Correlation of the bond provisions of CSA A23.3-94 with tests on tension lap splices in beams

Abstract

Data on beams with tension lap splices tested under static loading at the same institution over the past two decades are correlated with the bond provisions of the Canadian concrete standard CAN/CSA A23.3-94 (detailed design approach), as well as with the recommendations of ACI Committee 408, on which the Canadian standard appears to be largely based. The correlations show that transverse reinforcement is more effective than the new bond provisions allowed in cases where the bond failure is governed by splitting rather than bar pullout. Extending the effective limits for confinement provides a more accurate estimate of the bond resistance available at higher levels of confinement, resulting in a more uniform factor of safety over a wide range of confinements. Lap splices with no transverse confinement showed relatively poorer performance than lap splices with varying degrees of transverse confinement when correlated with resistances predicted on the basis of the new CAN/CSA A23.3-94 provisions. Weaker relative splice performance in the absence of transverse confinement raises a concern for the development lengths required by the CAN/CSA A23.3-94 provisions. With highly stressed lap splices, a class factor of 1.3 is applied to the basic development length to determine the lap length. Published information, on the other hand, has shown that lap splice lengths and development lengths should be the same for transferring or developing the same level of stress in tension reinforcement when the same level of confinement is provided along the anchorage. In contrast, the ACI Committee 408 recommendations use a larger factor of safety on development length and lap splice length, rather than applying class factors for splices only, making splice and development lengths the same for the same confinement and required strength transfer. For the data considered, required lap lengths are similar using both the CSA Standard CAN/CSA A23.3-94 (including the 1.3 class factor) and the ACI 408 recommendations, and only small differences in overall prediction accuracy were found. Differences in the definition of the concrete confinement term for close bar spacing by the two design models, different limits on the total confinement that can be considered effective, as well as a further modification factor for bar size in the CAN/CSA A23.3-94 provisions, result in only small differences in lap length requirements for most of the data considered. Key words: anchorage (structural), bond, confinement, lap splices, reinforced concrete, standards, static loading, tension.