PERFORMANCE OF DOUBLE-T PRESTRESSED CONCRETE BEAMS STRENGTHENED WITH STEEL REINFORCED POLYMER ABSTRACT

Transcription

1 PERFORMANCE OF DOUBLE-T PRESTRESSED CONCRETE BEAMS STRENGTHENED WITH STEEL REINFORCED POLYMER Paolo Casadei 1, Antonio Nanni 2, Tarek Alkhrdaji 3 and Jay Thomas 4 ABSTRACT In the fall of 2002, a two-storey parking garage in Bloomington, Indiana, built with preast prestrestressed onrete (PC) double-t beams, was deommissioned due to a need for inreased parking-spae. This led to the opportunity of investigating the flexural performane of the PC double-t beams, upgraded in the positive moment region with steel reinfored polymer (SRP) omposite materials, representing the first ase study where this material has been applied in the field. SRP makes use of highstrength steel ords embedded in an epoxy resin. This paper reports on the test results to failure of three beams: a ontrol speimen, a beam strengthened with one ply of SRP and a third beam strengthened with two plies of SRP anhored at both ends with SRP U- wraps. Results showed that SRP an signifiantly improve both flexural apaity and enhane pseudo-dutility. Preliminary analytial work shows that the same approah used for externally bonded fiber reinfored polymer (FRP) an be satisfatorly used for SRP. Keywords: double-t beams; dutility; flexure; in-situ load test; prestressed onrete; steel reinfored polymer; strengthening. 1 Leturer of Strutural Engineering, Department of Arhiteture and Civil Engineering, University of Bath, Bath, BA2 7AY, United Kingdom Tel , Fax , P.Casadei@bath.a.uk 2 V & M Jones Professor, Department of Civil, Arhitetural and Environmental Engineering 223 Engineering Researh Lab, University of Missouri-Rolla, Rolla, MO USA 3 Strutural Engineer, Strengthening Division of the Strutural Group. 4 Vie President, Strutural Preservation Systems In

2 1 INTRODUCTION The use of advaned omposite materials in the onstrution industry is nowdays a mainstream tehnology (Rizkalla and Nanni 2003), supported by design guidelines suh as the ACI 440.2R-02 (2002) in the United States and the Fib-Bullettin 14 (2001) in Europe. Fiber reinfored polymer (FRP) omposite materials, even though very attrative, may be hindered by lak of dutility and fire resistane. Both issues are urrently under study by the researh ommunity (Williams et al. 2004, Bisby et al. 2004, Seible et al. 1997), in order to provide on one hand, better knowledge in terms of overall strutural performane and, on the other, remedies suh as oatings that ould prolong fire resistane. A new family of omposite materials based on high strength twisted steel wires (about 7 times stronger than typial ommon reinforing bars) of fine diameter (0.20~0.35 mm (0.0079~ in) see Figure 1), that an be impregnated with thermo-set or ementitious resin systems is presented in this paper (Hardwire 2002). SRP has the potential to address the two shortomings mentioned for FRP, in fat: a) steel ords have some inherent dutility; and b) impregnation with ementitious paste may overome the problems of fire endurane. The steel ords used in SRP are idential to those used for making the reinforement of automotive tires, and manufatured to obtain the shape of the fabri tape prior to impregnation (Hardwire, 2002). The twisting of the wires allows some mehanial interlok between the ords and the matrix, and may also indue an overall dutile behavior upon strething. Charaterization work is urrently in progress as neessary for implementation in future design guidelines

3 Limited researh results have been published on this new generation of omposite materials. Huang et al. (2004) investigated the mehanial properties of SRP, testing different kinds of matries, epoxy resin and ementitious grout, inluding a omparison between theoretial and experimental results needed for design. Test results showed that the material does not experiene a substantial yielding, but rather a similar behavior to the one experiened by high-strength steel used in prestressed onrete (PC) onstrution, with a slightly non-linear range prior to rupture of the ords. The opportunity for experimenting this new material in the field, beame available in the winter of 2003 when the City of Bloomington, Indiana, deommissioned an existing parking garage near the downtown area, built with double-t PC beams. The onrete repair ontrator, Strutural Preservation Systems, Hanover, MD, strengthened in flexure the bottom stem of several double-t beams with with epoxy-based SRP. This paper reports on the experimental as well as analytial results of tests to failure onduted on three beams: a ontrol speimen, a beam strengthened with one ply of SRP and a third beam strengthened with two plies of SRP anhored at both ends with U- wraps. 2 EXPERIMENTAL PROGRAM 2.1 Building Charateristis The parking garage used for the tests was a two storey struture onstruted in the 1980s (see Figure 3). It onsisted of a reinfored onrete (RC) frame, ast in plae olumns and preast reversed-t PC beams, supporting double-t PC beams, of span length varying from 4.66 m (15.3 ft) to m (44 ft)

