Vibration Proof Mechanical Design Questions

[leeave96]

I've been tasked to design an enclosure that is to be subjected to vibration. That vibration will be looked at by our FEA guy and at the end of the day, on a vibration table. I have never really designed a mechanical enclosure for vibration - but have design lots of stuff for static loads, etc. Before I start designing this thing, I want to get as much info, up front, in terms of good design technique - rather than just guess at it.

I've done some research and I find all kinds of stuff about analysis for vibration AFTER a part/assembly is designed.

My question is - at the START of a mechanical design, what are some "rules of thumb" to get one going in their design prior to analysis.

What are good techniques in terms of corner construction, fasteners (bolts or rivets), welding, etc.

Any ideas/suggestions would be most helpful.

Thanks!
Bill

 

[Vaporware]

I've found that the best way to design things with vibration in mind are know where your vibrational grounds are and where your large weights are.

I've put in time on designing motor mounts, and typically they are large masses at the top of an adjustable table. The shorter the "moment arm" of the weight, the better. The closer the weights are to the grounds, the better.

Thickening members is not always better. You need to keep weight low and stiffness high. Chop the middle out of large plates to reduce weight without losing stiffness.

You don't really have a good way to design for vibration up front other than design it stiff and light. Most refinement will come in testing/FEA. Analyse the stress concentrations due to natural frequencies to see where you can remove material.

 

[IRstuff]

What might make or break you as an engineer in this field is whether you are able to eventually estimate the natural frequency of a structure by inspection only. I've only met a couple of engineers who could do that. One guy walks into our building just before a qual test and takes one look at our structure and said that the natural frequency was about 30 Hz, which was too low. Our engineers needed to run FEA to verify and the guy was within 10%.

TTFN

 

[ajamnia]

Now what IRStuff says is real engineering.

Now, Bill, you say that you are designing an enclosure. One thing to consider is what you will be placing into this enclosure will have a direct impact on your design approach. For example, if you are designing a rack system for electronic equipment, you may want to try to reduce transmibility of the vibratory loads inside your system.

In my opinion, you need to think inside-out if you are at the start of your design.

Ali

 

[GregLocock]

It can be very handy to run a rough FEA (beam elements, lumped masses) on a current design, and have a think about what that implies.

So far as getting the biggest bang per buck goes - many designers throw stiffness away by poor (or expedient)joint design, particularly in torsion. You should be able to find some guidelines on the web, or again your FE guy should come up with some insights. A typical joint only has about half the torsional stiffness of the sections it joins, there is no real excuse for this if you are going for efficiency.

Yu haven't mentioned what frequency range is important to you - is it of the same order as the frequency response of the structure?





Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.