BOND OF REINFORCEMENT IN FIBRE REINFORCED CONCRETE

BOND OF REINFORCEMENT IN FIBRE REINFORCED CONCRETE Klaus Holschemacher, Dirk Weiße Leipzig University of Applied Sciences (HTWK Leipzig), Germany Abstract This paper outlines the experimental programme and its results on the bond performance of conventional rebars in steel fibre reinforced normal strength concrete (NSC) and high strength concrete (HSC). The two main tasks were the time development of the bond properties and the influence of fibres on bond. So, the hardened concrete properties and the bond behaviour were experimentally investigated at 3 and 28 days and collated with plain concrete. One of the main conclusions is that steel fibres reduce the risk of splitting of the concrete cover. Also the ductility of the bond is enlarged. 1. Introduction Steel fibres are successfully applied in concrete for a long time. Due to these fibres especially the following properties are positively influenced: enlargement of the solidity of the curve in compressive and uniaxial tensile loading as well as for flexural bending, in particular in the descending branch, increased toughness, improved material behaviour for repeated, shock and impact loading, increase of the shear and punching shear capacity, higher rotations/twists can be tolerated in torsion Furthermore the crack development is hindered, the crack width is reduced and the cracks are more evenly distributed. In this context it makes sense to look into the bond of rebars in steel fibre reinforced concrete, because the bond mainly affects the cracking behaviour of reinforced concrete members. 349

2. Bond in Steel Fibre Reinforced Concrete 2.1 General Behaviour The bond behaviour of deformed reinforcing bars in concrete is influenced by many parameters. These factors can be divided into four main groups: concrete properties such as compressive strength, tensile strength, fracture energy, reinforcement like bar diameter, relative rib area, loading regime e.g. monotonic or cyclic loading, loading rate and history, system parameter like concrete cover, confinement and rebar location. For the fibre reinforced concrete the first and the last mentioned group play the most important role (compressive and tensile strength as well as confinement). In plain concrete a radial tensile stress is present around the rebar (Fig. 1). If this radial tensile stress exceeds the tensile strength of the concrete, longitudinal cracking and therefore splitting failure occurs. In case of transversal reinforcement this crack development is hindered. If there are steel fibres present in the surrounding concrete, the initial cracking can be delayed and when the concrete is cracked, it is still possible to transfer forces over the crack plains. Therefore a brittle failure due to splitting of the concrete cover (c 1 in Fig. 1) could be avoided. radial longitudinal cracks transverse cracks c 1 c 1 Ø tensile stress in the surrounding concrete before cracking Figure 1: Longitudinal and transversal cracking in the concrete due to bond action after Tepfers [2] 350

2.2 State of the Art The bond behaviour of rebars in conventional concrete is very extensively presented in the literature [1] - [4]. Also the bond properties of reinforcement in high performance concrete as in high strength [5] - [6] and self compacting concrete were investigated [7] - [9]. In this context the influence of steel fibres has been examined to a lesser extent [10] - [13]. Ezeldin and Balaguru [10] recount a wide test programme about bond of reinforcing bars in steel fibre reinforced concrete. Variables were the compressive strength (35-76 N/mm²), the bar size (Ø 9, 16, 19 and 25 mm), the fibre content (0, 30, 45 and 60 kg/m³) and the fibre length (30-60 mm). The following conclusions were drawn: Due to the use of fibres the bond strength was increased in comparison to plain concrete. This effect was stronger for large rebar diameter than for smaller ones. With an increasing fibre content in the mix also the ultimate bond stresses slightly raised. The contribution of fibres in the ascending branch in negligible. The reached slip values at maximum bond stress increased for higher fibre contents, on the other side the fibre length has no significant influence in this context. The descending branch of the bond stress-slip-relationship was improved by a higher fibre content and longer fibres. Hota and Naaman [11] carried out comparative investigations on plain and fibre reinforced concrete. Using the same steel fibre (wire fibre with hooked ends) in all experiments, the fibre proportion and the concrete compressive strength differed. Pullout specimens made of plain concrete failed by splitting of the cover, whereas the fibre reinforced ones showed a ductile pull-out failure. The influence of different steel fibre contents (0; 0.5; 1 and 2 vol%) on splices has been shown by Hamad et al. [12]. 12 beams with different rebar diameter (Ø 20, 25 and 32 mm) were cast with splices in centre without transversal reinforcement. All beams made of plain concrete failed in a brittle manner after longitudinal cracking in the splice zone. With fibre reinforced concrete the failure was more ductile, but still longitudinal cracks were present. The load for at initial cracking was not changed by the fibres, but the bending stiffness increased. It is clearly evident by these summarised results, that steel fibres improve the local bond behaviour as well as other bond influenced aspects like splices. Nevertheless no design guideline allows at the moment the use of steel fibres in order to reduce anchorage length or transversal reinforcement in anchorage and splice zones. 3. Test Programme The programme was divided into two main series, series 1 for normal strength and series 2 for high strength concrete. In order to assess the time development of the bond properties in relation to the strength development of the concrete, the tests were performed at a concrete age of 3 and 28 days in the second series and only 28 days in the first one. Both types of concretes were tested with and without steel fibres. 351

