Probabilistic residual displacement-based design for enhancing seismic resilience of BRBFs using self-centering braces

Abstract

Buckling-restrained braces (BRBs) can effectively control the structural maximum displacements. However, their full hysteresis tends to cause significant residual drifts after strong earthquakes, resulting in massive economic loss. To improve the resilience of the buckling-restrained braced frames (BRBFs) by reducing the residual drifts, self-centering braces (SCBs) are introduced to BRBFs based on the probabilistic residual displacement-based design (PRDBD) method in this study. Parametric analyses are first carried out on single-degree-of-freedom (SDOF) systems to represent the retrofitted BRBFs subjected to near-fault pulse-like earthquakes, wherein the isotropic strain hardening of BRBs is considered. The probabilistic prediction models for maximum and residual displacement are subsequently developed by employing an artificial neural network (ANN) machine learning algorithm. Subsequently, the PRDBD method is proposed. A benchmark BRBF is retrofitted by SCBs to demonstrate the PRDBD’s effectiveness. System-level nonlinear time-history analyses (THAs) are performed to study the dynamic performance of the enhanced BRBFs. It can be observed from the analysis results that the proposed PRDBD is efficient in controlling residual displacements of the enhanced BRBFs within the prescribed guarantee rate. Moreover, the analysis results confirm that the retrofitted BRBF with partial self-centering behavior can achieve remarkable post-earthquake repairability, as evidenced by residual inter-story drifts below 0.5% with 99% probability subjected to the maximum considered earthquakes.