Strengthening of reinforced concrete members with lap splices using steel collars

Abstract

Bond splitting failure of substandard lap splices have caused extensive damage to many structures during earthquakes. Existing methods to strengthen substandard lap splices primarily involve wrapping the substandard lap spliced regions with jackets possessing mainly in-plane stiffness. Hence, the efficiency of such techniques is compromised due to the lateral bulging of concrete and flexural bending of jackets. This study investigated hollow steel section (HSS) collars which offer axial and flexural stiffness in mitigating splitting failures associated with lap splices not conforming to current design codes. Three lap splice lengths, mainly 20, 28, and 35 times the bar diameter (db) were studied. Experiments were conducted on 19 beams subjected to four-point bending, with substandard lap splices within the constant moment region. Test variables involved the lap splice length, concrete cover, and spacing of HSS collars. Results revealed that HSS collars successfully prevented splitting failures resulting in ductile behavior. It was found that the bond stress of the lap-spliced bars increased till the onset of their yielding. Beyond that, the bond stress was maintained close to their peak value in the beams sufficiently confined by HSS collars. By applying HSS collars, improvements in the bond strength for beams with 20db lap splices were evident. The control beams having longer lap splice lengths exhibited yielding, and the improvement in their bond strength was limited despite showing ductile behavior. In order to capture the contribution of HSS collars to the improvement from brittle behavior to ductile behavior, the interfacial fracture energy was also used instead of the bond strength alone. An equation for the increase in the interfacial fracture energy due to the confinement by HSS collars was obtained using nonlinear regression. This equation was utilized to determine the required confinement ratio of HSS collars for a given substandard lap splice. The required confinement ratios of HSS collars obtained from the proposed equation were found to agree with the experiment while overestimation was found in some cases. Nonlinear fiber modeling using OpenSees was performed to predict the experimental load-deflection curves of beams. Several existing approaches were compared to estimate the strength of substandard lap-splices, whereas the splice strength in the case of insufficiently confined beams was predicted by an analytical approach based on interfacial fracture energy dissipated by lap-splices. An existing approach was adopted to estimate the compressive stress-strain response of HSS collar confined concrete by considering the axial and flexural stiffness of HSS collars. The proposed modeling strategy for beams resulted in close agreement with the experimental results. The same modeling concept as that for beams was adopted to model HSS collar strengthened RC columns in combination with pinching and strength degradation rules. An energy-based calibration was adopted to calibrate the predicted hysteretic response of RC columns. The pinching and hysteretic damage parameters for sufficiently confined columns were proposed. A good agreement between the predicted and experimental hysteretic response was obtained.