4 Sine no maintenane or onstrution reords were available for the materials and the layout of the prestressing tendons, a field investigation was arried out. Based on the survey, it was determined that the double-t PC beams were of type 8DT32 aording to the Prestressed Conrete Institute (1999) speifiations (see Figure 4) with onrete topping of 76 mm (3in), and with an arrangement of the tendons different from urrent speifiations. For the span of 4.66 m (15.3 ft), two straight 7-wire strands were found in eah stem, eah with a diameter of 12.7 mm (0.5 in), orresponding to an area of 112 mm 2 (0.174 in 2 ), the first at 248 mm (9.75 in) from the bottom of the stem and the seond spaed 305 mm (1 ft) from the first one (see Figure 4). No mild reinforement was found at any loation. Welded pokets, onneting two adjaent beams, were positioned every 910 mm (3ft) at a depth of 76 mm (3 in) from top surfae. Conrete properties were evaluated using three ores taken from three different beams at the loation of the stem and an avarage onrete ylinder strength of f =34 N/mm 2 (f =5000 psi) was found and its modulus of elastiity was determined aording to ACI Setion provisions (see Table 1). The strands properties were assumed to be onventional 1861 MPa (270 ksi) strength and summarized in Table Speimens and Installation of Steel Reinfored Polymer A total of three double-t PC beams were tested (see Figure 5): beam DT-C is the ontrol beam, beam DT-1 represents the beam strengthened with one ply of SRP and DT-2U the one strengthened with 2 plies of SRP anhored with SRP U-wraps. The epoxy resin for both strengthened beams was SikaDur Resin 330. Table 2 reports the resin properties supplied by the manufaturer and verified by testing aording to ASTM standards by Huang et al. (2004). Figure 6a shows the mixing prior to installation. The hoie of the resin was based on onstrutability so that it ould be rolled - 4 -

5 onto the surfae for overhead appliations, while having enough onsisteny, even before uring, to be able to hold the weight of the steel tape during ure. The tape was medium density onsisting of 6.3 ords per m (12 WPI), with material properties defined in Table 3 (Huang et al. 2004). The typial stress-strain diagram for an impregnated medium density tape, tested following the ASTM D 3039 reommendations, is reported in Figure 2 (properties based on steel net area). SRP was installed following the reomendations of ACI 440.2R-02 (ACI 440) provisions for FRP materials. The sequene of installation steps is reported in Figure 6. The bottom stem of the double-t beams was first abrasive-blasted to ensure proper bonding of the SRP system. With the surfae roughened and leaned, the first layer of epoxy was diretly applied (see Figure 6b), without primer oating. The steel tape was ut to length of 4.57 m (15 ft) and width of 102 mm (4 in), overing the bottom of the stem length and width entirely. A rib-roller was then utilized to press onto the tape to ensure epoxy impregnation and enapsulation of eah ord and allow exess resin to be squeezed out. The exess resin was spread with a putty-knife to reate an even surfae (see Figure 6) and a syntheti srim was applied to avoid any dripping of the resin (see Figure 6d). For the two ply appliation, one the first ply was in plae and the exess resin leveled, the seond ply was installed, following an idential proedure. This time the ply started 152 mm (6 in) away from the terminations of the first ply, making it 4.27 m (14 ft) long. To provide a mehanial anorage for the two longitudinal plies, an SRP U-wrap 914 mm (3 ft) wide was installed at both ends of the stems (see Figure 6e). Due to the stiffness of the steel tape, pre-forming is done with a standard sheet metal bender before installation. For this reason, the U-wrap was obtained by overlapping two L-shaped wraps

