Non-intrusive uncertainty quantification, that does not require the CFD code to be modified, has been identified as an enabling technology by NASA to advance the role of computational fluid dynamics codes in the Design Development Research and Engineering community developed by the aerospace industry, ultimately leading to utilization for flight certification. The proposed computational product offers a direct solution to link the various sources of uncertainties to predictions made by CFD tools, thereby enabling the usability of CFD tools for making risk-informed design decisions. The adaptive sparse grid method offers a significant advantage over other uncertainty quantification methods due to the ability to handle non-smooth system response with complex probability density distributions and much smaller number of required CFD simulations. This product can be a highly effective tool for wider applications requiring aerothermodynamics calculations where the lack of confidence in modeling parameters and predictive capability of the CFD codes has limited their impact.
The work established in this project including the Uncertainty Quantification workflow process and automated software tool can be generalized for potential applications to a wide range of applications utilizing CFD software. More specifically, this work can be transitioned to support a significant number of other non-NASA applications where reacting CFD modeling tools are utilized. Energy and propulsion applications such as gas-turbine combustors, augmentors, rockets, and many others can benefit from the product developed in the proposed work.
More »