An Efficient O(N) Algorithm for Computer Simulation of Rigid Body Molecular Dynamics

Abstract

Molecular dynamics is effective for a nano-scale phenomenon analysis. There are two major computational costs associated with computer simulation of atomistic molecular dynamics. They are calculation of the interaction forces and formation/solution of equations of motion. In this paper, an O(N) (order N) procedure is presented for calculation of the interaction forces and formation/solution of equations of motion. For computational costs associated with potentials or interaction forces, an internal coordinate method is used. Use of the internal coordinate method makes application of multi-rigid body molecular dynamics to an atomistic molecular system become possible. The algorithm based on the method makes the calculation considerably more practical for large-scale problems encountered in molecular dynamics such as conformation dynamics of polymers. For computational costs associated with formation/solution of equations of motion, Kane method and the internal coordinate method are used for recursive formation and solution of equations of motion of an atomistic molecular system. However, in computer simulation of atomistic molecular dynamics, the inclusion of lightly excited all degrees of freedom of an atom, such as inter-atomic oscillations and rotation about double bonds with high frequencies, introduces limitations to the simulation. The high frequencies of these degrees of freedom force the use of very small integration step sizes, which severely limit the time scales for the atomic molecular simulation over long periods of time. To improve this, holonomic constraints such as strictly constant bond lengths and bond angles are introduced to freeze these high frequency degrees of freedom since they have insignificant effect on long time scale processes in conformational dynamics. In this way, the procedure developed in multibody dynamics can be utilized to achieve higher computing efficiency and an O(N) computational performance can be realized for formation/solution of equations of motion.