November 17, 2006 from 16:00 to 18:00 (Montreal/Miami time) On location
In computational physics, molecular dynamics refers to the computer simulation of a material at the atomic level. In principle the only inputs to a molecular dynamics simulation are a set of particles and a description of the forces between them. An initial condition is chosen, and the evolution in time of the system is simulated by applying a simple numerical technique to Newton's equations. Given its scientific importance there is very little rigorous justification of molecular dynamics. From the viewpoint of numerical analysis it is surprising that it works at all. The problem is that individual trajectories computed by molecular dynamics are accurate for only small time intervals, whereas researchers trust the results over very long time intervals. It has been conjectured that molecular dynamics trajectories are accurate over long time intervals in some weak statistical sense. However, no one has been able to rigourously establish this for any systems of interest. I will present an overview of different approaches to the problem and some recent work which clarifies some of the issues.
AddressUQAM, Pav. Sherbrooke, 200, rue Sherbrooke O., Salle / Room SH-3420