will continue displace without gaining additional capacity until it becomes unstable.









Load


Figure 33: Idealized load-displacement curve for a URM pier (Holberg and Hamilton
2002).


Figure 34: URM pier with rocking load of Por.

Adding a ductile connection to the base will increase the lateral capacity to P = Por

+ Pc (Figure 35). The additional capacity, Pc is a function of the tensile force Tin the

ductile connection and pier geometry (Figure 36). Energy dissipation will also increase,

as shown by the shaded area in Figure 35. The shaded area is a result of the energy

dissipation from the yielding ductile reinforcement.









Load




|P rDnfl









Figure 35: Idealized load-displacement curve for a pier strengthened with a ductile
connection to the base (Holberg and Hamilton 2002).

P















T

Figure 36: Pier strengthened with a ductile connection with an overturning load of P.

Research was conducted by Holberg and Hamilton (2001) on URM walls

retrofitted with a hybrid strengthening system consisting of FRP composites and steel.

Two different types of steel connections were tested, an internal and an external. The

internal connection was a steel reinforcing bar placed in the outermost cells of the wall

and fully grouted. The external connection was a steel angle-plate assembly attached to

the foundation. The drift capacities of the reinforced specimens reached up to 1.7%.






64


There was an almost doubling of the reinforced specimens' lateral capacity when

compared to an unreinforced specimen's capacity.

















EXPERIMENTAL PROGRAM

Test Setup

Chapter B The test set up is shown in Figure 37. A 55 kip (240 kN) hydraulic actuator is

located on the reaction frame and displaces the concrete cap on the specimen. Positive

load and displacement correspond to actuator tension (Figure 38). Negative direction is

actuator compression. The reaction frame that supports the actuator is constructed of steel

and is prevented from overturning by its connection to the laboratory strong floor. A

concrete cap and base were constructed for the test fixture and designed to handle loads

of 50 kips (220 kN). The concrete base was tied to the laboratory strong floor. The

concrete cap was braced to the laboratory wall by steel angles to prevent out-of-plane

twisting (Figure 39).


Spring Syslem
55 kip Aclualor



Concrete Cap Reaclion Frame




DAQ Sys[em

Concrete Base


Figure 37: Test set up with specimen ready for testing (looking at North face of
specimen).












tNorth


Displacement Sign Convention
-- +


Concrete Cap


Figure 38: Plan view of test set up. The North direction in the laboratory is defined.


Figure 39: Angles used to prevent out-of-plane movement.

The specimens were constructed separately and placed into the test fixture at the

time of testing (Figure 40). The specimens were placed on the concrete base on a bed of

hydrostone to allow for an even bearing surface. A layer of fresh mortar was spread on

the top lintel of the specimen and the concrete cap was then immediately placed onto the

specimen. The actuator was then connected to the concrete cap by four threaded rods

running through the concrete cap.

A steel bearing plate was placed on plywood over a vertical head joint in the base

of the specimen where it would be tied to the concrete base. Steel channels were placed

on the bearing plate and along the side of the base. The threaded rod connecting the sill to

the base was post-tensioned using a hydraulic jack at a force of 42 kips (187 kN). The use










of bearing plates allowed for the force of the post-tensioned threaded rod to be directed

over any vertical head joint desired.

The concrete cap and the top lintel were connected using threaded rod and steel

channels. The nuts on these threaded rods were hand tightened. Figure 40 shows the

confinement details for the specimen.


8 inches
(200 mm)


--- MTS Load





Downward Force = 42 kips
| (187 kN)
y


o Specimen


SConcrete Base

Sb///////oratory Floor


Figure 40: Schematic of specimen in the test set up.

A dial gauge was placed between the cap and the specimen's top lintel to monitor

slip between the lintel and the concrete cap (Figure 41). Throughout testing, no slipping

occurred between the concrete cap and the top lintel of any of the specimens.


