Robin Ball

Research in Statistical Physics is wide ranging and has close links to our interdisciplinary Complexity Complex and its associated Doctoral Training Centre. Here are some of the recent objects of investigation.

Powders There is a real (albeit somewhat over-hyped) opportunity here to sort an outstanding state of matter. Whilst models are coming into surplus, the challenge remains to sort out fundamental principals and make contact with/impact on the established Engineering discipline of soil mechanics. We are making progress on how the statics depend on the microscopic powder geometry; a harder challenge is to close the theory by relating that geometry to material history.

Colloid Rheology New computer simulation techniques developed to cope with hydrodynamically coupled particles have opened up the rich field of how concentrated colloids flow. The sensitivity of shear thickening to near contact interactions is a new discovery which has prompted new theoretical ideas as well as having practical consequences. A more complete theory of flow induced structure is now a priority as the phase diagram emerges from simulation.

Fracture The serious theoretical work in this subject has been mainly in two dimensions, and addressing the onset of fracture. The challenge today is to understand propagating cracks and their three dimensional evolution - where we are near to completing the stability analysis. We are also trying to remedy the fact that Theoretical Physics still says almost nothing about the crack tip.

Protein Folding Computer Simulation (around the world) is beginning to solve the forwards folding problem with realistic potentials. This opens up a corresponding new dimension in protein engineering: how can we design real protein sequences to fold in a given way? Direct attack on this problem has spin-off in optimisation more generally. We are investigating how efficiently to optimise a design whose performance can only be sampled statistically.

Fractal Structure The days of hunting Fractal Dimensions are over, and the challenge is to find richer quantitative science in this field of beautiful images. I am exploring the information encoded in the three point correlation function, with the claim that it quantifies the linearised renormalisation group.