[Colloquium] Seminar Announcement: Scalable Visualization and Analysis for Atomistic Materials Science - TODAY!

[Ninfa Mayorga]

Computation Institute Presentation

Speaker: Aaron Knoll, Mathematics and Computer Science Division, Argonne National Laboratory
Date: November 10, 2011
Time: 1:00 PM - 2:00 PM
Location: University of Chicago, Searle 240, 5735 S. Ellis Ave., Argonne National Lab, TCS Building, Room 5172

Scalable Visualization and Analysis for Atomistic Materials Science

Abstract:
Modern materials and energy research is increasingly driven by computational chemistry. Many phenomena are computed using molecular dynamics (MD), at increasingly large scales. At the same time, atom geometry computed in MD does not wholly define the surface and structure of molecules. Conventionally, molecular graphics has employed abstractions, such as ball-and-stick, ribbons, and molecular surfaces, to represent, visualize and analyze atomistic data. While intuitive, these models often bear little resemblance to molecular structure suggested by first principles, namely self-consistent electron density fields from ab initio molecular dynamics (AIMD) or density functional theory (DFT) computation. In this talk, we identify visualization and analysis needs of chemical materials data, and demonstrate a framework for large-scale MD visualization based on approximate self-consistent fields. We employ a physically-driven uncertainty classification technique for both volume rende ring and analysis, on self-consistent fields computed from DFT/AIMD or approximated using our technique. Finally, we discuss problems with volumetric representation of molecular data, and how topological analyses may provide solutions.


Information: This talk will be broadcast via the Access Grid at both locations - (Computation Institute) Searle/Room 240A and (Argonne) TCS Building 240/Room 5172