BEHAVIOR OF STANDARD HOOK ANCHORAGE MADE WITH CORROSION RESISTANT REINFORCEMENT

Transcription

1 BEHAVIOR OF STANDARD HOOK ANCHORAGE MADE WITH CORROSION RESISTANT REINFORCEMENT By GIANNI T. CIANCONE 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 27 1

2 27 Gianni T. Ciancone 2

3 This thesis is dedicated to my loving wife Hilda and my daughter Alessandra for their support and caring throughout my academic endeavors 3

4 ACKNOWLEDGMENTS The author would like to thank my graduate advisor, committee chairman, Dr. H.R. Hamilton III, for his patience, advice, and support throughout this research. Also, I would to acknowledge the rest of the committee, Dr. Ronald A. Cook, and Dr. John M. Lybas. Their extensive knowledge, and experience in the Department of Civil and Coastal Engineering is greatly respected. The author would like to thank Florida Department of Transportation (FDOT) State Materials Office and Structural Lab for their support testing materials, and bending the bars. Special thanks go to the University of Florida-Structural Laboratory personnel, and to all the members of the Dr. Hamilton Group for their support constructing the specimens. The author would also like to thank VALBRUNA stainless steel, MMFX Technologies Corp, FLORIDA ROCK Industries, and BARSPLICE Products Inc. for their contributions to this research. Finally, I would like to thank my wife, daughter and close friends who have supported me during this research. 4

5 TABLE OF CONTENTS ACKNOWLEDGMENTS...4 LIST OF TABLES...7 LIST OF FIGURES...9 ABSTRACT...12 CHAPTER 1 INTRODUCTION LITERATURE REVIEW...15 page Hook Behavior and Geometry...15 Current Hook Design Practice...16 High-Strength Steel Reinforcement...21 Strut and Tie Evaluation of Anchorage EXPERIMENTAL PROGRAM...29 Specimen Design...29 Concrete Mixture Designs...32 Specimen Construction...33 Formwork...33 Casting...34 Test Setup...34 Data Acquisition Setup RESULTS AND DISCUSSION...47 Materials Properties...47 Concrete...47 Steel...47 Grade 6 Steel...48 Stainless Steel...48 MMFX Steel...49 Specimens Test Results...49 Behavior and Failure Modes...49 Mild Steel Specimens...51 Stainless Steel Specimens...55 MMFX Specimens ANALYSIS OF RESULTS

6 Anchorage Capacity...72 Bond Stress...73 Ductility...75 K-Factor CONCLUSIONS...86 APPENDIX A B CONCRETE COMPRESSIVE STRENGTH AND TENSILE RESULTS...88 CRACKS PATTERNS, LOAD-SLIP, AND LOAD-DISPLACEMENT...9 LIST OF REFERENCES BIOGRAPHICAL SKETCH

7 LIST OF TABLES Table page 2-1 Minimum hook dimensions Specimen design details for series Specimen design details for series 2 through Concrete mixture proportions (quantities are per cubic yard) Compressive concrete strengths Tension test results for ASTM A615 reinforcement Tension test result for stainless steel (316LN) Tension test results for MMFX steel Test results for mild steel #5 and #7 specimens Test results for stainless steel 16 mm and 2 mm specimens Test results for MMFX steel #5 and #7 specimens Anchorage capacity ratio for mild steel Anchorage capacity ratio stainless steel Anchorage capacity ratio for MMFX steel Bond stress normalized for mild steel Bond stress normalized for stainless steel Bond stress normalized for MMFX steel Ductility ratio for mild steel Ductility ratio for stainless steel Ductility ratio for MMFX steel K-factor for #5 and #7 mild steel bars K-factor for 16 mm and 2 mm stainless steel bars K-factor for #5 and #7 MMFX bars

8 A-1 Compressive concrete strength results age (days)...88 A-2 Tensile test results

9 LIST OF FIGURES Figure page 2-1 Cantilever beam Normal bar stresses #7 9 deg standard hook Standard hook details Points where slip was measured Recommended ϕ factor Comparison of proposed and ACI hook provisions Typical uses of a standard hook anchorage and F.B.D Extended nodal zone for standard hook anchorage Strut and tie model of specimen used in Marques and Jirsa research Specimen design with idealized boundary conditions Specimen design for series Specimen design for series 2 through Formwork schematics Formwork details Ready-mixed concrete being discharged into the container for transporting Slump of ready-mixed concrete Casting and compaction of the specimen Curing of the specimens Load test setup Coupler system Specimen schematic reactions Slip wire position in hooked bar Bond slip instrumentation

10 3-15 Linear potentiometer placed at the top face of the specimen Data acquisition system Stress-strain curve Stress-strain comparison Cracks Crack pattern for concrete splitting failure Concrete crushed inside of bend radius Load-displacement for mild steel Mild steel results in terms of hook capacity Load-slip for specimens Locations where relative slip was measured Load-slip for specimen Typical load-slip behavior for #5 mild steel specimens with 18-degree hook (6_5_18_35_2 shown) Relative slip at locations D1 and D2 for unconfined specimens with debonded length Typical load-slip behavior for #7 mild steel specimens with 18-degree hook (6_7_18_35_4 shown) Load - displacement for stainless steel Stainless steel results in terms of hook capacity Load-slip for specimens Typical load-slip behavior for 16mm stainless steel specimens with both 9 and 18- degree hooks (SS_16_18_35_4 show) Typical load-slip behavior for 2mm stainless steel specimens with both 9 and 18- degree hooks (SS_2_9_35_2 shown) Load-displacement for MMFX steel MMFX results in terms of hook capacity

11 4-21 Typical load-slip behavior for #5 MMFX specimens with both 9 and 18-degree hooks (MM_5_9_25_2 shown) Typical load-slip behavior for #7 MMFX specimens with both 9 and 18-degree hooks (MM_7_18_35_4 shown) Anchorage capacity ratios Comparison of normalized bond stress at capacity Comparison of ductility ratios...85 B-1 Crack patterns, load-slip, and stress-strain curves for mild steel hooked bars....9 B-2 Crack patterns, load-slip, and stress-strain curves for stainless steel hooked bars...97 B-3 Crack patterns, load-slip, and stress-strain curves for MMFX hooked bars

12 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 BEHAVIOR OF STANDARD HOOK ANCHORAGE MADE WITH CORROSION RESISTANT REINFORCEMENT Chair: H. R. Hamilton Major: Civil Engineering By Gianni T. Ciancone December 27 The objective of this study was to evaluate the behavior of standard hooks that are made using high strength reinforcing bars and tested in tension. The bars evaluated were ASTM A615, 316LN Stainless Steel and MMFX microcomposite steel. The impetus is that the current ACI/AASHTO equations for the development length of standard hooks do not address the use of high-strength and corrosion resistant steel bars. The development length of standard hooks was evaluated in terms of concrete strength, bar size, hook geometry, concrete covers, debonded length, and lateral reinforcement. Forty-eight specimens with different development length of standard hooks were constructed in accordance with ACI 318 and AASHTO Bridge Design Specifications. Four specimen design configurations were used as unconfined, confined with stirrups, unconfined with debonded length for 9 degree hooked bar and unconfined with debonded length for 18 degree hooked bar. Compressive cylinders tests were conducted in order to determine the target of average concrete strength of 55 psi. Also, rebar samples were tested in tension to obtain the yield, and tensile strength. 12

