ACI CODE VERIFICATION FOR FRP EXTERNALLY REINFORCED SLABS Usama Ebead Department of Civil Engineering, University of Sherbrooke 2500 boulevard de l’Université, Quebec, Canada, J1K 2R1 Hesham Marzouk Department of Civil Engineering, Memorial University of Newfoundland 300 Prince Philip Dr., St. John’s, Newfoundland, Canada, A1B 3X5 ABSTRACT An ACI code verification of FRP externally reinforced two-way slabs is introduced. An implementation of the ACI-318  and the ACI-440 F  is presented for the purpose of verification against experimental results. In the experimental work, two different types of FRP materials were evaluated; namely carbon FRP (CFRP) strips and glass FRP (GFRP) laminates. The externally reinforced or strengthened slabs had steel reinforcement ratios of 0.35% and 0.5%. Results show that the flexural capacity of two-way slabs can be increased to an average of 35.5% over that of the reference (unstrengthened) specimen. An increase of the initial stiffness is achieved; however, an apparent decrease in the overall ductility is evident. In addition, an average decrease in the values of the energy absorption of about 30% is observed. The estimated ultimate load capacity using the ACI code is in an accepted level of agreement with the experimental results. Keywords: concrete; FRP; external reinforcement, strengthening; reinforcement; two-way slabs. INTRODUCTION FRP materials have been used for strengthening reinforced concrete beams, columns and one-way slabs. The flexural capacity of concrete beams can be increased by bonding FRP sheets, strips or laminates to the tension side [3,4,5]. In addition, the shear strength of concrete beam can be increased by gluing FRP laminates to the concrete web at locations of high shear stresses [6,7]. The ease of handling FRP materials provides the means to the extension of their applications for strengthening other structural elements. A few research works has been conducted on FRP strengthening of reinforced concrete slabs especially two-way slabs. Some research works dealt with the strengthening of one-way slabs using FRP materials in which slabs were treated in a very similar way to beams [8,9]. Using FRP materials to enhance two-way slabs in flexure is very desirable from the applicability point of view due to the ease of handling and installing FRP materials. FRP materials are not subject to either corrosion or rust on the long term. A drawback of using FRP materials for flexural strengthening is the potential for brittle failure that results in a decrease of the overall ductility. In the following sections, a summary of an experimental program on two-way slabs strengthened using FRP is presented. Following that summary, ACI code verification is presented with a comparison between the experimental and the code implementation results. SUMMARY OF EXPERIMENTAL PROGRAM Materials The concrete mix was designed for an average target cylinder compressive strength of 35 MPa after 28 days. The mix proportion of 1.0 m3 is 1160 kg of gravel, 690 kg of sand, 350 kg of Portland cement and 175 litres of water. The actual cylinder compressive strengths ranged from 29 to 38 MPa as shown in Table 1. The steel reinforcement bars were CSA grade 400 deformed bars. The actual yield strength of the steel reinforcement ranged from 435 to 450 MPa. Two different diameters were used, namely 10 and 20mm for the slab and column, respectively. Unidirectional GFRP laminates and CFRP strips were used for strengthening. The Sika CarboDur was used for the CFRP strips and the glass fibre fabric, SikaWrap Hex 100G, was used for the GFRP laminates. The thickness of one layer of the GFRP laminates and CFRP strips was 1.0 mm and 1.2 mm, respectively. Two different types of two-component adhesive epoxy resins were used for each type of the FRPs as per the manufacturer’s specifications. Sikadur 30 and Sikadur Hex 300 epoxy resins were used for the CFRP strips and the GFRP laminates, respectively. The properties of FRP materials and the epoxy resins provided by the manufacturers are listed in Tables 2 and 3, respectively. Table 1: Compressive strength of concrete Specimen Title Cylinder Strengthening material Reinforcement Preload, compressive ratio kN strength, MPa Ref-0.35% 30.0 - 0.35% - Ref-0.5% 35.0 - 0.5% - GFRP-F-0.35% 29.0 GFRP laminates 0.35% 125 GFRP-F-0.5% 38.0 GFRP laminates 0.5% 165 CFRP-F-0.35% 35.0 CFRP strips 0.35% 125 CFRP-F-0.5% 34.0 CFRP strips 0.35% 165 