Stress-based topology optimization with discrete geometric components

In this paper, we introduce a framework for the stress-based topology optimization of structures made by the assembly of discrete geometric components, such as bars and plates, that are described by explicit geometry representations. To circumvent re-meshing upon design changes, we employ the geomet...

Full description

Saved in:
Bibliographic Details
Published in:Computer methods in applied mechanics and engineering Vol. 325; pp. 1 - 21
Main Authors: Zhang, Shanglong, Gain, Arun L., Norato, Julián A.
Format: Journal Article
Language:English
Published: Amsterdam Elsevier B.V 01.10.2017
Elsevier BV
Subjects:
Bar structures
Bars
Design for manufacturing
Discrete element method
Finite element method
Geometry projection
Meshing
Optimization algorithms
Plate structures
Stress constraints
Stress state
Stresses
Studies
Topology
Topology optimization
Geometry projection
Bar structures
Topology optimization
Design for manufacturing
Plate structures
Stress constraints
ISSN:0045-7825, 1879-2138
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper, we introduce a framework for the stress-based topology optimization of structures made by the assembly of discrete geometric components, such as bars and plates, that are described by explicit geometry representations. To circumvent re-meshing upon design changes, we employ the geometry projection method to smoothly map the geometric components onto a continuous density field defined over a uniform finite element grid for analysis. The geometry projection is defined in a manner that prevents the singular optima phenomenon widely reported in the literature, and that effectively considers stresses only on the geometric components and not on the void region. As in previous work, a size variable is ascribed to each geometry component and penalized in the spirit of solid isotropic material with penalization (SIMP), allowing the optimizer to entirely remove geometric components from the design. We demonstrate our method on the L-bracket benchmark for stress-based optimization problems in 2-d and 3-d.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0045-7825
1879-2138
DOI:10.1016/j.cma.2017.06.025