Hi,
I am modelling the IEA 22MW turbine with a fixed base and running a stepped wind case from 4m/s to 25m/s (first 120 seconds is 4m/s then the speed is stepped up in 1 m/s increments every 40 seconds) similar to Orcina’s setup in https://www.orcina.com/news/iea-22-mw-rwt-validation-study/ for load case 3.4. I am getting good agreement in general between OpenFAST and Orcaflex. However the blade flapwise tip deflections (I’ve transformed OrcaFlex’s ‘out of plane/in-plane deflections’ to ‘Flapwise/Edgewise deflection’ correctly in postprocessing) are showing a lot more fluctuations with OpenFAST than Orcaflex (see plot). Is this level of fluctuation normal/expected at the blade tip - would anyone from NREL be able offer some advice/insight as to why this might be happening?
Here are my results - comparing OpenFAST & OrcaFlex:
Here are Orcina’s results:
For more information I’ve followed Orcina’s OpenFAST model changes as follows:
I’ve tried to adjust the Blade flapwise damping in OpenFAST by adjusting the 5th damping component from the original setup:
mu1 mu2 mu3 mu4 mu5 mu6
(-) (-) (-) (-) (-) (-)
0.004048 0.003153 0.000140 0.003153 0.004048 0.000140
I multiplied the 5th coefficient by 5 and 10 and saw no visible differences in my results - would I need to also adjust the 1st damping coefficient to reduce the flapwise deflection fluctuations?
Flapwise Tip deflection with modified damping coefficient:
I also ran a turbulent wind case as specified in Orcina’s report (section 3.5). For this my period is off in OrcaFlex for some reason, but the fluctuations seem to align & my OpenFAST results agree very closely with those presented in the Orcina report:
Hi, I suspect that your blade is suffering from some aeroelastic instabilities. We ran into similar issues while working on this publication last year ttps://iopscience.iop.org/article/10.1088/1742-6596/2767/5/052042/meta
Maybe a first question: did you change UAMod to 2? If so, can you try rerunning with UAMod=4 as in the original OpenFAST model (IEA-22-280-RWT/OpenFAST/IEA-22-280-RWT-Monopile/IEA-22-280-RWT_AeroDyn15.dat at 9d47ed0a6665d5b3488689ccd855c4499c153901 · IEAWindSystems/IEA-22-280-RWT · GitHub)?
I ask because I saw instabilities with UAMod=3, and we have been looking into the source of that, so far without a final conclusion.
As for artificially increasing structural damping, I would apply a 10x on the torsional term mu6 more than on the flap term (mu5)
Let us know, and happy to see that the 22MW is being used!
Hi Pietro,
Thanks for your reply. I tried running with UAMOD=4 but it crashed, checked my files against Orcina’s and they had PLExp as zero (I had 0.14 in OpenFAST) so I tried this and it improved the flapwise deflection & moment:
Thanks,
Barry
1 Like
Happy to hear that the match improved. I’m a little puzzled to hear about UAMod=4 crashing… Could you please paste here the error message?
Yes, turns out UAMod=4 in OpenFAST is fine (it didn’t crash) - but it doesn’t provide as much improvement in terms of reducing the fluctuations in flapwise tip deflection/root moment as making shear exp=0.
- green - Shear Exp=0.14 and UAMod=2 ( same results as my original post above )
- yellow - Shear Exp=0.14 and UAMod=4 ( as Pietro instructed )
- red - Shear Exp=0 and UAMod=2 ( same results as my previous post )
- blue - OrcaFlex - shear exp=0, UAMod = Gonzalez variant (2)
Figure 1: key structural results:
Figure 2: key rotor performance results:
Figure 3: This plot isolates the flapwise deflections - OrcaFlex is in green (plotting mistake)
Let me know if there’s anything else you would like me to share.
Thanks,
Barry
1 Like
