My research is on flows of dry granular materials, such as sand and freshly-fallen snow. Despite their ubiquity in nature and industry, granular flows are surprisingly poorly understood and we know relatively little about the relationship between their bulk behaviour and the basic material properties of the grains.
I particularly concentrate on gravity currents, where the flow is driven by gravity and resisted by friction. Common examples are chute flows and avalanches. In some ways granular gravity currents are similar to a river or canal: for example, they can form hydraulic jumps and drops, and waves can form on their surface.
I use a combination of mathematical modelling and discrete particle simulations (see below).
It doesn’t reflect my more recent work, but I presented a poster at the Gordon Research Conference on Granular Matter in July 2016. (A higher-resolution version is available here.) I also made a poster that displays some of the work by other people in the granular group.
- Topographical effects on granular gravity currents, given to the UK Fluids Network’s Special Interest Group on granular flows, on Friday 31 March 2017, at Cambridge.
- Topographical effects on granular gravity currents, identically titled but substantially different, given at the APS DFD meeting on Sunday 20 November 2016, in Portland, OR.
- Not-so-continuum mechanics, given to the London Mathematical Society on 11 November 2016. A brief introduction that’s meant to be accessible to a general audience (with some basic physical knowledge).
Discrete particle method (DPM) simulations are a class of methods for studying granular materials or suspensions. In a DPM simulation, one calculates the motions of and interactions between individual particles; one specifies ‘microscopic’ contact laws between particles to determine the forces, and then time-evolve the system according to Newton’s second law. The advantage of DPM over practical experiments is that the former is often much cheaper, and gives us information that would be difficult to measure in reality. Unfortunately, real granular systems contain millions of grains and it would be impractical to simulate all of them; DPM simulations therefore need one to choose sensible approximations, as well as many algorithmic tricks.
We use the package MercuryDPM, which is developed by MercuryLab, based at the University of Twente. In May 2017, I became a developer for MercuryDPM.