Robust topology optimization of frame structures under geometric or material properties uncertainties

A robust topology optimization algorithm is proposed for frame structures in the presence of geometric or material properties uncertainties. While geometric uncertainties were modeled with uncorrelated random variables expressing the node locations of the structure, material properties uncertainties...

Full description

Saved in:
Bibliographic Details
Published in:Structural and multidisciplinary optimization 2017-10, Vol.56 (4), p.791-807
Main Authors: Changizi, Navid, Jalalpour, Mehdi
Format: Article
Language:English
Subjects:
Algorithms
Computational Mathematics and Numerical Analysis
Computer simulation
Engineering
Engineering Design
Fields (mathematics)
Frame structures
Material properties
Modulus of elasticity
Monte Carlo simulation
Perturbation methods
Properties (attributes)
Random variables
Research Paper
Robust design
Robustness
Stochastic processes
Theoretical and Applied Mechanics
Topology optimization
Uncertainty
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A robust topology optimization algorithm is proposed for frame structures in the presence of geometric or material properties uncertainties. While geometric uncertainties were modeled with uncorrelated random variables expressing the node locations of the structure, material properties uncertainties were modeled with a correlated random field of the material Young’s modulus with an exponentially decaying correlation structure throughout the domain. The proposed algorithm uses stochastic perturbation method for propagating these uncertainties to the structural response level, measured in terms of compliance, and optimizes the expected value plus multiple factors of the standard deviation of the response. A comparison between the resulting robust designs and deterministic designs is made, and changes to the final topologies are discussed. Moreover, using Monte Carlo simulation, it was shown that the robust designs outperform the deterministic designs under real-world situations that are accompanied with uncertainties.
ISSN:1615-147X
1615-1488
DOI:10.1007/s00158-017-1686-4