Hi,
I am basically doing shape optimisation on airfoils using AI, and then I struggle with some convergence issues when trying to simulate back the optimised wing on flow5.
What I noticed first of all is that sometimes it wouldn’t converge with the script, but would converge when I load the model in flow5 interactively and simulate using the GUI.
Secondly, I noticed that I can run into errors ( out of the flight envelop) when I have the fuselage of my plane in and they disappear when I remove the fuselage. Could it be that the airfoils at the root of the wing, where it intersects with the fuselage, are cropped in a way that the software cannot link them to their respective polars anymore?
If you have encountered the same issues, and have means to solve them please let me know.
Best,
Benoît
Hello
Difference between scripted and UI results: likely due to an option in the script which is not set in the same way in the UI. Note that you can run the script from the UI (Ctrl+X)
Root airfoils: with the fuselage included, it is likely that calculated lift coefficient at the sections close to the wing’s root exceed the stall limit and prevent interpolation. These are areas of high pressure and velocity gradients which would likely be damped in a viscous analysis. The potential + interpolation method implemented in flow5 reaches its limit in such a case.
André