Hello everyone!
I recently used OpenFast 4.0.0 to simulate a semi-submersible floating wind turbine with a single-point mooring system. During the simulation, I encountered an issue: whenever the difference between the platform’s yaw motion (PtfmYaw) and its initial yaw position (PtfmYaw in ElastoDyn) reaches 10 degrees, the error” UA_CalcOutput:Mach number exceeds 1.0. Equations cannot be evaluated.” Will occurred and the simulation was shut down. When I checked the output file I found the load on blade rise very sharply. And regardless of whether the yaw response rate is fast or slow—as long as the 10° deviation threshold, this error was occurred.
I would greatly appreciate any guidance or suggestions to address this problem. Thank you!
Dear @JiYuan.Men,
From your description, I would guess your OpenFAST model is experiencing an instability (either numerical or a physical aero-elastic instability) at certain yaw misalignments, but I’m not sure I can comment more without knowing more about your simulation set-up.
Best regards,
Dear Jason.Jonkman
I am very glad for you replying my question! I am tying simulation a kind of FOWT with single point mooring system and I want to analysis the tower root and blade root load under extreme conditions for idling turbine. I have uploaded several images that demonstrate some input parameters and partial output results from one of the simulation cases where errors occurred.
In this case, the turbine is idling; the mean wind speed is about 50m/s, the wind direction is about 45°; the wind file is created by Turbsim. The wave is generated by JONSWAP spectrum with Hs=6m, Tp =10s. the wave direction is about 0°. The surface speed of current is about 1.5m/s and the depth is about 1m. In order to avoid error in the topic as much as possible, the initial ptfmYaw= 35°,PtfmSure =-30m and PtfmSway =50m (I had tried many initial position such as ptfmYaw=45°-30°,But it didn’t work).
This case cannot be complete due to error shown in picture. We can find that the PtfmYaw is about 45° which is about 10° different from initial PtfmYaw. The blade tip displacement seems not diverge, but the force of blade root rose very very shaply.
This is just one of error cases I meet. All error cases are similar and when the amplitude of PtfmYaw reaches 10° the error will occur. I hope this information can provide support for solving the problem.
Thank you very much,
Dear @JiYuan.Men,
It looks like you are modeling an idling rotor with the generator disabled and blades feathered. You haven’t shared your AeroDyn input file, but from your messages and errors, it sounds like you have wake and unsteady airfoil aerodynamics enabled. But these models are only valid for operational conditions. I would suggest disabling these models in AeroDyn (Wake_Mod
= 0, UA_Mod
= 0), leaving only table-look up airfoil aerodynamics. Is your model stable then?
Best regards,
Dear Jason.Jonkman
Thank you for your help! As you suggested, I set both of Wake_Mod and UA_Mod as 0, so that the unsteady aerodynamic is inoperative. The simulation became a bit of more stable than early case. But when I change the initial position of FOWT the error is occurred as shown in the pictures. Although the error is different from former case, I think the reasons are same.
In this case, the FOWT is same as yesterday’s case. The WaveMod is 0 (still water), the wind field is generated by turbsim and mean wind speed is about 50m/s, the wind direction is about 30°. During simulation, the wind turbine is idling. I upload some result shown in pictures. We could find that the acceleration of bladed tip is diverging in a second. And the period is about 0.001s twice as many as time step (0.0005s). The time domain result of PtfmRVzi is diverging too. On the other hand, the motion response is not oscillation as velocity or acceleration. After this, I set Wake_Mod=1 and UA_Mod=3 which is same as former case. As expected, the error occurred and the result in time domain is diverging too (shown in pictures).
Besides those two cases, I also tird to close all Dofs except the Dofs of platform. Finally, the error occurred and the velocity result of platform is diverging with 0.001period.
I think, this is numerical instable rather than aeroelastic. And the reason is maybe large yaw motion. Is there any way to solve this problem?
Thank you
Dear @JiYuan.Men,
I’m having a hard time following all of the variations you are trying and how those variations impact your solution. As with an instability, I would try to isolate it by finding the simplest model in terms of limited DOFs and features enabled that still has the instability. Which DOF or feature, when enabled, makes a model that otherwise runs stable, to run unstable?
Best regards,