Past Event: Oden Institute Seminar

What type of model order reduction should I use?

Patrick Blonigan, Senior Member of Technical Staff, Sandia National Laboratories

Abstract

Computational modeling and simulation plays a crucial role in a number of high consequence application spaces studied at Sandia National Laboratories. Verification, validation, and uncertainty quantification (UQ) is needed to establish trust in these simulations. Unfortunately, outer loop analyses such as UQ require many model evaluations. These simulations can be very computationally expensive for even a single model evaluation, making UQ for the simulations costly or even intractable in some cases. This high cost of outer loop analyses has motivated the study of model order reduction at Sandia and beyond, resulting in a wide range of different types of reduced-order models (ROMs). Despite all of this progress it is not always clear which ROM should be used in a given situation. 

In this talk I will discuss my experiences deciding what type of model order reduction to use, looking at three different applications in high-speed aerothermodyamics, a discipline that requires computationally expensive multi-physics simulations. These examples show that there is not one model order reduction approach that works best in all cases, and that practitioners should be familiar with a range of approaches. 

Biography

Patrick Blonigan is a principal member of the technical staff at Sandia National Laboratories. He currently leads or contributes to several research projects on model order reduction with application to computational models of high-speed aerodynamics and thermal protection systems. Prior to joining Sandia in 2018, Patrick was a postdoctoral fellow in the advanced supercomputing division at NASA Ames Research Center. Patrick holds a B.S. in mechanical engineering from Cornell University, a M.S. in aeronautics and astronautics from MIT, and a PhD in aerospace computational engineering from MIT.