13 A test frame was constructed in the University of Florida-Structures Lab to test specimens in tension by means of a center hole hydraulic jack. During the test, cracks pattern were observed, and load-displacement were recorded. Test results were compared in function of anchorage capacity, bond stress, ductility, and K-factor. Also, test results indicated that mild steel was consistent and agreeable with ACI and AASHTO requirements for development lengths. For #7 MMFX hooked bars, however, further investigation need to be conducted to evaluate the proper development length. Based on the results obtained from this research the test setup and the procedures using the strut and tie approach appear to provide an adequate basis to evaluate the unconfined anchorage capacities of grade 6 hooked bars. 13

14 CHAPTER 1 INTRODUCTION Mild steel reinforcing bars have been used for decades in buildings, bridges, highways, and other construction projects. One weakness of reinforcement is its lack of corrosion resistance if the concrete cover is breached or penetrated by corrosive elements such as chlorides. This issue can drastically reduce the service life of the structure requiring costly repairs or even replacement early in the life of the structure. One potential solution that has been explored is the use of corrosion resistant steels such as stainless steel, and MMFX. These materials typically have higher strengths than that of mild steel. However, the use of high-strength and corrosion resistant bars has been presented as a substitute for coated and uncoated Grade 6 bars. On the other hand, high-strength reinforcing steel bar reduces not only the use of steel in structural elements but also the labor costs. The main objective of this research was to evaluate the behavior of standard hook anchorages made with high-strength bars as Stainless Steel and MMFX microcomposite steel relative to Grade 6 steel. Since the current ACI/AASHTO Code specifications do not address the use of these kinds of materials, equations for the development length of standard hooks made with high-strength and corrosion resistant steel bars need to be evaluated. The development length of standard hook was evaluated in terms of concrete strength, bar size, hook geometry, concrete covers, slip, anchorage capacity, ductility, bond stress, and K-factor. Also, cracks pattern were evaluated with respect to the failure modes. 14

15 CHAPTER 2 LITERATURE REVIEW The structural performance and flexural behavior of high-strength steel reinforcement has been evaluated as a substitute for Grade 6 bars. Limited research, however, has been conducted dealing with the behavior of standard hook anchorages made with high-strength reinforcement. Hook Behavior and Geometry The structural concrete codes are designed so that, wherever possible, the reinforcement will yield before the concrete crushes when the nominal strength of a reinforced concrete element is reached. Development of the yield strength of a reinforcing bar requires that a sufficient length of bond is available on either side of the critical section where capacity is expected to occur. In locations where space is limited, insufficient space may be available to allow a reinforcing bar to develop. In these cases, it is common to bend the bar to form either a 9-degree or 18-degree hook. Figure 2-1 gives an example of one possible situation where a concentrated load is located near the end of a cantilever beam. The critical section for flexural strength is located at the face of the support. If the required straight development length is longer than the cantilever, then the bar would protrude from the concrete. The typical method to deal with this situation is to turn the bar down into the section, creating a 9-degree hook. The required length to develop the hook is shorter due to the mechanical advantage provided by the concrete located at the inside radius of the bend. Figure 2-2 shows the normal bar stresses in a #7 9-degree hook as reported by Marques and Jirsa (1975). The stresses in the bar increase dramatically around the bend of the hook (from 13 ksi to 57 ksi), indicating that the bearing of the inside of the hook against the concrete provides a significant portion of the anchorage. These bearing stresses cause significant lateral tensile stresses, which can result in a splitting failure when confinement reinforcement is not present. 15

16 Because the strength of hooked anchorages is determined empirically, it was necessary to create a standard geometry for hooks. Figure 2-3 shows the dimensions for standard hooks that are the same in both ACI and AASHTO design specifications. The development length approach was first proposed by Pinc, Watkins, and Jirsa (1977). Table 2-1 shows the minimum hook dimensions proposed in this research. Current Hook Design Practice Standard hook anchorages are currently designed using either the provisions of AASHTO Bridge Design Specifications (24) for bridges or ACI Building Code and Commentary (25) for buildings. The ACI Equation is l dh ψ e.2 λd f b y = f ' (2-1) c and AASHTO LRFD Specifications equation is: 38d f b y l = 6 (2-2) dh ' f c where l dh is the hook development length in in., ψ e is the coating factor, λ is the lightweight aggregate concrete factor, d b is the bar diameter in in., f c is the specified concrete strength in psi, and f y is the specified yield strength of the bar in psi. These provisions were developed in the early 197s and were finally implemented into the code in their present form in

17 Minor and Jirsa (1975) studied the factors that affect the anchorage capacity of bent deformed bars. Specimen geometry was varied to determine the effect of bond length, bar diameter, inside radius of bend, and angle included in the bend. Slip between the bar and the concrete was measured at several points along the bar as load was applied. Load-slip curves were used to compare different bar geometries. The results indicated that most of the slip occurred in the straight and curve portion of the hook. Marques and Jirsa (1975) investigated the anchorage capacity of hooked bars in beamcolumn joints and the effect of the confinement at the joint. The variables considered were size of anchored bars, hook geometry, embedment length, confinement, and column axial load. Full scale beam-columns specimens were designed in order to allow the use of large diameter hooked bars in accordance with ACI code hook geometry standards. The test used #7 and #11 mild steel bars anchored in the columns. ACI specifications were used for 9 or 18 degree standard hooks. Also, for 9 and 18 degree standard hooks, slip of the bar relative to the surrounding concrete was measured at five points along the anchored bar (Figure 2-4). As results, the slip measured on the tail extension of the hook was very small in comparison with slip measured at the point (1H) and the point (2H). The slip measured at the lead was greatest in most of the cases. Also, the slip at point (2H) was similar to the slip at point (1H) when the lead straight embedment was short. In addition, the strength of the bars was evaluated using the ACI design provisions for hooked bar. The strength was determined by calculating the stress developed by the hook (f h ) plus an additional straight lead embedment (l l ). It was found that the straight lead embedment calculated using the basic equation for development length was not enough to develop the yield stress in the hooked bar. On the other 17

18 hand, the use of shorter straight embedment did not improve the stress transferring from the bar to the concrete. Marques and Jirsa (1975) found that the equations from ACI underestimated the anchorage capacity of the hooks. They found that for their test specimens the tensile stress in the bar when the bond capacity was reached was: f h = 7(1.3d ) ψ f ' (2-3) b c where f h can not be greater than f y in psi, d b is the diameter of the bar in in., f c is the average concrete strength in psi, and ψ is a coefficient factor which depends on the size of the bar, the lead straight embedment, side concrete cover and cover extension of the tail. They also determined the straight lead embedment length (l l ) between the critical section and the hook could be expressed as follows: l + ' ' = [.4A (f l b y f )/ f h c ] l (2-4) where l is 4db or 4 in., whichever is greater, A b is the bar area in sq. in., fy the yield strength of the bar in psi, f h the tensile stress of the bar in psi, and f c is the average concrete strength in psi. Pinc, Watkins, and Jirsa (1977) also studied beam-column joints to determine the effect of lead embedment and lightweight aggregate concrete on the anchorage capacity of the hook. The first approach consisted in examining the hook and lead embedment separately. Variables as f l /f.5 c and l l /d b were correlated to obtain the straight embedment strength (f l ). The total strength of the anchored bar (f u ) resulted by adding the straight embedment strength (f l ) and the hook strength (f h ) equation: 18

