I using FAST v8 with AeroDyn v15 to simulate the NREL 5MW turbine (with Sandia 5MW blade definition) in TurbSim-generated turbulence fields (normal and extreme turbulence) throughout the operating range. Simulation setup is based on Certification Test #18. I am using the steady airfoil aerodynamics model (AFAeroMod = 1), an uncoupled skewed wake correction model (SkewMod = 1), and I do not include tower aerodynamic effects.
For some wind conditions (extreme turbulence model at 3, 5, 7, and 9 m/s and normal turbulence model at 3 m/s), the BEM solution is failing well into the simulation (anywhere from 100 s to 1200 s in) with a negative Vx. For example:
Code: Select all
FAST_Solution:FAST_AdvanceStates:AD_UpdateStates:BEMT_UpdateStates(node 7, blade
3):BEMT_UnCoupledSolve:DeterminePhiBounds:There is no valid value of phi for these operating
conditions! Vx = -1.50015E-02, Vy = 14.886, rlocal = 19.936, theta = 0.17736
FAST encountered an error at simulation time 154.6 of 200 seconds.
Simulation error level: FATAL ERROR
Although I wouldn't expect these steps to be helpful in this case, I have tried adding a correction step, making DT smaller, and changing InterpOrder, all to no avail. I set the initial conditions according to the discussion here https://wind.nrel.gov/forum/wind/viewtopic.php?f=4&t=1728&p=8211#p8211 (that post had to do with simulations failing right away, which was solved by setting correct initial conditions--now the failure happens well into simulation). I understand that Vx is the relative wind speed (wind speed minus structural speed); the structural solution seems reasonable up until failure for my simulations.
Based on the discussion with Yannick in https://wind.nrel.gov/forum/wind/viewtopic.php?f=4&t=1514&p=7513&hilit=negative+vx#p7513 it appears that problems stemming from small/negative Vx may be a known issue. Is that the case? Is there anything else I might try? Otherwise I may try to see if there is any difference using AeroDyn v14.