发表刊物：Thin-Walled Structures
关键字：Sandwich structures; Multi-layered corrugated core; Underwater explosion; Multi-fidelity surrogate modeling; Surrogate-based optimization
摘要：The present study investigates the application of a multi-fidelity surrogate-based optimization framework in enhancing the resistance of multi-layered corrugated core sandwich panels subjected to underwater explosion. A high-fidelity finite element (FE) model considering nonlinear material behavior, complex contact and fully coupled fluid-structure interaction (FSI) was established and validated against experimental results. On this basis, medium- and low-fidelity FE models were further constructed through coarsening the mesh and neglecting FSI effects, respectively. Although the simplified FE models significantly improved the simulation efficiency, they inevitably leaded to deviation amplifications. Only the medium-fidelity model preserved trend consistency with the high-fidelity responses. A co-Kriging surrogate model trained on high- and medium-fidelity datasets could achieve desirable prediction accuracy with a substantial reduction in computational cost. Based upon the surrogate model, multi-objective optimization was conducted to minimize back face deflection (MaxD) and maximize specific energy absorption (SEA). The optimal solution, which offered a well-balanced trade-off between MaxD and SEA, should be configured with thicker front face and first core layer, thinner back face and remaining core layers, as well as a lower uniform corrugation angle. Compared with the baseline panel, the optimized designs reached up to 30% improvement in MaxD or SEA without compromising the other index, demonstrating the effectiveness of the proposed framework for efficiently designing explosion-resistant structures.
论文类型：期刊论文
卷号：231
是否译文：否
发表时间：2026-07-24
收录刊物：SCI
DOI码：10.1016/j.tws.2026.115425
影响因子：8.6
第一作者：蔡思培