Available student project - Force networks in granular materials: imaging, pattern recognition and data mining

Research fields

  • Topological and Structural Science
  • Materials Science and Engineering

Project details

A joint ANU-UoM PhD Project

ANU Supervisor Mohammad Saadatfar (mohammad.saadatfar@anu.edu.au)

UoM Supervisor Antoinette Tordesillas (atordesi@unimelb.edu.au)

This project employs an integrated experimental and analytical approach to interrogate granular materials (e.g., soil, sand and sedimentary rocks, powder, colloidal systems, coal, snow etc.).  The experimental part, undertaken at ANU, involves the accurate experimental measurement and 3D visualisation of contact forces at the contacts between particles. The analytical part, undertaken at UoM, focuses on “mining” hidden patterns in the experimental data, using new tools from mathematics and statistics of complex systems.

To give an idea of the challenges, consider an idealised assembly of spherical grains (see Fig.1 above).  Even for the simplest rules of interaction between grains in contact (e.g., Hooke’s law and Coulomb’s law), a remarkably rich array of phenomena often emerges across different scales in space and time. Many of these manifest themselves as patterns in grain motions and inter-grain forces. Uncovering these patterns and understanding what they mean is far from straightforward.  For many real granular materials, this challenge is further compounded by multiphysics behaviours arising from: complex shaped grains (see Fig.2), and non-linear and highly dissipative grain interactions.   Advanced data analytics tools, designed specifically for multiscale and multiphysics phenomena germane to granular systems, are needed to “mine” hidden patterns in these experimental data sets.  Detailed knowledge of these patterns is crucial, since these yield clues on the mechanisms underlying material strength, failure and permeability.

This project will expose you to some of the most exciting frontiers in high-resolution experimental and big data science. It will provide training in

  • Multi-scale approaches
  • Multi-physics and coupled phenomena
  • Computational and mathematical tools in big-data analytics

Project suitability

This research project can be tailored to suit students of the following type(s)
  • Phd or Masters

Contact supervisor

Saadatfar, Mohammad profile

Updated:  15 January 2019/ Responsible Officer:  Director, RSPE/ Page Contact:  Physics Webmaster