Big properties from small simulations

Kofke Lab

Configuration generated during calculation of 7th virial coefficient of a polarizable water model.

Engineers need good physical property data to design new materials and processes. With decades of advances in computing hardware and algorithms behind us, the ability to compute these properties from fundamental physics is coming within reach.

Overview

It makes sense to advance this capability first for the simpler (but extremely important) thermodynamic phases — solids and gases. Remarkably, for the gas we can accurately compute “big” macroscopic properties from “small” simulations, considering interactions of just two, three, four, etc. molecules at a time. The project has many facets:

* developing efficient computational methods accommodating complications such as nuclear quantum effects

* improving the ability of the approach to describe properties well while modeling fewer molecules

* deriving formulations for a broader range of properties, such as dielectric constants

* understanding the approach in the context of phase transitions, such as condensation and vapor-liquid criticality

* applying these methods to problems of broad relevance, from energy to chemical technology to the environmentProject description