19 f u = 55(1.4d +.8l /d ) ψ f (2-5) b l b ' c Also, the variables f u /f c.5 and l dh /d b were plotted to obtain the following equation: f u ' = 5ψl f / (2-6) c dh d b As results, it was found that Equation 2-5 and Equation 2-6 were practically the same except for the number of terms in each equation. Equation 2-6 can be rearranged into a form that gives the development length, a parameter that is more useful in design: l dh.2d f b y = (2-7) ' ψ fc where l dh represents the development length for a hooked bar in in., d b is the diameter of the bar in in., f y is the yield strength of the bar, f c is the average concrete strength in psi, and ψ is a coefficient factor which depends on the size of the bar. The ACI 48.1R-79 presented recommendations for standard hook provisions for deformed bars in tension based on the study reported by Pinc, Watkins, and Jirsa (1977), and those recommendations were discussed and explained by Jirsa, Lutz, and Gergely (1979). The development length (l dh ) for standard hook proposed for the ACI 48 committee was the result of the product of the basic development length (l hb ) and the applicable factors. The basic development length was computed as: l hb 96d b = (2-8) φ f ' c 19

20 where l hb represents the basic development length for a hooked bar in in., d b is the diameter of the bar in in., f c is the average concrete strength in psi, and ϕ represents the factor for anchorage which was incorporated in the design equation. The applicable factors included in ACI 48 committee were f y /6, for reinforcement having yield strength over 6, psi,.7 for side cover,.8 for use of stirrups, 1.25 for use of lightweight aggregate, and A sr /A sp for reinforcement in flexural members in excess. Figure 2-5 shows the recommended ϕ factor not only for splices but also for hooked bar, and it compares the test/calculated values for ACI with proposed ϕ factor of.8. Figure 2-6 shows a comparison between the development length proposed and ACI The proposed development length was computed as a lineal function of the diameter of the bar (Figure 2-6), the greater the diameter of the bar the greater the development length. For ACI , the development length was underestimated from #3 until #8 bars and overestimated for bars greater than #8 in comparison with the proposed. Basically, the ACI 318 for basic development length for hooked bar has not changed since Also, most of the applicable factors have not changed except for the inclusion of the epoxy-coated factor of 1.2 which was proposed by Hamad, Jirsa, and D Abreu de Paulo (1993) and included in the ACI For ACI 318-2, the basic development length equation changed in the way as the terms were arranged. Applicable factors as epoxy-coated (β), lightweight concrete (λ) and the yield strength of the bar (f y ) were included in the equation rather than being multiplier factors. Additionally, in this code was included a factor of.8 for 18 degree hook enclosed within ties or stirrups. 2

21 Finally, the development length and the factors included in the current ACI 318 code are the same as ACI High-Strength Steel Reinforcement High strength steel reinforcement has been introduced as a material which is more durable than steel reinforcing bars. The use of high strength reinforcing bars is increasingly rapidly due to the advantages that can offer over conventional reinforcing steel such as fatigue resistance, corrosion resistant, toughness, and ductility. Also, high strength reinforcing bars can be used in bridges and other structures where the high seismic activity is prevalent. Stainless Steel and MMFX are one of the materials categorized as high strength steel due to they do not have well-defined yield points and do not exhibit a yield plateau. Stainless Steel reinforcing bars can be used in reinforced concrete structures where very high durability is required and the life cost analysis is justified. Also, stainless rebar has been used thoroughly in North America and Europe. Stainless rebar might be considered to be used in marine structures where chloride ion is present. As Stainless Steel, MMFX reinforcement is a corrosion-resistant material and stronger than conventional steel. MMFX reinforcing bars have been also used in structures across North America including bridges, highways, parking garage, and residential and commercial projects. Several researches using stainless steel and MMFX reinforcing bars have been conducted and published by universities throughout the United States and sponsored for the Federal Highway Administration (FHWA), and State Departments of Transportation (DOTs). These third parties have conducted studies investigating bond stress behavior, corrosion evaluation, tensile tests, and bending behavior in concrete structures. Strut and Tie Evaluation of Anchorage The strut-and-tie method was proposed by Schlaich, Schäfer, and Jennewein (1987). This method was incorporated in AASHTO LRFD Specifications in 1994 and in ACI Appendix 21

22 A in 22. The design basis of the strut-and-tie method is based on a truss model. The truss model has been used in beams loaded in bending, shear and torsion. However, this model just takes into account certain parts of the structure. The strut-and-tie method consists of struts and ties connected by means of nodes as a real truss. The struts represent the compressive member (concrete) and they serve either as the compression chord in the truss or as the diagonal struts. Diagonal struts use to be oriented parallel to the expected axis of cracking. The ties represent the tension member (stirrups and longitudinal reinforcement) where the anchorage of the ties is crucial to avoid anchorage failure. In order to apply correctly the strut-and-tie model, the structure is classified in B and D regions. The B-regions (B for Bernoulli or beam) are based on the Bernoulli hypothesis which facilitates the flexural design of reinforced concrete structures by allowing a linear strain distribution for any loading stages (bending, shear, axial forces and torsional moments). On the other hand, D-regions (D for discontinuity, or disturbance) are portions of a structure where the strain distribution is nonlinear. D-regions are characterized for changes in geometry of a structural portion (geometrical discontinuities) or concentrated forces (statical discontinuities). For most types of D-regions as retaining walls, pier cap, and deep beam, the use of standard hooks are common as anchorage (Figure 2-7). Additionally, the strut-and-tie model is based on the lower bound theorem of plasticity which allows yielding the bar (ties or stirrups) before crushing of concrete (struts and nodes). The nodes can be classified according with the sign of the forces. At least three forces should act on the node for equilibrium. A C-C-C node represents three compressive forces, a C- C-T node represents two compressive forces and one tensile force, a C-T-T node represents two tensile forces and one compressive force, and a T-T-T node represents three tensile forces. A C- 22

23 C-T node (Figure 2-8) show the nodal zone and extended nodal zone which serve to transfer strut-and-tie forces. The extended nodal zone is defined as the portion limited by the intersection of the strut width (w s ) and the tie width (w t ). The anchorage length (l anchorage ) as shown in Figure 2-8 represents the development length of the hooked bar which is anchored in the nodal and extended nodal zone. Figure 2-9a shows the beam-column specimen used for Marques and Jirsa (1975) and Figure 2-9b shows the strut-and-tie behavior of the hooked bar. 23

24 Table 2-1. Minimum hook dimensions. 18 degree 9 degree Diameter Extension Bar No. d b (in) Head (in.) Tail (in.) (in.) (in.) 4d 6d b 12d b 4d b b Ratio (in.) 3d b mm mm

25 Critical Section for Flexure Figure 2-1. Cantilever beam. 57 ksi 75 ksi 45 kips 13 ksi Figure 2-2. Normal bar stresses #7 9 deg. standard hook. Figure 2-3. Standard hook details. 25

26 3V 2H 3H 4H 4V Slip { 1H column face Vert. Horiz. Figure 2-4. Points where slip was measured. Figure 2-5. Recommended ϕ factor. 26

27 Figure 2-6. Comparison of proposed and ACI hook provisions. A B STRUT C Figure 2-7. Typical uses of a standard hook anchorage and F.B.D. A) Pier cap, B) Deep beam, and C) Retaining wall. 27

28 C w s w t Nodal zone Extended nodal zone T C l anchorage Figure 2-8. Extended nodal zone for standard hook anchorage. Figure 2-9. Strut and tie model of specimen used in Marques and Jirsa research. 28

