Azimuth Slew Bearing Vibration/Frequency Response

[alross]

I have already posted this question on the "Bearing design, manufacture and maintenance Forum"
--------------------------

I would be grateful to get some ideas for the simplest way to calculate the vibration/frequency response for an azimuth slew bearing.

The bearing is quite a large slew bearing which has a payload of 155kg with a defined C of G wrt the bearing. The bearing has ball bearings in a wire race. The manufacturer is currently calculating the bearing stiffness for me.

The size of the bearing is an outer diamter of ~400mm and an inner diameter of ~300mm. The bearing is ~30mm thick.

The bearing sits on top of the horizontal roof of a vehicle. The payload is the hung from (and bolted onto) the inner race of the bearing so that hangs into the vehicle. The bearing outer race is bolted onto the outside of the vehicle roof. The bearing is the only support for the payload.

After such a lengthy description (!) my question is:
Does anyone know a simple method to calculate the anticipated frequency response of the bearing? I'm willing to try either hand calcs or run a FEA model in Mechanica or Ansys. This is for the bearing in a static mode when it is not rotating, i.e., I'm interested in the non-rotational modes of the bearing only.

The frequency range that I am interested in is from 0Hz up 2kHz.

Thank you.

Alasdair.

 

[alexit]

If you know the stiffness of the static bearing and the CG of the payload, distribute the stiffness as a massless spring to each bearing ball location and calculate wrt any axis passing through the CG of the payload and ball (use pendulum calculation.)


Alex

 

[alross]

Alex,

Do you have any further details about the pendulum calculation as I'm not sure what you mean?

When you say to distribute the stiffness, do you mean that I simply divide it by the number of balls in the bearing?

Thank you.

Alasdair.

 

[alexit]

I am sorry, I was assuming you are interested in oscillation frequencies because for pure vertical modes a simple mass - spring calculation should suffice.

For horizontal oscillation use a cross-section, then determine a pivot point. This way 50% stiffness is assigned to two springs (left and right.) Translate horizontal motion of the payload into resultant forces at the two springs, the resonance frequencies can be determined by sum of the equations before integration.

Alex

 

[vgarzani]

I assume some sort of lubrication is involved. So, you are going to do some testing for validation if it really is important - right? My thinking is that with a test payload and a well defined mounting some preliminary tests could be quickly done to get your model of the bearing tuned in. But you would likely find that the most difficult aspect of the whole effort is accounting for the actual field or vehicle roof mounting details.
And to be devils advocate - if you are avoiding really bad interactions associated with resonant amplification, how much margin is required over time. Will the roof mount change, will lubrication condition changing become at all significant. Argh, sorry I'll just stop.

 

[alross]

Alex,

Sorry that I have not responded before now - rather busy with other stuff!

Could you tell me what you mean by "sum of the equations before integration"? What should be integrated?

Thanks a lot.

Alasdair.

 

[alross]

Vgarzani,

I am only thinking of doing some basic analysis (either FEA or manual calcs) to determine resonant frequencies to see if there are any resonances in the operational frequency spectrum - I'm sure that there are resonances.

I agree that there are many variables that will affect the resonant frequencies in the field, but I think that some basic analysis may tell me if I have a problem before I move onto physical validation.

Alasdair.

 

[ccw]

Hi Alross,
As your posts indicate, the large turntable bearing vendors, such as Rotek(TM), normally have the stiffness available since they are often asked questions similar to yours.

As a simplifying first cut (I am envisioning something like a tank turret or HMMMV roof turret) take the supplied stiffness and construct a simple 2-D model. The model would be a horizontal platform supported on two springs; or one spring and a fulcrum or hinge mount; sized to give the same stiffness (the rotation of the platform is in the vertical or elevation plane). Then locate the mass CG on the platform as it relates to the center of rotation of the bearing. Using some elemental equations from mechanical vibration theory, (pendelum equations, etc.) determine the undamped natural frequency. See if it is in your range of interest.

Be sure and check the stiffness of the roof support. I am betting that the horizontal roof support of the vehicle, the frame, the vehicle suspension, outriggers, etc., may be order magnitude softer than the bearing stiffness, and your fundamental structural vibration frequency of the bearing supported load (pitching or rotation in the vertical plane or bouncing) may not influential in the fundamental.