Skip to main content
NASA Advanced Supercomputing (NAS) Division World-class modeling, simulation, and supercomputing in support of the NASA mission

To Mars! Machine Learning with Supersonic Retropropulsion Wind Tunnel Test Data

AMS Seminar Series

Speaker: Matthias Ihme, Stanford University

Presentation

Seminar Slide Deck (PDF-8.2MB)

Abstract

Human missions to Mars will require rocket-powered descent through the Martian atmosphere starting at supersonic conditions because parachutes alone will not provide sufficient deceleration. This presentation explores machine-learning methods to advance the fundamental understanding of multi-nozzle plume physics, to quantify uncertainties, and to inform deployment of supersonic retropropulsion (SRP) technology. The data was collected from NASA wind-tunnel tests for a wide range of operating conditions. We analyze this data to generate knowledge to then develop robust and efficient physics-informed tools which will support the design of SRP vehicles. We present a hierarchical data-driven modeling framework that combines physics-based models with residual learning to account for uncertainties and non-linear complexities not contained in the low-order physics-based model. The resulting framework shows promising results and is capable of efficiently processing new data. Wind tunnel testing, currently performed at NASA Langley, will provide more data for us to process through our framework to further the understanding of SRP.

Acknowledgments: David Wu, Wai Tong Chung, Karl Edquist, Ashley Korzun, NASA ESI Program

Biography

Photo of Matthias Ihme
Matthias Ihme is Professor in the Department of Mechanical Engineering and the Department of Photon Sciences at Stanford University. He holds a BSc. degree in Mechanical Engineering and a MSc. degree in Computational Engineering. In 2008, he received his Ph.D. in Mechanical Engineering from Stanford. After being on the faculty of the Aerospace Engineering Department at the University of Michigan for five years, he returned to Stanford in 2013. He is a recipient of the NSF CAREER Award (2009), the ONR Young Investigator Award (2010), the AFOSR Young Investigator Award (2010), the NASA Early Career Faculty Award (2015), the Hiroshi Tsuji Early Career Research Award (2017), and the Bessel Award of Alexander von Humboldt Foundation (2021). His research interests are broadly on the computational modeling of reacting flows, the development of numerical methods, and the investigation of advanced energy conversion concepts and molecular processes.