Why Computational Materials Science?

Michael Glazoff
Idaho National Laboratory
Advanced Process and Decision Systems

Computational Materials Science is used more and more extensively because of the confluence of several factors:
(a) growing and readily available HPC clusters; (b) development and maturation of theoretical concepts (DFT);
(c) dramatic reduction in experimentation required for discovering new materials

We will consider the following topics:  
Self-consistent evaluation of phase equilibria and thermodynamic properties of materials (ThermoCalc);
Diffusion modeling - a case study of diffusion welding modeling using DICTRA
Development of corrosion resistant thermo-mechanical treatment for nuclear fuel elements at Advanced test Reaction in INL using such tools as PRISMA and JMatPro;
Microstructure evolution modeling using MICRESS (a  case study of dendritic solidification)
Atomistic modeling and atomic and electronic structure of matter using DFT
Mechanics of finite deformations modeling using ABAQUS and plasticity research (case study - hcp materials)
Artificial neuron networks and deep data mining as an instrument of material characterization and discovery

Finally we will draw some conclusions and discuss future developments.