6 2.3 Test Setup and Instrumentation The experimental setup is shown in Figure 7a and Figure 7b. The beams were tested under simply supported onditions and subjet to a single onentrated load spread over both stems at mid-span, that is, 3-point bending at mid-span (see Figure 7). All three tests were onduted using a lose-loop load onfiguration, where no external reation is required. The load was applied in yles by one hydrauli jak of 890 kn (200 kip) apaity onneted to a hand-pump. The load was transferred to the PC beam in two points through one spreader steel beam (see Figure 7b). The reverse-t PC- Ledger beams, on whih the double-t beam rests, supplied the reation. As the hydrauli jak extended, it pulled on the high-strength steel bars, whih lifted the reation bailey-truss below. The reation truss was built with three bailey-truss frames 6.09 m (20 ft) long assembled as per manufaturer s speifiations (Mabey Bridge and Shore, Baltimore, MD), and properly designed to arry the test load (see Figure 7a). Plywood was plaed at eah ontat point to protet the onrete. The load was measured using a 890 kn (200 kip) load ell plaed on top of the jak (see Figure 7). The preparation work onsisted of drilling one hole of small diameter (~50 mm (2 in)) neessary for passing the high-strength steel bar through the flange of the double-t PC beam and isolating eah test speimen from the adjaent beams originally joined by the welded-pokets. An eletroni data aquisition system (see Figure 8a) reorded data from four linear variable differential transduers (LVDTs) and two eletrial strain-gages applied to the SRP in beams DT-1 and DT-2U. Two LVDTs were plaed at mid-span (see Figure 8b), and the remaining two LVDTs, were plaed under the reverse-t ledger beams to verify - 6 -

7 potential support settlements. Strain gages were installed at mid-span on the bottom flange of the two strengthened double-t beams, diretly onto the SRP material. 2.4 On-Site Safety Safety proedures were adopted during the performane of the tests. The parking garage areas affeted by eah test were fened and no one allowed within suh areas. Shoring was provided and designed to arry the weight of the beam tested (multiplied by a safety fator equal to 2.0 to aount for impat) and the additional weight of the testing equipment. Shoring was not in diret ontat with the beam stems to allow unobstruted defletion. 3 RESULTS AND DISCUSSION All beams failed in flexure and had a similar behavior up to the raking load. Beam DT-C failed due to frature of the lowest tendon. In beam DT-1, sine the SRP ply was not mehanially anhored, failure was ditated by peeling off of the ply from eah stem almost simultanuously. Beam DT-2U, strengthened with two anhored plies per stem, failed due to rupture of the lower tendon. Table 4 reports the test results. In beam DT-C flexural raks were onentrated in the mid-span region where the point load was applied. As soon as raking ourred, sine no mild reinforement was present and tendons were plaed far away from the bottom of the stem, raks developed throughout the entire stem. In beams DT-1 and DT-2U a similar behavior ourred with the differene that the presene of the SRP allowed the formation of additional flexural raks (see Figure 9). In beam DT-1 the SRP laminate started debonding at mid-span initiated by the widening of mid-span raks (see Figure 9a) and then progressed towards the supports (see Figure 9b). Complete detahment of the - 7 -

8 laminate ourred at one end of the beam with part of the onrete substrate attahed to the laminate, denoting a good interfae bond between the onrete and the SRP. In beam DT-2, SRP ould not ompletely peel off due to the presene of U-wraps. Delamination propagated from mid-span towards the supports similarly to Beam DT-1, until rupture of the lower tendon ourred, whih was immediately followed by SRP rupture exatly at the loation where the SRP U-wrap started. No shear raks were noted on any of the three beams. Figure 10 through Figure 12 shows the Load-vs-mid-pan Defletion urves for all three beams. The apaities of beams DT-1 and DT-2U inreased by approximately 12 and 26% with respet to the ontrol speimen DT-C. Figure 13 and Figure 14 report the Load-vs-Mid-Span Strain responses for beams DT-1 and DT-2U. Two distint phases, pre- and post-raking, haraterize the behavior of eah speimen. Up to raking there was pratially no strain in the SRP. Past the raking load, the presene of the SRP signifiantly affeted performane. Beam DT-C (see Figure 10) raked at a onsiderably lower load (250.8 kn (56.4 kip)), with respet to the other two strengthened speimens. The ourrene of the first rak, at mid-span only, orresponds to the load drop in the Load-vs-Displaement plot. Upon unloading, the beam remained almost perfetly elasti, reovering almost all defletion. At the third loading yle the lower strand suddenly fratured at a load of kn (77.4 kip). For beams DT-1 and DT-2U the raking load inreased of approximately 23% and 17% with respet to DT-C (see Figure 11 and Figure 12). The lower raking load for DT-2U may be explained by the fat that the beam had been previously repaired by means of epoxy injetion

