Changing the baseline airfoil : FOAM FATAL IO ERROR

[usmanshahid90]

I am changing the baseline airfoil for optimization from NACA0012 to NACA6 series. Saving the coordinates as per the guidance of DAFoam FAQ, I am having this error.

FOAM FATAL IO ERROR
Attempt to get back from bad stream
file : volumemesh.xyz at line 1
From function bool foam : : Istream : : getback (Foam: :token&)
in file db/IOstreams/IOstreams/Istream.C at line 56.
FOAM exiting

Generating Mesh .. Done

Solution is running even after this error but why this error? Plus, after the solution run, when i see export the baseline coordinates, they again are of NACA0012. Not of 6 series.
Kindly guide me in this regard

[friedenhe]

The error message should be in logMeshGeneration.txt

[friedenhe]

Can you increase to autoPatch 60 -overwrite >> logMeshGeneration.txt? If it is still not working, upload the case here.

[usmanshahid90]

hy,
I tried it too but it didnt work. I will upload the files soon.
Before that, can u please tell me would changing the baseline airfoil would make any difference in the optimized result if constraints and objective functions are same?

Kindly guide me.

[friedenhe]

If you use a very large number of FFD points and converge the optimization tightly (like this paper), the results will be the same. But because we usually use like 20 FFD points, the result will be slightly different.

[usmanshahid90]

Alright, just need to discuss it ; I've read in surrogate modelling research papers that gradient based methods are prone to stuck in local minima. But as per the research paper you mentioned, it seem like starting from circle and ending to airfoil, it doesnt stuck in local minima. what you'll say about it?

[friedenhe]

This is an issue only if your optimization problem has multiple local minima. We have not found a multiple local minima case for RANS-based wing/airfoil aerodynamic shape optimization. If you find one, let us know, and we will be more than happy to try it.