#18. Multiscale Materials Modeling using Ab-initio Accuracy Methods


  • Efthimios Kaxiras, Harvard University, USA
  • Ruth Pachter, Air Force Research Laboratory, USA
  • Vikram Gavini, University of Michigan, USA (vikramg@umich.edu)
  • Markus Stricker, École Polytechnique Fédérale de Lausanne, Switzerland
  • Till Junge, École Polytechnique Fédérale de Lausanne, Switzerland
  • Max Veit, École Polytechnique Fédérale de Lausanne, Switzerland
  • Felix Musil, École Polytechnique Fédérale de Lausanne, Switzerland
  • Francesco Maresca, University of Groningen, Netherlands


Ab-initio methods, especially those using density functional theory, have provided many useful insights into bulk properties of a wide range materials. However, a large class of problems—which include extended defects in materials, incommensurate 2D materials, emergent quantum phenomena, multiple principal component alloys—need new approaches for conducting large-scale electronic
structure calculations or systematic and efficient methods that can accurately upscale the quantum interactions.

This symposium aims to bring together researchers from Physics, Applied Mathematics, Materials Science, Engineering Mechanics and Scientific Computing to discuss the forefront of methodological developments, mathematical analysis and numerical implementations of ab-initio accuracy methods as well as their current practical application in quantitative predictive materials modeling.

Topics of interest, include, but are not limited to:
• Hierarchical formulations of materials behavior informed by ab-initio calculations
• Recent numerical advances in large-scale electronic structure calculations, including DFT, TDDFT and many-body quantum methods
• Advances in the broad field of machine learned potentials (SOAP, GAP, Neural Network Potentials, etc.) that can provide ab-initio accuracy
• Mathematical analysis of electronic-structure of defects
• Consistent mathematical formulations and numerical methods for spatial and temporal coarsegraining of ab-initio calculations
• Machine-learning approaches to improve approximations in electronic structure theory, and accelerate electronic structure calculations
• Electronic structure calculations elucidating mechanical, electronic and optical properties of materials
• Applications to materials discovery and development, e.g. predictive computational metallurgy