Computational annealing of simulated unimodal and bimodal networks

A conjugate gradient Monte Carlo algorithm was used to simulate the anneali ng of two and three dimensional end-linked unimodal and bimodal polydimethy lsiloxane networks. Equilibrium is satisfied at every crosslink during netw ork energy minimization resulting in distinct differences in network charac teristics from classical assumptions. Annealed unimodal networks were found to retain the uniformly dispersed arrangement of crosslinks generated duri ng the crosslinking algorithm. Radial distribution functions of chain vecto r lengths for various unimodal systems show a shift in the mean chain lengt h from the rms length prior to annealing to shorter lengths upon annealing. Short chains in bimodal networks cluster during the annealing process in a greement with experimental investigations of short chain agglomeration in t he literature. This work provides the first predictions of bimodal chain ne twork clustering via simulated network formation and demonstrates the criti cal role of network annealing in determining the initial configurations of deformable elastomeric networks. This information is extremely useful in th e development of accurate constitutive models of bimodal networks. (C) 2001 Elsevier Science Ltd. All rights reserved.