3.1 Materials In each case local sand and gravel were used as aggregates in the mixes. Two different cement types were chosen, an ordinary Portland cement for the HSC and a Portland composite cement (Table 1); furthermore superplasticizer as admixture in both concrete types. Table 1: Mix Design Material Series 1 - NSC Series 2 - HSC sand (0-2 mm) 727 584 gravel (2-8, 8-16 mm) 1102 1083 cement 270 2) 425 1) fly ash [kg/m³] 50 - silica suspension - 81 water 177 145 superplasticizer 1.3 11.9 w/c-ratio - 0.66 0.34 1) CEM I 42,5 R; 2) CEM II A/S 32,5 R In series 1 (NSC) two different steel fibre types were applied, wire fibres with hooked ends and crimped cut sheet fibres. The wire fibres used in series 2 (HSC) were different lengths from those in series 1, only a length of 35 mm. The reason for the shorter fibres in HSC is the required good workability. All fibre details are listed in Table 2. Table 2: Fibres used in both series Series 1 - NSC Series 2 - HSC fibre type cut sheet fibres wire fibres wire fibres geometry elliptical cross section; crimped round cross section, hooked ends round cross section, hooked ends length l [mm] 50 50 35 width [mm] 2 - - diameter d [mm] - 0.80 0.55 aspect ratio l/d - 62.5 65.0 tensile strength [N/mm²] 520 > 1100 > 1150 number of fibres per kg 2000 5100 14500 352

3.2 Pull-out Specimens The bond behaviour was tested on pull-out specimens under monotonic loading after the RILEM- recommendation [14]. According to this reference the bond length between steel and concrete measures 5-times the rebar diameter. This requirement was fulfilled for NSC, but modified for HSC to 2.5-times the bar diameter (Table 3). Since the bond strength increases with the compressive strength, the bond length must be reduced for HSC. This change is necessary because of the high transferable bond stresses, otherwise the rebar would yield beyond the bond length. In the unbonded area, the rebars were encased with a plastic tube and sealed with a silicone material in order to avoid an unplanned force transfer between the reinforcement and the concrete (Fig. 2). The monotonic loading was applied under displacement control with a loading rate of 0.001 mm/s, in order to study the descending branch of the bond stress-slip-relationship. The slip between rebar and concrete was measured both at the loaded and unloaded end of the specimen with three rotation-symmetrically around the rebar fixed LVDT's each. For the analysis only the values from the unloaded end were used. The orientation of the rebar with respect to the casting direction was varied in both series 1 and 2, so in horizontal and vertical rebar position (Fig. 2). In addition two rebar diameters of 10 and 16 mm were used for HSC in series 2. Table 3 gives an overview of the experimental programme, whereby fibres were included. Of course also pull-out tests were carried out for both concrete types without fibres. Table 3: Experimental Programme Series 1 - NSC Series 2 - HSC cut sheet fibres wire fibres wire fibres wire fibres rebar diameter [mm] 10 10 10 16 bond length [mm] 50 50 25 40 cover size [mm] 45 45 45 72 rebar orientation relative to the casting direction vertical vertical and horizontal vertical and horizontal vertical and horizontal age at testing [days] 28 28 28 28 number of tested specimens 6 6 6 6 353

10/10 cm Direction of Casting 30 cm 5 cm 5 cm 5 cm Direction of Casting 10 mm, concentrically mounted in the specimen 16/16 cm 16 mm, concentrically mounted in the specimen Direction of loading Direction of Loading 5 cm 4 cm 12 cm 30 cm Figure 2: Examples of pull-out specimens with different orientations with respect to the casting direction and rebar diameters 4. Results and Discussion 4.1 Hardened Concrete Properties The measured values are shown in Table 4, which are a mean out of 3 values each. The cylinder compressive strength and the E-Modulus were tested on cylinders ( 150/ h = 300 mm), the cube compressive and the splitting tensile strength on cubes (150x150x150 mm³). Table 4: Hardened Concrete Properties after 28 days Material Property Series 1 - NSC Series 1 - HSC Cylinder compressive strength f c,cyl [N/mm²] without fibres cut sheet fibres wire fibres without fibres wire fibres 29 30 26 95 88 Cube compressive strength f c,cube [N/mm²] Splitting tensile strength f ct,sp [N/mm²] Flexural tensile strength f ct,fl [N/mm²] Modulus of elasticity E c [N/mm²] 32 30 32 100 103 3.0 2.9 3.0 6.9 7.0 4.0 3.9 3.4-9.9 25,300 26,300 28,400 40,600 38,500 354