9 Beam DT-1 reahed the peak load of 387 kn (87 kip) and held it onstant with inreasing defletion, while SRP progressively delaminated from mid-span towards the support. The strain profile reported in Figure 13 shows how the SRP was not engaged until raking ourred and as soon as the first rak opened at mid-span, the SRP bridged the rak and strain suddenly inreased to approximately 5500 m e (strain-gauge was plaed at mid span where the first rak ourred). The maximum strain reorded in the steel tape (12300 m e), prior to omplete peeling-off, shows how the material was well bonded to the onrete substrate. The dutility reported in the load-defletion urve, is the result of the slow peeling propagation rather than to yielding of the reinforing steel tape itself. Figure 2 shows in fat an almost elasti behavior till rupture of the SRP laminate. Past the raking load (Figure 12), beam DT-2U behaved almost linearly, although with a lower stiffness, until it reahed the load of 400 kn (90 kip) then, stiffness dereased signifiantly till the peak load was reahed. When the load of 434 kn (97.6 kip) was reahed, the lower tendon ruptured and a sudden drop in the load-defletion urve was reorded. The strain in the SRP material when the tendon ruptured was 6400 m e. At this stage, one the lower tendon ruptured, the SRP laminate was ompletely debonded exept for the region where anhoring was provided by the U-wraps. The test was ontinued untill suddenly the SRP laminate ruptured at 388 kn (87.2 kip). The strain reorded in the SRP laminate at failure was me, similar to the values attained in beam DT

10 4 ANALYTICAL APPROACH The onventional analytial approah outlined in ACI (2002) was used in onjuntion with ACI 440.2R-02 (2002) provisions to ompute the ultimate apaity of the beams without onsidering safety fators normally inluded in design. The SRP behavior was approximated as illustrated in Figure 2 (Huang et al. 2004) and the values used for f fu_srp,e fu_srp,and E SRP are reported in Table 3. The moment apaity M n, inlusive of the SRP strengthening, an then be omputed following ACI 440 provisions, using the appropriate equations to ompute g and b 1 (Todeshini et al. 1998) so that a retangular stress blok suitable for the partiular level of strain in the onrete ould be used, as (see also Figure 15): M n _ SRP ApB f pb d β pb ApT f pt d β pt ASRP f fe _SRP h β = (1) where the first two terms of the equation represent the existing prestress steel reinforement, with the index pb and pt indiating the ontribution of the bottom and top tendons, and assuming the following: total losses in the prestress tendons = 30% in-plae moment, prior to testing, only due to beam self weight. The third term, of Eq.(1), represents the SRP ontribution with the following assumptions being made: the area of SRP is omputed as: ( ) A = n t w (2) SRP SRP SRP where the n represents the number of plies, t SRP the thikness of one ply (obtained by multiplying the area of one ord by the number of ords in the installed ply and dividing by the width of the ply) and w SRP the width of the ply;

11 the k m, bond redution fator used to ompute the effetive stress in the SRP, has been omputed aording to ACI 440 provisions, using SI units, as follow: 1 nesrptsrp κm = ε fu _ SRP 360, 000 (3) being nesrptsrp 180, 000 for both beams DT-1 and DT-2U. Table 5 reports on the analytial results. As reported in the seond olumn, none of the tested beams reahed the ultimate ompression strain of e u = Beam DT-C was found to fail in tension due to rupture of the lower tendon, as found experimentally, with a strain in the lower tendon of e pb =0.023 and the ultimate failure load was found to be less than the experimental by only 2%. Both Beam DT-1 and DT-2U were found to fail due to attainment of the effetive SRP strain value, that were and for beams DT-1 and DT-2U respetively. Even though the experimental and analytial apaity values are very lose, a onvining and exhaustive alibration of the k m fator and the orresponding delamination need to be undertaken in order to validate these findings. 5 CONCLUSIONS The following onlusions may be drawn from this experimental program: SRP omposite materials have shown to be effetive in inreasing the flexural apaity of the double-t PC beams. End anhors in the form of SRP U-wraps have shown to be effetive by preventing a omplete detahment, one debonding has ourred throughout the onrete-srp interfae

