Advancing CFD with AI: Surrogate Modeling Approaches in the NASA Software Ecosystem

Rescale's presentation in NASA's Advanced Modeling & Simulation Seminar Series highlighted the transformative potential of AI-powered surrogate models in aerospace design, addressing the limitations of traditional high-fidelity simulations like computational fluid dynamics (CFD) and finite element analysis (FEA) that often take days to complete. These surrogate models, built using AI physics methods, can predict simulation outcomes up to 1,000 times faster, enabling rapid exploration of the design tradespace and increasing the number of design iterations without additional computational resources. During the seminar, Viktor Rozsa from Rescale demonstrated how companies like Boom Supersonic are leveraging these models alongside NASA's simulation software to enhance digital engineering workflows. The series, hosted by NASA's Computational Aerosciences Branch, underscores the importance of integrating robust simulation data with optimized hyperparameters to ensure accurate and reliable predictions, ultimately expediting design processes across aerospace and other industries. Rescale's integration with NASA software, such as FUN3D and Cart3D, automates data extraction and standardization, facilitating the development of AI surrogates and offering significant efficiency improvements in complex engineering simulations.