Affordable Access

Access to the full text

A benchmark strategy for the experimental measurement of contact fabric

Authors
  • Wiebicke, Max1
  • Andò, Edward2
  • Šmilauer, Václav3
  • Herle, Ivo1
  • Viggiani, Gioacchino2
  • 1 Technische Universität Dresden, Institute of Geotechnical Engineering, Dresden, 01062, Germany , Dresden (Germany)
  • 2 Univ. Grenoble Alpes, CNRS, Grenoble INP, 3SR, Grenoble, 38000, France , Grenoble (France)
  • 3 woodem.eu, Prague, Czech Republic , Prague (Czechia)
Type
Published Article
Journal
Granular Matter
Publisher
Springer Berlin Heidelberg
Publication Date
Jun 14, 2019
Volume
21
Issue
3
Identifiers
DOI: 10.1007/s10035-019-0902-x
Source
Springer Nature
Keywords
License
Yellow

Abstract

The mechanics of granular materials can be better understood by experimental measurement of fabric and its evolution under load. X-ray tomography is a tool that is increasingly used to acquire three-dimensional images and thus, enables such measurements. Our previous study on the metrology of interparticle contacts revealed that the most common approaches either fail to accurately measure contact fabric or introduce a strong bias. Methods to improve these measurements (i.e., the detection and orientation of contacts) were proposed and validated. This work develops a strategy to benchmark image analysis tools that can be used for the determination of contact fabric from tomographic images. The discrete element method is used to create and load a reference specimen for which the fabric and its evolution is precisely known. Chosen states of this synthetic specimen are turned into realistic images taking into account inherent image properties, such as the partial volume effect, blur and noise. The application of the image analysis tools on these images validates the findings of the metrological study and highlights the importance of addressing the identified shortcomings, i.e., the systematical over-detection of contacts and the strong bias of orientations when using common watersheds.

Report this publication

Statistics

Seen <100 times