Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
How to calculate the eigenfrequency and modal shape of a rotor with rotational speed by using Solid Rotor(rotsld) in Rotor Dynamic Module
Posted Oct 11, 2019, 7:49 a.m. EDT Structural & Acoustics, Structural Mechanics Version 5.3a 1 Reply
Please login with a confirmed email address before reporting spam
Hi everyone,
I am trying to calculate the eigenfrequency and modal shape of a rotor by using Solid Rotor(rotsld, the rotor is modeled by solid elements) in Rotor Dynamic Module. When I evaluate the results, it turns out that when the rotational speed is low, the rigid body modal appears first and then bending modal, which is I am expecting. But when the rotational speed is high, some of the eigenfrequency are changed signicifantly and the order in which the modal appears is confusing, which is wrong according to rotor dynamic theory. I also carried out the same simulation in Strctural Mechanics Module, the results are the same. However, the results obatined from Beam Rotor(rotbm, the rotor is modeled by beam elements) in Rotor Dynamic Module do not have the problems mentioned above, and the results are much more reasonable. Here are the details about my simulation results.
The first 7 eigenfrequency with 0Hz rotational speed calculated by Solid Rotor are
52.476+1.698i(1)(1st order rigid body modal (backforward whirl)),
52.476+1.698i(2)(1st order rigid body modal (forward whirl)),
95.803+5.6335i(2nd order rigid body modal (backforward whirl)),
95.803+5.6336i(2nd order rigid body modal (forward whirl)),
368.01(torsional modal),
428.9+4.7071i(1st order bending modal (backforward whirl)),
428.91+4.7071i(1st order bending modal (forward whirl)).
When the rotational speed is 500Hz, the results are changed to
2.7758+0.009i(1st order rigid body modal(backforward whirl)),
9.2894+0.1i(2nd order rigid body modal(backforward whirl)),
184.43+1.3631i(1st order bending modal(backforward whirl)),
367.61(torsional modal),
757.2+1.8418i(2nd order bending modal(backforward whirl)),
812.88+0.873i(bending modal(forward whirl)),
1004.8+5.23i(rigid body modal(forward whirl)).
Clearly, not only the value of the eginfrequency has changed significantly but also the order in which the modal appears is confusing calculated by solid elements under high rotational speed , which is unreasonable according to rotor dynamic theory. But this problem will not appear when using beam elements to model the rotor. And the campbell plot of solid rotor and beam rotor are in the attachment.
So, has anyone encountered the problem like this and could please tell how to modify my simulation?
Thank you!
Please login with a confirmed email address before reporting spam
Hi,
If you just do the simple eigenfrequency analysis using the Solid Rotor interface there is one effect that is not accounted for --- Stress stiffening. This effect results into a geometric stiffness of the rotor due to its spin. You can incorporate this effect by adding a stationary analysis step before the eigenfrequency to get the stress state due to the spin of the rotor. In the eigenfrequency analysis use the solution from the stationary study as a linearization point. With this you should be able to get the eigenfrequencies closer to that obtained from the Beam Rotor interface.
In the Beam Rotor interface equations are formulated in space fixed frame, therefore, you do not model the frame acceleration forces as a load separtely. In the Solid Rotor interface, equations are formulated in Co-rotating frame. Due to this you have frame acceleration forces acting on the rotor. These frame acceleration forces (centrifugal force) cause the stress in the rotor due to the pure radial expansion. This stress gives additional geometric stiffness to the rotor (similar to that in buckling analysis) that is important to get the correct eigenfrequency.
I hope this helps. You can write to us at support@comsol.com for further clarification.
-------------------Thanks and regards,
Prashant Srivastava