Effects of lead core heating on the superstructure response of seismic isolated buildings

Temperature of the lead core of the lead rubber bearing (LRB) increases and accordingly strength of the isolator decreases under reverse cyclic motions. However, the widely used analysis methods (bounding analyses) for seismically isolated structures do not consider strength deterioration in the force displacement relationship of the LRB model unlike the actual behavior. The theoretical basis of an analysis (temperature dependent analysis) method that gives acceptably closer results to the actual behavior has been presented in recent years. This analysis method updates the force displacement relationship of LRB instantly with the increasing heat in the lead core. This study focuses on comparison of the bounding analyses that predict the results with approximate calculation methods and temperature dependent analysis, which gives acceptably closer results to the actual behavior in terms of engineering demand parameters of superstructure (peak floor accelerations and peak drift ratios). In other words, the reliability of the widely used analyses methods (bounding analyses) have been controlled with temperature dependent analysis method. Also, the effect of several parameters, which are considered to play an active role in the lead core heating, on the response of the superstructure were investigated. These parameters are the as the isolation period (T), Q/W ratio and peak ground velocity (PGV) of the ground motion. In this study, 5760 nonlinear time history analyses were conducted in OpenSees platform with two different base isolated structural systems, whose superstructures are composed of 23-story reinforced concrete and 3-story steel buildings. Analyses results indicate that the bounding analyses remain on the safe side but do not provide a cost-effective solution compared to temperature dependent analyses for the 3-storey steel building. Especially, for certain isolator types, floor acceleration values obtained by temperature analyses are less than the lower bound analyses. This can lead to overdesign calculations for the equipment in the stories. Because the results of temperature dependent analyses can be less than or very close the lower bound limits, the bounding analyses must be supported with the temperature dependent analyses.