Optimizations of Monte Carlo Particle Transport on CAD-based Geometry

Patrick Shriwise, Argonne National Laboratory
Shutterstock Data Image

Monte Carlo particle transport enables engineers to accurately predict the behavior of nuclear systems. Over the years, this simulation method has been applied to increasingly complex devices and analyses. Constructive Solid Geometry (CSG) has been the default geometry representation in these codes historically. While CSG is robust, generating complex models in this format requires large amounts of human time and often duplicates the efforts of CAD engineers developing the same model for other types of analysis.

This talk will review the origins and ray tracing capabilities of the Direct Accelerated Monte Carlo (DAGMC) toolkit, developed at the University of Wisconsin - Madison, which provides particle tracking on tessellated CAD geometry. DAGMC enables high geometric fidelity while streamlining CAD-based engineering analysis workflows. Pathways to increased ray tracing performance using alternative data structures oriented toward modern CPU architectures will be explored. Finally, applications and extensions of this work will be discussed.

Speaker Bio: Patrick Shriwise is an Assistant Computer Scientist at Argonne National Laboratory in the particle transport research group of the Computational Science division. He primarily performs research with OpenMC, an open source continuous energy Monte Carlo radiation transport code used widely in academia, national labs, and industry. His work involves visualization enhancement, advanced particle tracking techniques, and representation of arbitrarily complex 3D geometry for multiphysics applications.

Zoom Link:  https://argonne.zoomgov.com/j/1618035769

See all upcoming talks at http://wordpress.cels.anl.gov/lans-seminars/