||Advances in theoretical understanding, algorithms and computational power are enabling computational tools to play an increasing role in materials discovery, development and optimization. For example, recently developed data mining techniques and genetic algorithms enable the “virtual synthesis” of novel materials, with their properties being predicted on a computer before ever being synthesized in a laboratory. This symposium will cover recent applications and methodological developments at the
frontier of computational materials discovery, ranging from quantum-level prediction to macro-scale property optimization. Of particular interest is computational and theoretical work that features a strong connection to experiment.
This symposium will include a session emphasizing atomistic computational methodologies that represent one of the key tools required for the success of the Materials Genome’s Initiative. Large-scale computations for complex materials, that are needed to guide and complement novel experiments, are only as reliable as the energy models they are based upon. This session will cover the development, testing, and applications of novel empirical energy models from atomistic potentials to coarse-grained approaches to machine-learning techniques. Of equal interest are ab-initio properties determinations to be used in the optimization and testing of such potentials, discussion of novel potential forms for prediction of chemical, mechanical and other properties and/or for describing compounds and alloys not currently available, and testing methodologies for determining the range of applicability of such potentials.
• First principles materials discovery
• Development of empirical and semi-empirical energy models
• Algorithm to search structure-composition design space
• Data mining techniques
• Innovations that improve accuracy and efficiency of computational materials design