Table 2: Properties of one layer of the FRP materials FRP Thickness Tensile E, Elongation Weight Fibre Fibre , mm strength, MPa GPa at break , , content density % g/m2 g/cm3 CFRP strips* 1.2 2800 170 >1.7 2240 68% 1.5 GRFP laminates** 1.0 600 26.1 2.24 913 50-80% 2.54 * Sika CarboDur ** Sika Wrap Table 3: Properties of the epoxy adhesive Property Epoxy for strips* Epoxy for laminates** Tensile strength, Mpa 24.8 72.4 Elongation at break, % 1.00 4.8 Elastic modulus, Gpa 4.5 3.1 *Sikadur 30 **Sikadur Hex 300 Test slabs The tested specimens were square with 1900- mm side length and 150- mm thicknesses. The test specimens were simply supported along the four edges with corners free to lift and were loaded through the column stub. A layout of the tested slabs is shown in Fig. 1 that shows also reinforcement details of the tested specimens. Two unstrengthened specimens were used as reference specimens, namely, Ref-0.35% and Ref-0.5% of reinforcement ratios of 0.35 and 0.5%, respectively. Specimens CFRP-F-0.35% and CFRP-F-0.5% had steel reinforcement ratios of 0.35 and 0.5%, respectively, and were strengthened using CFRP strips. Similarly, specimens GFRP-F-0.35% and GFRP-F-0.5% had reinforcement ratios 0.35 and 0.5%, respectively and were strengthened using GFRP laminates. A minimum concrete cover of 25 mm was maintained for all specimens at compression and tension sides. Column stubs were square of 250- mm side dimension and were located at the slab center. The column stubs were extended on the compression side to a distance 850- mm from the concrete surface to allow for the application of the load. Fig. 1: Details of a typical test specimen Test set up and instrumentation The specimens were tested using a large reaction steel frame. A 10-ton capacity crane was used to lift and install the specimens vertically inside the frame. Rubber pieces were placed between the back surface of the tested slabs and the supporting edges of the frame. A hydraulic actuator facing the specimen was used to apply a uniform load through the column stub. A load cell was used to measure the load using four calibrated electrical resistance strain gages fixed to the inner cylinder of the load cell. The actuator had a maximum load capacity of 700 kN and a maximum stroke of 150 mm. Linear Variable Displacement Transformers (LVDT’s) were built in the front actuator to measure the deflection of slabs. The loads were applied using displacement control to avoid the uncontrolled failure at the maximum loads. The displacement rate for the actuator was 0.25 mm/min. A displacement function of the ramp type was applied through computerized function genera tor. Eight-mm length electrical resistance strain gages having a resistance of 120 ± 0.3% and a gage factor equal to 2.070 ± 0.5% were used to measure the steel reinforcement strains at locations shown in Fig. 1. The LVDT’s and the electrical strain gages were connected through a master panel to a data acquisition system. The analog electrical signals of loads, deflections and steel strains were converted through the data acquisition system to digital signals and were stored in digital computer files. Load application and testing procedure The unstrengthened reference specimens; Ref-0.35% and Ref-0.5% were loaded through the column stub until failure to estimate the ultimate load carrying capacity. The ultimate load carrying capacity of the reference specimens was 250 and 330 kN, respectively. Fifty percent of the ultimate load carrying capacity of the reference specimens was used as an initial loading for the specimens prior to strengthening. Hence, the specimens with reinforcement ratios of 0.35 and 0.5% to be strengthened were loaded prior to strengthening with initial loads of 125 and 165 kN as initial loading. Fifty percent of the load represents a level of load on a building in field where strengthening may be required. The applied loads were completely released to represent a state of shoring two-way slabs in the field prior to strengthening. Afterwards, the specimens were removed from the loading frame for strengthening according to the strengthening procedure detailed below. After one week of curing, the specimens were relocated at the loading frame and were subjected to the load until failure. Strengthening procedure The concrete surface to be strengthened was roughened carefully using a vibrating hammer to improve the bond characteristics between concrete and CFRP