Optimizing Vibration Control for Off-Road Vehicle Roll Cage Attachment

[anonME]

I am designing a plate that will attach to the roll cage of my off-road vehicle. The plate will be supported on each end by the frame.

I want to make sure that vibrations from the engine do not cause the plate to oscillate out of control.

How do I calculate this out?

 

[SteamKing]

It depends on how the plate is attached to the cage and whether the plate has any stiffening members attached to it.

 

[anonME]

Are these calculations that are done by hand or should I use FEA software?

 

[SteamKing]

I really couldn't say. You know more about your application than I do. I'm assuming you want to 1. make sure the plate stays attached. 2. don't want to encounter any natural frequency problems which may cause the plate to detach or to start drumming excessively.

 

[AlephZero]

Suppose the engine speed range (idle to max) is 600 to 6000 RPM, or 10 to 100 Hz. It's unlikely you can design a plate structure that has no vibration modes over the whole of that range, unless you make it ridiculously stiff (which probably means ridiculously heavy). So the problem then changes to "how to stop the engine vibrations getting into the plate", rather than worrying about the vibration frequencies of the plate.

You could think about doing that by fixing the plate on rubber mounts rather than metal-to-metal bolts or welds (possibly bolted joints with rubber washers, or something similar to rubber engine mounting blocks). Or think about ways to damp out the vibration of the plate, for example make a "sandwich" of two thin metal plates with a layer of rubber, polystyrene, etc, in between them.

(I'm assuming you are talking about a metal plate here - if not, wood will give more damping than metal for the same mass of material, if it satisfies your other design criteria.)

Choosing the best points to fix the plate to the frame will also make a difference. You want to fix it at the points where there is least vibration in the frame. (You may be able to judge that simply "by feel", if you don't have any vibration measuring equipment.)

Designing something like this entirely by calculations isn't very practical. As a starting point, you would probably want to measure the vibration levels in the frame without the plate at different engine RPMs (and probably different engine powers as well), rather than try to make a model of the engine and vehicle to predict them from first principles. Experimental "trial and error" is probably as good as way as any.

 

[xxChrisxx]

You aren't likely to have durability problems with a plate of metal if it welded or bolted to something. The worst case scenario is that it makes an annoying vibration/sound at whatever speed it goes off at. Don't bother trying to isolate something until you know you have a problem with it, especially if you need the structure to be rigid.

Choosing the best points to fix the plate to the frame will also make a difference. You want to fix it at the points where there is least vibration in the frame. (You may be able to judge that simply "by feel", if you don't have any vibration measuring equipment.) Designing something like this entirely by calculations isn't very practical. As a starting point, you would probably want to measure the vibration levels in the frame without the plate at different engine RPMs (and probably different engine powers as well)

Strictly speaking measuring the frame without the plate actually doesn't help, unless you are correlating a CAE model. As the stiffness and frequencies will change by having the plate there. Sticking the plate at an area of 'low vibration' isn't a reliable way to avoid problems. As you can end up sticking it at a node of frequency x (that you can see but won't affect the plate), but an anti-node of higher frequency that sets the plate off. The large areas of low fq vibration could be better, as the plate would move rigidly with the structure.

Experimental "trial and error" is probably as good as way as any.

+1 to this.

Stick the plate where you want it, run it and see if there is a problem. If there isn't; great. If there is; we can try to work out a solution.