Dear all,
I used openFAST v3.5.2.
I am simulating forced oscillations (through Extptfm module) on the IEA15MW floating wind turbine model, I tried to turn off tower DOF(FA、SS1STand2nd) to ignore the influence of the flexibility of the tower, but there were some errors that prevented the simulation from converging,(Generator speed and Pwr were NaN).
If I increase the period of the platform surge, this error will not occur and the simulation will proceed normally, So I think maybe the intensity of the platform motion prevents the computation from converging? I wonder if there is any other way to solve this error?
Dear @Jason.Jonkman ,
Thank you for your quick reply,
According to what you said, I make GenDOF=0, although there are still some warnings, the generator speed and power can be output,
But as you said, turning off this degree of freedom is equivalent to turning off the controller, then the power will continue to remain at 15000kW, many other physical quantities also deviate from the actual operation process,
I want to have the controller involved in the simulation, can there be another way to achieve?
Perhaps the bandwidth of the controller you using is not fast enough to react to forced high-frequency surge of the turbine, resulting in unstable response?
Dear @Jason.Jonkman ,
I am trying to make the output thrust equal to the aerodynamic force. In your previous answer, you mentioned that the thrust equals the aerodynamic load plus the inertial load and the gravitational load, so I keep the forced movement of the platform only Surge movement, and the other platformDOF is 0. The freedom of the tower is turned off and Shfttilt=0.
Here are my two questions:
a) when shfttilt=0, the previous issue of exporting NaN is solved, can you tell me the possible reason? I don’t understand why changing this Angle would make such a big difference, My understanding is that changing this Angle is equivalent to changing the relative Angle of the incoming wind
b) When the wind is steady 22m/s (bigger than rated windspeed), PtfmSurge period is 60s, the output thrust and aerodynamic load are shown in the figure. My Settings avoid the effects of gravity loads, Can you tell me the possible reasons why there is still a gap?
I’m sorry to bother you, I’m new to wind turbines, so my questions may be simple.
Best regards,
Regarding (a), are you saying that the controller is unstable with GenDOF = TRUE and ShftTilt is nonzero, but the controller is stable with GenDOF = TRUE and ShftTilt = 0?
Regarding (b), the surge motion will induce an inertia force in the shaft as a result of the surging rotor mass.
Dear @Jason.Jonkman ,
I’m not sure If the previous error was caused by controller instability, (The controller I’m using is ROSCO v2.8.0)
When GenDOF = TRUE and ShftTilt is nonzero, no matter how long I give the Surge period, The warning shown in the figure will appear and output NaN,
Thanks for confirming. I’m not sure why your model is going unstable by a small change in ShftTilt (from zero to nonzero). Does your model remain stable with nonzero ShftTilt if you disable the platform degrees of freedom in ElastoDyn, effectively eliminating the effect of the forced surge motion?
Dear @Jason.Jonkman ,
Sorry for taking so long to reply you. During this period, I am also looking for the cause of the problem. Now I am conducting the simulation of the 5MW wind turbine, and I think it may be because the initial value of the tower in Elastodyn is not set properly, so the simulation is not easy to astrate.
Thank you for your reply!
