An iterative method for the prediction of crack propagation on highly heterogeneous media

M.J. Patricio Dias, M.E. Hochstenbach

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

1 Downloads (Pure)

Abstract

The usage of composites in high performance products such as aerospace components, storage tanks or racing car bodies is increasing. Typically, a lightweight yet strong material is sought, capable of withstanding harsh loading conditions. Such conditions are prone to the occurrence of cracking. The prediction of this phenomenon on composite materials is therefore an important task. Many authors have proposed methods for modelling crack propagation on homogeneous linear elastic materials. Extending these for highly heterogeneous materials presents several issues. The equations describing the elastic behaviour of composites are quite complex due to their highly oscillating coefficients. Accounting for how the internal material interfaces influence the direction of crack propagation is an added difficulty. Here a multiscale model is proposed to deal with brittle crack propagation on highly heterogeneous elastic two-phase composites. An automated incremental algorithm is employed to predict the path of a pre-existent crack in a 2D plate. Elasticity problems are solved employing homogenisation and a finite element analysis. The maximum circumferential stress criterion is adopted and the interactions between the crack and the material interfaces are modelled. It is shown that this procedure allows virtually the same accuracy as a full mesoscopic analysis, requiring reasonable computational effort.
Original languageEnglish
Title of host publicationProceedings of the 18th European Conference on Fracture (ECF18, Dresden, Germany, August 30-September 3, 2010)
Pages1-6
Publication statusPublished - 2010

Fingerprint

Dive into the research topics of 'An iterative method for the prediction of crack propagation on highly heterogeneous media'. Together they form a unique fingerprint.

Cite this