29 CHAPTER 3 EXPERIMENTAL PROGRAM Specimen Design Figure 3-1 illustrates the typical hooked bar anchorage uses that were targeted with this research. The specimen and load configuration were designed to simulate the development conditions indicated in the figure. Reinforcing bars fabricated with steel that did not have a welldefined yield point were used to investigate the behavior of hooked bar anchorage designed using ACI/AASHTO equations. The effects of concrete strength, bar size, concrete cover, debonded length, and lateral reinforcement were considered. The bars evaluated were ASTM A615, 316LN Stainless Steel and MMFX microcomposite steel. Initial testing was conducted with the design shown in Figure 3-1a and b, which are denoted as unconfined and confined, respectively. The specimen configuration incorporated a single bar centered in a concrete block. The focus of this initial testing was to validate the test setup, specimen design, and loading configuration. Consequently, only ASTM A615 reinforcement was tested. Because the design complied with both design specifications, the expectation was that the specimens would be capable of reaching at least the yield strength of the mild steel reinforcement in both the confined and unconfined specimens. The test results, however, indicated that the confined specimens could reach yield, but that the unconfined specimens were well below yield when the concrete failed. Furthermore, the failure was generally spalling of a corner section of concrete under the reaction at the outside of the hook, which was not the targeted splitting of the specimen in the plane of the hook. In general, the mechanics of hooked bar anchorage can be defined using a strut and tie approach as indicated in the free body diagrams shown for each of the common hook uses. This approach is followed by ACI Appendix A and AASHTO LRFD (Sec ). In 29

30 fact, as indicated in Figure 8, the available development length for the anchorage is defined by the intersection of the reacting compression strut with straight portion of the hooked bar (Schlaich, Schäfer, and Jennewein, 1987). Adjustment to the specimen configuration to simulate the strut and tie behavior of the actual hook is shown in Figure 3-1c and d. The bearing over the hook was lengthened to ensure complete engagement of the bar over the design development length. Although the figure shows the bearing as uniform, it is likely that the actual bearing distribution varied along the length of the specimen. This was not expected to affect the results significantly. The embedded portion of the bar beyond the design development length was debonded to create strut angles between 25 and 47 degrees. The remainder of the testing was conducted with these two configurations using unconfined specimens. Forty eight specimens were cast and tested in five series, with each series representing the specimens cast with a single batch of concrete. The specimen details and testing configuration for the first series are given in Figure 3-2 and Table 3-1. Table 3-1 complied with both AASHTO and ACI design specifications for clear cover and spacing. A factor of.7 was applied because the specimen side cover and cover on bar extension beyond hook were not less than 2-1/2 in and 2 in., respectively. In addition, a factor.8 was applied to the confined specimens to account for the hooks being enclosed by ties or stirrups. Confined specimens used #3 stirrups spaced at 1.88 or 2.63 in. along the development length of the hook. The remaining four series are detailed in Figure 3-3 and and also complied with both AASHTO and ACI design specifications for clear cover and spacing. The specimen naming convention is as follows. The first term represents the type of steel where (6) indicates ASTM 3

31 A615, (SS) stainless steel, and (MM) microcomposite steel. The second term represents the bar size, #5, #7, 16 mm or 2 mm. The third term represents the hook bend angle of 9 or 18 degrees. The fourth term represents the strut angle 25, 35 or 47 degrees, and the last term represents the specimen number or the presence of # 3 stirrups in the hook region. The metric designation of the stainless steel bars was retained because they were manufactured in Italy under hard metric sizes. The 16 mm diameter and area are very near that of a U.S. Customary #5, the 2 mm has slightly smaller diameter and respective area than that of a #7. In Table 3-1 and Table 3-2 f y is the yield strength used to calculate the development length of the bars and does not necessarily represent the actual yield strength of the material. In the ASTM A615 specimens the specified yield strength was used to provide a basis of comparison for the subsequent high-strength steel bars. The values used for f y in determining the development lengths of the SS and MM specimens were taken from tests conducted on bars from the same heat as those used in the pullout tests. The yield strength for these bars was determined using the.2% offset method. Detailed results of these tests are in Chapter 4. The target concrete strength (f c ) used to calculate the development length is shown in these tables. Actual concrete strengths for each of the series varied somewhat from these target values. Actual concrete strengths are provided in Chapter 4. The remainder of the variables in the tables describe the specimen geometry including the development length of the hooked bar as measured from the back edge of the hook. The strut angles shown in the tables are a function of the specimen geometry and were varied to determine the effect of the strut angle on the hook capacity. 31

32 In series of specimens two and three, there were found that slips from specimens with 35 degree strut were greater than slips obtained for specimens with 25 degree strut. Therefore, the strut angle used in series four and five was 35 degree. Also, different development lengths were evaluated for the same kind of rebar. Specimens number 1 and 2 were tested in accordance with ACI Section 12.5 (development of standard hooks in tension), whereas specimens number 3 and 4 were tested with larger development lengths already used in series three (Figure 3-3 and Table 3-2). Concrete Mixture Designs Five batches were used during the research, which correspond to each series detailed in the previous section. The batch for the first series was prepared at Florida Department of Transportation State Materials Office (SMO) in Gainesville, and the last four batches were prepared by Florida Rocks Industries, a local ready-mix concrete supplier. The concrete mixture proportions per cubic yard are shown in Table 3-3. All mixtures used a maximum aggregate size of 3/8-in. (#89 crushed limestones) and silica sand as coarse and fine aggregates respectively. The first batch had a water to cement ratio of.44, and a slump of 5 in. The cement, fine and coarse proportion was 1:2.4:1.99. The second batch had a water to cement ratio of.28, and a slump of 7.5 in. The cement, fine and coarse proportion was 1:2.45:2.5. The last three batches had an average water to cement ratio of.19, and a slump of 7.5 in. The cement, fine, and coarse proportion for those three batches were 1:1.82:1.62. The size of the concrete batch for the first batch was nine cubic feet (.25 cubic meters), and for the last four batches was 81 cubic feet (2.29 cubic meter) per batch. Air-entrained admixture and high-range water reducer were included in the mixture proportions. The water to cement ratio was reduced in the last four batches by means of the inclusion of high-range water reducer (superplasticizer) in order to obtain high concrete strengths 32

33 at early age (14 days). Air-entraining admixture was also used to improve the workability of the concrete. The volume of concrete used in each batch included the specimens, extra examples and concrete for quality control testing. As quality control testing was used the Standard Test Method for Slump of Hydraulic Cement Concrete (ASTM C 143). About twenty standard cylinders 6 x 12-in (152 x 35-mm) were cast at the same time, and vibrated in two layers by means of a vibrating table which was used to assure the compaction. Also, the cylinders were cured at room temperature and under the same condition as the specimens for each concrete batch. Compressive tests were performed in accordance with the Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens (ASTM C39 1). All cylinders were loaded at a load rate of 35 pound square inch per second, and also they were loaded to failure. The maximum load obtained from the universal testing machine was used to calculate the maximum compressive strength. Specimen Construction Formwork The formwork design, shown in Figure 3-4, consisted of a base, two side forms, one front form, one back form, and two 2 x 4 pieces. The front and back forms were kept between the side forms to allow adjustment in the specimen length. This flexibility in the specimen length allowed the formwork to be reused for differing specimen configurations. The front form was built in two pieces to ease bar placement. Three pieces of 2 x 4 were attached below the base to allow forms to be moved either with the crane or the forklift. The long pieces of plywood were clamped together with two 2 x 4 and two threaded rods. The 2 x 4 braces maintained the shape of the forms and dimensions of the specimen. The forms were sealed with a water-based adhesive caulk. 33

