About building vibration evaluation

[zhouy7]

Hi all,

I don't have much background in vibration analysis.

We're looking for a new building for our laser measurement system lab. We used a few Kistler 8393B2 (0-250hz response) to collect the vibration data for these new buildings and did some fft analysis. We want set up some criteria to evaluate the data and choose one of the building to move in. But what's start point I need to looking at? Is only the FFT I can work on the data? Should I start from these ISO or ANSI test and analysis stardard? I think I need to find out the sensitivity of our laser system first, but we don't know where to start either. Thanks for any quick put!

I'll post some fft plot for these test result.

 

[zhouy7]

http://files.engineering.com/getfile.aspx?folder=4262ff72-117d-43fe-83e9-4239f1fc1605&file=3.JPG

 

[zhouy7]

Sorry for the last post.

Here are links for 3 FFT plots:

http://files.engineering.com/getfile.aspx?folder=e72a52e2-764d-4c93-97de-2a137371ae8b&file=1.JPG

http://files.engineering.com/getfile.aspx?folder=e2df265d-3ec1-4b47-b362-282695b111b0&file=2.JPG

http://files.engineering.com/getfile.aspx?folder=776c7cb9-b9f7-46f3-95d8-6ce2159c7776&file=3.JPG

 

[Tmoose]

back in the 90s Sound and Vibration magazine had an article on acceptable facility vibration levels for sensitive equipment like SEMs.

Does your equipment mfr have background vibration limits? They should.

Did I miss the units of the posted FFTs.

 

[zhouy7]

Thanks Tmoose!

I think the X units on FFTs is Hz to Y Amplitude(unit mm?). It's just a data plot after the FFT.

As for the magazine paper, do you have more details, such as title...

 

[GregLocock]

Judging by the first jpg you mostly seem to be measuring mains hum, or possibly a synchronous machine.

Since this is a reasonably expensive decision I suggest you get someone who knows how to measure vibration to do the job.



Cheers

Greg Locock

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

 

[MikeHalloran]

Choose the building that's farthest from existing or proposed major highways, drop forges, and especially, railroads.



Mike Halloran
Pembroke Pines, FL, USA

 

[zhouy7]

Yes, these buildings are all close to a busy highway.

My questions is not only focusing on the choosing the better one, but also to find out and set up some criteria for the vibration evaluation for our future product and project. We're trying to learning. Thanks and please input any suggestions.

 

[IRstuff]

Seems like you're putting the cart before the horse.

What laser measurements are you trying to make? Which ones are actually vibration sensitive?

TTFN

FAQ731-376

 

[zhouy7]

I have found some good papers from:

http://cganda.com/

My new questions:

1. I used accelerometers to get the vibration data and I get the acceleration signals. How can get the velocity signals?

2. How can I get the root-mean-square (rms) values for the FFT analysis? What's the benefit of rms compared to the plain FFT data?

Thanks.

 

[IRstuff]

velocity is the integral of acceleration.

Again, you need to explain what LASER measurements you're going to do and how. Measuring linear vibration might be interesting, but it might also be TOTALLY irrelevant to your laser test setup.

TTFN

FAQ731-376

 

[zhouy7]

My laser measurement system is a laser transmitter or tracker. Will sit on either the floor or roof truss.

My question now:

I have the data for Acceleration-Time from the accelermometer. How can get the RMS acceleration/Frequency plot? It might be 1/3 octave frequecy. Any suggestions?

Thanks.

 

[271828]

I think you need to look at the American Institute of Steel Construction Design Guide 11 and the Steel Construction Institute's equivalent document, Floor Vibrations, A New Approach or something like that. Floor vibration analysis methods exist in both of these and are relatively easy to compute. At least those would be good to use along with your tests.

 

[zhouy7]

Still the question about the rms.

If I did one trial and collected 100000 data (acceleration-time), can I break these data into 100 set (100 data/set) and calculate rms for each set. I get 100 data and they're all rms data. Then I do the FFT for these 100 rms data. Can I say it's the RMS(acceleration)-Frequency? Other way, I just to do 100 trial test and do the rms for 100 trial. The only difference is sample space. Am I right or wrong?

Please tell me how can I to get the rms(acceleration)-frequency. Many thanks.

 

[Transient1]

You may find Vibration Specturm Analysis by Goldman useful. It's a practical book of vibration measurement and if I remember correctly seems to have answers to a lot of the kinds of questions your having (like velocity from acceleration, etc).

 

[zhouy7]

Excellent! Thank you so much, Transient1!

I took a glance at the book. It's a perfect fit for my requirements.