Molecular Dynamics - Frequently Asked Questions


Question List


  1. Why can't I find an EAM potential for element X or alloy X-Y?
  2. What are some references about creating EAM potentials?
  3. When varying the lattice parameter, shouldn't the time averaged energy be at a minimum when the time averaged stress is zero?


Answers


1. Why can't I find an EAM potential for element X or alloy X-Y? (21 Dec 2009)

EAM potentials are best suited to free elecron metals, thus copper, silver and gold are probably the easiest metals to model with EAM; the closer the resemblence to a free electron metal, the better the fit to the mechanical properties. It is progressively difficult to create EAM models for systems that deviate from free electron metals.

XMD supports two non-EAM models for modeling Si (the Stillinger-Weber and Tersoff potentials) and Si-C (the Tersoff potential).



2. What are some references about creating EAM potentials? (4 April 1997)

Unfortunately, I know of no complete references for creating EAM potentials. And it is a long messy business. The best I can do is recommend papers which describe the creation of specific potentials. The relevant authors are M. I. Baskes, R. A. Johnson, A. F. Voter, D. Farkas.

A reference describing our own potentials:

Dislocation Generation and Crack Propagation in Metals Examined in Molecular Dynamics Simulations; Mat. Res. Soc. Proc. 278:173-178 (1992); Computational Methods in Materials Science, J. A. Rifkin, C. S. Becquart, D. Kim and P. C. Clapp.

Also, some references recommended by Furio Ercolessi of http://www.sissa.it/furio. Other recommendations are welcome.



3. When varying the lattice parameter, shouldn't the time averaged potential energy be at a minimum when the time averaged stress is zero? (Jan 8, 1998)

No. The Free Energy is a minimum when stress is zero, not the time-averaged potential energy - except at 0 Kelvin, when the two quantities are equal.

In particular, the time-averaged potential energy is a minimum when the stress is zero only when

< s E > = 0

where s is the stress and E is instantaneous potential energy, and < > represents either the time or ensemble average.

This page maintained by Jon Rifkin at jon.rifkin@uconn.edu.