strips and GFRP laminates. Dust and fine materials caused from the roughening process were removed carefully from the concrete surfaces. In addition, for CFRP strips, a special solvent was used to remove all grease, waxes, foreign partic les and other bond inhibiting materials from the bonded surface as specified by the manufacturer. The two-part epoxy resin was applied on the concrete surfaces and the strengthening materials. Afterwards, the FRP strengthening materials were bonded to the concrete surface according to the type of strengthening. The strengthening material was located at the tension side of the slab and was extended to a location 50 mm before the support. Two 300- mm width layers of GFRP laminates were bonded to the slab surface in both directions of specimens GFRP-F-0.35% and GFRP-F-0.5%. Specimens CFRP-F-0.35% and CFRP-F-0.5% were strengthened using three adjacent CFRP strips 100- mm width each so that the strengthened width is 300 mm. Additional transverse layers of CFRP strips were bonded at the end of the FRP materials to improve the end anchorage of the FRP strips or laminates with concrete surface. The anchorage layers were 100 mm wide and 500 mm long. Fig. 2 shows the strengthening details of the specimens. Additional FRP strips layer One layer of FRP Two layers of FRP strips at the tension laminates at the side tension side Fibres direction Fibres direction Fibres direction Fibres direction CFRP-F-0.5% GFRP-F-0.5% CFRP-F-0.35% GFRP-F-0.35% Fig. 2: The strengthening details of the specimens TEST RESULTS AND DISCUSSION Deformational prope rties and load capacity The load-deflection relationship was recorded using the data acquisition system. The slope of the load-deflection curve was higher than that of the corresponding reference specimens. Moreover, the average deflection at the ultimate load of the specimens was about 0.61 that of the corresponding reference specimens. In general, the strengthened specimens experienced smaller deformation compared to the corresponding reference specimens due to the effect of the FRP materials on the overall behaviour of the slabs. Fig.3 shows the load deflection relationship for the tested specimens. The ultimate load carrying capacity will be referred to as the load capacity. The strengthened specimens showed higher load capacity than that of the corresponding reference specimens. Specimens CFRP-F-0.35% and GFRP-F-0.35% showed an increase of 44.4 and 38%, respectively in the load capacity over that of the reference specimen, Ref-0.35%. Moreover, Specimens CFRP-F-0.5% and GFRP-F-0.5% showed an increase of 36.4 and 25.8 %, respectively in the load capacity over that of the reference specimen, Ref-0.5%. The load capacity of the corresponding reference specimens was influenced by the reinforcement ratio that is in accordance with earlier research work . The load capacity of specimen Ref-0.5% was 1.32 times that of specimen Ref-0.35%. The stiffness of a slab at any loading point is the slope of the load-deflection curve at that point. The initial stiffness, K, was evaluated numerically as the slope of the load-deflection curve within the first 5- mm deflection. This is an approximation made to avoid the misleading initial readings when there is a relaxation of the load actuator. The strengthened specimens showed higher initial stiffness over that of the reference specimens. The average initial stiffness of Specimens CFRP-F-0.35% and GFRP-F-0.35% was about 2.37 times that of the reference specimen, Ref-0.35%. Moreover, the average initial stiffness of Specimens CFRP-F-0.5% and GFRP-F-0.5% was about 1.99 times that of the reference specimen, Ref-0.5%. The energy absorption is the area under the load-deflection curve for a tested specimen. This area was evaluated numerically based on the available values of load and the corresponding values of deflection. At the maximum load, it was clearly noticed that the strengthening technique contributed to a decrease in the energy absorption of the strengthened specimens. An average decrease in the values of the energy absorption of about 30% for the strengthened specimens was observed. Values of deflection at the cracking load, cr , the ultimate load, Pu , the Deflection at the ultimate load, u , the initial stiffness, K, and energy absorption, for each slab are summarized in Table. 