Computational Fluid Dynamics Blogs

This past summer, Cadence began funding research at the Massachusetts Institute of Technology (MIT) and the University of Tennessee at Knoxville (UTK) on a topic of significant current interest to the aerospace CFD community: mesh adaptation for high-speed flows.

The work is being performed under the name “Rapid and Robust Analysis of High-Speed Environments Using Inviscid Output-Based Adaptive-Mesh Solvers.” It brings together a team of world-class researchers including Marshall Galbraith (MIT), Steve Allmaras (MIT), and James Coder (UTK). Coordinating the work on Cadence’s side is Nick Wyman.

“We greatly appreciate Cadence funding this research. It is an excellent opportunity to advance the state of the art in automation and reliability of CFD simulations by combining the ground-breaking flow solver and adaptation technology encapsulated in SANS with the proven Pointwise meshing capabilities. We are optimistic that we can make positive contributions in the difficult regimes of supersonic and hypersonic flow simulations.”
– Steve Allmaras, Research Engineer, MIT Department of Aeronautics and Astronautics.

The research concept involves use of MIT's SANS flow solver which combines high-order finite elements with output-based adaptation. The adaptation process iteratively modifies the mesh to minimize numerical errors in engineering quantities of interest, such as lift, drag, and heat transfer. For each mesh, numerical errors are estimated, and an optimization is performed to minimize those errors with respect to mesh scales. The output of this optimization is a fine-grain point cloud of mesh scales that is then used by the Cadence Pointwise Meshing software to generate a better mesh for the next iteration.

Can you optimize an airfoil for drag starting from a crude shape defined with just four points? This early result (shown above) of the proposed adaptation procedure demonstrates that you can do so for a Joukowski airfoil. (Mesh cells colored by pressure.)

The entire process is automated and produces a solution with appropriate resolution of all relevant flow features, including shock waves and boundary layers. The coupling with SANS allows the industry leading meshing capabilities in Pointwise to be driven in an automated manner, resulting in flow solutions with minimal numerical error.

“Grid generation is still a major bottleneck in high-speed flow simulations. The process is usually artisanal, and you need a person in the loop to iterate between the solver and the mesh to refine regions like shock waves and entropy layers. Output-based mesh adaption brings in much-needed mathematical rigor, and assures that all salient physics are considered. This also includes real-gas effects along with thermal and chemical non-equilibrium.”
– Jim Coder, Asst. Prof. Mechanical, Aerospace, and Biomedical Engineering Dept., University of Tennessee Knoxville.

This research runs through June 2022 and we’ll be back with news about what we’ve learned and how it will influence new capabilities in Pointwise.

If you’d like to try Cadence Pointwise Meshing and see its current mesh adaptation and other capabilities in action, request a free trial today.