Hysteresis model for beam-to-column connections of steel storage racks

Abstract

© 2019 Elsevier Ltd Steel storage racks, especially pallet ones, have gained significant scientific attention in the last few years. Even though the rack structural concept is elementary, due to thin-walled cold formed sections (also punched or slotted) and highly non-linear (often boltless) beam-to- column connections (BCC), the design of steel racks is demanding. This is especially the case when it comes to seismic loading, where hysteretic behaviour with significant pinching of the connections dictates the structural response. Even though several experiments on BCCs were performed in the past, it is difficult to cross-correlate the results because of a wide range of locking systems and geometric diversities. The component method, widely accepted in the design of steel joints, is an anticipated approach to overcome the above mentioned difficulties, even though it has not yet been established for the entire cyclic response of a BCC. The main obstacle to incorporating of the component method in the rack design lies in the fact that the load transfer between a beam-end connector (BEC) and a column has still not been explicitly investigated or mathematically described, e.g. stress in tabs has not been measured and documented. A proper estimation of the seismic response of the whole structure demands non-linear time history (NLTH) analyses with appropriate connection models. Two such models, both developed for concrete joints, have predominantly been used in recent numerical investigation of racks' response to seismic loading: the Pivot model and the Pinching4 model (integrated in OpenSees). In this paper, a new constitutive model is developed. By using only four parameters, the model can simulate both pinching and low-cycle fatigue of a BCC. This constitutive model of joint is based primarily on the Pivot model, which is altered to simulate peculiarities observed in all available documented moment-rotation curves of racks' BCCs, so it can be considered as the model principally developed for steel racks' joints. Results of ten recent experiments on BCCs are used to evaluate the model's capability and accuracy. The model predictions, which are evaluated through a comparison of dissipated energy and resisting moment through cycles, have been proven satisfactory. In the case study, NLTH analyses are performed on three identical racks with different BCC constitutive models, including the proposed one. These analyses validated the proposed model as computationally effective under dynamic loading. The BCC response estimations of three analysed constitutive models were compared in terms of both local (joint) and global (rack) response to seismic input, and the proposed model was found to be acceptable.