Name: Mitchell Wood
Pronouns: he/him/his
Biography:
Mitchell Wood is a Principal Member of the Technical Staff at Sandia National Labs in the Center for Computing Research. He holds a Ph.D. in Materials Science and Engineering from Purdue University, a B.S. in Physics from Michigan State University and has been at Sandia since 2016 first as a postdoc, then a staff member since 2018. His work has focused on multiscale modeling and simulation methods of matter in extreme conditions, and in the utilization of machine learning in predictive simulations. He currently leads multiple research efforts focused on multiscale modeling, including the joint DOE/DoD effort for predictive simulation of energetic materials. Being a willing collaborator, he has contributed on wide ranging topics from high pressure phase transitions in Carbon, data-driven design of high entropy alloys, plasma facing materials and the development of accurate and efficient constitutive models for shock hydrodynamic codes. He currently serves on the executive committee of the APS Shock Compression of Condensed Matter topical group, and is an active developer of the open source LAMMPS and FitSNAP software projects.
Institution/Lab: Sandia National Laboratories
Website:
SRP Collaboration Topic/Title: Data-driven Models for Material Science and Beyond
Field or research area: Multiscale Materials Modeling
Please select all the topical areas that apply to your project:
Computational Science Applications (i.e., bioscience, cosmology, chemistry, environmental science, nanotechnology, climate, etc.); High-Performance Computing; Machine Learning and AI; National Security
Brief Abstract:
Experiments to study materials extreme environments (temperature, pressure, strain rate) are challenging and time consumptive, therefore we turn to modeling tools to refine and predict outcomes beforehand. Ideal computational models balance absolute physical accuracy against approximate but computationally lightweight constitutive inputs. In addition, with exascale super computers arriving in the near future, it is timely to ask whether our simulation software is capable of matching this unprecedented computing capability. While many research challenges in material physics, chemistry and biology lie just out of reach on peta-scale machines due to length and time restrictions inherent to Molecular Dynamics and electronic structure methods, questions of the accuracy of our predictions will continue to linger. This is particularly true for complex alloys, composites of disparate components as well as materials in extremes of temperature, pressure and radiation exposure. Here we aim to break the normal accuracy-cost tradeoffs by using machine learned models that scale to the largest supercomputing platforms in the world.
Desired relevant skills, background, or interests:
Material Science, Physics, Chemistry, Python, High Performance Computing
Other comments:
Do any special requirements apply? In-Person Only; U.S. Citizen Only
Other, specify:
Keywords:
Materials; Simulation; Machine Learning
Lightning Talk Title: Data-driven Models for Material Science and Beyond