12 SRP is similar to FRP in terms of ease of installation, although self weight should not be ignored when seleting the resin system in overhead appliations. Epoxy resin behaved well in bonding the steel tape to the onrete substrate. The analytial validation, using ACI 440 provisions has proven to be effetive in antiipating the ultimate apaity, although further investigation in a ontrolled laboratory environment is need to properly alibrate the bond fator k m. 6 ACKNOWLEDGMENTS This researh study was sponsored by the National Siene Foundation Industry/University Cooperative Researh Center on Repair of Buildings and Bridges (RB 2 C) at the University of Missouri Rolla. Hardwire LLC., Poomoke City, MD, provided the steel tapes and Sika Corporation, Lyndhurst, NJ, the resins for the installation. The City of Bloomington, IN, provided the opportunity for testing the struture

13 REFERENCES ACI , 2002: Building Code Requirements for Strutural Conrete and Commentary (318R-02), Published by the Amerian Conrete Institute, Farmington Hills, MI, pp ACI 440.2R-02, 2002: Guide for the Design and Constrution of Externally Bonded FRP Systems for Strengthening Conrete Strutures, Published by the Amerian Conrete Institute, Farmington Hills, MI, pp. 45. ASTM D 3039, 2002: Test Method for Tensile Properties of Fiber Resin Composites Published by the Amerian Soiety for Testing and Materials, West Conshohoken, PA, pp. 13. Bisby, L.A., Kodur, V.K.R., and Green, M.F. Performane in Fire of FRP-Confined Reinfored Conrete Columns, Fourth International Conferene on Advaned Composite Materials in Bridges and Strutures - ACMBS-IV July 20-23, 2004 The Westin Hotel, Calgary, Alberta, Canada. FIB Bullettin 14 (2001). Design and use of externally bonded fibre reinfored polymer reinforement (FRP EBR) for reinfored onrete strutures, by 'EBR' working party of FIB TG 9.3, July 2001, 138 pp. Hardwire LLC, 2002, What is Hardwire, Poomoke City, MD. Huang, X., Birman, V., Nanni, A., and Tunis, G., Properties and potential for appliation of steel reinfored polymer and steel reinfored grout omposites, Composites, Part B: Engineering, Volume 36, Issue 1, January 2004, Pages Mabey Bridge & Shore, In., Baltimore, MD. Nawy, G. E., Prestressed Conrete, Prentie Hall, Upper Saddle River, NJ, 2002, 789 pp

14 PCI (1999): PCI Design Handbook: Preast and Prestressed Conrete, Published by the Preast/ Prestressed Conrete Institute, Chiago, IL. Rizkalla, S. and Nanni, A. (2003) Field Appliations of FRP Reinforement: Case Studies ACI Speial Publiation 215, Published by the Amerian Conrete Institute, Farmington Hills, MI. Seible, F.; Priestley, M. J. N.; Hegemier, G. A.; and Innamorato, D., 1997, Seismi Retrofit of RC Columns with Continuous Carbon Fiber Jakets, Journal of Composites for Constrution, No. 1, pp Sika, 2004, Sikadur 330, Lyndhurst, NJ. Todeshini, C., Bianhini, A, and Kesler, C. (1982) "Behavior of Conrete Columns Reinfored with High Strength Steels." ACI Journal, Proeedings, Vol. 61, No. 6, pp , November-Deember Williams, B.K., Kodur, V.K.R., Bisby, L.A., and Green, M.F. The Performane of FRP-Strengthened Conrete Slabs in Fire, Fourth International Conferene on Advaned Composite Materials in Bridges and Strutures - ACMBS-IV July 20-23, 2004 The Westin Hotel, Calgary, Alberta, Canada

15 LIST OF TABLES Table 1 - Properties of Constrution Materials Table 2 - Mehanial Properties of Epoxy Resin Table 3 - Material Properties of Steel Tape Table 4 Beam Test Results Table 5 Analytial Beam Results at Ultimate

