PolyDyna: a Matlab implementation for topology optimization of structures subjected to dynamic loads

Oliver Giraldo-Londoño, Glaucio H. Paulino

Research output: Contribution to journalArticlepeer-review

34 Scopus citations

Abstract

We present a Matlab implementation for topology optimization of structures subjected to dynamic loads. The code, which we name PolyDyna, is built on top of PolyTop—a Matlab code for static compliance minimization based on polygonal finite elements. To solve the structural dynamics problem, we use the HHT-α method, which is a generalization of the classical Newmark-β method. In order to handle multiple regional volume constraints efficiently, PolyDyna uses a variation of the ZPR design variable update scheme enhanced by a sensitivity separation technique, which enables it to solve non-self-adjoint topology optimization problems. We conduct the sensitivity analysis using the adjoint method with the “discretize-then-differentiate” approach, such that the sensitivity analysis is consistently evaluated on the discretized system (both in space and time). We present several numerical examples, which are explained in detail and summarized in a library of benchmark problems. PolyDyna is intended for educational purposes and the complete Matlab code is provided as electronic supplementary material.

Original languageEnglish (US)
Pages (from-to)957-990
Number of pages34
JournalStructural and Multidisciplinary Optimization
Volume64
Issue number2
DOIs
StatePublished - Aug 2021
Externally publishedYes

All Science Journal Classification (ASJC) codes

  • Software
  • Control and Optimization
  • Control and Systems Engineering
  • Computer Science Applications
  • Computer Graphics and Computer-Aided Design

Keywords

  • Compliance minimization
  • Elastodynamics
  • HHT-α method
  • Newmark-β method
  • Sensitivity separation
  • Topology optimization
  • ZPR update scheme

Fingerprint

Dive into the research topics of 'PolyDyna: a Matlab implementation for topology optimization of structures subjected to dynamic loads'. Together they form a unique fingerprint.

Cite this