Adaptive mesh generation procedures for thin-walled tubular structures

In this paper, a family of adaptive mesh generation schemes specially designed for finite element modelling of structural hollow section (SHS) tubular joint is presented. This family of adaptive mesh generation schemes is implemented based on a series of realistic and consistent geometrical models w...

Full description

Saved in:
Bibliographic Details
Main Authors: Lie, S. T., Nguyen, T. B. N., Chiew, Sing Ping, Lee, Chi King
Other Authors: School of Civil and Environmental Engineering
Format: Article
Language:English
Published: 2014
Subjects:
Online Access:https://hdl.handle.net/10356/102287
http://hdl.handle.net/10220/19047
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:In this paper, a family of adaptive mesh generation schemes specially designed for finite element modelling of structural hollow section (SHS) tubular joint is presented. This family of adaptive mesh generation schemes is implemented based on a series of realistic and consistent geometrical models which is founded on measurements obtained from real structures. The underlying geometrical models provide the definitions of different levels of geometrical details and special features that could appear at different stages of the life cycle of the structure. The adaptive mesh generation schemes accompanying the geometrical models are capable of discretizing the SHS tubular joints into different forms of finite element meshes including pure surface meshes, hybrid meshes with surface and solid elements, and full 3D solid element meshes with or without welding and crack details. As a result, a hierarchical adaptive modelling procedure could be developed to assess the performance of the structures for their whole life cycle from quai-static failure strength analysis to long term fatigue and fracture behaviours under cyclic loadings. In addition, all the mesh generators in this family are adaptive mesh generators such that the discretization error of the corresponding FE models could be effectively controlled by combining them with appropriate adaptive refinement schemes.