Design of ultra-thin shell structures in the stochastic post-buckling range using Bayesian machine learning and optimization

M. A. Bessa, S. Pellegrino

Research output: Contribution to journalArticleScientificpeer-review

15 Citations (Scopus)

Abstract

A data-driven computational framework combining Bayesian regression for imperfection-sensitive quantities of interest, uncertainty quantification and multi-objective optimization is developed for the design of complex structures. The framework is used to design ultra-thin carbon fiber deployable shells subjected to two bending conditions. Significant increases in the ultimate buckling loads are shown to be possible, with potential gains on the order of 100% as compared to a previously proposed design. The key to this result is the existence of a large load reserve capability after the initial bifurcation point and well into the post-buckling range that can be effectively explored by the data-driven approach. The computational strategy here presented is general and can be applied to different problems in structural and materials design, with the potential of finding relevant designs within high-dimensional spaces.

Original languageEnglish
Pages (from-to)174-188
JournalInternational Journal of Solids and Structures
Volume139-140
DOIs
Publication statusPublished - 2018
Externally publishedYes

Keywords

  • Buckling
  • Data mining
  • Design charts
  • Evolutionary optimization
  • Heteroscedastic Gaussian process
  • Post-buckling
  • Ultra-thin composites

Fingerprint Dive into the research topics of 'Design of ultra-thin shell structures in the stochastic post-buckling range using Bayesian machine learning and optimization'. Together they form a unique fingerprint.

Cite this