Oc
IC]L
CC
'0-


I
























Figure 41: Dial gauge to monitor slipping between the concrete cap and the specimen's
top lintel.

Lifting Frame

A lifting frame was constructed to move the specimens (Figure 42). It consisted of

a channel placed on the top of the specimen and a series of threaded rod that were tied to

angles. Prior to specimen construction, 5/8 inch holes were drilled two inches from the

top of the lintels. Steel pipes were then inserted in the holes and cast into place when the

concrete was poured. The specimen was constructed on what was the bottom of the lintel.

Threaded rod was run through the pipes and attached to steel angles. Hooks at the top of

the lifting frame were attached to the laboratory crane.








































Figure 42: Lifting Frame.

Axial Loading

A gravity load of 9 kips (40 kN) was applied to the specimen. This translated to an

axial stress of 75 psi (0.5 MPa) at the base of the pier. This gravity load included load

from two rail car springs compressed to develop 6 kips (27 kN), the weight of the

concrete cap (2 kips, 9 kN) and the self-weight of the pier and lintel (1 kip, 4 kN). The

two springs were compressed by a hydraulic jack pushing upward on the top steel plate

and held in compression by tightening the nuts on the top of two threaded rods tied

between the steel angle system on the concrete cap and the steel connection post









tensioned to the concrete base (Figure 43). The hydraulic jack was pressurized to 500 psi.

With a 12 inch2 cross-sectional area, 500 psi in the jack translated to 6 kips. This

procedure was repeated for each specimen. The compressed springs allowed the

specimen to move laterally without large changes in the axial load. If the specimen lost

the ability to carry axial load, the threaded rod would lose tension as the spring loses

compression. No instrument was placed to monitor the amount of axial load sustained

during testing. The only way to tell if any axial load remained is by checking the tension

in the rod and monitoring the amount of pressure in the hydraulic jack needed to release

the spring.

The springs were purchased from Barber Spring, Co. in Pittsburgh, PA. The

specified spring stiffness was 4,288 lbs per inch of deflection (751 kN/m). A force of 6

kips was reached at a 1.4 inches (35 mm) deflection. The spring compression was

monitored visually with measured tick marks that were attached to the spring system.

Also, wooden blocks cut to allow only a 1.4 inch deflection were attached to the middle

steel plate (flush with the top of the springs) to maintain a consistent spring compression

among all the specimens.


























Specimen


A B

Figure 43: Axial load spring system. A) Schematic. B) Photo.

Lateral Loading

A 55 kip (245 kN) MTS hydraulic actuator was used for testing (Figure 44). The

hydraulic actuator was part of a closed-loop hydraulic loading system. The system was

controlled by an MTS 407 controller. The controller used a sinusoidal voltage output

from the data acquisition system to impose the displacements. The actuator also

contained a load cell that was used to monitor the lateral load capacity of the specimen

during testing.
























Figure 44: MTS actuator in place.

The ICBO Acceptance Criteriafor Concrete and Reinforced and Unreinforced

Masonry Sti eigithenlig Using Fiber-Reinforced Composite Systems (AC125) (ICBO

1997) was followed for cycling in this testing program (

Figure 45). The displacement at which the reinforcement was expected to yield is

marked as the first yield point, t = 1. For specimens that did not contain any steel (CMU

1 and CMU 6), the yield point was taken as the displacement that was expected to caused

failure of an unreinforced specimen. The specimens were loaded in displacement control

with three complete cycles for each displacement level, [t. The displacements were [t = /4,

12, 34, 1, 2, 3, 4, 6, 8 and 10. The displacements were increased until the specimen lost its

lateral load carrying capacity. The loading rate was approximately 20 seconds per cycle.

12-
i-10


4-s


4


Figure 45: ICBO test sequence of imposed displacement.









Data Acquisition

A personal computer, National Instruments LabVIEW software and a 16-bit data

acquisition card were used for data acquisition. Linear potentiometers and string pots

were used for measuring displacements. Strain gauges were placed on the steel rebar and

FRP strips.

