Numerical study of an innovative SMA based beam-column connection in reducing the seismic response of steel MRF structures


1 Department of Civil Engineering, Sharif University of Technology, Tehran, Iran.

2 Building and Housing Research Center, Tehran, Iran.


During recent years, Shape Memory Alloys (SMAs) have been e ectively used to control the seismic response of structural systems. Recovering the residual strains upon unloading (super-elasticity) or by heating (shape memory e ect) is the main characteristic of SMA materials. This paper explores the eciency of a proposed application of steel & SMA bolts in improving the seismic behavior of connections in steel Moment Resisting Frame (MRF) structures. A new semi-rigid, end-plate, beamcolumn connection is introduced, which is based on parallel application of steel and austenite-phase SMA bolts. A number of 3-D steel MRF structural models with di erent numbers of stories and mass eccentricities are considered. These structural models that are designed by employing the conventional rigid and the proposed semi-rigid connections are subjected to bi-directional earthquake records that have been scaled according to ASCE7-05 procedure. Extensive nonlinear dynamic time history analyses are conducted to evaluate the eciency of the steel & SMA based connections in enhancing the seismic performance of the structural models. The obtained results indicate that the proposed steel-SMA connection is highly promising and can eciently reduce the seismic response parameters of the structural systems. Steel bolts of the end-plate connections contribute both to shear force transfer (from beam to column) and dissipation of the input seismic energy through hysteresis behavior. On the other hand, parallel application of steel and SMA bolts in the connections reduces the story drifts as well as the induced forces in the structural elements while decreasing the permanent story displacements. The main goal obtained is to keep the beams away from the formation of plastic deformation. OpenSees platform is used for generating the structural models and performing the numerical analyses.