Simulation error small wind turbine


I am using FAST to make the certification analysis of a small wind turbine (Rotor diameter=1.7m) which has a full GM15 airfoil with a low Reynolds (0.1million), in order to find the bending moments and tip deflection. I’m using Test09 from the CertTest examples as this is also a downwind turbine. I’ve run tests with Elastodyn and Aerodyn active, ServoDyn inactive and InflowWind file from Test09, nevertheless I get the following error.

Command Prompt
angle assumption violated in SUBROUTINE SmllRotTrans() due to a large blade deflection (ElastoDyn
SetCoordSy). The solution may be inaccurate. Simulation continuing, but future warnings from
SmllRotTrans() will be suppressed.
Additional debugging message from SUBROUTINE SmllRotTrans(): 1.50000E-03 s

I’ve read through several posts and the most useful one was this
I’ve tried the suggested approaches, when I reduced the time step DT to 0.001 the simulation works but all the output values are NaN. Then I tried to reach the root of the problem and I thought it could be the stiffness. I made the blade much more stiffer and got the new mode shapes but the problem remained. Then I thought the problem could be the extrapolation of the airfoil data (angle of attack, cl, cd, cm) but I also tried the Test09 using GM15 airfoil and it worked with good numerical results.

Now I’m wondering if the problem is due to the size of the turbine, since the rotor diameter is very small and I’m not aware of FAST boundaries to validate a simulation. I followed the IEC standard and already performed the simplified loading calculations but now we would like to test the turbine under turbulence and specific cases due to the high centrifugal forces that the rotor experiences (the rotational speed is 540 rpm with direct drive generator).

I finally looked and did some trials using Test17 as is the smallest rotor in FAST (5.8m) but also got NaN values or didn’t run because of the previous error.

Thanks in advance,

Lizet Ramirez
MSc SET student

Dear Lizet,

From what you describe, I would guess that your model is going numerically unstable because of too large of a time step. The smaller and stiffer the turbine, the smaller the time step requirement is. I would try a time step of 0.0001 s or 0.00001 s to see if you can get the model to run. Obviously, you want to choose as large of time step as possible while keeping the solution numerically stable. My rule of thumb for selecting the time step for the structural module is given in the following forum topic:

Best regards,