Possible Research Projects in Computational Science

Definition of Computational Science.

Computational activities from whatever discipline with the following characteristics:

• Precise mathematical statement
• Intractable by traditional methods
• in-depth knowledge of science, engineering or arts
• significant in scope.

All of the projects below can be approached at the level of summer project, honours or PhD thesis.
The main difference being the depth to which the material is covered.

Background Required:

Some of the following would be useful as background:

• Graphics and image/signal processing
• Numerical computation/integration
• Travellnig Salesman - combinatorial
• N-body problems
• Multidimen6~6~6~sional data visualisation
• Chaos and non-linear dynamics
• Distributed and Parallel Computation
• Monte-Carlo and other statistical simulations
• Bioinformatics
• Fluid and atmospheric dynamics

Possible Projects:

• Resonance effects in large non-linear spring-mass chains and lattices

  When a single particle in a long chain of such particles - all interconnected by non-linear springs - is moved and then released, one might expect the particles to oscillate for a while, with this oscillation gradually reducing as the particles in the chain return to their equilibrium position. Under certain conditions, this does not happen Instead, some particles oscillate out of proportion to their neighbours, and continue to do so.

  This project will attempt to classify the situations under which this behaviour persists. Perhaps a delineation of the bounds of the problem can aid in its understanding.

  Req: Materials physics/engineering, Parallel computation

• Other projects in Science/Engineering

  We are also open to other projects in collaboration with other departments or faculties.

All code produced is to be under the GPL.

Last Updated: 7 October 2003
Maintained by: Ron van Schyndel