I’m trying to model a Vestas 660KW blade in Numad, which is an interface software between BPE,PreComp, BModes and Fast. I have some problems in Bmodes inputs. The defaults of Numad for BModes inputs are rotor rpm as 15, hub radius and other inputs as 0,when I change the rotor rpm to 0 the predicted natural frequencies by BModes are close to Ansys natural frequencies blade, but by changing the values to real values of my blade ( rotor rpm =28.5 and hub=0.6m ) the predicted natural frequencies change a lot and have about 15% error with Ansys model. Can anyone help me about this that why rotor rpm and hub radius have such effects on natural frequencies and what should I do to reduce this error?
I have no idea about other parameters such as precone (Blade built-in precone angle; the angle between the blade reference axis, usually the pitch axis, and the plane of rotation normal to the shaft.) and bl_thp (Blade pitch setting. This may be a built-in or a control pitch setting for the whole blade. The actual twist at any section is this pitch setting plus a section twist angle). Are these parameters are as important as rotor rpm? (Since Numad defauls for these values is 0 and when I changed them randomly, there was no change in natural frequencies).
My another question is about the results of BModes which gives overall natural frequencies(not flapwise or edgewise) and flap and edgewise mode shapes. How should I find out which mode shape is flap or edge and which one is first flap or edge mode shape as inputs of Fast.
Thanks for all guides in advance;
The natural frequencies and mode shapes of a spinning rotor blade are influenced by centrifugal stiffening, which depends on the rotor speed, the radial length of the blade, and the blade mass.
If you are using BModes to derive mode shapes for FAST, please see my post dated Mar 18, 2015 in the following forum topic for guidance: http://forums.nrel.gov/t/nrel-5mw-blade-mode-shape/1092/1.
Thanks for your answer. According to your answer, it should be common that the BModes results have considerable error with Ansys modal analysis results or natural frequency experimental tests, since neither Ansys nor experiments do not consider centrifugal factors. I would be thankful if you could help me to get more information about how these factors influence natural frequencies and mode shapes.
When I was studying centrifugal stiffening myself, I wrote a paper in grad school about the effect. You can find this paper attached. Hopefully it provides suitable background.
RotatingBeamFinalReport.pdf (638 KB)