Abstract

Particle simulation studies the time evolution of the dynamical state of a system of interacting particles by the computation of trajectories of particles based on the numerical integration of the Newton equations of motion. This approach is widely used in biomolecular modeling to study, for example, proteins; and in astrophysics to study gravitational N-body system. The dimension of this system of equations of motion is in the order of tens of thousands. To solve this system involves an enormous amount of computer time. Several techniques have been advanced to reduce the computational complexity of this problem. Some of which are: distance cut-off, multiple time-stepping, use of appropriate integration algorithms, fast multipole algorithms, stochastic dynamics and a use of parallel and distributed computing. This work focuses on the the design of suitable algorithms for the molecular dynamics simulations exploring, as much as possible, different strategies mentioned above.

Department(s)

Computer Science

Comments

The Author is a Graduate Student.

This work is partly supported by National Science Foundation Grant CCR-9408973.

To appear in the Proceedings of 2nd. International Conference on Dynamic Systems and Applications Atlanta, USA, May 24-27, 1995.

Report Number

CSc-95-05 & CSc-95-06

Document Type

Technical Report

Document Version

Final Version

File Type

text

Language(s)

English

Rights

© 1995 University of Missouri - Rolla, All rights reserved

Publication Date

02 June, 1995

Download

Share

 
COinS