Random end-switching configurational bias Monte Carlo for long chain molecules

Abstract

The exponential attrition of configurational bias Monte Carlo for long chains can be reduced to almost quadratic by a simple modification of the basic move. Each trial move begins on the side of the remaining sub chain opposite to the cut location. This type of move is akin to large-scale reptation and in the limit of one-segment cut is reptation. The extension is shown to require the same Rosenbluth weighting scheme as the original algorithm. Several examples are used to demonstrate the reliability and the improved performance of the proposed method.