ENABLING THE STRUCTURAL DESIGN OF HIGH-MACH, HIGH-ALTITUDE VEHICLES THROUGH A HOLISTIC APPROACH TO UNCERTAINTY MODELING
Abstract
The fielding of hypersonic vehicles has long been a goal of the U.S. Air Force to increase its operational envelope. The design of such vehicles is a new paradigm in which safety factors are no longer appropriate to address the uncertainties resulting from vehicle-to-vehicle variability (aleatoric uncertainty) and from modeling assumptions (epistemic uncertainty); they would lead to overweight vehicles unable to carry out the planned missions. The successful design of hypersonic vehicles will require the modeling of all aleatoric and epistemic uncertainties and their propagation to the structural response in a computationally viable framework.The focus of the proposed effort is thus on formulating and assessing a holistic uncertainty modeling of all three disciplines of the coupled structural-thermal-aerodynamic system which incorporates both aleatoric and epistemic uncertainties and permits the balancing of computational efficiency versus overall uncertainty, a process referred to here as uncertainty management. The enabling framework for this effort is the class of nonlinear coupled structural-thermal reduced order models (ROMs) developed and extensively validated by the P.I. These ROMs provide a computationally efficient and accurate platform to predict the response but also condense both aleatoric and epistemic uncertainties into the ROM parameters and a limited number of other terms the modeling of which can then be addressed stochastically. Three specific research thrusts are envisioned. The first one focuses on the formulation, identification, and validation of uncertain structural-thermal ROMs combining random matrix theory and polynomial chaos representation. The second thrust addresses computational aspects of the uncertain ROMs by introducing sparsity in the ROM governing equations which decreases dramatically the computational effort for predicting the response. Finally, Thrust 3 focuses on the aerodynamic problem, seeking to replace high order models (e.g., from computational fluid dynamics) with low-order ones in which the associated epistemic uncertainty has been explicitly quantified and modeled.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Mar 24, 2016
- Source ID
- FA95501610021
Entities
People
- Marc Mignolet
Organizations
- Air Force Office of Scientific Research
- Arizona State University
- United States Air Force