Modeling bearings 1

[WARose]

I know this topic has come up a few times over the years (in fact there is a similar one going on the structural board)....today something has come up that has me wondering....

The situation is this: I need to put in some pads to change the vertical stiffness of a machine base. But for the analysis, that would give it a different compressive stiffness than its uplift stiffness. So, for modeling purposes, it has me wondering: do I do 2 models (one with the pad vertical stiffness and the other with the bolt hold down stiffness)? How do I know which was is applicable to which situation? Possibly by looking at the mode shapes? (something I am leery of)

Before anyone asks: there is a net uplift at some bolt anchors.

Thanks in advance.

 

[GregLocock]

No, mode shapes are the result of your analysis.

So as I understand it you have a non linear stiffness curve for the pad. If the load bears down sufficiently it will overcome the bolt preload and just rely on the rubber's compressive stiffness. It seems to me you have to know a great deal about the system to get a good handle on this, for instance you need the static load on each pad, the details of the bolt preload, and the amplitude of the vibration. I'd add that the pad is doing you no good until the bolt preload is overcome.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[WARose]

So as I understand it you have a non linear stiffness curve for the pad. If the load bears down sufficiently it will overcome the bolt preload and just rely on the rubber's compressive stiffness. It seems to me you have to know a great deal about the system to get a good handle on this, for instance you need the static load on each pad, the details of the bolt preload, and the amplitude of the vibration. I'd add that the pad is doing you no good until the bolt preload is overcome.

And after knowing all that....then what? Is it a case of: the uplift cannot be "engaged" (for a lack of a better way of putting it)?

Just a FYI: very little pre-load is in the bolts. (Very little is desired for some reason that has not been made clear to me.)

 

[GregLocock]

Oh, I stopped because I wanted you to confirm that I was on the right track. If there is little preload I'd be inclined to ignore the bolt and just use the pads stiffness.

However the fundamental issue is, what is the dynamic response of a mass on a non linear spring? The answer is amplitude dependent. I don't know of an analytical equation to solve it, I've just done a time based simulation.

A more strictly analytical approach leads quickly into deep water, chaotic systems and the like.

https://en.wikipedia.org/wiki/Duffing_equation

is a place to start

Sorry, that isn't much help, as I said, I do a time based sim.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[WARose]

However the fundamental issue is, what is the dynamic response of a mass on a non linear spring? The answer is amplitude dependent. I don't know of an analytical equation to solve it, I've just done a time based simulation.

The software I use has a multilinear spring support command.....just can't use it in dynamic load cases.

I'll look into the link you gave. Thanks.

 

[GregLocock]

Of course basically the idea is wrong. What you should do is have a compliance between the head of the bolt and the foot of the machine, as well as between the foot of the machine and the floor, or else a proper designed bushing rather than a pad.

Like this

https://www.blackwoods.com.au/part/05128217/mount-machinery-isolator-mackay-m11060

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?