报告题目:The Scaled Boundary Finite Element Method: Towards Fully Automated Engineering Analysis
报告时间:2023年12月1日(周五)9:00—11:00
报告地点:澳门尼威斯人3号楼(厚兴楼)307会议室
报告人简介:
Dr. Chongmin Song is a Professor of Civil Engineering and Director of the Centre for Infrastructure Safety and Engineering, University of New South Wales, Sydney, Australia. He obtained the degree of Bachelor of Engineering from Tsinghua University, China and the degree of Doctor of Engineering from the University of Tokyo, Japan. His current research interests are on the development of advanced numerical methods and their engineering applications. He is one of the two original creators of the scaled boundary finite element method.
内容简介:
The process of computational engineering analysis includes the discretization of geometric models and the solution of partial differential equations using numerical methods. In the popular finite element method, a geometric model is discretized into a mesh of elements of simple geometries (triangles and quadrilaterals in 2D, and tetrahedrons and hexahedrons in 3D). With increasingly affordable computer power, the human effort required in mesh generation becomes increasingly critical in terms of both cost and time. Furthermore, geometric models in digital image, STL format and point clouds are becoming more and more popular in engineering applications and present challenges to well-established numerical methods.
This presentation covers the development of the scaled boundary finite element method, aiming to fully automate the process of engineering analysis directly from common formats of geometric models. The scaled boundary finite elements require the discretization of boundary only and can have any number of faces, edges and vortices, leading to a much higher degree of flexibility in mesh generation than standard finite elements. This allows the use of simple and efficient quatree/octree algorithm for fully automatic mesh generation of digital images, STL models, point clouds and traditional CAD models in a unified approach. Moreover, the algorithm is suitable to high-performance computing (HPC). Some salient features and HPC performance of the proposed technique will be demonstrated by numerical examples.