16 Table 1 - Properties of Constrution Materials Material Cylinder Compressive Strength, MPa (psi) Yield Strength MPa (ksi) Rupture Strength MPa (ksi) Elasti modulus (2) MPa (ksi) Conrete (1) 27, (5,000) - - (4,000) ,000 Steel - (230) (270) (29,000) (1) Average of 3 speimens [76.2 mm mm (3 in 6 in) ylinders]. 7 wire Tendon Cross Setion, Ap mm 2 (in 2 ) (0.174) (2) ' E = 4700 f ACI 318 Setion Matrix Table 2 - Mehanial Properties of Epoxy Resin Tensile Strength, MPa (psi) Ultimate Rupture Strain ε fu (%) Tensile Modulus of Elastiity, MPa (ksi) SikaDur 330 (1) 30 (4350) (551) (1) Values provided by the manufaturer (Sika, 2004) Cord Coating Cord Area per 12 Wires, mm 2 (in 2 ) Table 3 - Material Properties of Steel Tape Cords per m (in) Nominal Thikness (1), t SRP mm (in) Tensile Strength f fu_srp, MPa (ksi) Ultimate Rupture Strain ε fu_srp (mm/mm) Tensile Modulus of Elastiity, GPa (ksi) Brass 3070 (447) (26700) ( ) (9.5) (0.0058) (1) The nominal thikness has been omputed assuming the area of eah ord and ounting the number of ords in eah ply, reported in ords per m Beam Failure load kn (kip) Table 4 Beam Test Results Load SRP Strain Capaity at Failure Inrease ε SRP ( m e) Failure Mode DT-C 344 (77.4) 1 - Rupture of Lower Tendon DT (87) SRP Delamination DT-2U 434 (97.6) Rupture of Lower Tendon

17 Beam Conrete Strain ε DT-C DT DT-2U Neutral Axis Position mm (in) (0.83) 34.8 (1.37) 37.3 (1.47) Effetive Stress in the Tendons after Losses MPa (ksi) 1303 (189) Top Tendon Strain ε pb Table 5 Analytial Beam Results at Ultimate Bottom Tendon Strain ε pb κ m Bond Fator Existing Substrate Strain ε bi (1) SRP Strain ε SRP M n kn-m (kip-ft) N/A * N/A * N/A * 393 (290) * N/A = Not Appliable (1) Determined from an elasti analysis onsidering only the self weight of the beams, at time of SRP installation 454 (335) 513 (380) P u kn (kip) 337 (75.8) 389 (87.5) 442 (99.4) Failure Mode Attainment of Limit Tendon Strain P u-experimental / P u-analytial 0.98 Attainment of SRP 1.00 Effetive Strain Limit

18 LIST OF FIGURES Figure 1 Example of Steel Cord and Tape Figure 2 SRP Laminate Stress vs Strain Behavior Figure 3 Bloomington Parking Garage Figure 4 Double-T Geometry Details (SI units 1 mm = in) Figure 5 Test Beams (SI units 1 mm = in) Figure 6 SRP Installation Proedure Figure 7 Test Set Up Figure 8 Installed Instrumentation Figure 9 Failure Mehanisms in Strengthened Beams Figure 10 Load vs Mid-Span Defletion (Beam DT-C) Figure 11 Load vs Mid-Span Defletion (Beam DT-1) Figure 12 Load vs Mid-Span Defletion (Beam DT-2U) Figure 13 Load vs Mid-Span Strain (Beam DT-1) Figure 14 Load vs Mid-Span Strain (Beam DT-2U) Figure 15 Strain and Stress Distribution Aross Beam Depth

19 a) Steel Cord with Wires Wrapped by One Wire b) Tape with Cords Held Together by Polyester and Copper Knits Figure 1 Example of Steel Cord and Tape Average Wire Stress (ksi) f fu_srp = 3070 MPa (447 ksi) Experimental Stress-Strain Curve Design Approximation E SRP = 184 GPa (26.7 msi) e fu_srp =16700 me Average Wire Strain (me) Average Wire Stress (MPa) Figure 2 SRP Laminate Stress vs Strain Behavior

20 a) Side View of Parking Garage b) Top View of the Dek ) Bottom View of the Dek Figure 3 Bloomington Parking Garage Topping of ast in-plae onrete 51 Strand Loations Figure 4 Double-T Geometry Details (SI units 1 mm = in)

21 DT-C DT-2U DT a) Saw-Cut Marks on Top of Dek b) Plan View Neoprene Pad PC Reversed-T Beam DT ply ply, 102-mm wide ) Beam Strengthened with 1 ply (DT-1) DT-2U U-wrap 1 ply, L-Shape 2 plies, 102-mm wide 2 plies, 102 mm wide nd ply st ply ply of U-wrap, 914 mm wide d) Beam Strengthened with 2 plies + U-wrap (DT-2U) Detail of Ply Arrangement Figure 5 Test Beams (SI units 1 mm = in)

