CFD settings
Indeed, without proper information on the CFD settings, it's difficult to judge the quality of these results. This includes:
- Mesh resolution: smallest cell size, refinement strategy, ...
- Mesh type: hexagonal? Prism layers? ...
- Solver: which solver was used? Steady or unsteady? RANS or LES or ...? Were the results averaged? Were wall functions used? Y+ values?
- Wheel rotation: which technique was used for wheel rotation? Just a "tangential wall velocity", or MRF, or a sliding mesh, or ...?
Judging by the resolution of the flow pattern at the bottom, by the fact there is enough mesh to capture the details & gaps between wing elements, ... I'd say there is at least a decent amount of cells. But then again, that's just a very indirect estimation :)
3D model
- Rake: the car is running at quite a high rake. This is just one of the "infinite" number of possible combinations for rear and front ride height and so on. So to get a full aero map, you would need to run the car at various positions.
- Tyre compression: the tyres seem to have been cut-off at ground level, which is a common technique to add in the effect of compression. But the "height" of cut-off is subjective and has a large effect on the aerodynamics, as it influences the ride height and thus the underfloor aerodynamics. Also, there is no tyre deformation in the 3D model (or perhaps I missed it, difficult to tell from the screenshots).
Analysis versus optimization
This is an analysis, which tells us quite a lot. But it doesn't teach us how the team interprets results or, even better, how it translates this into improved designs. It would be cool to see how an adjoint solver would improve this car towards more downforce for example!
(if you don't know what adjoint is, check this video - https://airshaper.com/videos/aerodynamic-shape-optimization-...).
Overall, I find it very interesting - even if the model, CFD setup, ... aren't an exact match with what the guys at Williams are doing, it does provide a lot of interesting insight into existing flow structures on such cars - which is an interesting tool for learning.
CFD settings Indeed, without proper information on the CFD settings, it's difficult to judge the quality of these results. This includes: - Mesh resolution: smallest cell size, refinement strategy, ... - Mesh type: hexagonal? Prism layers? ... - Solver: which solver was used? Steady or unsteady? RANS or LES or ...? Were the results averaged? Were wall functions used? Y+ values? - Wheel rotation: which technique was used for wheel rotation? Just a "tangential wall velocity", or MRF, or a sliding mesh, or ...? Judging by the resolution of the flow pattern at the bottom, by the fact there is enough mesh to capture the details & gaps between wing elements, ... I'd say there is at least a decent amount of cells. But then again, that's just a very indirect estimation :)
3D model - Rake: the car is running at quite a high rake. This is just one of the "infinite" number of possible combinations for rear and front ride height and so on. So to get a full aero map, you would need to run the car at various positions. - Tyre compression: the tyres seem to have been cut-off at ground level, which is a common technique to add in the effect of compression. But the "height" of cut-off is subjective and has a large effect on the aerodynamics, as it influences the ride height and thus the underfloor aerodynamics. Also, there is no tyre deformation in the 3D model (or perhaps I missed it, difficult to tell from the screenshots).
Analysis versus optimization This is an analysis, which tells us quite a lot. But it doesn't teach us how the team interprets results or, even better, how it translates this into improved designs. It would be cool to see how an adjoint solver would improve this car towards more downforce for example! (if you don't know what adjoint is, check this video - https://airshaper.com/videos/aerodynamic-shape-optimization-...).
Overall, I find it very interesting - even if the model, CFD setup, ... aren't an exact match with what the guys at Williams are doing, it does provide a lot of interesting insight into existing flow structures on such cars - which is an interesting tool for learning.