Interpolation scheme for fictitious domain techniques and topology optimization of finite strain elastic problems

Research output: Contribution to journalJournal articleResearchpeer-review

Abstract

The focus of this paper is on interpolation schemes for fictitious domain and topology optimization approaches with structures undergoing large displacements. Numerical instability in the finite element simulations can often be observed, due to excessive distortion in low stiffness regions. A new energy interpolation scheme is proposed in order to stabilize the numerical simulations. The elastic energy density in the solid and void regions is interpolated using the elastic energy densities for large and small deformation theory, respectively. The performance of the proposed method is demonstrated for a challenging test geometry as well as for topology optimization of minimum compliance and compliant mechanisms. The effect of combining the proposed interpolation scheme with different hyperelastic material models is investigated as well. Numerical results show that the proposed approach alleviates the problems in the low stiffness regions and for the simulated cases, results in stable topology optimization of structures undergoing large displacements. © 2014 Elsevier B.V.
Original languageEnglish
JournalComputer Methods in Applied Mechanics and Engineering
Volume276
Pages (from-to)453-472
ISSN0045-7825
DOIs
Publication statusPublished - 2014

Keywords

  • Energy interpolation
  • Ersatz material models
  • Fictitious domain
  • Hyperelastic material model
  • Large deformation
  • Topology optimization
  • Deformation
  • Elasticity
  • Mechanisms
  • Shape optimization
  • Stiffness
  • Elastic energy density
  • Fictitious domains
  • Finite element simulations
  • Hyperelastic material models
  • Interpolation schemes
  • Material models
  • Numerical instability
  • Optimization approach
  • Interpolation

Fingerprint Dive into the research topics of 'Interpolation scheme for fictitious domain techniques and topology optimization of finite strain elastic problems'. Together they form a unique fingerprint.

Cite this