# Computational Materials Science

Materials science research has become a data-intensive science, requiring high-throughput computational and experimental methods and multiscale methods for simulating complex behavior.

PNNL's approach to materials science is a paradigm shift, integrating computational methods, statistical structure, and mathematical approaches to handle diverse heterogeneous data to elicit information of interest. Our computational capabilities and established informatics programs provide a foundation to bridge the gap between purely statistically based approaches and materials discovery.

Computational materials science is concerned with modeling and simulation of materials' macroscopic constitutive behaviors by incorporating the influences of microstructure and processing parameters. We employ modeling tools to better understand material behavior. We combine it with an approach to materials informatics that seeks to ascertain how materials properties vary with structure by employing theory, experimental data, calculated data, statistical and pattern-analysis techniques.

Integration of materials informatics with computational materials modeling and simulation provides

- Structure, function, and dynamics of multifunctional materials in appropriate environments
- Comparative analysis of large-scale molecular simulation
- Multiscale modeling of materials response, linking atomistic and macroscopic spaces.

The resulting information is utilized to form a mechanistic basis for a more highly efficient materials development process that focuses on structure-property relationship development with targeted data structures, to design better materials and accelerate their integration to commercial applications.

Contact: Kim Ferris