The main difference should be the Aoa, therefore I plotted the angle at last four nodes at 200s and at 350 s.
Dear Moritz,
Clearly enabling DBEMT (WakeMod=2) is resulting in a bad solution.
What DBEMT settings are you using in AeroDyn?
Also, I noticed that you are using (some branch?) of OpenFAST v2.1. When DBEMT was first introduced into OpenFAST, there was an issue in the OpenFAST glue code that would require that a correction step be added (NumCrctn>0) in some cases when DBEMT was enabled. This was fixed in pull request #164: github.com/OpenFAST/openfast/pull/164, which was merged into OpenFAST v2.2. So, you may need to add this correction step when enabling DBEMT in v2.1, but you shouldn’t need to add this correction step if you upgrade to v2.2.
I hope that helps.
Best regards,