Multifidelity Modeling of Rotating Detonation Rocket Engines
Abstract
Detonation-based combustion is being increasingly seen as a game-changing technology for realizing novel propulsion systems. Specifically, compact and unconventional combustor designs with high energy density can be obtained. However, such gains are feasible only if the complex interactions between shocks, chemical reactions and turbulence that drive energy release are fully understood. Due to the multiscale nature of this problem, detailed numerical simulations remain a useful, albeit challenging approach to gaining physics insight. In this proposed work, state-of-the-art computational tools that leverage emerging supercomputing frameworks, utilize machine learning based tools to accelerate simulations, and resolve the detonation structure are used to study a canonical rotating detonation rocket engine (RDRE). In particular, the interaction of fuel/air injection process, subsequent mixing and shock-driven chemical reactions will be studied. Based on these simulations, a reduced-order model for predicting performance characteristics will be developed. Using data assimilation procedures for uncertainty quantification, a Bayesian design framework is then proposed. When fully integrated, this multifidelity design approach can be used to optimize device performance for a given nominal geometry. Through close interaction with AFRL, candidate geometries and operating conditions will be used to exercise this design framework.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 21, 2022
- Source ID
- FA95502110079XX0
Entities
People
- Venkatramanan Raman
Organizations
- Air Force Office of Scientific Research
- Board of Regents of the University of Michigan
- United States Air Force