34 Casting Four specimens were cast in series one and two, twelve specimens in series three and four, and sixteen specimens in series five. All specimens were cast with the bar placed in the bottom of the forms with the tail of the bend pointed upward (Figure 3-4b and Figure 3-5). A thin wire was attached to the side forms and to the tail of the hook to hold the bar level, and to maintain the side cover required. The debonded part of the bar was composed of a plastic tube which was sealed with electric tape to prevent cement paste from entering the tube. Since most of the formwork as placed inside of University of Florida-Structural Laboratory, the concrete from the ready mix truck was poured directly to a galvanized steel container (Figure 3-6). Afterward, the container was moved to be near to the formworks, and a slump test was performed as stated in ASTM C143- (Figure 3-7). To ensure that the instrumentation and bar position were not disturbed, concrete was delivered to the forms from the container by hand (Figure 3-8A). Each specimen was cast in two lifts, which were compacted using mechanical vibrators. As concrete was placed in the forms, standard 6x12-in (152 x 35-mm) cylinders were cast, and also vibrated in two layers. Once finished with the casting procedure, the top surfaces of the specimens were smoothed with a finishing trowel (Figure 3-8B). Finally, a plastic sheet was placed over the specimens to minimize the evaporation of the water (Figure 3-9). The specimens and cylinders were left to cure in the same environment until they were tested. Test Setup A test frame was constructed with back-to-back structural channels. Each two structural channels were connected and stiffened by.5-in. thick plates. A double C15x4, and C15x4 were welded together to form a 9 degree frame. Each end of the frame was then welded to C12x3 shapes, which were attached to the strong floor and wall. Stiffeners were added to stiffen 34

35 the frame against the heavy concentrated loads from the specimen (see Figure 3-1A, and Figure 3-1C). After fabrication, the test frame was connected to the strong wall and floor by means of eight 5/8 bolts, and eight 1-1/4 bolts respectively (Figure 3-1B). The specimen was seated in a 22 x 22-in. steel base. Tension was applied to the bar extension by means of a center hole hydraulic jack. The threaded rod passed through the 2C15x4 beam, and the center hole hydraulic jack (Figure 3-1B). A coupler system was used to connect the anchored bar to a threaded rod (Figure 3-11). This load was reacted with a strut placed between the specimen and the horizontal member of the reaction frame. The moment generated by the couple was reacted horizontally with the vertical member of the reaction frame. The reaction on the left face of the specimen shown in Figure 3-12 was distributed over the development length of the hook. The remaining portion above the bar was debonded to ensure that only the portion of the hook under the reaction contributed to the bar development. Data Acquisition Setup Slip between the hooked bar and the concrete was measured by a procedure developed and used by Minor and Jirsa (1975). Figure 3-13 shows the locations along the hook where relative slip was measured. Location 1 was at the loaded end and location 2 was at the beginning of the bend. A.625 in. diameter hole was drilled in the hooked bar. A.16 in. diameter wire was attached to the anchored bar at points 1 and 2 by inserting part of the wire to the ¼-in deep holes and securing with a small brass screw. The wire was placed inside of a thin plastic conduit of.42 in. diameter along the entire length in order to prevent bonding and to allow free movement of the wire relative to the surrounding concrete (Figure 3-14). The conduit containing the wire was extended from the bar attachment point through the concrete and exited the specimen on the side opposite to the straight portion of the bar. The 35

36 exposed conduit and wire was then connected to a linear pot placed in a 1 x 1-in. frame (Figure 3-14B). The linear pots were used to measure the relative movement between the wire and the conduit, which is nearly a direct measure of the relative movement of the bar and concrete at attachment point of the wire. Bar displacement was also measured relative to the top side of the specimen using a linear pot clamped to the bar (Figure 3-14A, Figure 3-15). The purpose of this linear pot was to measure the strain of the debonded portion of the bar and any slip that might occur before failure. The data acquisition system consisted in a LabView virtual instrument which was programmed to read and record data points from linear pots, and a load cell (Figure 3-16). 36

37 Table 3-1. Specimen design details for series 1. Specimen f y (ksi) f'c (psi) W (in) H (in) B (in) Strut Angle l dh tested (in) d L (in) 6_5_9_S _5_9_ _7_9_S _7_9_

38 Table 3-2. Specimen design details for series 2 through 5. Series f Specimen y f' c W H B Strut l dh tested d L Number (ksi) (psi) (in) (in) (in) Angle (in) (in) 6_5_9_25_ Two 6_5_9_25_ _7_9_47_ _7_9_47_ SS_16_9_25_ SS_16_9_25_ SS_16_9_35_ SS_16_9_35_ MM_5_9_25_ Three MM_5_9_25_ MM_5_9_35_ MM_5_9_35_ MM_7_9_25_ MM_7_9_25_ MM_7_9_35_ MM_7_9_35_ SS_16_18_35_ SS_16_18_35_ SS_16_18_35_ SS_16_18_35_ MM_5_18_35_ Four MM_5_18_35_ MM_5_18_35_ MM_5_18_35_ MM_7_18_35_ MM_7_18_35_ MM_7_18_35_ MM_7_18_35_ _5_18_35_ _5_18_35_ _7_18_35_ _7_18_35_ _7_18_35_ _7_18_35_ SS_2_9_35_ Five SS_2_9_35_ SS_2_9_35_ SS_2_9_35_ SS_2_18_35_ SS_2_18_35_ SS_2_18_35_ SS_2_18_35_ MM_7_9_35_ MM_7_9_35_

39 Table 3-3. Concrete mixture proportions (quantities are per cubic yard). Series and Mixing Dates Materials 2/1/27 3/9/27 4/9/27 5/9/27 6/8/27 W/C Cement (lb) Fly Ash (lb) Water (lb) Fine Aggregate (lb) Coarse Aggregate (lb) Air-entrained (oz) Admixture (oz) Slump (in.)

40 l dh l dh A B Debonded Debonded l dh STRUT l dh STRUT C D Figure 3-1. Specimen design with idealized boundary conditions. A) Unconfined, B) Confined with stirrups, C) 9 deg. hook, unconfined with debonded length, and D) 18 deg. hook, unconfined with debonded length. Ct Cs Ct Cs A A A A Ct Cb Ctail H B Section A - A W A Se Cb W Ss H Ctail W No. 3 stirrup B Section A - A Cb B Figure 3-2. Specimen design for series 1: A) Unconfined specimen details and B) Confined specimen details. 4

41 Ct Cb Ct Cb dl dl A l dh a A Ctail H W A l dh a Ctail A H W Cb W B Section A - A A W B Section A - A B Figure 3-3. Specimen design for series 2 through 5: A) Unconfined specimen details for 9 degree bend and B) Unconfined specimen details for 18 degree bend. 5/8" Thread Rod 1 x 1 Lumber 2 x 4 Lumber Coupler A A 2.5" Plywood 3/4" A 2 pieces of 3/4" of Plywood placed above and below the bar Coupler 2 x 4 Lumber Section A - A B Figure 3-4. Formwork schematics A) Plan view, and B) Section. 41

42 Figure 3-5. Formwork details. Figure 3-6. Ready-mixed concrete being discharged into the container for transporting. 42

43 Figure 3-7. Slump of ready-mixed concrete. A B Figure 3-8. Casting and compaction of the specimen A), and B) Finishing of specimens. Figure 3-9. Curing of the specimens. 43

44 Load Cell Hydraulic Jack A Strong Wall 2C15x4 C12x3 1 17" 4 " 5 8 " Bolts 7" Open holes " 12" 5' - 3" 2C15x4 A HSS 4x3x1/4 Coupler Thread Rod 4' 2C6x13 Strong Floor C12x3 4' - 2" A Section A-A 22" x 22" Base B C Figure 3-1. Load test setup A) Plan view schematic, B) Section schematic, and C) Photo. 44

45 Figure Coupler system. Plate 12x3x1 Bearing length varied as needed to create target development length Shims 6x12x1/4 T HSS 4x3x1/4 Shims 6x1x1/4 Neoprene 6x8x1/4 Neoprene 6x8x1/4 2C6x13 l dh STRUT Neoprene 6x12x1/4 Figure Specimen schematic reactions. 1 2 Figure Slip wire position in hooked bar. 45

