Real-Time Parameterized Reduced-Order-Model (ROM)-Based Aeroservoelastic Simulator

ABSTRACT: The main objective of this SBIR Phase I proposal is three-fold: to design an innovativecomputational technology based on parametric, linearized and nonlinear aeroelastic/aeroservoelastic Reduced-Order Models (ROMs); to develop its core components; and to verify and validate them. The overall technical approach will be based on the state-of-the-art of Projection based Model Order Reduction (PMOR) methodology. Specifically, adaptive and non-adaptive greedy algorithms will be developed for sampling the parameter domain. For linearized problems, the PMOR methodology will be based on the concept of a database of point-wise ROMs and its associated real-time algorithms for interpolation on matrix manifolds. For nonlinear ROMs, it will be based on the concept of global ROMs with local reduced-order bases and equipped with a structure-preserving hyperreduction procedure. To account for model and parametric uncertainties and any other sources of error, ROMadaptation will be performed based on test data using a novel nonparametric probabilistic approach. The proposed computational technology will be operable as: a predictive tool to estimate aeroelastic/aeroservoelastic stability and aerodynamic loads prior to testing; a real-time system that takes inputs from a given live flight test and outputs a corresponding estimate of the aeroelastic/aeroservoelastic stability and aerodynamic loads; a pilot-in-the-loopaeroservoelastic/aeroelastic simulator.; BENEFIT: The end product of this SBIR effort will be a computational technology for the reduction of parameterized, linear and nonlinear, high-dimensional, aeroelastic --- that is, fluid-structure --- computational models based on the coupling of Computational Fluid Dynamics (CFD) and Computational Structural Dynamics (CSD) representations of fluid and structural systems, respectively, in view of enabling parameterized aeroelastic analyses in real time. For a flight test center, it will enable the online estimation of aeroelastic/aeroservoelastic stability and aerodynamic loads prior to testing, and therefore improve the efficiency of flight testing. It will also enable the predictive simulation of "as flown" rather than "as planned" sorties, which will minimize the sources of error in flight testing and reduce the number of needed test points. It will be also operable as a pilot-in-the-loop aeroservoelastic/aeroelastic simulator: therefore, it will upgrade existing flight simulators with flexibility effects and provide a broader base for educating and training both discipline engineers and Test Pilot School students. The benefits of this anticipated end product will also extend to the aeronautics industry at large. Indeed, because it will enable real-time predictive aeroelastic analyses, this product will pave the way for introducing aeroelasticity into the design process at an earlier stage than currently done, thereby contributing to bridging the gap between analysis and design.More importantly, the market for the end product of this SBIR effort will be much larger than that of flight test centers and the aeronautics industry. Indeed, there is a pressing need today for predictive, CFD-CSD simulations for the solution of many important engineering problems. These include, to name only a few: the aerodynamic rehabilitation of suspension bridges where finding a way to avoid repeating the cycle of deterioration and crash repairs continues to challenge civil engineers; the aeroelastic tailoring of front and rear wings of Formula 1 and other race cars; the control of flow-induced noise in various military and commercial systems; and the design and test support of biomedical micro devices for drug delivery. Just like the design and test support of high-performance aircraft (where flutter, LCO, and buffeting remain among the most important considerations) are mission critical for the military and many commercial aerospace companies, these problems and applications are of significant concern to the high-end civil and marine engineering companies, automotive industry (automobile, train, ...), and an important segment of the emerging bioengineering industry such as that focusing on diagnostics in cardiology. Unfortunately, in all of these areas, high-fidelity CFD-CSD simulations remain so computationally intensive that they cannot be performed as often as needed, are more often performed in special circumstances than routinely, and are avoided for time-critical applications. However, because it will enable the construction of parametric CFD-CSD Reduced-Order Models (ROMs) that are operable in real time, the end product of this SBIR software will change this situation. Because its underlying technology for constructing such ROMs will be based on technologies for constructing individual CFD and CSD ROMs and coupling them, it will be a game changer not only for coupled CFD-CSD, but also CFD and CSD. For all these reasons, it will disrupt the markets of all of CFD, CSD, and CFD-CSD.