Discovering & Modeling Turbulence and Chemistry Interactions in High Speed Reactive Flows
Abstract
High-speed reacting flows, such as those in scramjet and ramjet engines, are characterized by highlevels of compressibility, localized chemical heat release, and strong thermal and pressure/densitygradients. Turbulence and chemistry mutually interact with each other, and small-scale processessuch as baroclinic torque and dilatation can transfer energy to the large scales, directly affectingbulk motion. High-speed flows are also impacted by the geometry and flow conditions that dictateshock dynamics. These factors can lead to departure of the flow from homogeneous, isotropic, andequilibrium turbulence. Clearly, many of the above features are applicable also to external flows ofhypersonic vehicles. Under these conditions, turbulence and turbulent mixing, chemistry, and theirinteractions are tightly coupled with the boundary conditions. The coupling renders direct numericalsimulation (DNS) of canonical, equilibrium turbulent flows not useful for yielding physical insightsrequired to model high-speed reacting flows.This MURI effort will leverage recent advances in DNS of turbulent reacting flows, laser/opticaldiagnostics of internal and external flows, mathematical theories, high performance computing, anddata science to build a multiscale modeling framework to systematically unravel the underlyingphysics, and to define and provide a path to quantitative predictions ing framework features a heterogeneous modeling method in which an array ofDNS domains (high resolution simulation) are embedded as sensors in a coarse modeling schemefor probing and quantifying turbulence-chemistry interactions. Combined with reliable data, to becollected in a tightly coupled experimental program, and mathematical and data assimilation tools,the physics revealed by embedded DNS will be used to inform and refine model structures. Thecoarse model, in turn, provides the relevant boundary and initial conditions for the embedded DNS.Mathematical approaches will be developed and implemented to assimilate the embedded DNS andexperimental data seamlessly into the model. Through iterations, this modeling approach is expectedto reveal physical insights into turbulence-chemistry interactions with increased accuracy and toprovide modeling capabilities with improved fidelity.We specifically focus our effort on utilizing geometry and boundary conditions that replicate highspeedflight trajectories. The iterative modeling process enables us to explore new flow paths whilefacilitating a broad range of model validation. Our emphasis will be to probe and create flow conditionsthat would target certain small-scale processes, e.g., heat release, to directly impact bulk flowmotions.The research team consists of experts in computational modeling (V. Raman (UM), M. Panesi(UIUC), C. Scalo (Purdue), H. Wang (Stanford)), experiments (T. Lee (UIUC)), and applied mathematics(R. Ghanem (USC)). We will also collaborate extensively with University of Queensland inAustralia to leverage their external flow facilities. The outcomes of this project include a) a theoreticalunderstanding of the small-scale processes and their influence on large scale flow motion, b)a set of computational tools for predictive modeling of the nonequilibrium flow paths, including awell-trained large eddy simulation (LES) framework; c) a comprehensive database of experimentsof a scramjet with axisymmetric/perisymmetric design and a modified BOLT-type geometry basedexternal flows; and d) a mathematical framework for integrating experiments and high-resolutionsimulations through data assimilation, which allows extraction of model structures and uncertaintyquantification.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- May 05, 2021
- Source ID
- N000142112475
Entities
People
- Venkatramanan Raman
Organizations
- Board of Regents of the University of Michigan
- Office of Naval Research
- United States Navy