DynSyst_Special_Topics: Temporal Coarse-Graining of Molecular Dynamics Using the Parametrized Locally Invariant Manifolds Method and Data Warehousing Techniques

Carnegie Mellon University, Pittsburgh PA

Investigators

Abstract

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5). A systematic computational methodology for temporal coarse-graining of Molecular Dynamics (MD) will be developed. Our primary interest is the mechanics of nanostructures and complex materials, and the aim is to define the constitutive response of these systems at technologically relevant timescales. Our strategy builds on ideas from three separate disciplines: dynamical systems, data warehousing, and nonequilibrium MD. In particular, dynamical systems concepts will be used to formulate the calculation in terms of the phase-space structure rather than time-stepping; data warehousing techniques will enable the storage and manipulation of the phase-space structure to make this practical; and nonequilibrium MD will provide a framework to formulate and interpret these efforts. Molecular dynamics is a powerful tool to predict material behavior at the nano- and micro-meterscales. However, there is an immense separation in time-scales between atomic motions (femtoseconds) and engineering interest (greater than microseconds). Hence, numerical simulations are unable to reach timescales of practical importance. It is this gap that our work addresses with a combination of theory and computation, informed by the needs of practical engineering. Our work is inherently interdisciplinary and closely ties together the subjects of dynamical systems, data warehousing, and materials science. The work has the potential for a transformative impact on the cost-effective design of material behavior for diverse applications, ranging from electronic to civil infrastructure.