College of Science & Engineering
This project focuses on the development and application of high-performance methods for spanning multiple length and time scales in atomistic simulations of materials and nanodevices. Efforts focus on a number of directions:
- Development of a high-performance 3D implementation of the spatial multiscale quasicontinuum (QC) method with temporal acceleration (Hyper-QC) that greatly reduces the computational cost of atomistic simulations by only retaining atomistic resolution where necessary and using a continuum approximation elsewhere. MSI resources are used to test different parallelization strategies and to perform QC production runs in several projects related to the fundamentals of 3D fracture.
- Study of the fracture of single and polycrystalline silicon for MEMS applications, amorphous carbon thermal protection systems, refractory high-entropy alloys (RHEAs) for high-temperature space applications, and for studying the fundamental physics of fracture. The simulations will be performed using both the LAMMPS molecular dynamic package and QC3D as noted above.
- Development of methods within the Knowledgebase of Interatomic Models (KIM) project for assessing and improving transferability of interatomic potentials used in atomistic and multiscale simulations by comparing their predictions to density functional theory (DFT) calculations. This includes the development of machine learning (ML) interatomic potentials for 2D material systems. MSI resources are used to perform DFT calculations to obtain high-quality reference data and to train ML potentials.
- Studying the static and dynamic mechanics of 2D heterostructures comprised of a stack of 2D materials interacting via weak long-range van der Waals forces. 2D heterostructures are a new and active field of research that has emerged from recent advances in producing single layers of semi-metals (graphene), insulators (boron nitride), and semiconductors (transition metal dichalcogenides). The prospect of combining the properties of these layered materials opens almost unlimited possibilities for novel devices with desirable, tailor-made electronic, optical, magnetic, thermal and mechanical properties. MSI resources are used to perform large-scale MD simulations and multiscale simulations of 2D heterostructures.
This research was featured on the MSI website in January 2020: A New Model for Multilayer Graphene.