Linear Displacement Measurement

Instrumentation was applied to the specimens and the actuator to monitor linear

displacements and applied loads (Figure 46). Instruments labeled Carol, Diane, Flo and

Gina were string potentiometers (SP) attached between a solid steel frame and hooks

attached to the specimen (Figure 47). The SP's were BEI Duncan miniature spring return

linear motion sensors, models 9610 and 9615. The string was tied to a wire and pulled out

to the halfway point before being attached to the specimen, allowing positive and

negative measurements. The calibration factor used for the SP was taken from the value

imprinted on each instrument. The data acquisition program measured the SP supply

voltage as well as the output of the SP so that an accurate linear measurement could be

computed.



























Figure 46: Locations of instruments for linear displacement measurement.


B








A C

Figure 47: String pots between steel frame and specimen. A) Overall photo of two of the
string pots. B) Closer view of a string pot that measure horizontally. C) Closer
view of a string pot that measures vertically.

The rest of the displacement instruments were linear potentiometers. The linear

potentiometers were calibrated using a calibration tool and a LabVIEW program

developed by the researchers. The tool held the instrument in place while the stroke









length was measured and recorded by the LabVIEW program. Values recorded were used

to calculate the calibration factor.

Each instrument was glued to a small metal box. This metal box protected the

instrument during application and removal. The metal box was adhered to the specimen

using hot glue and measured off the steel frames located on either side of the pier (Figure

48). These steel frames were constructed by the researchers and stood solidly on the

laboratory floor. The stroke was compressed to the halfway point when it was placed

onto the specimen. This enabled the instrument to read in the positive and negative

directions, depending on the movement of the wall.
















Figure 48: Linear potentiometers on the specimen

Strain Gauges

The strain gauges used on the steel were purchased from Texas Measurements, Inc.

and were type FLA-5-11-1L. The dowel was ground, polished flat and cleaned where the

strain gauge was to be placed. The strain gauge was glued in place using ethyl

cyanoacrylate (as recommended by the strain gauge manufacturer). After the glue dried, a

thin layer of silicone sealant was placed over the strain gauge to protect it from the

masonry grout. Electrical tape was bound two-inches above and below the strain gauge.









The tape was then covered with a shrink rubber tube to provide a four inch (101.6 mm)

unbonded length of bar, which was centered on the pier/base interface during installation

(Figure 49).

Having a four inch gauge length allows for the strain at that location to be

distributed over the unbonded length. The advantage of this is that the elongation of the

bar occurs over the four inch length and not over a smaller length.
















Figure 49: Rubber tube installation on dowel over the strain gauge.

Strain gauges were placed into the wet resin during FRP composite application

(Figure 50). The strain gauges used in the FRP composite were also purchased from

Texas Measurements, Inc. and were type PFL-30-11. See Figure 51 for strain gauge

locations. All strain gauges were placed in the center of the FRP strap that they were

measuring. One of the strain gauges on the FRP (Linda) was initially placed immediately

above the pier/base intersection and then moved to the FRP/steel transition point (Linda

(2)) after it was noticed that there was very little strain read above the pier/base

intersection point. On those specimens that had the vertical FRP strip running

continuously across the pier/base interface, a strain gauge (Marie) was placed

immediately below the interface. On those that had vertical FRP strips terminating at the









bottom of the pier no strain gauge was placed below the pier/base intersection. For all

specimens, a strain gauge (Nancy) was placed on the horizontal strip that ran along the

top of the base at the corer where the pier met the base.


Figure 50: FRP Strain Gauge.










Strain Measurement
Direction


Figure 51: FRP strain gauge locations.

LabVIEW Programs

A LabVIEW program was developed to obtain the measurement readings. On the

main screen (Figure 52), the user could specify file names for recording the data, identify

cycle number, monitor the strain gauges and view a load vs. displacement graph for the

test. The load and displacement shown on the screen are from the actuator load cell and

LVDT. At the beginning of every test, a Zero Scan was run before any specimen

displacements occurred. The Zero Scan measured the initial positions of each of the

instruments. A Regular Scan was run while the specimen was displacing. The Regular

Scan recorded the value of the current position minus the value of the zero scan. The

value from the Regular Scan was also shown on the screen.








