Techno Press

Structural Engineering and Mechanics   Volume 17, Number 3, March 2004, pages 379-391
Adaptive mesh refinement/recovery strategy for FEA
Chang-Koon Choi, Eun-Jin Lee and Won-Jin Yu

Abstract     [Full Text]
    This paper deals with the development of h-version adaptive mesh refinement and recovery strategy using variable-node elements and its application to various engineering field problems with 2D quadrilateral and 3D hexahedral models. The variable-node elements which have variable mid-side nodes on edges or faces are effectively used in overcoming some problems in connecting the different layer patterns of the transition zone between the refined and coarse mesh. A modified recovery technique of gradients adequate for variable-node elements and proper selection of error norms for each engineering field problems are proposed. In the region in which the error is greater than the permissible refinement error, the mesh is locally refined by subdivision. Reversely, in some parts of the domain having the error smaller than the permissible recovery error, the mesh is locally recovered (coarsened) by combination. Hierarchical structures (e.g. quadtrees and octrees) and element-based storage structures are composed to perform this adaptive process of refinement and recovery. Some numerical examples of a 3D heat conduction analysis of the concrete with hydration heat and a 2D flow analysis of vortex shedding show effectiveness and validity of the proposed scheme.
Key Words
    adaptive FEA; mesh refinement/recovery; variable-node element; quadrilateral element; hexahedral element.
Chang-Koon Choi; Department of Civil and Environmental Engineering, KAIST, Daejeon 305-701, Korea
Eun-Jin Lee; DPRI, Kyoto University, Gokasho, Uji, Kyoto 611-0011, Japan
Won-Jin Yu; Samsung Engineering & Construction, Sungnam-Si, 463-721, Korea

Techno-Press: Publishers of international journals and conference proceedings.       Copyright © 2019 Techno Press
P.O. Box 33, Yuseong, Daejeon 305-600 Korea, Tel: +82-42-828-7996, Fax : +82-42-828-7997, Email: