You are here

Multicomponent Reduced Order Modeling of Hypersonic Boundary Layers

Description:

OUSD (R&E) CRITICAL TECHNOLOGY AREA(S): Trusted AI and Autonomy

 

The technology within this topic is restricted under the International Traffic in Arms Regulation (ITAR), 22 CFR Parts 120-130, which controls the export and import of defense-related material and services, including export of sensitive technical data, or the Export Administration Regulation (EAR), 15 CFR Parts 730-774, which controls dual use items. Offerors must disclose any proposed use of foreign nationals (FNs), their country(ies) of origin, the type of visa or work permit possessed, and the statement of work (SOW) tasks intended for accomplishment by the FN(s) in accordance with the Announcement. Offerors are advised foreign nationals proposed to perform on this topic may be restricted due to the technical data under US Export Control Laws.

 

OBJECTIVE: To develop multi-physics component-based reduced order models (ROMs) and associated interfaces to accelerate high-fidelity design tools for predicting detailed time-accurate hypersonic vehicle flow-fields.

 

DESCRIPTION: The Army is interested in designing next-generation hypersonic flight vehicles with enhanced system speed, reach, and lethality addressing Army’s and DoD’s Priorities in Long Range Precision Fires and Hypersonics. Revolutionary systems must meet new tactical requirements for performance, reach, and lethality. Computational fluid dynamics (CFD) has played a central role in the design and development of hypersonic vehicles, in part due to prohibitive costs associated with testing facilities. However, existing CFD approaches have prohibitive computational costs when attempting to predict high Reynolds number hypersonic aerothermodynamics and its interactions with fully resolved physical processes. Hypersonic modeling under realistic flight conditions is complicated by the nonlinearity and multiphysics nature present that acts across a wide range of scales [1]. Variations in atmospheric conditions, chemical kinetics, vibrational excitation, ablation products, and gas-surface interactions further complicate high enthalpy flow and plasma [2]. Recent detailed direct molecular simulations have also demonstrated macroscopic impacts of complex transport phenomena often omitted in coarse grained models such as Large Eddy Simulation (LES) and Reynolds Averaged Navier Stokes [3,4].  The computational expense associated with solving complex coupled fluid, thermal, kinetic, and structural problems currently significantly limit the rate at which design space can be accurately explored. Recent success accelerating modeling of complex flows of similar computational complexity through component-based ROMs[5] suggests potential for model acceleration strategies that exploit coupling of local mesoscale ROM domains. Newly developed localized ROM domain partitioning, nonlinear compression, and adaptivity suggest potential to attain greater efficiency and scalability than start-of-the-art models through the mitigation of high Kolmogorov n-width complexity associated with device scale transient turbulent flows.  To shorten design cycles, revolutionary capabilities for accelerating high fidelity external and internal aerothermodynamics such as these must also be integrated with associated multi-physics couplings. The Army is therefore soliciting scalable adaptive model order reduction technologies capable of recovering high-fidelity predictive power for the flight environment of a hypersonic vehicle, along with associated gas-flow chemistry, detailed transport, shock induced heating, and their associated material-responses. The goal will be to achieve at least an order of magnitude reduction of computational cost versus existing wall-resolved LES (WR-LES) techniques while recovering full Direct Numerical Simulation (DNS) accuracy levels on transitional flows where wall-modeled (WM-LES) results diverge from WR-LES and DNS solutions.  While either non-invasive data-driven model order reduction or fully invasive ROM technologies will be considered, priority will be given to approaches that develop modular compressed bidirectional data interfaces that enable tight coupling among diverse physics tools with improved scalability.  The new tools should be able to handle realistic glide body, missile geometries, and scramjet propulsion systems for sustained powered flight in the Mach 6 to 20 range. Models that reduce sensitivity to near-wall mesh quality are particularly encouraged. Tools must have the ability to be deployed in traditional/emerging high performance computing architectures (CPU GPU) efficiently and demonstrate efficient weak scaling at least competitive with LES models. Compressed interfaces for in-situ visualization and data extraction techniques enabling seamless navigation of the sea of data encountered in real-time analysis is also encouraged.

 

PHASE I: Develop component-based reduced order model (ROM) technologies that, once trained, demonstrate accurate 3D high-fidelity prediction of transient hypersonic boundary layer flows with transition for novel flow conditions within training set bounds.  Attain order of magnitude reduction in memory footprint compared to existing state-of-art WR-LES of commensurate accuracy without explicitly defined wall models. Ability to recover DNS level solution accuracy in geometries incompatible with existing wall models should also be demonstrated with order of magnitude speedup relative to state of the art fully resolved high-order DNS solutions.  Performance scaling for high fidelity reacting turbulence commensurate with DNS solutions with finite rate chemical kinetics and detailed transport should also be demonstrated.  The company should identify strengths/weaknesses associated with alternative solutions, methods, and new concepts. Demonstrate theoretical credibility of proposed computational methods. Computational vetting and demonstration of concepts to be conducted using canonical blunt-nose single or double cone hypersonic shapes and simple flameholder geometries at minimal Reynolds number required to demonstrate transitional flow behaviors is suitable in this phase.  Solutions capable of maintaining order of magnitude speedups with ROM training or adaptation time included without loss of predictive accuracy across parametrically varying geometric configurations are highly encouraged.

 

PHASE II: During Phase-II, the framework developed in Phase-I will be extended and validated to support hypersonic design of potential applications in air-breathing missiles, boost-glide missiles, and high-maneuver interceptors. Tools should demonstrate ability to model complex aerothermochemistry, transport, thermoacoustics, shock induced heating, and structural material responses with statistical properties shown to converge towards DNS and canonical experimental data with computational cost at least one order of magnitude below WR-LES models for equivalent conditions. Fluid-structure component-based ROM coupling that enables conjugate heat transfer and fluid structure interaction calculations that accurately model sharp features resulting from shock-heating are also encouraged. The tools should inherit the ability to capture in detail non-equilibrium processes including boundary layer transition to turbulence, onset of material ablation, finite-rate non-equilibrium chemistry, and gas-surface interactions responsible for surface deformation from baseline full-order DNS models. Tight-coupling of time-accurate predictions through compressed component-based ROM interfaces for fluid structure interactions for parametric variations of both flow boundary conditions and design properties for external hypersonic vehicle flows should be demonstrated. Complete model, multi-physics ROM interface application programming interfaces (APIs), and executable code for deployment on state-of-the-art high performance computing systems with demonstrable performance on existing or emerging computing architectures is expected.

 

Teams must demonstrate model validation by comparison with experiments, reference DNS databases in the open literature, or data from the Army or DoD laboratories. Capturing turbulent transition at dramatically reduced computational complexity for arbitrary geometries and flow conditions is emphasized. The complete software package shall be available to ARL during all phases of the project to conduct independent assessment and vetting of the developed tools. Development efforts will be coordinated with the government and potential prime-contractor partners to ensure product relevance and compatibility with missile defense projects and government modeling and simulation systems. While compatibility with specific production codes is not required, selection will favor projects with viable transition strategies for either enhancing or supplanting production codes in existing high-cost Multiphysics analysis pipelines.  The developed complete computational tool sets along with user guide(s) at the end of Phase-II shall be delivered to ARL for government use on HPC platforms to conduct mission projects.

 

PHASE III DUAL USE APPLICATIONS: This work will enable collaboration with high-fidelity simulation model developer(s) and/or user(s) on integration of product(s) into accelerated missile defense application pipelines. Long-term optimization of toolsets and APIs to accommodate new advances in the technology of tracking and prediction of glide body or cruise missile flight will continue. Technology will transition to an appropriate government or defense contractor for integration and testing. Integration and validation into design cycles for a real-world missile defense application will continue.

 

REFERENCES:

  1. Candler, G.V., “Rate effects in hypersonic flows”, Annual Review of Fluid Mechanics, vol. 51, pp. 379-402, 2019.;
  2. Bisek, N. J., I.D. Boyd, J. Poggie, "Numerical Study of Plasma-Assisted Aerodynamic Control for Hypersonic Vehicles", AIAA J. Spacecraft and Rockets, vol. 46 (3), 2009;
  3. Grover, M.S., A. M. Verhoff, P. Valentini, N. J. Bisek, “First principles simulation of reacting hypersonic flow over a blunt wedge”, Phys Fluids 35, 086106, 2023;
  4. Morreale, B.J., J. Shine, R. D. Bowersox, N. Bitter, R. Wagnild, “Hypersonic Multi-Fidelity Turbulence Modeling on a Mach 5 Blunt Ogive with Cool Walls”, AIAA 2023-0455, 2023;
  5. Huang, C., K. Duraisamy, C. Merkle, “Component-Based Reduced Order Modeling of Large-Scale Complex Systems”, Front. Phys., 2022;
  6. US3D: Aerodynamic and Aerothermodynamic Simulations Software (20110126, Dr. Graham Candler)

 

KEYWORDS: Reduced order modeling, hypersonics, aerothermochemistry, computational fluid dynamics

US Flag An Official Website of the United States Government