Computational Fluid Dynamics
Willem Toet
![Willem Toet]()
Willem Toet
F1 and Aerodynamics Expert. Over 30 years in F1 Aerodynamics, now Professor of Motorsport Engineering (UoB), Aerodynamics Advisor & Consultant, former F1 Head of Aerodynamics at Sauber, Ferrari, Benetton, BAR Honda
OK, I've started it now, I'm giving away old secrets. Here is a 3 part series on how to process the information you get from CFD cases. I'd love to create a follow up with input from other experts out there in any area of this science - particularly those working in the fields of aerospace, yachting / boating and biology. Great tools for understanding fluid flow these and we should promote them more!
Please direct any comments to me here or via LinkedIn messaging if we are connected. I log in most days. Sauber Aerodynamics also offer CFD services to customers if there is an interest but this post is for students of aerodynamics interested in using CFD.
More posts from me here
More posts also on RaceTech
Project Lead | Project Management | Team Leadership | Risk Management
5 年Great article, definetly sending the fsae-a students here.
Customer Reliability Engineering Lead
5 年Willem, another great resource for all the new engineers out there! You're a great asset to the industry. From experience, I would not recommend using surface contours of CpT. Total pressure is a volumetric property and makes little sense plotting it on a surface. Also CpT on a surface is just a projection of the first cell, so it depends heavily on the local meshing strategy. It's the same thing as velocity/dynamic pressure; surface velocity should be 0 unless a moving wall BC, but CFD sometimes reports first cell velocity instead. This obviously affects the dynamic pressure value, and therefore the surface CpT. Obviously, the same goes for wall velocity components. For looking at near surface flow, tau seems to be a wise choice as it is less mesh dependent (law of the wall is still very mesh dependent) Interesting to hear your opinion!
Global Adoption Manager at ESI Group
6 年Nice article! I would add y+ value on the car surface. When using wall model, you want to know in which range you are. I also like to visualize the grid with the flow structure. For example, I want to make sure that around the wheel I have not been to conservative with my grid refinement/coarsening. Concerning the laminar/turbulence issue with RANS, do you know if F1 teams are now using the Langtry/Menter model (https://turbmodels.larc.nasa.gov/langtrymenter_4eqn.html)? In addition of FLUENT, It is now in STARCCM and OpenFOAM as well. Something that is becoming more popular is the html report, a little bit nicer that the statics pdf report and easier to compare cases.?
Aerodynamicist at Williams Racing
6 年Thanks for sharing your experience and knowledge! Tried out a few of the tips and found it much easier to understand the CFD results. I am definitely bookmarking this post and going to share it with others interested in learning CFD analysis.
RBF Morph Founder - Associate Professor of Machine Design
6 年https://www.ansys.com/blog/nascar-rescale-high-performance-computing/