Vibration Test Fixtures Design 2

[eli28]

Hello,

This is the first time I design a fixture for a vibration test, so I guess my question might be basic to most of you.
I have an assembly that should be carried by a certain platform.
The Vibration Spectrum of the carrying platform was measured and is given as part of the requirements.
As part of the assembly design an engineer who is specialized in dynamic analysis made sure that the natural frequencies of the assembly are far enough (higher) and out of the input spectrum.
Now I have to plan a vibration test, and shake the assembly through a fixture that is connected to the shaker table. I am attaching a schematic description of the test:

I read that the guideline of fixtures design is making them as light and stiff as possible for achieving a natural frequency as high as possible, so no resonance may occur in the test due to the fixture.
Now I wonder how should I check by analysis the fixture I design.

I initially thought that if we have a separate analysis of the tested assembly we only have to check separately the test fixture alone (and make sure its natural frequency is high enough) as described below:

But I was told that this isn't right and that I have to include the Tested Assembly mass in its C.G without taking into account its stiffness etc. as described below:

Remark: I am sure that the most accurate way is an analysis of the entire assembly, but I would like to know what is the right way of design when you don't have an access to the tested assembly, and there is a necessity to simplify.

I will be happy if you refer me to a simplified example in which there is a calculation that illustrate the importance of each factor in the calculation.

Thank you

 

[IRstuff]

If you are designing a test fixture, then you know where the cg of the UUT is, since you're designing something that fits it.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[3DDave]

It would not occur to me to have flanges on the adapter. That looks like a place to have unwanted flexibility and looks to depend on the stiffness of the UUT to control a vibration mode in the adapter.

 

[eli28]

You are right, but I am not sure you answered my question :/

 

[eli28]

Can someone please answer my principal question what is the right method to analyze my case without concentrating on the schematic pictures that were meant to generally illustrate.

 

[BrianE22]

You can't design the fixture vibration-wise without knowing the mass of the specimen. If you can't get it from your customer then you have to make a guess. Then you need to put in writing that the fixturing was designed based on a specimen mass of such and such and make sure the customer signs off on that before you begin.

What is your specimen? I'll bet some of the people here have some experience with similar items. Another consideration is if the specimen has its own natural frequencies. If so then some of its mass will decouple above those frequencies helping you out a bit.

Another approach would be to look at what is driving your shaker table. It will have a certain natural frequency depending on your fixture mass and the specimen mass. Just design you fixturing so its natural frequency is above the table/drive natural frequency.

 

[3DDave]

Everyone with familiarity with vibration fixtures knows what the schematic looks like. If you are not illustrating exactly how the fixture is designed then it is misleading.

You already have instruction from someone to analyze the fixture with the UUT mass but not with any structural contribution.

That is the right way.

Why not compromise and do both?

 

[IRstuff]

A typical environmental spec will essentially describe the mounting interface to a UUT as being infinitely stiff, so that's what you must duplicate with reasonable practicality, and why fixtures are often substantially beefier than anything in the UUT. That's not to say that crap doesn't happen. You show something that's possibly top-heavy, and if you are planning to test this on a conventional slip table, you'll might see a resonance in the 30-Hz regime, but that's the resonance of the oil film/slip table rocking from the moment generated by the UUT/fixture's high cg. If that's the case, you need to test with a bearing table, that's vertically constrained by hydrostatic bearings. We once spent several months trying to figure where a resonance was occurring within the UUT, only to find out that it was the slip table interacting with the high cg of our UUT.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[eli28]

Thank you all for giving me a starting point and tips.
As for the question why I don't include the true tested body in the analysis, there are 2 reasons:
1. The tested body is still being designed by another group in the company I work in, and for making the design phase shorter- there is a decision of working in parallel.
2. I am not familiar at all with dynamic simulations, and I am afraid of garbage in garbage out when dealing with big assemblies, on the contrary to simulation of one single part in which the chance to have a mistake is significantly smaller.

 

[3DDave]

"I am not familiar at all with dynamic simulation"

The shaker table engineers should be telling you what to do to get useful results.

 

[eli28]

3D Dave, theoretically you are right...but actually I have to take action and know what I do for avoiding a failure... after all I am responsible for that jig.

 

[GregLocock]

I'd wag that our test fixtures are something like ten to one hundred times the mass of the UUT. You have to include the UUT's mass - the fixture for a 2 ton vehicle is different to the fixture for an 18 kg wheel, even if the input acceleration spectrum is the same.

Cheers

Greg Locock


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

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

 

[eli28]

Hey GregLocock, thank you for your comments!