22 a) Mixing of the Epoxy Resin b) Appliation of Longitudinal Ply ) Squeezing Out the Resin Exess d) Appliation of Srim on Longitudinal Ply e) Appliation of U-Wraps f) Appliation of Epoxy on U-Wrap Figure 6 SRP Installation Proedure

23 Spreader Steel beam 1)Steel plate 2)Plywood Load Cell Hydrauli Jak Saw ut (depth 3 in) Dywidag bar a) Bottom View Bailey Truss Steel plate Dywidag nut Crippling b) Top View ) Cross Setion at Mid-Span Figure 7 Test Set Up a) Data Aquisition System b) LVDT Loations Figure 8 Installed Instrumentation

24 a) Crak Propagation Prior to Complete Peeling b) Debonding Propagation from Mid-Span Beam DT-1 ) SRP Rupture d) Rupture of the Lower Tendon Beam DT-2U Figure 9 Failure Mehanisms in Strengthened Beams

25 100 Defletion (mm) Load, P (kip) Load, P (kn) Rupture of the Lower Strand Defletion (in) Figure 10 Load vs Mid-Span Defletion (Beam DT-C)

26 100 Defletion (mm) DT-C Failure Load Load, P (kip) Load, P (kn) SRP Delamination Defletion (in) Figure 11 Load vs Mid-Span Defletion (Beam DT-1)

27 100 Defletion (mm) DT-C Failure Load Load, P (kip) Load, P (kn) 20 Rupture of the Lower Strand Rupture of the SRP Laminate Defletion (in) Figure 12 Load vs Mid-Span Defletion (Beam DT-2U)

28 Load, P (kip) Load, P (kn) SRP Delamination Strain (me) 0 Figure 13 Load vs Mid-Span Strain (Beam DT-1)

29 Load, P (kip) Load, P (kn) 20 Rupture of the Lower Strand Rupture of the SRP Laminate Strain (me) 0 Figure 14 Load vs Mid-Span Strain (Beam DT-2U)

30 b = 1 a f' Conrete Strength neutral axis d d A pt Top Tendon pt f pt h A pb Bottom Tendon pb f pb A SRP SRP, e bi f SRP,e=ESRP SRP,e SRP Ply Strain Distribution Stress Distribution Figure 15 Strain and Stress Distribution Aross Beam Depth

31 NOTATION A SRP = n ( t w SRP SRP ) area of SRP reinforement [mm2 ] A pb = area of bottom steel tendon reinforement [mm 2 ] A pt = area of top steel tendon reinforement [mm 2 ] = depth of the neutral axis [mm] E SRP = f fu _ SRP ε fu _ SRP tensile modulus of elastiity of SRP [MPa] E = ' 4700 f tensile modulus of elastiity of onrete (ACI 318 Setion 8.5.1) [MPa] d pb = depth of bottom steel tendon [mm] d pt = depth of top steel tendon [mm] ' f = ultimate ompressive strength of onrete [MPa] f fe_srp = effetive stress in the SRP; stress level attained at setion failure [MPa] f fu_srp = f 3σ ultimate design tensile strength in the SRP [MPa] f fu _ SRP fu _ SRP mean ultimate tensile strength of SRP based upon a population of tests as per ASTM D 3039 [MPa] f pb = stress in bottom steel tendon at ultimate [MPa] f pt = stress in top steel tendon at ultimate [MPa] h = height of the ross setion [mm] t SRP = nominal thikness of one ply of SRP reinforement [mm] w SRP = width of one ply of SRP [mm] ε ' ε = = strain level in the onrete [mm/mm] 1 ultimate ompressive strain of onrete (Todeshini et al. 1998) [mm/mm].71f E ε fu_srp = ε fu _ SRP 3σ design rupture strain in the SRP [mm/mm] mean rupture strain of SRP based upon a population of tests as per ASTM D 3039 ε = fu _ SRP [mm/mm] 1 4[ ( ε ε ) tan ( ε ε )] ratio of the depth of the equivalent retangular stress β 1 = blok to the depth of the neutral axis (Todeshini et al. ( ε ε ) ln( 1+ ε ε ) 1998) γ = 0.9ln ε 1 + ε ε β1 ε ' 2 ' 2 k m = Bond dependent oeffiient for flexure ' multiplier on f to determine the intensity of an equivalent retangular stress distribution for onrete (Todeshini et al. 1998)