46 Load Cell Displacement 1 Bond Slip 2 Bond Slip A B Figure Bond slip instrumentation A) Displacement and slip position, B) Linear potentiometers. Figure Linear potentiometer placed at the top face of the specimen. Figure Data acquisition system. 46

47 CHAPTER 4 RESULTS AND DISCUSSION Materials Properties Concrete About twenty standard cylinders 6 x 12-in (152 x 35-mm) per batch were tested in accordance with the Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens (ASTM C39 1). Compressive strengths of each batch are shown in Table 4-1. The first batch was mixed at Florida Department of Transportation State Materials Office (SMO) in Gainesville, and the last four batches were delivered by Florida Rocks Industries, a local readymix concrete supplier. Compressive strengths were tested after 7, 14, 21, and 28 days of continuous lab cured for all the concrete mixes (APPENDIX A). Steel ACI indicates that for bars exceeding a specified yield strength of 6 ksi (413 MPa), the yield strength is to be determined using the stress corresponding to a.35% strain. The.2% offset method (ASTM A37-7), however, is more generally applicable to high strength steel that have no well-defined yield point. Consequently, for the stainless steel and MMFX bars that do not have well-defined yield points and do not exhibit a yield plateau, the.2% offset method was used in lieu of the.35% strain method. All the tension tests were conducted at Florida Department of Transportation State Materials Office (SMO) in Gainesville. Four coupons were tested for each Grade 6, Stainless Steel, and MMFX bars. The load rate used was.2 inches per minute per in. of distance between the grips (in/min/in) until the yield point was determined. After yielding, the rate used was 3.5 in/min/in until bar rupture occurred. 47

48 Stainless steel and MMFX bars do not have a well-defined yield point and do not exhibit a yielding plateau; therefore, the.2% offset method (ASTM A37-7) was used to determine the yield strength of the bar. This method is illustrated in Figure 4-1 where the intersection of the stress-strain curve with a line parallel to the slope equal to the initial tangent modulus and which intercept at.2 strain defines the yield point. Data gathered during tension tests included strain at.2% offset, load at.2% offset, and ultimate strength. Complete tension test results are given in APPENDIX A. Grade 6 Steel All mild steel bars came from the same heat and were purchased locally at a building supply center. The #5 bar had yield strength measured at.35% (.35 in/in) strain of 63 ksi, and a tensile strength of 15 ksi. The #7 bar had yield strength measured at.35% (.35 in/in) strain of 64 ksi, and a tensile strength of 16 ksi (Table 4-2). The two samples of each size exceeded and complied with the ASTM A615 (Grade 6) standard which established the minimum yield strength, and tensile strength of 6 ksi, and 9 ksi, respectively. Stainless Steel The stainless steel 316LN bars were made in Italy and were provided by Valbruna Stainless Steel. Valbruna Stainless Steel is a company specialized in supplying and producing stainless steel and special metal alloys. The company has several braches in United States and Canada. Their stainless steel bars have been used worldwide in different applications as bridges, highway and roads, viaducts, and ports. After testing, the 16 mm bar had a yield strength measured at.2% (.2 in/in) strain offset of 16 ksi, and a tensile strength of 124 ksi. The 2 mm bar had a yield strength measured at.2% (.2 in/in) strain offset of 96 ksi, and tensile strength of 12 ksi (Table 4-3). The yield and tensile strengths measured in the two samples of 48

49 each size exceeded and complied with the minimum yield strength of 75 ksi and minimum tensile strength of 1 ksi required for ASTM A955 and Valbruna product specifications. MMFX Steel The MMFX bars were provided for MMFX Steel Corporation of America. MMFX Steel Corporation of America is a subsidiary of MMFX Technologies, a company that has invented the MMFX 2 steel bar which has a microstructure different to the conventional steel. The MMFX 2 steel rebar is a corrosion resistant and a high grade steel which has been used nationwide in different construction applications as bridge decks, bridge structures, and residential. After testing, the #5 and #7 bars had yield strengths measured at.2% (.2 in/in) strain offset of 122 ksi, and 128 ksi, respectively (Table 4-4). The yield strength measured in the two samples of each size exceeded and complied with the minimum yield strength of 12 ksi required for ASTM A135 and MMFX product specifications. Specimens Test Results Behavior and Failure Modes Figure 4-2 shows the stress-strain plot of three pullout specimens to illustrate the typical behavior of each type of steel. Load-slip and stress-strain curves for all specimens are shown in APPENDIX B. The stress was obtained by dividing the measured load by the nominal area of the reinforcing bar. The strain was obtained by dividing the measured bar displacement by the debonded length. In general, as load was applied the specimen remained uncracked and linear elastic until the yield point was reached. In some of the specimens cracking occurred, this caused a loss of bond and a premature failure. This failure mode was deemed concrete splitting which occurred suddenly when the peak load was reached. This type of failure was characterized by cracks that split the specimen from the front to the right face (Figure 4-3A). Also, diagonal cracks formed on 49

50 the right and left side of the specimen confirming the strut behavior of the specimens (Figure 4-3B). The front face of the specimen presented the typical Y crack which is seen in bond test using beam end specimens (Ahlborn and DenHartigh, 22). The rear face exhibited an inverted Y crack which split the specimen in three parts (Figure 4-3C and D). Crack pattern of this kind of failure was seen in specimen MM_7_18_35_3 as it is shown in Figure 4-4. After testing, a larger portion of the side cover was easy to remove. During the specimen examination, it was found crushing of the concrete inside radius of the hook. This kind of behavior was seen not only in 9 degree but also in 18 degree hooks (Figure 4-5). Moreover, crushing of the concrete near to the radius of the bend was because of the higher tensile force applied to the bar producing mini cracks between the bar and the concrete and resulting in loss of bond. This type of behavior was also observed and reported by Marques and Jirsa (1975) and Hamad, Jirsa, and D Abreu de Paulo (1993). The main objective of those studies was to evaluate bond characteristics and anchorage capacity of uncoated (mild steel) and epoxy-coated hooked bars for 9 and 18-degree bend angle. If the specimen was able to sustain load beyond yield, one of two possible failure modes occurred. The bar yield with concrete splitting, occurred after the bar had yielded indicating that the anchorage was able to hold load at least to the yield point. Cracks pattern are similar to the concrete splitting failure. Bar yield was characterized by continued deformation of the bar without concrete splitting or bar rupture. This typically occurred on the stainless steel specimens when the hydraulic jack stroke limit was reached. Specimens SS_16_9_25_1, SS_16_9_25_2, SS_16_9_35_1, 5

51 SS_16_9_35_2, SS_16_18_35_1, and SS_16_18_35_4 were loaded until the stroke of the hydraulic jack reached its limit, however; the bar reached the yielding point before the test was terminated. After testing, cracks were not seen on the faces of the specimen. Finally, several specimens failed due to bar yield and rupture. This occurred when the full rupture strength of the bar was reached before the concrete failed. The bar yield and rupture failure was mainly observed in MMFX specimens. Mild Steel Specimens In this section the detailed results of the mild steel specimens are presented and discussed. Failure modes for each specimen are documented as well as the load displacement and load slip behavior. Figure 4-6 shows the load displacement behavior for all of the #5 and #7 mild steel specimens. Also, Figure 4-6 show the coupon yield load (P yt ) for #5 and #7 which confirms that the bars reached yield. The plots for each are shown with different scales to accentuate the differences in behavior among the specimens with the same size bar. The 25-degree strut specimens appear to have a larger initial stiffness than that of the 35-degree strut specimens when comparing the results for the #5 bar. This is likely due to the manner in which the displacements were measured. The linear potentiometer was attached to the bar at the point where it exits the concrete and measured the relative movement between the bar and concrete. The 25-degree strut specimens had shorter debonded lengths than that of the 35-degree strut specimens resulting in larger elastic deformations under the same load. The sudden change in slope of the load displacement plots indicate yielding of the bars and generally agreed well with the measured yield strength of the bare bars. The anchorage strength of #5 specimens with 18-degree hook improved about 23% with respect to #5 specimens with 9-degree hooked bar as the concrete strength and the strut angle increased (Figure 4-6a). 51

