Optimization of distribution patterns of link beams in a double‐column tall pier bent subjected to earthquake excitations

Abstract

Seismic performance of double‐column tall pier bents has to date scarcely been investigated. The deployment of link beams between columns, including numbers and locations, is generally designed according to the experience of engineers, without any sophisticated procedure. This paper proposes a response surface‐based optimization procedure that searches for optimal configurations of link beams of double‐column tall pier bents, considering multiple performance objectives. The response surface model is well‐developed by considering training samples from uniform design methods, boundary conditions, and iteration procedures. With a typical double‐column tall pier bent in southwest China as an example, the correctness of the proposed procedure is validated by comparing seismic responses with those from conventional optimization methods. Furthermore, this response surface‐based procedure possesses a substantially higher time efficiency than is the case with conventional methods, due to less nonlinear time history analyses as required. The Pareto optimal front is employed to trade off competing objectives in this example and to provide potential optimal design options for engineers.

An efficient and accurate fragility approach for seismic performance assessment of structures

“…Due to its computational efficiency, Cloud has been widely used for seismic fragility analysis in the existing literature 17–27 . Some studies have been conducted to improve the accuracy of the Cloud and estimate the IDA‐based fragility curves based on Cloud results.…”

“…Due to its computational efficiency, Cloud has been widely used for seismic fragility analysis in the existing literature. [17][18][19][20][21][22][23][24][25][26][27] Some studies have been conducted to improve the accuracy of the Cloud and estimate the IDA-based fragility curves based on Cloud results. Miano et al 28 explored an efficient Cloud-to-IDA approach with limited scaling to performance IDA.…”