79









A Lab VIEW program was also written to control the MTS actuator (Figure 53).
------ -.-- 1, -
-, [










































buter size il i'n'' ii| .,l i '1 .1 ... 11W)
01000-






S-11 4 O1111"






















Iu I3 i
Figure 52: LabVIEW program main screen.

A LabVIEW program was also written to control the MTS actuator (Figure 53).


The user inputs the displacement value, frequency and the number of cycles for each


displacement value. The program sent a voltage output corresponding to the particular


displacement to the MTS controller. The MTS controller controlled the amount of


hydraulic fluid entering the actuator. A calibration factor of 0.5 inches per volt output


remained constant throughout the entire testing program.







Figuel5i MTS Signal Generationp Ir screen.
vd -h'H-1
--_ Il l l . I I

I I II- IIT






H II







Figure 53: MTS Signal Generation program screen.














APPENDIX C
SPECIMEN CONSTRUCTION

Eight concrete masonry specimens were constructed in running bond of medium

weight 8-in. (200-mm.) concrete masonry units by Painter Masonry, Inc. in Gainesville,

FL. Type N mortar was used in faceshell bedding for each of the specimens. The ASTM

C90 units were purchased locally from Florida Rock Industries, Inc. in Gainesville, FL.

All test specimens were single wythe and consisted of a pier 48-in. (1200-mm.) tall by

48-in. (1200-mm.) wide and a base 9-ft. 4-in. (2800-mm.) long and 16-in. (400-mm.) tall

(Figure 54). Since the test specimen consisted of full scale CMU, no scaling effects were

required. The masons installed the joint reinforcement every other course in the

specimen. This was done in accordance to common practice in the field. The location of

the joint reinforcement is located in Figure 54.

The specimens were constructed to simulate typical piers between windows of a

building. Tests at the Georgia Institute of Technology (Moon et al. 2002) and at the U.S.

Army Engineer Research and Development Center Construction Engineering Research

Laboratory (Marshall and Sweeney 2002) used a similar pier configuration.

The specimen was constructed on a precast concrete lintel 9-ft. 4-in. (2800-mm.)

long. A second precast concrete 5-ft. 4-in. (1600-mm) lintel was placed on the top course

of the pier. Both lintels were placed with the same mortar used for constructing the

specimen. Lintels were filled with concrete and the bottom lintel also contained two #6

reinforcing bars embedded into the concrete along its length. The precast lintels

contained reinforcement in the bottom. To take advantage of this reinforcement for lifting










the specimen, the lintels were inverted before specimen construction began. The bottom

lintel facilitated moving the specimen and the top lintel was used to attach the specimen

to the concrete cap for loading. The outer cells of the base (at the tie down locations)

were filled with grout (approximately one month after construction) to prevent

overstressing the masonry during post-tensioning of the tie down.

5'-4" (1630 mm) Concrete fill placed after
4'-0" (1220 mm) Precast installation.
Concrete Lintel Precast
SConcrete Lintel

#6 Rebar

I Joint /
Reinforcement 7-
ReinforcementGrouted Cells C



Precast
Concrete Lintel


Lifting Rings Precast
SConcrete Lintel
9'-4" (2845 mm)


Figure 54: Specimen dimensions and location of joint reinforcement.

Three specimens were constructed with #4 reinforcing bars in the jambs and sills

(Figure 55). The existing steel was placed and grouted approximately one month after

specimen construction. The grout for the existing steel was deposited with a shovel and

consolidated with a vibrator. Many concrete masonry structures contain this type of

prescriptive reinforcement around openings. These specimens allowed evaluation of the

performance of partially reinforced masonry with composites added.