Skip to content

An iterative multiscale modelling approach for nonlinear analysis of 3D composites

Research output: Contribution to journalArticle

Original languageEnglish
Pages (from-to)42-58
Number of pages17
JournalInternational Journal of Solids and Structures
Early online date17 Aug 2017
DateAccepted/In press - 15 Aug 2017
DateE-pub ahead of print - 17 Aug 2017
DatePublished (current) - Feb 2018


The advent of new more complex classes of strongly heterogeneous materials, such as 3D woven composites, introduces new challenges for well-established finite element multiscale modelling approaches. These materials'internal architecture dominates the local stress concentrations, damage initiation and damage progression. Additionally, the material loses periodicity during manufacture and conventional homogenization approaches become inapplicable. In this paper, a multiscale modelling approach based on domain decomposition and homogenization is proposed to model the mechanical behaviour of these materials. The proposed model formulates a set of displacement and force compatibility conditions between the various subdomains. The compatibility conditions are formulated by limiting the set of kinematically admissible solutions on the smaller scales in order to satisfy the larger scale basis functions at the interfaces. The different subdomains are solved alongside the compatibility conditions in an iterative process. The proposed multiscale framework reduces the stress artefacts on the subdomains' boundaries. This feature allows the 3D woven material internal architecture represented at the smaller scales to control the structural response at the global scale. Additionally, this framework allows for selective application of nonlinear material models to the subdomains of interest through its ability to redistribute stresses across the subdomains boundaries.

    Structured keywords

  • Composites UTC

    Research areas

  • Multi-scale modelling, Domain decomposition, 3D Woven Composites, Progressive Damage

Download statistics

No data available



  • Full-text PDF (final published version)

    Rights statement: This is the final published version of the article (version of record). It first appeared online via Elsevier at . Please refer to any applicable terms of use of the publisher.

    Final published version, 5 MB, PDF document

    Licence: CC BY


View research connections

Related faculties, schools or groups