4. Fig. 3: Load deflection relationship of the tested slabs Table 4: Ultimate capacities and deflection characterestics of the tested slabs Title Deflection at Ultimate Deflection at Energy Stiffness, cracking load, load, ultimate load, Absorption’ cr , Pu , u, , K, mm KN mm kN.mm kN/mm Ref-0.35% 7.00 250 42.01 9346 8.42 Ref-0.5% 6.25 330 35.57 9445 12.54 CFRP-F-0.35% 7.25 361 18.08 7821 15.54 GFRP-F-0.35% 7.69 345 27.72 4597 24.42 CFRP-F-0.5% 6.03 450 21.03 6686 26.76 GFRP-F-0.5% 6.35 415 26.71 7475 23.15 Failure characte ristics Reference specimens, Ref-0.35% and Ref-0.5%, failure mode was classified as flexural-ductile. Flexural reinforcement yielded and the two specimens showed re latively large deflection values before reaching the ultimate load. Fig. 4 shows a typical flexural failure mode of GFRP and CFRP strengthened specimens after failure. It is evident that the FRP materials contributed to an increase of the capacity until t he bond between the FRP material and concrete failed. De-bonding cracks appeared at a late stage of loading that resulted in a separation of the strengthening materials. These cracks were located along the edges of the strengthening material length. This indicates that end anchorage severed to a certain extent preventing a premature bond failure at the cut-off end of FRP materials. Followed the appearance of these cracks, the specimens failed due to accelerated concrete flexural failure after the FRP de-bonded from the slabs without rupture of the FRP material. (a) GFRP strengthened specimen (b) CFRP strengthened specimen Fig. 4: Typical layout of GFRP and CFRP strengthened specimens at failure ACI CODE EVALUATION A simplified method for the code evaluation of the load carrying capacity for two-way slab strengthened with steel plates and bolts was introduced . This method is based on the analysis of two-way slabs recommended by Rankin and Long . This approach is based on the following formula: S Pflex 8 M b 0.172 . (1) l c Equation (1) is based on the virtual work done by the action of the yield lines. The value of M b in Equation (1) is the radial moment capacity of the slabs. For the original concrete specimens, the ACI 318-1995 is used to evaluate M b . For the strengthened FRP specimens, the ACI 318-1995 is used to evaluate M b . The same equation can be used for the analysis of two-way slab strengthened with FRP using approximated evaluations of some of the parameters of Equation (1). The contribution of the strengthening materials is taken into account when evaluating the radial moment capacity, M b . M b M b1 M b 2 , (2) where M b1 is the unstrengthened capacity of the slab. According to ACI 318-1995, M b1 is evaluated according to the following Equation: M b1 bd f y 1 0.59 2 , , f y , f yd d d , , (3) f c, where M b 2 is the contribution of the strengthening material and is evaluated according to the following equation assuming full bonds between FRP and concrete, a w M b 2 E FRP t FRP FRP h FRP . (4) 2 l Equation (4) is based on the FRP section analysis as recommended by ACI 440 F Repair and w Rehabilitation guidelines . The term FRP is introduced for the two-way slab equation to take into l account the ratio between the width of the strengthening material and the slab width, l. The factor is the strengthening efficiency factor and is taken as 0.75 for two-way slabs as recommended by Ebead . The strain in FRP strips layer can be evaluated as: h h FRP 1 cu s (5) d d The distance of the neutral axis from the top of slab a can be calculated as follows: d a 0.8 cu (6) cu y The test results indicated for CFRP strips, it could be assumed that at failure concrete reached the ultimate strain and the steel reached the yield strain. However, for the specimen strengthened with GFRP, the concrete strain reached the ultimate value and the steel strain at failure reached four times the yield strain. Once the FRP strain and the location of neutral axis are determined, the strengthened moment contribution to the slab can be evaluated from Equation 4. For the evaluation of the total load capacity, Equation 1 is used and replacing the value of c by the FRP width, wFRP , Hence in the case of FRP strengthening, S l w 0.172 Ptheo 8 M b (7) FRP The implementation of the prescribed analytical method showed a good agreement with the experimental results as shown in Table 5. Table 7: Comparison with code evaluation Specimen Pexp Ptheo Ptheo /Pexp CFRP–F–0.35% 361 332 0.92 GFRP–F–0.35% 345 323 0.94 CFRP–F–0.5% 450 420 0.93 GFRP–F–0.5% 415 411 0.99 CONCLUSION Strengthened two-way