Constructing analysis-suitable parameterization and curvature-based r-adaptive parameterization for isogeometric analysis
Date:日期：
Forum talk论坛报告 at ，Yuelu Forum 2022, Mathematics Sub-forum (2022 岳麓论坛 · 数学分论坛), Hunan University, Changsha, China
Slides, Photo 1, Photo 2, Photo 3, Conference Link
In this talk, I presented our work on constructing analysis-suitable parameterizations for isogeometric analysis (IGA) from boundary representations, together with curvature-based r-adaptive parameterization. The first part covered the barrier function-based and penalty function-based NURBS parameterization methods, their comparison with the methods of Pan et al. (2020) and Liu et al. (2020), and an elliptic grid generation method based on a scaled harmonic map whose nonlinear system is solved with preconditioned Anderson acceleration. The second part introduced a curvature-based r-adaptive parameterization that uses the absolute principal curvature of the isogeometric solution as an error estimator within a bi-level strategy (a coarse design model and a fine analysis model), applied to a heat transfer problem with a moving Gaussian heat source.
Keywords: Isogeometric Analysis, Analysis-Suitable Parameterization, r-Adaptivity, Curvature-Based Parameterization