OpenFAST dev-tc: Floating system yaw instability (not consistent results)

[RBergua]

Bug description
As part of the USFLOWT project, we are using OpenFAST tight coupling for our floating system.

We were running OpenFAST compiled from dev-tc commit 36f0d55 (July 9, 2025).

For a still water condition without wind, we were obtaining the expected behavior. The system is oscillating around the platform pitch equilibrium position.

However, when taking the exact same model and boundary conditions and trying to run it with commit a1a9a12 (August 15, 2025), the OpenFAST solution diverges in the first 0.5 seconds of simulation.

As it can be observed in the below plot, the platform yaw angle increases significantly in a very short amount of time:

We tried to reach a converged solution by reducing the time step and increasing the number of correction iterations. But we always get this wrong behavior.

Note for myself for traceability: this occurs when the pretensioned cables are included in MoorDyn.

[gatosardina]

Are you using different compilers?
Seems like the more recent commit is compiled with a 4 year old compiler while the older commit was compiled with an updated compiler
Worth a try, I can attest that the intel fortran compiler is finicky, specially when they transitioned from ifort to ifx

[RBergua]

That's a good point, @gatosardina

The commit 36f0d55 (July 9, 2025) was compiled by @deslaughter. The Visual Studio solution of OpenFAST dev-tc had a problem that was solved with commit a1a9a12 (August 1, 2025). I was only able to compile using Visual Studio after that. Accordingly, the commit a1a9a12 (August 15, 2025) was compiled at my side. So, the 2 results shown above come from two different compilers.

@deslaughter: Could you compile commit a1a9a12 to rule out the influence of the compiler in these two runs?