For the determination of the flexural tensile strength path-controlled 4-point bending tests on prisms measuring 700x150x150 mm were conducted after the German recommendation for steel fibre reinforced concrete [15]. All specimens were cured under water until testing, except the prisms for the bending test that were cured sealed in plastic film. The concrete compressive strength is nearly not influenced by the steel fibres, nevertheless the flexural tensile strength for NSC is decreased. Unfortunately the values for HSC without fibres were not recorded due to problems with the measurement. 4.2 Bond Properties Normal Strength Concrete - Series 1 The performed pull-out tests are characterised by a relative soft bond behaviour after reaching the maximum bond stress. Independently from the fibre type and accordingly without fibres the main failure mode was pulling out of the rebar associated with a shearing-off of the concrete keys. Figure 3 shows the results of the pull-out tests for NSC with and without fibres in vertical rebar orientation. It can be seen that the bond properties are almost not influenced by both fibre types. 0,8 Relative Bond Stress b / f c,cyl without fibres - rebar Ø 10 0,6 wire fibre - rebar Ø 10 0,4 cut sheet fibre - rebar Ø 10 0,2 0,0 0,0 0,3 0,6 0,9 1,2 1,5 1,8 2,1 2,4 2,7 3,0 Slip s [mm] Figure 3: Relative bond stress-slip-relationships for the NSC - vertical rebar orientation, loaded in casting direction, rebar Ø 10 mm (each curve represents an average out of 3 pull-out specimens) For horizontal rebar orientation two specimens failed by splitting of the concrete cover (VB 2 and 3 in Fig. 4) in contrast to the above mentioned results (curves VB 1 and wire fibre - rebar 10 - vertical orientation in Fig. 4). In this case the action of the fibres becomes apparent. After cracking the released forces can be transferred over the crack plains. The bond stress falls rapidly, but due to the fibres the crack opening is hindered 355

and therefore forces still can be transferred. A brittle failure by splitting is prevented, the residual bond stress is kept 1/3 lower than for an uncracked concrete. Relative Bond Stress b / f c,cyl 0,8 wire fibre VB 2+3 - rebar Ø 10 - horizontal orientation wire fibre - rebar Ø 10 - vertical orientation 0,6 0,4 0,2 wire fibre VB 1 - rebar Ø 10 - horizontal orientation 0,0 0,0 0,3 0,6 0,9 1,2 1,5 1,8 2,1 2,4 2,7 3,0 Slip s [mm] Figure 4: Relative bond stress-slip-relationships for the NSC with wire fibres - horizontal rebar orientation, loaded 90 to casting direction, rebar Ø 10 mm High Strength Concrete - Series 2 The use of steel fibres in high strength concrete led to higher maximum bond stresses. Furthermore the displacements reached at ultimate bond stress were higher than those for the plain HSC (Fig. 5). In the area of the ascending branch of the bond stress-sliprelationship high local stresses at the ribs of the rebar cause damaging of the concrete and led finally to the shearing-off of the concrete keys. It can be believed that the fibres positively influence the redistribution of the stresses and therefore higher bond stresses can be transferred. This assumption is backed up by the fact of the low decrease of the bond stresses after reaching the maximum. It is noticeable that the bond stiffness in the ascending branch for the larger diameter (Ø 16 mm) is lower than for the smaller ones (Ø 10 mm). 356

