(348b) A Reassessment of the Precision and Accuracy of Atomistic Calculations of Elastic Constants of Amorphous Engineering Polymer Glasses

A Reassessment of the Precision and Accuracy of Atomistic Calculations of Elastic Constants
of Amorphous Engineering Polymer Glasses

Efficient and cost effective development of new polymer-based materials for aerospace,
automotive and related engineering applications relies on the availability of accurate
values of material constitutive properties such as thermal and mechanical behavior.
In the case of mechanical properties of amorphous polymers below their glass transition
temperature, theoretical methods required to obtain predictions of elastic constants
from atomistic simulations were developed early in the evolution of computational
materials science [1, 2]. These methods were shown to give reasonable estimates of
stiffness coefficients and the associated moduli [2, 3].  However, since the size of typical
simulated systems is at most a few tens of Angstroms, and since amorphous polymers pack
molecularly in a large number of configurations with comparable probabilities, the
resulting structures tend to give rise to a broad distribution of calculated elastic
constants, leading to considerable uncertainty in the elastic constants of macroscopic
samples of material.  This problem was addressed by Suter and Eichinger, who visualized the
macroscopic structure of glassy polymers as collections of nanoscopic domains and applied
theoretical methods developed by Hill and Walpole for composites to derive relatively
narrow bounds estimates for the macroscopic material elastic constants [4]. Surprisingly
however, since this work was first published, there have been few reported attempts to
examine the effective precision achievable in prediction of elastic constants and to
apply the methods to important engineering polymers.

In recent work, we have performed a detailed investigation of the precision and accuracy with
which elastic constants can be predicted, and have begun to perform calculations on a variety
of materials.  In particular, in this presentation we discuss the scale of calculations
required to obtain acceptable precision of elastic constants, and also illustrate the
accuracy achievable by comparison of predicted results with experimental values for experimentally
well-studied common amorphous polymers. In addition, we report on application of the methods developed
to calculation of the elastic constants of known compatible engineering polymer blend systems
containing polystyrene and poly(phenylene oxide), and further examine the results to determine
the extent to which the precision is adequate to probe synergistic behavior in the blend systems.

References

1. Parrinello, M. and Rahman, A., J. Chem. Phys. 76, 2662 (1982).
2. Theodorou, D.N. and Suter, U.W., Macromolecules 9, 139 (1986).
3. see, Hutnik, M, Argon, A.S. and Suter, U.W., Macromolecules 24, 5956 (1991);
   Fan, C.F., Cagin, T., Chen, Z.M. and Smith, K.A. Macromolecules 27, 2383 (1994).
4. Suter, U.W. and Eichinger, B.E., Polymer 43, 575 (2002).
5. MacKnight, W.J., Karasz, F.E. and Fried, J.R., in "Polymer Blends, Vol I",
   D.R. Paul and S. Newman (eds), Academic Press, New York (1978).