报告题目 (Title)： A Monolithic Immersed Boundary Framework for ThermalFSI with Surrogate Model-based Uncertainty Quantification (面向热流固耦合的整体式浸入边界框架及基于代理模型的不确定性量化)
报告人 (Speaker)：Jung-Il Choi，徐甜甜 (韩国延世大学)
报告时间：2026年05月 30 日（周六）11：00-12：00
报告地点 (Place)：GJ303
邀请人(Inviter)：潘晓敏
报告摘要：This talk presents a monolithic immersed boundary framework for thermal fluid–structure interaction (thermal FSI) problems with surrogate model-based uncertainty quantification. Thermal FSI describes the coupled dynamics among fluid flow, structural deformation, and heat transfer, and is relevant to applications such as bimetallic strips, vascular stents, and soft robotic systems. In practical simulations, model inputs such as material properties, boundary conditions, and external forces are often uncertain, while deterministic CFD simulations provide only a single response and cannot fully characterize system variability or reliability. To address these issues, the proposed framework treats the fluid, structure, and thermal fields within a unified computational formulation. The immersed boundary method is used to handle complex geometries and structural motion without requiring body-fitted mesh generation. In addition, surrogate model-based uncertainty quantification is incorporated to analyze parameter uncertainty, sensitivity, optimization, and parameter identification. The framework provides an efficient computational approach for the simulation and reliability analysis of complex thermal FSI systems.