slabs using GFRP laminates and CFRP strips showed an average gain in the load capacity of about 40 and 31%, respectively over that of the reference specimens. In addition, the strengthened specimens showed a stiffer behaviour than that of the reference specimens. An ACI code verification of FRP externally reinforced two-way slabs is introduced. An implementation of the ACI- 318 and the ACI-440 F is presented for the purpose of verification against expe rimental results. The recommended theoretical analysis used for FRP strengthening of two-way slabs showed a good agreement with experimental test results. NOMENCLATURE a = distance from the top of slab to the neutral axis, mm c = side length of the square column, mm cequ = equivalent side length due to strengthening, mm d= distance from the compression face to the center of the tension reinforcement, mm d`= distance from the compression face to the center of the compression reinforcement, mm Ec = the modulus of elasticity of concrete, MPa EFRP = the modulus of elasticity of FRP materials, MPa f c`= Compressive strength of concrete, MPa f y = yield stress of the slab reinforcement, MPa h = Overall slab thickness, mm K = the initial stiffness of specimen, kN/mm l = side length of the square slab, mm lp = length of the strengthening steel plates, mm Mb = radial moment of resistance of the strengthened section, N.mm/mm Mb1 = radial moment of resistance of the unstrengthened section, N.mm/mm Pcr = first crack load of a slab before strengthening, kN Pflex = flexural load carrying capacity, kN Pu = ultimate load of a specimen, kN t FRP = total thickness of FRP material, mm wFRP = the width of FRP materials, mm cr = the deflection at a slab center at the first crack load, mm u = the deflection at a slab center at the ultimate load, mm = the energy absorption of a specimen, kN/mm tension reinforcement ratio of the slab compression reinforcement ratio of the slab = Effective strengthening width coefficient FRP the strain in FRP material REFERENCES 1- ACI-440F. (2002), “Repair and Rehabilitation – Guide for the Design and Construction of Externally Bonded FRP Systems for Strengthening Concrete Structures, Amer ican Concrete Institute, Farmington Hills, Mich. 2- ACI-318. (1999), “Building Code Requirements for Structural Concrete,” Detroit. USA. 3- Al-Sulaimani, J., Sherif, A., Basunbul, I. A., Blunch, M. H. and Ghalab, B. N. (1994), “Shear Repair for Concrete by Fibreglass Plate Bonding,” ACI Structural Journal, Vol. 92, No. 3, pp. 458-464. 4- Ritchie, P. A., Thomas, D. A., Lu, L. and Connelly, M. (1991), “External Reinforcement of Concrete Beams Using Fibre Reinforced Plastics,” ACI Structural Journal, Vol. 88, No. 6, pp. 490-500. 5- Chaallal, O., Nollet, M. and Perraton, D. (1998), “Strengthening of Reinforced Concrete Beams with Externally Bonded Reinforced Plastic Plates: Design Guidelines for Shear and Flexure,” Canadian Journal for Civil Engineering, Vol. 25, pp. 692-704. 6- Triantafillou, T. C. (1998), “Shear Strengthening of Reinforced Concrete Beams Using Epoxy- Bonded FRP Composites,” ACI Structural Journal, Vol. 95, No. 2, pp. 107-115. 7- Norris, T. and Saadatmanesh. (1997), “Shear and Flexural-Strengthening of R/C Beams with Carbon Fibre Sheets,” Journal of Structural Engineering, Vol. 123, No. 7, pp. 903-911. 8- Karbhari, V. M., Seible, F., Seim, W., Vasquez, A. (1994), “Strengthening of Concrete Slabs,” In Proceedings, Infrastructure: A New Materials and Methods of Repair, Proceedings of the Third Materials Engineering Conference, pp. 1163-1172, San Diego, CA, USA. 9- Kikukawa, K, Mutoh, K., Ohya, H., Ohyama, Y. and Tanaka, H. (1998), “Flexural Reinforcement of Concrete Floor Slabs by Carbon Fibre Textiles,” Composite Interfaces, Vol. 5, No. 5, pp. 469-478. 10- Marzouk, H. and Hussein, A. (1991), “Experimental investigation on The Behaviour of High Strength Concrete Slabs,” ACI Structural Journal, Vol. 88, No. 6, pp. 701-713. 11- Ebead, U. and Marzouk, H (2002), “Strengthening of Two-Way Slabs Using Steel Plates” ACI Structural Journal, Vol. 99, No. 1, pp. 26-31. 12- Rankin, G. I. B. and Long, A. E. (1987), “Predictiong the Punching Strength of Conventional Slab - Column Specimen,” Proceedings, Part1: Design and Construction, Vol. 82, The Institution of Civil Engineers. pp. 327-346. 13- Ebead, U. (2002), “Strengthening of Reinforced Concrete Two-Way Slabs,” PhD. Thesis, Memorial University of Newfoundland, St. John’s, NF, Canada.