52 Post-yield slopes are not likely to provide useful information because the measurement of bar displacement is made relative to the concrete surface around the bar. Microcracking is likely to occur near yield, which will result in movement of the concrete along with the bar as ultimate strength is approached. This behavior is described more fully when the slip data are presented. Figure 4-7 summarizes the results of the tests in terms of the hook capacity. The complete test results for mild steel specimens are shown in the Table 4-5. f c shows the average concrete strength of the specimen concrete as tested on the day of the pullout test. P u is the peak measured load applied to the bar. To allow comparison of the peak measured loads among the specimens that contained varying concrete strength, P u was normalized to the square root of the ratio of the design strength (55 psi) to the measured strength. P ye is the load at which the bar yielded using the.35% strain. Δ u is the displacement corresponding to P u and Δ y is the displacement corresponding to P ye. The bar stress based on the peak measured load is also given (P u /A b ). D1 and D2 represent the total measured slip of the bar when the load in the bar is P u. The load slip data gathered during the testing provides interesting insight into the behavior of the hooked bar anchorages. Figure 4-8 show two graphs that compare the confined and unconfined #5 bar specimens from the first series of testing. Recall that this testing was conducted with the original test configuration. It is readily apparent that the unconfined specimen (which did not reach yield) has a shallower load-slip slope than that of the confined specimen with stirrups, indicating that the lack of stirrups allowed greater bar movement prior to reaching ultimate capacity. This confirms observations by Hamad, Jirsa, and D Abreu de Paulo (1993). Hamad, Jirsa, and D Abreu de Paulo evaluated beam-column joints with mild steel and epoxy-coated hooked bars. 52

53 Their results concluded that for #7 uncoated specimens with 9 degree hooked bar, the anchorage strength increased about 51% with the inclusion of stirrups. However, for #7 specimens tested in this research with 9 degree hooked bar, the anchorage strength increased about 69% with the inclusion of stirrups. The differences between the results of comparative tests are based on the test setup, the use of high concrete strength, strut-and-tie approach, and stirrups spacing. Further examination of the plots indicates that the slip at D1 is greater than that of D2 until higher loads are reached where the plots cross. This occurs in both the confined and unconfined specimens. D1 was expected to remain greater than D2 up to failure since the bar exits the specimen near where D1 is measured. The cross-over of the plots is likely due to cracking late in the loading process and is a function of the slip measurement technique and not an indication of peculiar behavior. Figure 4-9 shows the idealized location of cracks in unconfined and confined specimens, which are similar to those observed during and after the testing. As load is applied, the slip at D1 is greater than that of D2. As additional load is applied, diagonal cracks form perhaps along line 2-3. When these cracks occur, a spall in the shape of forms and moves with the bar as further load is applied resulting in zero bond stress in this area. Because the slip measurement device measures relative movement between the concrete and steel, less (or zero) slip will register after the spall occurs. These cracks likely form when the specimen is near capacity, which confirms the crossing locations in the plots. For unconfined specimens, initial slip located at D1 was greater than slip located at D2 until diagonal cracks formed as shown in Figure 4-9a. For confined specimens, the use of transverse reinforcement not only improved the anchorage capacity of the hooked bar but also 53

54 controlled crack propagation. The inclusion of transverse reinforcement was sufficient to yield the bar and to achieve the bar rupture failure. Figure 4-1 shows the relative behavior of the confined and unconfined #7 tests. The unconfined test is similar to that of the #5 with failure occurring before bar yield and with a crossing of the slip plots near the specimen ultimate capacity. In contrast, however, the confined specimen never exhibits the cross-over of the slip plots. This is probably due to the confinement restricting the formation of the spall in the region of D1. Slip behavior of the series 2 through 5 tests was similar to that of the unconfined specimen from series 1 except that most of the specimens tested with the revised setup reached yield before failure. Figure 4-11 provides an example of the load slip behavior for a #5 bar with a 18-deg. hook. As expected, D1 remained greater than D2 for the entire test, and never crossed D2 as the load approached capacity. Recall that the slip D1 was measured at the end of the debonded length (dl), which placed it closer to the bend than in the previous test setup (Figure 4-12). Figure 4-12 shows two possible locations where diagonal cracks formed at the edge of the strut. Crack 2-3 is shown above D1 and Crack 4-5 is shown below. It is believed that the reason there was no cross-over is that the cracking occurred primarily along line 2-3, which formed spall and allowed the relative slip D1 to continue to be measured up to failure. Furthermore, the D2 plot shows a plateau forming while D1 remains linear up until failure of the concrete indicating that the bar was well beyond its yield point at D1. Typical behavior of a #7 mild steel bar with a 18-degree hook is shown in Figure The behavior illustrated is similar to that of the #5 specimen in that D1 remains larger than D2 until failure. 54

55 Stainless Steel Specimens Detailed results of the stainless steel specimens are presented and discussed. Failure modes for each specimen are documented as well as the load displacement and load slip behavior. Figure 4-14 shows the load displacement behavior for all of the 16 and 2-mm stainless steel specimens. Also, Figure 4-14 show the coupon yield load (P yt ) for 16mm and 2mm, which confirms that the bars reached yield. The plots for each are shown with different scales to accentuate the differences in behavior among the specimens with the same size bar. All of the specimens with 16 mm bars reached their yield point with no bar rupture. In many cases, the test was terminated when the stroke of the hydraulic jack reached its limit. In contrast, most specimens with 2 mm bars reached their yield point but then failed by splitting of the concrete. During this portion of the testing program it was discovered that stainless steel bars from two different heats had used (P yt1 and P yt2 ), which explains the difference in the yield loads exhibited in Figure 4-14a for the 16 mm bars. For 16 mm and 2 mm specimens, the bond between the bar and the concrete made the tangent modulus slopes steeper (Figure 4-14). For 2 mm specimens, load-displacement curves were quite similar despite of different development lengths, strut angles, and hook geometries (Figure 4-14b). Figure 4-15 summarizes the results of the tests in terms of the hook capacity. The test results for stainless steel specimens are shown in the Table 4-3. f c shows the average concrete strength of the specimen concrete as tested on the day of the pullout test. P u is the peak measured load applied to the bar. To allow comparison of the peak measured loads among the specimens that contained varying concrete strength, P u was normalized to the square root of the ratio of the design strength (55 psi) to the measured strength. P ye is the load at 55

56 which the bar yielded using the.2% offset strain. Δ u is the displacement corresponding to P u and Δ y is the displacement corresponding to P ye. The bar stress based on the peak measured load is also given (P u /A b ). D1 and D2 represent the total measured slip of the bar when the load in the bar is P u. Because of the 25-degree strut specimens had shorter debonded lengths than that of the 35- degree strut specimens resulting in larger elastic deformations under the same load (Figure 4-16). As a result, it was found that the maximum slip for specimen SS_16_9_35_2 increased about 56% as the strut angle increased in comparison with the specimen SS_16_9_25_2 ( Table 4-6). Typical load-slip behavior is illustrated in Figure 4-17 for 16 mm stainless steel specimens. Initial slip is larger for D1 than for D2. As the load nears yield, however, the plots cross, indicating that the diagonal crack formed the spall (Figure 4-12) in the debonded region of the bar. Figure 4-18 indicates that the 2 mm stainless steel specimens behave more like the #7 mild steel specimens than that of the 16 stainless steel specimens. This may be due to the difference in the failure mode. Recall that the 16 mm stainless steel specimens did not split while both the #7 mild steel and 2 mm stainless steel specimens yielded and then split. MMFX Specimens In this section the detailed results of the MMFX specimens are presented and discussed. Failure modes for each specimen are documented as well as the load displacement and load slip behavior. Figure 4-19 shows the load displacement behavior for all of the #5 and #7 MMFX specimens. Also, Figure 4-19 show the coupon yield load (P yt ) for #5 and #7 which confirms that the bars reached yield. The plots for each are shown with different scales to accentuate the 56

