Philip VargheseProfessor Philip Varghese is one of six faculty members to receive the Institute for Computational Engineering and Sciences (ICES) 2015 W.A. “Tex” Moncrief Grand Challenge Award.

The Grand Challenge Award Program provides resources to cover release time from teaching for UT-Austin faculty for one or more semesters to work on ICES research and academic programs related to the Grand Challenges in computational engineering and sciences that affect the competitiveness and international standing of the nation. Varghese will receive a stipend of approximately $55,000 to cover salary and other expenses necessary to further his research.

Chemists and engineers need to calculate the forces acting on individual nuclei when molecules collide because these forces determine energy transfer and chemical reaction rates. The force information is encoded in a potential energy surface (PES), and even for relatively simple systems it is a function of many variables.

For example, the surface needed to describe collisions of nitrogen and oxygen molecules in air is a function of at least six variables. If the surface is known accurately, and approximating the nuclei as “classical” particles (i.e. obeying Newton’s laws of motion) one can calculate the dynamics of molecular collisions. The accuracy of representation of the PES is not well characterized and often we do not know where in the high-dimensional space of several variables it needs to be known accurately in order to calculate chemical reactions rates reliably. Varghese’s project seeks to answer these questions.

PES trajectories

This example shows a potential energy surface (PES) analysis for H+O2. For non-reactive trajectories (left), the reactants are typically scattered before HO2  formation. For reactive trajectories (right), long HO2 residence time before OH formation.

Working with Prof. Venkat Raman (now at the University of Michigan) and graduate student Stephen Voelkel, Varghese recently developed an enabling technology that allows him to address these questions and answer them quantitatively for the first time. Previously, researchers could not frame these questions because they could not conceive of running the computations to answer them in a reasonable time. Their newly developed computer code will run very large numbers of quasi-classical trajectory calculations very efficiently (peta-scale computing).

Varghese holds the Stanley P. Finch Centennial Professorship in Engineering and serves as director of the Center for Aeromechanics Research.