Accelerating topology optimization codes using mesh refinement continuation
Name
chen-achn-MEng-CEE-2023-thesis.pdf
Description
Thesis PDF
Size
2.31 MB
Format
Adobe PDF
Checksum (MD5)
6d9db54455c637e1de4eb3917092e214
Author(s)
Chen, Austin
Advisor(s)
Carstensen, Josephine V.
Date Issued
June 2023
Publisher
Massachusetts Institute of Technology
Abstract
A new concept for an algorithm accelerating topology optimization programs is introduced and explained in detail, which involves a continuation of increasing mesh resolutions to achieve low-compliance solutions with largely reduced computation times. Comparisons with examples from relevant literature show speedups of up to approximately 60% on discretizations up to the order of 106 elements for common benchmark problems. Improvements in speed can be attributed to taking advantage of running code on coarse meshes as a faster way to generate smart initial guesses to be reused as inputs for subsequent runs on finer meshes. A MATLAB script for the new algorithm and associated modifications to existing topology optimization code is included.
Keywords: Topology Optimization, Mesh Refinement, Computational Efficiency, MATLAB
MIT Department
Massachusetts Institute of Technology. Department of Civil and Environmental Engineering
Terms of Use
In Copyright - Educational Use Permitted
Copyright retained by author(s)
Persistent DSpace Link
