Analysis of the Performance of Two Distance Geometry Algorithms in Simulation of Polymer Systems

Melnik, R.V.N., Uhlherr, A. and Hodgkin, J.

Proceeding of the IV International Engineering Mathematics and Applications Conference,  Eds. May, R.L., Fitz-Gerald, G.F. and Grundy, I.H., Melbourne, ISBN 085825 705X, 223--226, 2000

Abstract:

We propose efficent distance geometry algorithms capable of dealing effectively with individual polymer chains isolated in space, as well as with large material structures, including bulk materials. By providing good approximations to topological molecular structures of materials, such algorithms can increase the efficiency of available molecular modelling codes and therefore can lead to the possibility of calculating structural and mechanical properties of polymer and composite structures using available methodologies. As a development of this workwe are now undertaking a comparison of the performance of several distance geometry algorithms for simulating bulk molecular cross-linked structures using end-to-end distributions, radii of gyration, and radial distribution functions.

Key words: polymer modeling; Molecular Dynamics and Monte Carlo simulations; polymers and composites; objective functions and algorithms based on their smoothing; Gaussian transform; quasi-Newton algorithms; distance geometry problems; initial samplings; constrained optimization; complex systems and topology.