A Loci-based High-Order Production Solver for Aerothermochemical Modeling of Hypersonic Entry Systems on Modern Heterogeneous Supercomputers

Background: NASA’s stated goals include crewed missions to the Moon by 2024 and to Mars in the 2030s. This growth of mission complexity imposes new challenges on spacecraft design. The entry system, which is of direct relevance to the survival of spacecraft, must be designed with in-depth knowledge of the aerothermochemical environment of the spacecraft in flight. In reality, the entry environment cannot be replicated on the earth ground; hence, computer simulation techniques play a major role in modeling the entry environment and assisting with entry-system design.

A number of computer codes have been deployed in NASA for entry-system modeling (e.g DPLR, FUN3D and Loci-CHEM). These codes share one common characteristic—they are all based on the second-order finite-volume discretization of reactive-flow governing equations. The computer codes of such type struggle to provide high-accuracy and robust predictions of flow-field and heat-transfer at hypersonic speeds when a strong shock (typically above Mach 6) forms ahead of space vehicles. Specifically, the numerical solutions often exhibit deficiencies, including i) unreliable surface heating predictions due to the nonphysical post-shock entropy fluctuations; ii) extreme solution sensitivities to the choice of flux formulation and the setup of the computational grid; and iii) slow solution convergence, which hinders the timely delivery of useful results.

Goals: The proposed effort will address the limitations of NASA’s current software and modeling tools and establish a prediction capability for aerothermochemical modeling of entry systems based on the high-order accurate discontinuous Galerkin (DG) scheme and the heterogeneous programming model. Recent studies showed that DG can offer superior accuracy when used for shock simulations, capable of providing oscillation-free shock profiles and highly-resolved thin thermal boundary layers. The DFEM method will address deficiencies i) and ii) mentioned above. Moreover, issue iii) will be addressed by enabling a Loci/CUDA or Loci/OpenCL hybrid programming model to fully utilize the computing resources on modern heterogeneous supercomputers.

This research will leverage our unique high-order DG solver—Loci-THRUST, developed for chemically reacting flows and shock-capturing (Lv & Ihme, J. Comput. Phys., 2014, 2015; Lv et al., AIAA 2019-2168). A list of tasks will be implemented: i) Develop efficient, robust and time-stable algorithms to incorporate the physical models of vibrational-energy relaxation, thermal radiation, and ionization reaction into the reactive-flow DG formulation. ii) Extend the Loci-THRUST framework by hybridizing the Loci-programming model with OpenCL and CUDA languages to utilize the heterogeneous processors and complex memory hierarchy. iii) Implement OpenCL and CUDA algorithmic kernels and enable proper interfacing with Loci. iv) Carry out code verification and demonstrate the code speedup in classical test cases, such as hypersonic-flow-over-cylinder and double-cone configurations. v) Benchmark the new solver against NASA’s production code, DPLR, and provide best practice guidelines for using the new entry-system modeling capabilities.

Impacts: The proposed project aims to develop a robust entry modeling research program in the State of Mississippi. The impacts of the proposed research will include 1) establishment of a multi-institutional research program that can successfully compete for extramural funding, 2) enhancement of the research tracks of the involved researchers by producing high-quality conference papers and journal articles, 3) delivery of a upgraded version of Loci-Thrust solver with advanced aerothermochemical modeling capabilities to NASA, and 4) enhancing the STEM educational infrastructure through funding graduate students and helping train next generation of researchers and engineers in a timely research topic.

This is a mature technology (TRL 7+) — the realistic path in is usually NASA's Technology Transfer Program: licensing an existing NASA patent, or a Space Act Agreement to use NASA facilities/expertise directly. NASA also runs a startup licensing program with no upfront fee for companies formed to commercialize a specific NASA technology.

None of these are guaranteed paths for this specific project — TechPort itself doesn't have an "apply" button. Reaching out to the contact(s) above with a specific question is usually the fastest way to find out what's actually open.