NextSim coordinator Oriol Lehmkuhl (BSC) participates in 2021 AIAA Aviation Forum as a panelist at the CFD2030: Physical Modeling: Essential for Successful Model-Based Engineering in Aerospace panel discussion on 2 August 2021.

More about CFD2030:

Advances in high performance computing capabilities have enabled higher resolution simulations that significantly reduce the numerical errors associated with the application of CFD to aerospace design. Increased computing capacity enables expanded use of scale resolving turbulent simulation methods as well as higher fidelity models of combustion and transition to turbulence. Currently, the large number of simulations required for conceptual design, detailed design, and certification preclude the use of wall resolved large eddy simulation (LES) for the bulk of CFD applications related to product development. Despite improvements enabled by increased computational power, errors associated with physical modeling of transition, turbulence and combustion processes remain an obstacle to rapidly accelerating the use of CFD to reduce design cycle duration and development testing requirements. The lack of accurate, robust, and automated methods for prediction of transition remains a challenge, particularly for vehicles designed for use with significant extents of laminar flow. Models employed to account for the effects of turbulence including RANS, hybrid RANS-LES and wall modeled LES each have unique accuracy limitations relative to wall resolved LES and test derived data. In addition, they each have very different computational requirements. A particularly challenging problem is quantification of physical model uncertainty for vehicles without extensive existing test data. To realize the CFD 2030 Vision, significant advances in both accuracy and computational efficiency of physical modeling approaches are required. This panel will discuss the critical physical modeling issues that need to be addressed and potential solutions to enable the virtual model-based engineering design paradigm.