Feb_March_AMP_Digital

HIGHL IGHTS A D V A N C E D M A T E R I A L S & P R O C E S S E S | F E B R U A R Y / M A R C H 2 0 2 0 5 2 IMAT 2020 Update PSDK and Machine Learning Sessions at IMAT Richard Otis and James Saal—IMAT 2020 Session Chairs PSDK Celebrates 15th Anniversary at IMAT 2020 The PSDK XV: Phase Stability and Diffusion Kinetics 2020 symposium sponsored by ASM International’s Alloy Phase Diagram Committee will take place at IMAT 2020 in Cleveland, in September. The understanding and modeling of phase stability and diffusion kinetics have been central themes of materials science and engineering, with merits in both science and technology perspectives. Advances in computational and instrumentation techniques, cross-scale modeling, and novel experimental observations continue to refine our understanding, resulting in design-capable tools for industry and education. This symposium will focus on the temporal and spatial evolution of phase constituents and compositions with due consideration for materials chemistry, processing, and applications. Topics include: • Physics- and mechanism-based theory, modeling, and simulations • Fundamental and applications-oriented experimental observations • Microstructural modeling with an emphasis on using fundamental data • Phenomenological expressions and atomistic mecha- nisms • Methodology development and interpretation of data and databases • Multi-scale models and integration methods • Energetics and diffusivity of crystalline defects such as grain boundaries • High-throughput phase diagram determination • High-throughput tracer diffusion studies • Machine learning and artificial intelligence utilizing thermodynamic and kinetic data • Application of high-performance computing to Calphad • Effects of pressure on phase transformations and micro- structure • Evolution of metastable/non-equilibrium states Invited sessions in honor of the 2020 J. Willard Gibbs Phase Equilibria awardee Mark Asta will headline this symposium. Artificial Intelligence and Machine Learning for Materials Materials informatics is becoming an increasingly common and powerful approach to accelerate the develop- ment of materials solutions. The past decade has seen rapid development of novel methods in artificial intelligence, machine learning, and other data- driven methods in the physical sciences. Improvements in materials data handling, aggregation, and database con- struction have enabled use of these methods for a range of materials applications of practical industrial importance, such as the discovery and design of novel materials, devel- opment of materials processing, and optimization of man- ufacturing. While promising, materials informatics faces numerous hurdles to wider adoption and success, such as the need for larger and more comprehensive materials databases, the development of best practices for informat- ics as applied to materials data, improved interpretability of machine learning models, and effective use of uncertainty in the design of experiments. IMAT 2020 will provide a forum for researchers and practitioners to learn and share the state-of-the-art in materials informatics, its capabilities and shortcomings, and discuss its future. Topics include: • Use case studies and exemplar applications of materials informatics • Sequential learning, active learning, and Bayesian de- sign of experiments • Surrogate modeling, reduced-order modeling, and forming systems of models • Physics-informed and “black box” learning, model interpretability • Uncertainty quantification and propagation • Verification and validation of results, including best practices for reproducibility • Reinforcement learning for materials and manufacturing • Design space construction across structure, chemistry, and processing, their exploration, and multi-objective optimization • Materials databases, data schema, and automated data generation (experimental and computational) Otis Saal IMAT 2020 UPDATE