Lab Home | Phone | Search | ||||||||
![]() |
|
|||||||
![]() |
![]() |
![]() |
Behavior of materials under extreme high temperature and pressure conditions has far reaching implications that extend from the study of planetary bodies to materials synthesis science. However, experimentally driving a sample to 1000s of K and 10s to 100s of GPa and characterizing the ensuing changes in structure and properties can be exceedingly challenging in many cases. Atomic resolution simulations have played a critical role in aiding in interpretation of such experiments and can provide valuable microscopic insights into behavior of matter under extreme conditions. However, these simulations have traditionally been confined to either highly accurate yet computationally intensive first-principles-based approaches that constrain studies to approximately nm and O(10) ps spatiotemporal scales, or highly efficient classical interatomic models that struggle to accurately describe interatomic interactions under far from earth-ambient conditions. In this presentation, I will discuss ChIMES, our unique machine-learned interatomic model framework designed to fill this capability gap by enabling “quantum-accurate” simulation on large spatiotemporal scales. I will provide examples of how we are using ChIMES to advance fundamental understanding of chemistry and phase transformations in carbon-containing systems under extreme and far-from equilibrium conditions, and how we are using those insights to develop and refine ultra-fast laser-driven-shock strategies for rapid synthesis of nanocarbon materials. Host: Brenden Hamilton(brenden@lanl.gov) and Romain Perriot(rperriot@lanl.gov) |