Thermoelastic Reduced Order Models For Representative Hypersonic Structures

Project: Research project

Project Details


Accurate and computationally efficient prediction of the dynamic response of hypersonic aircrafts panels is a challenge. The complexity of the problem stems from (1) the severity of the thermal and aerodynamic loading that induces large, geometrically nonlinear motions of the structure, and (2) the multi-disciplinary coupling. In recent years, the adoption of reduced order models for the prediction of the structural response [1-13] has emerged as the option of choice to avoid the computational burden associated with full finite element computations while still maintaining accuracy. The application of reduced order modeling strategies to the aerodynamic computations and the thermal analyses have also been proposed primarily separately (e.g. see [16,17] and references therein) but also integrally coupled with a structural reduced order model [14,15,18-20]. These investigations have demonstrated both accuracy and computational benefits. There are extensions and further validation of reduced order modeling methods that must be carried out to transition these novel techniques to routine design and analysis tools. For example, it was recognized that the thermal-structural problem is in principle fully coupled but the feedback of the structural motions on the thermal problem, i.e. the latency effect and the change in geometry, were neglected resulting in a one-way thermal to structural coupling. The implications of this assumption are not clearly understood at this point and ought to be clarified. Further, a coupling with the aerodynamics of both structural and thermal reduced order model is necessary to complete the three-discipline interaction problem The focus of the present investigation is to apply the present body of work towards representative hypersonic structure with accompanying realistic environments. It is proposed to validate this extension of the methodology on a representative hypersonic panel design, with quasi-steady thermal, aerodynamic, and acoustic loading obtained from the aerothermodynamic analysis of [16]. This investigation provided full-fidelity thermal and structural results, serving as the necessary baseline to validate the thermoelastic reduced order models. Note that full dynamic analyses could not be carried out in [16] because of computational limitation induced by full finite element models. A first structural reduced order model of the cold panel has recently been obtained and its complete validation, and fine tuning is in process a first step in demonstrating the applicability of reduced order models to such panels. The development of a reduced order model of the temperature field that permits the accurate representation of the data presented in [16] has also been recently achieved and is being coupled with its structural counterpart including temperature dependent properties. The availability of a combined structural-thermal reduced order model will permit the assessment of the validity of one way, thermal-structural coupling but is missing a key component of the problem, i.e. the aerodynamics. Indeed, the aerodynamics affects both the structural response and the temperature field, the former directly through the aerodynamic forces and the latter through the heat flux. The focus of the planned effort is accordingly on the extension of the current computational models to include the aerodynamic and its coupling to the structural and thermal reduced order models. Possible choices for the aerodynamic modeling are piston theory with flux prediction, full order CFD codes (such as CFL3D), or aerodynamic reduced order models. The focus of this coupling effort will be to recover and extend results presented in particular in [16] with full order models.
Effective start/end date9/16/134/7/14


  • DOD-USAF-AFRL: Air Force Office of Scientific Research (AFOSR): $74,849.00


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.