Finite element analysis on post-buckling of thin-walled structures using ANSYS

This project studies the feasibility of applying finite element analysis (FEA) to simulate the buckling and post-buckling behaviours of cold-formed thin-walled steel structures, thereby predicting their ultimate load bearing capability in axial compression. ANSYS, a commercial finite element softwar...

Full description

Saved in:
Bibliographic Details
Main Author: Kong Kah Hon Danny.
Other Authors: Chou Siaw Meng
Format: Final Year Project
Language:English
Published: 2009
Subjects:
Online Access:http://hdl.handle.net/10356/16966
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:This project studies the feasibility of applying finite element analysis (FEA) to simulate the buckling and post-buckling behaviours of cold-formed thin-walled steel structures, thereby predicting their ultimate load bearing capability in axial compression. ANSYS, a commercial finite element software, was used in this study. This report documents the improvements made on the design procedure for ultimate load prediction, with the aim of obtaining load results which are accurate, robust and conservative for real-time utilisation. To validate the robustness of the design procedures, specimens of four different cross-sections were tested in this phase of study, namely, hat channel, hat channel with compound lip, lipped channel and lipped channel with compound lip. The FEA results obtained from the simulations were verified by checking the meshing of the elements and nodes, characteristics of the load against displacement graphs, stress contour plots and data convergence. These were followed by a comparison with the experimental results and new design factors were proposed eventually. During the course of investigation and study, previous design factors for fixed-ended hat channels and hat channels with compound lip were found to be invalid due to the presence of significant errors. Rectifications were made and new design factors were provided. In addition, a new loop command was introduced to save the hassle of the user having to duplicate long and repetitive commands in the FEA scripts. Finally, to reduce simulation time, it was discovered to be more appropriate to reduce the number of load steps instead of increasing the element size, as this will not compromise as much on the accuracy of the results. SHELL181 element type was also found to be effective in resolving convergence difficulties encountered during simulations.