Computational Topological Investigations and Constitutive Modeling of Bimodal Elastomer Systems

Professor Paris R. von Lockette
Department of Mechanical Engineering and Applied Mechanics
University of Michigan
Ann Arbor, MI

Sponsored by the Department of Engineering Mechanics

Date:  Tuesday, April 14, 1998
Time:  3:30 p.m.
Place:  W128 Nebraska Hall


Elastomers are widely used in engineering applications which utilize their ability to recover completely from large deformations.  Recently special types of elastomers, termed bimodal in view of their bimodal distribution of starting oligometers, have gained increasing attention due to the anomalous toughening and ultimate strength properties exhibited by certain mixture concentrations.  Researchers investigating the source of these increased properties have used experimental as well as numerical methods in their investigations.  These investigations  provide insight into several possible mechanisms for this behavior, but none adequately addresses the effect of molecular network structure, or topology, on the physical properties of the bimodal system.   The work presented here describes the development of a program, NETSIM, which simulates the formation of an elastomer network from di-functional starting oligomers and f-functional reactive agents.  The program is based upon an algorithm, first used by Eichinger, which has been modified by the author to reduce computational intensity, errors due to "edge effects", and possible biasing towards certain topologies.   The improvements arise from the use of novel periodic boundary conditions.   Results of the program are compared to previous works.  Mechanism of self-reinforcement in a bimodal system arising from the simulated topologies are discussed and used to motivate the development of a large deformation bimodal elastomer constitutive model.  For a given molecular weight combination, the model is capable of predicting mechanical response over the entire spectrum of mixture concentrations using a single set of parameters.  Results of the model are compared to data from Mark.