Research Repository

Bi-directional evolutionary method for stiffness and displacement optimisation

Yang, Xiaoying (1999) Bi-directional evolutionary method for stiffness and displacement optimisation. Research Master thesis, Victoria University of Technology.

[img] Text
YANG_1999compressed.pdf

Download (3MB)

Abstract

This thesis presents a method for structural optimisation called bi-directional evolutionary structural optimisation (BESO). It is an extension of the systematic research on the evolutionary method. The basic concept of evolutionary structural optimisation (ESO) is that by slowly removing the inefficient material, the structure evolves towards an optimum. BESO extends the concept by allowing for the efficient material to be added while the inefficient material is removed. The formulation of BESO is motivated to improve the reliability and efficiency of the ESO method. The BESO method for topological optimisation of 2D continua subject to stiffness and displacement constraints is the major task of this thesis. The theoretical aspects are explored by following the optimality criteria algorithm for problems of discrete design variables. These aspects include the optimality criteria, sensitivity analysis, displacement extrapolation and evolutionary procedure. The bi-directional evolutionary procedure is incorporated with the finite element analysis to realise an automatic optimisation process. A wide range of examples are tested by using the proposed BESO procedure. Different design conditions are considered including stiffness optimisation and single or multiple displacement optimisation under single and multiple loading conditions. The solution reliability and parametric effect are further studied to improve the BESO performance. The comparison of results by BESO and ESO are attempted and the satisfactory agreement demonstrates the validity of the proposed procedure. Two major conclusions are derived from the work in this thesis. The first one is that BESO is as effective as ESO, and the second one is that BESO can be computationally more efficient in most cases.

Item Type: Thesis (Research Master thesis)
Additional Information:

Master of Engineering

Uncontrolled Keywords: optimal structural designs, structural optimization, static and dynamic behaviour, composite material
Subjects: FOR Classification > 1203 Design Practice and Management
FOR Classification > 1204 Engineering Design
FOR Classification > 1299 Other Built Environment and Design
Faculty/School/Research Centre/Department > School of Engineering and Science
Depositing User: VU Library
Date Deposited: 21 Dec 2011 00:52
Last Modified: 23 May 2013 16:56
URI: http://vuir.vu.edu.au/id/eprint/18230
ePrint Statistics: View download statistics for this item

Repository staff only

View Item View Item

Search Google Scholar