Multi-material topology optimization of viscoelastically damped structures using a parametric level set method

Max van der Kolk, Gijs van der Veen, J de Vreugd, Matthijs Langelaar

Research output: Contribution to journalArticleScientificpeer-review

9 Citations (Scopus)

Abstract

The design of high performance instruments often involves the attenuation of poorly damped resonant modes. Current design practices typically rely on informed trial and error based modifications to improve dynamic performance. In this article, a multi-material topology optimization approach is presented as a systematic methodology to develop structures with optimal damping characteristics. The proposed method applies a multi-material, parametric, level set-based topology optimization to simultaneously distribute structural and viscoelastic material to optimize damping characteristics. The viscoelastic behavior is represented by a complex-valued material modulus resulting in a complex-valued eigenvalue problem. The structural loss factor is used as objective function during the optimization and is calculated using the complex-valued eigenmodes. An adjoint sensitivity analysis is presented that provides an analytical expression for the corresponding sensitivities. Multiple numerical examples are treated to illustrate the effectiveness of the approach and the influence of different viscoelastic material models on the optimized designs is studied. The optimization routine is able to generate designs for a number of eigenmodes and to attenuate a resonant mode of an existing structure.
Original languageEnglish
Pages (from-to)2430-2443
JournalJournal of Vibration and Control
Volume23
Issue number15
DOIs
Publication statusPublished - 2017

Keywords

  • Viscoelastic damping
  • topology optimization
  • multi-material optimization
  • level set method
  • loss factor
  • modal analysis
  • constrained layer damping

Fingerprint Dive into the research topics of 'Multi-material topology optimization of viscoelastically damped structures using a parametric level set method'. Together they form a unique fingerprint.

Cite this