I am trying to run openFAST in which the waves and wind field have a propagation direction of 180 deg. At the same time, I am also running the exact same model, but for directions of both 0 deg and -90 deg (wave and wind are colinear).
The latter two directions are working just fine, but the 180 deg simulations keeps throwing an excess violation error. I have narrowed it down to the use of the turbulent wind field: for a steady wind (WindType = 1) the simulation works, once switching to WindType = 4 the simulation does not run, with the error message as shown in the attachment:
The wind field is exactly the same as for 0 deg and -90 deg and generated in TurbSim. All other simulation parameters are the same as well, except for the wind and wave direction. Would you have any idea what could be the cause for this error?
I’m not sure (as I haven’t seen this error before), but perhaps there is small bug? Does the simulation run with wind/wave propagation directions slightly smaller than 180 degrees, e.g. 179 degrees? If so, that may be a good work around until the problem is found and fixed.
I see you are getting warnings because of the second-order potential-flow hydrodynamic terms that have been enabled. This issue has recently been fixed in a branch of OpenFAST–see: github.com/OpenFAST/openfast/issues/9. But unless you are using the source code from this branch, I would suggest disabling second-order potential-flow hydrodynamic capability when simulating with waves that are at 0 deg propagation.
Thanks for your help. I indeed tried already small variations around the 180 deg, but these also did not run. I extended this range of angle to check at what point the simulations starts to fail. I changed the following settings:
- Turned off waves/current in order to isolate the wind contribution;
- Starting from a propagation direction of 0 deg (which runs fine) I step-wise increased the propagation angle.
At propagation directions higher than 10 deg, I get the error as shown in the opening post. I also tried switching off the wake computation (WakeMod = 0). This marginally improved the situation, since the simulations now fail at angles higher than 15 deg.
Just to be sure, am I correct in my assumption that I do not need to specify the initial yaw angle in the ElastoDyn file? I did try to set this value (NacYaw) such that the rotor is aligned with the wind, but this didn’t help matters (I get errors related to small angle rotations instead).
Are you referring to the access violation error or the warnings? I thought in your first post you said a direction of -90deg worked fine (now I’m not sure I understand)? Did you disable the second-order potential-flow solution?
If you are changing the wind direction, then I would expect that you’d also change the nacelle-yaw angle, unless you are trying simulate large yaw errors. But changing the nacelle-yaw angle requires more than just changing the initial nacelle-yaw angle (NacYaw) in ElastoDyn. Presumably you also have a yaw spring and damper set in ServoDyn, so, the neutral yaw angle (YawNeut) set in ServoDyn determines the commanded yaw position. And if NacYaw doesn’t equal YawNeut, the nacelle-yaw angle will quickly snap back to YawNeut and cause strong / unrealistic yaw excitation.
Ok, then I was effectively simulating very large yaw angle misalignment with a large restoring moment applied at t = 0 (which was not my intention). Apparently the 90 deg propagation didn’t immediately throw an error, but the same faulty simulation set-up was used.
Your guidance to the YawNeut parameter in ServoDyn solved the issue: once I set YawNeut and NacYaw to 180 deg, the simulations runs stable.