Detecting precursors of localization by strain-field analysis

O. Gorodetskyi, M. Hütter, M.G.D. Geers

Research output: Contribution to journalArticleAcademicpeer-review

5 Citations (Scopus)
144 Downloads (Pure)


Localization in the deformation field, even though initiated locally at the microscopic scale, leads upon increased deformation to fracture at the macroscopic scale, thereby violating the separation of length scales. Localization and damage can be accounted for in macroscopic modeling by appropriate enrichments at that level, however doing so requires (i) detecting the onset of localization prior to its actual occurrence and (ii) quantifying the kinematical characteristics of the localization band. This paper serves that goal. A methodology is developed to analyze the evolution of strain- and displacement-fields during deformation. A key ingredient in this analysis is the use of the Minkowski functionals (also known as intrinsic volumes, quermass integrals, or curvature integrals) from integral geometry, to detect emerging patterns in thresholded strain- and displacement-fields. Doing so, the onset of localization in the microstructure is detected as the emergence of a correlated and narrow pattern of high strains, prior to the actual loss of material stability. Furthermore, the developed localization band is characterized in terms of a weak displacement discontinuity, incorporating the width and direction of the band. The developed methodology uses kinematical fields only, and is therefore applicable to both numerical and experimental deformation-field data. For illustration purposes, numerical data from a finite-element simulation of a deformed voided microstructure is used, without any loss of generality.
Original languageEnglish
Pages (from-to)84–97
Number of pages14
JournalMechanics of Materials
Publication statusPublished - 1 Jul 2017


  • Kinematic enrichment
  • Localization
  • Micro-fluctuation field
  • Minkowski functionals
  • Pattern recognition


Dive into the research topics of 'Detecting precursors of localization by strain-field analysis'. Together they form a unique fingerprint.

Cite this