Relative Bond Stress b / f c,cyl 0,8 wire fibre - rebar Ø 10 - horizontal orientation wire fibre - rebar Ø 10 - vertical orientation 0,6 0,4 without fibres - rebar Ø 10 - horizontal orientation wire fibre - rebar Ø 16 - vertical orientation 0,2 wire fibre - rebar Ø 16 - horizontal orientation 0,0 0,0 0,3 0,6 0,9 1,2 1,5 1,8 2,1 2,4 2,7 3,0 Slip s [mm] Figure 5: Relative bond stress-slip-relationships for the HSC with and without wire fibres vertical and horizontal rebar orientation, rebar Ø 10 and 16 mm (each curve represents an average out of 3 pull-out specimens) 5. Summary and Conclusions The use of steel fibres in concrete leads to a higher loading capacity in bending, shear and punching shear. With the new German Recommendation DBV-Merkblatt Steel Fibre Reinforced Concrete [15] it is possible to use steel fibre reinforced concrete in the structural design for the above mentioned ultimate limit states. Regarding the anchorage of rebars and for lapped splices no special design rules are included, only the transversal reinforcement can be omitted in certain situations. Based on the presented tests it could be shown, that the bond behaviour is positively influenced by the steel fibres. A brittle failure by splitting of the concrete cover was avoided in both normal and high strength concrete. Additionally for HSC the maximum bond strength was increased by about 10 to 15%. Furthermore bond stiffness was higher due to the steel fibres, so the ascending branch of the bond stress-slip-relationship was influenced. In contrast fibres do not have a significant effect on the ultimate bond stresses in normal strength concrete. Therefore for structural design, the contribution of steel fibres to the bond strength is only interesting for HSC. However, for defining possible design values further tests are certainly necessary. References 1. Bond of reinforcement in concrete. State-of-art report. fib-bulletin 10, Switzerland, 2000. 357

2. Tepfers, R.: 'A theory of bond applied to overlapped tensile reinforcement splices of deformed bars'. Report 73-2, Chalmers University of Technology, Göteborg, 1973. 3. Rehm, G.: 'Über die Grundlagen des Verbundes zwischen Stahl und Beton'. Deutscher Ausschuss für Stahlbeton, Booklet 138, 1961, (in German). 4. Martin, H.: 'Bond Performance of Ribbed Bars (Pull-out-Tests) Influence of Concrete Composition and Consistency'. Proceedings of the International Conference on Bond in Concrete, Paisley, Scotland 1982, pp. 289-299. 5. Kurz, W.: 'Ein mechanisches Modell zur Beschreibung des Verbundes zwischen Stahl und Beton'. Dissertation, Darmstadt 1997, (in German). 6. Esfahani, M. R. and Rangan, B. V.: 'Local Bond Strength of Reinforcement in Normal Strength and High-Strength Concrete (HSC) '. ACI Structural Journal, Sep.-Oct. 1998, pp. 96-106. 7. König, G.; Holschemacher, K.; Dehn, F. and Weiße, D.: 'Bond of Reinforcement in Self-Compacting Concrete (SCC) under monotonic and cyclic loading'. Proceedings of the 3rd International RILEM Symposium on Self-Compacting Concrete, Reykjavik, 2003, pp. 939-947. 8. Sonebi, M.; Bartos, P. J. M.: 'Hardened SCC and its bond with reinforcement'. Proceedings of the First International RILEM Symposium on Self-Compacting Concrete, Stockholm 1999, pp. 275-289. 9. Lorrain, M.; Daoud, A.: 'Bond in Self-Compacting Concrete'. Proceedings of the 3rd International Symposium on Bond in Concrete from research to standards. Budapest, 2002, pp. 529-536. 10. Ezeldin, A. S. and Balaguru, P. N.: 'Bond Behavior of Normal and High-Strength Fibre Reinforced Concrete'. ACI Materials Journal, Sep.-Oct. 1989, pp. 515-524. 11. Hota, S. and Naaman, A. E.: 'Bond Stress-Slip Response of Reinforcing Bars Embedded in FRC Matrices under Monotonic and Cyclic Loading'. ACI Structural Journal, Sep.-Oct. 1997, pp. 525-537. 12. Hamad, B.S.; Harajli, M. H. and Jumaa, G.: 'Effect of Fibre Reinforced on Bond Strength of Tension Lap Splices in High-Strength Concrete'. ACI Structural Journal, Sep.-Oct. 2001, pp. 638-647. 13. Dupont, D.; Vandewalle, L. and De Bonte, F.: 'Influence of Steel Fibres on Local Bond Behaviour'. Proceedings of the 3rd International Symposium on Bond in Concrete from research to standards. Budapest, 2002, pp. 783-790. 14. RILEM. 'Technical Recommendations for the Testing and Use of Construction Materials: RC 6, Bond Test for Reinforcement Steel. 2. Pull-out Test', 1970. 15. DBV-Merkblatt Stahlfaserbeton, Deutscher Beton- und Bautechnik-Verein e.v., Oct. 2001, German Recommendation for Steel Fibre Reinforced Concrete (in German). 358