Skip to main content

Research Themes

Coarse-graining and molecular recognition (Tiffany Walsh)

Coarse graining and mesoscale modelling are central to the development of modelling techniques that bridge the length and time scales - a key theme of CSC. This involves collaborations between chemists, physicists and biologists, underpinned by the mathematics of dynamical systems and stochastic processes. Examples are wetting and spreading behaviour of liquid films, lubrication, flow of colloidal suspensions, granular materials and liquid crystals, elastomers, and foams. (Postscript file) (Postscript file)

Molecular Modelling (Mike Allen)

The computer may be used as an experimental tool, to simulate behaviour at the molecular level which would otherwise be difficult to determine experimentally. It is a kind of molecular virtual reality, like computer-aided design or architecture, but at a much smaller scale. Work in progress in the Physics department includes the modelling of liquid crystalline colloidal systems which have the potential to become technologically important materials, if we can understand their behaviour over many length scales. A key part of these activities is the efficient implementation of simulation programs for millions of particles on massively parallel computer architectures. (Postscript file) (Postscript file)

Cell biology (Markus Kirkilionis)

Based on the detailed modelling of genetic, signalling and metabolic networks including sub-cellular transport processes we study the up-scaled behaviour of cellular regulation. Multi-scale methods are used to derive models for cellular interaction describing biological, biomedical and biotechnological applications. Computer simulations allow the in-silico study and prediction of complex biological systems where various data sources (data bases, GFP-based confocal microscopy etc.) have to be consistently combined. (Postscript file)

Fluid Turbulence (Bob Kerr)

Obtaining a complete picture of turbulence in the laboratory or in the atmosphere is very difficult. There is just too much happening at any one time. There are big eddies billowing out in all directions, on top of which there are all sizes of smaller ripples and puffs. We use simulations because they can provide a complete picture as a turbulent fluid evolves in time. Nonetheless, to capture all the fine features requires a simulation that uses all the computational power that can be obtained. (Postscript file) (Postscript file) (Postscript file) (Postscript file)

Visualisation and Models (Markus Kirkilionis)

Tissues and organs, such as the liver etc. are highly topologically structured. How can we model their function? Solving the transport problem alone must take into account the complex geometry of such objects using adaptive, unstructured grids. The program package UG, developed at the University of Heidelberg, is a reliable code which has already been parallelised efficiently. (Postscript file)

Image Analysis (Elke Thönnes)

Recent years have seen the use of stochastic geometry models in the statistical analysis of high level image attributes. A Bayesian approach using these models as prior distributions not only allows the introduction of prior knowledge, such as smoothness or connectivity characteristics, but also provides the means of assessing the variability of estimates. (Postscript file) (Postscript file) (Postscript file)