57 differences in behavior among the specimens with the same size bar. All of the specimens with #5 bars reached yield, which appears to be at approximately the same load. In contrast, just a few specimens with #7 bars reached their yield point before failure by concrete splitting occurred, indicating that the bond strength was not sufficient to develop the #7 bars as fully as the #5 bars. It was found that the anchorage strength at failure of #5 specimens with 18-degree hook improved about 9% as the development length increased ( Table 4-7). Figure 4-2 summarizes the results of the tests in terms of the hook capacity. Also, in Figure 4-2, it was not noticed any difference between the average strength of 9 and 18-degree hook for #5 and #7 specimens. The test results for MMFX specimens are shown in the Table 4-7. f c shows the average concrete strength of the specimen concrete as tested on the day of the pullout test. P u is the peak measured load applied to the bar. To allow comparison of the peak measured loads among the specimens that contained varying concrete strength, P u was normalized to the square root of the ratio of the design strength (55 psi) to the measured strength. P ye is the load at which the bar yielded using the.2% offset strain. Δ u is the displacement corresponding to P u and Δ y is the displacement corresponding to P ye. The bar stress based on the peak measured load is also given (P u /A b ). D1 and D2 represent the total measured slip of the bar when the load in the bar is P u. Typical behavior of a #5 and #7 mild steel bar with a 9 and 18-degree hooks is shown in Figure 4-21, and Figure The behavior illustrated is similar to that of the #5 and #7 mild steel specimens with 18 degree hook in that D1 remains larger than D2 until failure. The maximum slip for specimen MM_5_9_35_2 increased about 114% as the strut angle increased 57

58 in comparison with the specimen MM_5_9_25_2. Also, it was found that the maximum slip for specimen MM_5_18_35_2 increased about 116% as the development length increased in comparison with the specimen MM_5_18_35_4. 58

59 Table 4-1. Compressive concrete strengths. Series Average Concrete Strength Coefficient of Variation (%) Table 4-2. Tension test results for ASTM A615 reinforcement. Grade 6 Yield Strength at.35% strain (ksi) Strain at.35% yield (in/in) Load at.35% (kip) Ultimate Strength (ksi) #5 Average COV (%) < 1. < 1.11 #7 Average COV (%) < 1. < 1 < 1 Table 4-3. Tension test result for stainless steel (316LN). Stainless Steel Yield Strength Strain at.2% at.2% offset offset yield (ksi) (in/in) Load at.2% offset (kip) Ultimate Strength (ksi) 16 mm (.625 in) Average COV (%) < < 1 < 1 2 mm (.787 in) Average COV (%) < 1 Table 4-4. Tension test results for MMFX steel. MMFX Yield Strength at.2% offset (ksi) Strain at.2% offset yield (in/in) Load at.2% offset (kip) Ultimate Strength (ksi) #5 Average COV (%) < 1 < 1 < 1 < 1 #7 Average COV (%) < < 1 < 1 59

60 6 Table 4-5. Test results for mild steel #5 and #7 specimens. 55 Pu P u P ye D1 u D2 u P u /A b f ' c Specimen notation f' c (psi) (kips) (kips) Δ u (in) Δ y (in) (in) (in) (ksi) (kips) Failure Modes 6_5_9_ N.A.85 NA Bar yield with concrete splitting 6_5_9_S N.A.289 NA Bar yield and rupture 6_5_9_25_ NA NA Bar yield with concrete splitting 6_5_9_25_ Bar yield with concrete splitting 6_5_18_35_ Bar yield and rupture 6_5_18_35_ Bar yield and rupture 6_7_9_ N.A.37 N.A Concrete splitting 6_7_9_S N.A.89 N.A Bar yield with concrete splitting 6_7_9_47_ N.A N.A Bar yield 6_7_9_47_ Bar yield 6_7_18_35_ Bar yield with concrete splitting 6_7_18_35_ Bar yield with concrete splitting 6_7_18_35_ Bar yield with concrete splitting 6_7_18_35_ Bar yield with concrete splitting

61 61 Table 4-6. Test results for stainless steel 16 mm and 2 mm specimens. 55 Pu P u P ye P u /A b f ' c Specimen notation f' c (psi) (kips) (kips) Δ u (in) Δ y (in) D1 u (in) D2 u (in) (ksi) (kips) Failure Modes SS_16_9_25_ Bar yield SS_16_9_25_ Bar yield SS_16_9_35_ Bar yield SS_16_9_35_ Bar yield SS_16_18_35_ Bar yield SS_16_18_35_ Bar yield with concrete splitting SS_16_18_35_ Bar yield and rupture SS_16_18_35_ Bar yield SS_2_9_35_ Bar yield with concrete splitting SS_2_9_35_ Bar yield with concrete splitting SS_2_9_35_ N.A.11 N.A Bar yield with concrete splitting SS_2_9_35_ Bar yield with concrete splitting SS_2_18_35_ Bar yield with concrete splitting SS_2_18_35_ Bar yield with concrete splitting SS_2_18_35_ Bar yield with concrete splitting SS_2_18_35_ Bar yield with concrete splitting

62 62 Table 4-7. Test results for MMFX steel #5 and #7 specimens. 55 Pu P u P ye P u /A b f ' c Specimen notation f' c (psi) (kips) (kips) Δ u (in) Δ y (in) D1 u (in) D2 u (in) (ksi) (kips) Failure Modes MM_5_9_25_ Bar rupture MM_5_9_25_ Bar rupture MM_5_9_35_ Bar yield with concrete splitting MM_5_9_35_ Bar yield with concrete splitting MM_5_18_35_ Bar yield with concrete splitting MM_5_18_35_ Bar yield with concrete splitting MM_5_18_35_ Bar yield with concrete splitting MM_5_18_35_ Bar rupture MM_7_9_25_ N.A.21 N.A Concrete splitting MM_7_9_25_ N.A.29 N.A Bar cast out of position MM_7_9_35_ N.A.1 N.A Bar cast out of position MM_7_9_35_ N.A.29 N.A Bar cast out of position MM_7_9_35_ N.A.44 N.A Concrete splitting MM_7_9_35_ Bar yield with concrete splitting MM_7_18_35_ N.A.35 N.A Concrete splitting MM_7_18_35_ Bar yield with concrete splitting MM_7_18_35_ N.A.14 N.A Concrete splitting MM_7_18_35_ Bar yield with concrete splitting

63 f f y.2 % ε y ε Figure 4-1. Stress-strain curve. Stress-Strain Comparison _5_9_25_1 MM_5_9_25_1 SS_16_9_25_ Strain (in/in) Stress (MPa) Figure 4-2. Stress-strain comparison. STRUT A B Figure 4-3. Cracks A) on the Top, B) on the side faces, C) on the rear and D) on the front faces. 63

64 C D Figure 4-3. Continued. Top Front Rear Bottom Right Left Figure 4-4. Crack pattern for concrete splitting failure. A B Figure 4-5. Concrete crushed inside of bend radius A) 9 deg. hook and B) 18 deg. hook. 64