Relaxation in a disordered network of bonds
Dr. Vanessa de Souza (School of Chemistry, University of Sydney)
MSI Computational Mathematics Seminar SeriesDATE: 2009-03-16
TIME: 11:00:00 - 12:00:00
LOCATION: JD G35
CONTACT: JavaScript must be enabled to display this email address.
ABSTRACT:
In modelling glassy systems, we are interested in how the arrangement of particles determines the spatial pattern of motions in the system. We have chosen a simple model of structure in which the bonds between particles are treated as constraints. Once these constraints have been added, the remaining degrees of freedom, or unconstrained motions, describe the relaxation of the structure. However, we cannot assume that every constraint removes a degree of freedom, as an added constraint may be redundant if the distance between two particles is already fixed. This problem has been solved in 2D using a Pebble Game algorithm [D. J. Jacobs and M. F. Thorpe, Phys. Rev. Lett. 75, 4051 (1995)]. The algorithm locates redundant bonds, decomposes the system into separate rigid clusters and calculates the remaining degrees of freedom in a system.
We are interested in locating these degrees of freedom in terms of
motions available to the system. We show how the spatial character of unconstrained motions in a random
network of bonds can be directly inferred from the spatial arrangement
of constraints and the resulting rigid clusters. Relaxation timescales can be
associated with each unconstrained motion and hence timescales can be assigned to each particle in the random network. The spatial distribution of these various motions of the disordered network is
studied with relationship to the well studied percolation rigidity
transition. The sensitivity of the spatial distribution of unconstrained motions to minor configuration changes depends dramatically upon how close the bond configuration is to the rigidity percolation point.
BIO:
http://www.srcf.ucam.org/~vkd21/
