Modeling Unbalance force in solid rotating shaft

[iboby]

I am working on a rotating shaft with an unbalanced disk. The center of unbalanced disk is located on the mid-span of the shaft with offset=r_shaft. I used SOLID186 to model the shaft. I defined one mass point (MASS21) on the center of the shaft in APDL to model the disk. I know that the unbalance force components (f_x and f_y) cannot applied on a single node as we usually do in the case of beam element. I would like to know how can I properly model the unbalance force on the shaft in transient analysis? My desired output is the elliptic orbit of mid-span point of the shaft.

 

[Erik Panos Kostson]

I have very little to no experience with rotor dynamics, so this relates only on how to add a rotating force. Below is just one possible way to add an unbalanced/rotating force in harmonic response FEA (e.g., F= Mass_unbalanced*radius*omega^2).

It can for instance be achieved by having two forces with directions perpendicular to each other, and with a phase shift of 90 degrees. That means in ansys wb, that one force has a pure real part (see below, 100 N real part in Y), and the other one has a pure imaginary part (see below, 100 N imaginary part in Z).

These forces can be applied to a single individual node (not part of the shaft mesh) in the centre of the shaft that also has a mass21 on it. It is linked and connects to the mesh of the cross section via links (see RBE3 in ansys apdl). Thus it is a point for remote load applications, and the force is distributed on the the mesh via the RBE3.

I have written something about remote load application (not including MPC and contacts though), in another post that might give some guidance.

Link

This is just my idea (taken from my Strand7 experience - Strand7 does not have any rotor-dynamic capabilities). There must be a lot of material on the Internet about this so it is good to have a search (e.g., Example: Harmonic Response to Unbalanced Force using Component Mode Synthesis (CMS)). There is also a rotating force in workbench (harmonic), see if that helps also. At least in workbench this is doing a similar/same thing as described in the first and second paragraph here at the top.