Abstract
Important aspects of collective behaviour in granular media, such as formation of force chains and shear localisation, are related to kinematics at the underlying particle scale. Particle kinematics consists of translation and rotation. While considerable effort has been dedicated to resolving the translational component of particle displacements, rotational components have received less attention that is due partly to the difficulty associated with distinguishing rotation from translation at the particle scale. In this paper, three new methods based on digital image correlation, image registration and point least-square estimation are introduced to quantify planar particle rotation. The underlying theory behind each method is described and the performance of the methods is assessed through detailed error analysis. The strengths and weaknesses of each method are discussed based on their applicability to study particle kinematics in natural sand. The methods are presented in the form of open-source codes that can be integrated with other analytical tools to investigate particle kinematics. The application of the three methods is demonstrated through the study of particle rotation in plane strain assemblies of granular media under active and passive conditions.
Original language | English (US) |
---|---|
Pages (from-to) | 23-40 |
Number of pages | 18 |
Journal | International Journal of Physical Modelling in Geotechnics |
Volume | 17 |
Issue number | 1 |
DOIs | |
State | Published - Mar 1 2017 |
Keywords
- Computational mechanics
- Granular materials
- Retaining walls
ASJC Scopus subject areas
- Geotechnical Engineering and Engineering Geology