Studies on the pull-out strength of ribbed bars in high-strength concrete

The transfer of forces from reinforcing bars to surrounding concrete in reinforced concrete (RC) is influenced by many parameters. Several efforts were made to understand the influence of bond on global behaviour of RC members. However, the information on bond strength of high strength concrete (HSC) is lacking. An attempt was made to study the influence of various parameters on bond such as bar diameter, strength of concrete, lateral confinement and embedment length. The bond lengths were 50mm and 150mm with different bar diameters, strength of concrete and type of confinement. The bar diameters were 16mm and 20mm. The bars were embedded in concrete without confinement and with confinement using spirals and ties. The casting was done keeping the bars in the horizontal position. The anchorage bond specimens were tested using displacement control system and the slip of the bars was controlled at a rate of 1.51mm/minute (0.025mm/second). The bond stress-slip response was studied by varying the variables. As the strength of concrete increases the slip at failure decreases in the descending branch. With smaller bond length, the bond stress was found to be higher. The bond strength was found to decrease as the bar diameter increased. Splitting failure was observed in unconfined specimens, whereas pullout failure in confined specimens. The ultimate bond strength ranges between 10.8 MPa and 19 MPa with spiral confinement, whereas it ranges between 9.2 to 16 MPa with tied reinforcement. The ductility was found to increase with spiral reinforcement. and beam width, end anchorage, flexural bond and anchorage bond were reported by Ferguson et al. (1966). Lutz and Gergely (1967) studied the action of bond forces and the associated slip and cracking using rebars with different surface properties. The slip was found to be due primarily to the relative movement between concrete and steel along the surface of the ribs and also due to crushing of mortar. Goto (1971) studied the primary and secondary cracking by injecting ink around the deformed rebars in axially loaded tests. Nilson (1972) estimated the bond stress from the slope of the steel strain curve. The strain in concrete and steel was measured internally and the bond slip was calculated from the displacement functions obtained by numerical integration of strains. Jiang et al. (1984) developed new test method by cutting the reinforcing bars into two halves and placing in two opposite sides of the cross section to study the local slip, secondary cracking and strain distribution in concrete surrounding the interface. A simple one-dimensional analysis predicts the stresses in steel and concrete, local bondslip, tensile stiffening and total elongation of the reinforcing bar. Ueda et al. (1988) studied the beam bar anchorage in exterior beam-column joints. A model has been proposed to predict the load-lead end deformation and anchorage length of rebars extended from beams into exterior columns and subjected to large inelastic loadings. Effects of anchored bar diameter, confinement of joint and compressive strength of concrete on the hook behaviour in exterior beam-column joints have been studied (Soroushian, 1988)). An analytical model has been developed for predicting the overall pullout behaviour of rebars, which has been recognised by ACI–318-83 for development of standard hooks in tension. Soroushian and Choi (1989) reported on local bond strength of deformed bars with different diameters in confined concrete. The bond strength decreases as the bar diameter increases. Soroushian et al. (1991) studied the influence of strength of concrete with different confinements. Confinement influences local bond of deformed bars. The ultimate bond strength increases as square root of concrete compressive strength. Abrishami and Mitchell (1992) formulated a new testing technique to simulate uniform bond stress distribution along a rebar to determine bond stressslip response. Malvar (1992) tested specimens with varying confining pressure using confining rings with rebar ribs normal and inclined to the surface and obtained consistent bond-slip response over a short embedded length. Mathematical model for bond-slip behaviour of a reinforcing steel bar embedded in concrete subjected to cyclic loading was reported by Yankelevsky et al. (1992). Bortolotti (2003) proposed models to predict the tensile strength of concrete from pullout load. The confinement improved the bond strength slightly but ductility was improved significantly (Harajli et al. 2004). Somyaji et al. (1981) and Jiang et al. (1984) conducted several experimental and theoretical studies on bond in NSC. The secondary cracks as well as the distribution of strain in concrete in the vicinity of rebar have been studied. Darwin et al. (1996) reported development length criteria for conventional and high relative rib area of reinforcement. On the basis of a statistically based expression, the development length of reinforcement and splice strength in concrete for compressive strength varying between 17 and 110MPa with and without confinement have been investigated. The effects of cover, spacing, development/spliced reinforcement were incorporated in design equation. The effects of concrete compressive strength, splice length and casting position on the bond strength of rebars have been studied (Azizinamini et al. 1993; 1999a)). Increasing the development length in HSC in tension does not seem to increase the bond strength of deformed rebars, when concrete cover is small. Concrete crushing occurred in front of the ribs in NSC, whereas there was no indication of concrete crushing in front of the ribs in HSC with the first few ribs being more active. In HSC with small cover, failure occurred due to splitting of concrete prior to achieving uniform load distribution Azizinamini et al. (1993). Azizinamini et al. (1999b) in another study reported that when calculating the development length in HSC for tension splice, a minimum number of stirrups should be provided over the splice region. Statically based on the experimental data an expression has been proposed to calculate the extra number of stirrups required. Eligehausen et al. (1983) reported comprehensive study on the effect of bar diameter embedded in NSC. The maximum bond capacity decreases slightly with increasing bar diameter. The frictional bond resistance was not influenced by the bar diameter, lug spacing or relative rib area. Larrard et al. (1993) investigated the effect of bar diameter on bond strength. The bond strength increases with tensile strength of concrete at a higher rate with smaller bar diameters. A parameter which accounts for the ratios of side cover and bottom faces, and spacing of the spliced bars was introduced. CEB-FIB report (2000) presented a general description of the local bond law for tensile forces. Six main stages have been recognized in local bond stress-slip response. Goto (1971) carried out tests to clarify the propagation of different types of cracks around the tensile reinforcing bars. The internal cracks develop around the reinforcing bars in concrete cylinders as shown in Figure 1. The inclination of internal cracks and the direction of compressive forces on the bar ribs vary between 45 and 80. Figure 1. Internal cracks around the reinforcing bar embedded in concrete (Goto, 1971). Tepfers (1973) showed the radial components of bond forces balanced against tensile rings in concrete in Figure 2, using a two dimensional finite element analysis (FEA). The angle “α” is 45 degrees along a perimeter touching the ribs of reinforcing bars independent of rib face angle. Figure 2. View of tensile ring (Tepfers, 1973). Three mechanisms for bond resistance i.e. (i) chemical adhesion, (ii) friction, and (iii) mechanical interaction between concrete and deformed bars are responsible (Lutz and Gergely, 1967). According to Rehm (1961), and Lutz and Gergely (1967), slip of deformed bars occurs due to (i) splitting of concrete by wedging action, and (ii) crushing of concrete in front of the ribs. For the face angles between 40 and 105 degrees, the slip seems to be not influenced. However, the slip is mostly due to crushing of concrete in front of the ribs. This in effect produces a rib with face angles of 30 to 40 degrees (Lutz and Gergely, 1967). 2 EXPERIMENTAL PROGRAMME