Studying Motion of <100> Tilt Grain Boundaries Using Molecular Dynamics Simulation
- PDF / 949,322 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 113 Downloads / 206 Views
Studying Motion of
Tilt Grain Boundaries Using Molecular Dynamics Simulation
Shijing Lu and Donald W. Brenner Department of Material Science and Engineering, Raleigh, NC 27695-7907, U.S.A. ABSTRACT In the study of grain boundary migration of metallic materials using molecular dynamics simulation (MDS), grain boundary mobilities and activation energies are often found to be different from experimentally observed values. To reconcile the discrepancies, tremendous effort has been made to replicate experiment conditions in MDS, e.g.as low a driving force as possible, near zero grain boundary velocity. In the present study, we propose an analytic method that removes effects from non-physical conditions such as high driving force or high temperature. The analytic model presumes that two types of rate limiting events coexist during grain boundary migration. Kinetics parameters, such as activation energies, of the rare events are different and therefore should be modeled separately. Activation energies from this model are closer to experiment than previously reported values. Further, by analyzing the evolution of atomic structures, these two types of rate limiting events correspond to shear coupled migration and grain boundary sliding mechanisms, respectively. INTRODUCTION Grain boundary (GB) migration is of essential importance in the science and engineering of polycrystalline materials. Molecular dynamics simulations have been used to model the dynamic behavior of grain boundaries. However, despite of continuous efforts in this area during the past decades1–8, our ability to simulate GB motion and to predict microstructure evolution from these simulations is still lacking. In principle, molecular dynamics simulation (MDS) with embedded atom method (EAM) potentials can provide a relatively accurate description of the atomistic energy of metal systems. However, GB properties such as GB mobilities and activation energies predicted by MDS are often different from experimentaally measured values. To reconcile the discrepancies between molecular dynamics simulation and experiments, methods have been developed that are intended to make experimental conditions such as grain boundary velocity of m/s and driving force about MPa accessible to MDS 9,10. Published in 2006, Janssens et al proposed an orientation dependent driving force method to study grain boundary migration (GBM) in f.c.c. aluminum bicrystals. Driving force used in their work is 0.025eV per atom or 244MPa on the interfaces at a simulation temperature of 800K (~0.8Tm). Despite the driving force being several orders higher than the driving force used in experiments, some boundaries, including all of the 111 twist boundaries were found to be immobile with this method. Soon after this work, Zhou and Mohles 4 claimed that driving force used in the above mentioned work is too large, and that boundary migration velocity and driving force were nonlinear. To obtain a GB velocity at the limit of a zero driving force, they computed a series of GB velocities under different dr
Data Loading...
