Vibration Comparison

[frankjconway]

I'm an ME that has very limited experience in vibration analysis. I was recently handed a project to evaluate the effectiveness of a vibration damping mount.

We had two tests done to the UUT per MIL-STD-810F(Fig 514.5C-1). The first test had the UUT mounted directly to the shaker table. The second test had the UUT mounted to the vibration damping mount, which was mounted directly to the shaker table.

All in all, I got 6 sets of data (3 axes with vibe mount, 3 axes without). The plots I got were in g^2/hz as a function of hz. Also included in each chart is a g(rms) value. It is my understanding that this value is the area under this curve. Is that correct? Also, is the relationship between the g(rms) values a good indicator of the effectiveness of the mount? For example, one axis had a value of 2.70 g(rms) without the mount and 1.66 g(rms) with the mount. Can I use these numbers as a valid comparison to present to others? Thank you.

 

[SomeYahoo]

It's part of the story, but overlaying the plots and looking at key frequencies will also yeild useful data (reduction of peaks, shifting of resonances, etc.).

Good luck.

 

[CESSNA1]

FRANK: Not surprisingly this may not be as easy as it seems. First what is meant by "evaluate the effectivenss of a vibration mount"? Overlaying the PSD plots may give an indication but they will vary with time so it will not be a strictly direct comparison. Typically they have a 3 db (50 % power) tolerance, but the shaker does not shutdown until they reach typically 6 db. Therefore there can be a significant variation. Each plot should have the overall Grms printied on it. That is the the overall energy where the accelerometer is placed. Again there will be difference between the plots. How was the location of the acclerometers detremined? Where is the contol accelerometer on the table of the UUT? Can you post the PSD plots to a website? I assume that the same UUT was used with the same accelerometers in the same location of the UUT.

Regards
Dave

 

[rob768]

A vibration mount should isolate vibrations from the machine/engine/vibrating part from the rest of the installation. So, basiscally, there should be a considerate difference between vibration on and below the mounts.
Some basic aspects to consider:
- no resonance in the resiliently mounted instlalation.
- Displacements within certain specified tolerances (which can be altererd by changing stiffness of the support.
- Allowable displacements during transients.

 

[denis81]

don't forget that in this kinda test it all depends on frequency (if it's low - no worries), when you mount your UUT directly to the shaker table, there is some damping occurs(which is gonna depend on coupled diameters and tipe of mounting... I assume you used greese). Funny you should mention this, but just couple of weeks ago, this guy from B&K gave me this paper. long story short it says that if your standard reference's resonant frequency is lower than 33kHz, then at 10kHz during the test your error might be up to 30%. Mounting is really important aspect of any vibration data aquisition.

 

[rob768]

Denis81, if the frequencies are low, there are still things to worry about. Low frequencies means large amplitudes. This isn't favorable for piping and other components connected to the a resiliently mounted installations. We have have examined an engine fire which started due to broken feul lines, despite the use(or in fact because the use) of bellows. This because of resonance of a resiliently mounted diesel engine which was excited during engine misfiring.
The mean thing is: avoid resonance by tuning mounts and possible exciting forces.