Noise converted to vibration 4

[WARose]

First off, let me say: I'm (obviously) not a noise guy. Most of my work in vibration has been (low frequency) structural design for machinery vibration control. That being said, I am looking for a sanity check that some noise is not inducing some level of vibration I am measuring. (This is answering a suggestion some people who work at this facility have made....and I have to answer.)

More specifically: a piece of machinery in a room is believed to possibly be producing vibration in an adjacent area from its noise level. I am seeking to say it is likely more from unbalanced forces transmitted through the structure.

I think I can track down the absorption coefficients and all that.....but rather than make this a complicated acoustics problem....is there a way to just figure: hey, this is the max. noise this thing is producing....it can't cause more vibration in [widget x] than [this]...ergo we know....

Only one problem (possibly among many): I don't know how to go from a sound level to induced vibration. Can anyone help me with that? Thanks.

 

[GregLocock]

It isn't easy except in simple cases. For instance suppose you measure the noise near a vibrating panel. Is the panel driving the air or vice versa? So you need to measure acoustic intensity, ie the energy flux. Here's a somewhat relevant paper

http://www.sea-acustica.es/fileadmin/Coimbra08/id161.pdf

I do have a few acoustic books still, I'll see if I can come up with a formula for the vibration in a panel due to a plane wave. It'll depend on the modal properties of the panel



Cheers

Greg Locock


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

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

 

[GregLocock]

No luck there. A google search on "acoustic excitation of structures" gave a large number of hits. The dangerous frequencies are those where the surface wave speed of the panel, which varies with frequency (and a whole lot else) match the speed of sound in air. That's called the coincidence frequency. Many years ago two of us spent a long time working out the inverse problem, the sound power generated by the various surfaces of a vibrating engine. There is a useful paper on that, which I have not been able to identify, written in the early 80s.

Cheers

Greg Locock


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

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

 

[Erik Panos Kostson]

I believe it is perhaps possible for simple cases like a vibrating panel (baffled plate) excited in a certain mode (or multi-modal), to predict the sound field from the plate. Books in acoustic might have some solution for that. Now for more complex general cases I think it will be hard, and a vibro acoustic analysis/measurements are necessary. See here for a relevant paper from ISVR, which is also about what GregLocock explained, the coincidence freq. which is the frequency where the bending wave phase velocity is equal to the speed of sound (~ 340 m/s), and hence they radiate very well into that medium:

https://www.sciencedirect.com/science/article/abs/pii/S0003682X10001490

You are also right that sound can be structure born, and travel via structures from one place to another, via bending (e.g., A0) and quasi-longitudinal (S0) waves, and thus generate sound in adjacent rooms. Also it can travel in the air (air born), and then be transmitted to other rooms. So many paths and very/quite complex phenomena - very specialised area. An acoustic engineer/consultant is a specialist in this field, and if this is a serious problem, they will be able to assess that (they will typically perform something like this, so a typical task, including noise reduction implementation, especially when the noise levels are above certain criteria depending on the legislation in the country one is in).

 

[WARose]

Thanks for the replies guys. (And the papers.)

It isn't easy except in simple cases. For instance suppose you measure the noise near a vibrating panel. Is the panel driving the air or vice versa?

Seems to me like the worst could be assumed and run from there. (That's my ignorance on this subject talking perhaps.) If you have a (max) measured noise level at the source....and know the frequencies/modal shapes of a object receiving/in question......I would think/hope you could back out of that with a worse possible displacement. (Maybe?)

 

[Tmoose]

I missed what kind of instrumentation you have available.

Can you hear the machinery in the locations where vibration is a concern?

Please elaborate on "adjacent."

Please elaborate upon the vibration complaints in the "adjacent area.
The plants on the VP's private secretary's desk can be seen a-quivering?
The conference room's ceiling mounted projector throws a shaky image during presentations?
The sales director's power naps are too deep as a result of futon vibration?

What is the construction of the building between the machinery and the "adjacent" areas.
Is the machinery isolation mounted?
If so, is there any perceptible vibration at the "grounded" end of the isolators? The floor, the electrical support connections and conduit, piping and ductwork wall, ceiling and floor penetrations?

My starting point would be confirming the machinery's sound frequencies match the vibration frequencies in the "adjacent" area.
And similarly, quantifying the "adjacent region's" vibration amplitude and frequency.
Octave band frequency resolution is not enough for this. I'd be tossing the audio thru my signal analyzer.

Once again those practical folks at ASHRAE provide some wonderfully useful plainly written information.
The attached image is from ASHRAE handbook FUNDAMENTALS. Chapter 8 in the 2009 edition.

 

[btrueblood]

"More specifically: a piece of machinery in a room is believed to possibly be producing vibration in an adjacent area from its noise level."

Along the lines of Tmoose, I'd be tempted to haul some sensitive microphones or accelerometers, and their associated recorders/analysers to the adjacent area, and make some measurements; specifically (a) when the machinery in question is running, and (b) when it isn't. Should be kinda cut and dried from there...

 

[WARose]

Thanks for the ASHRAE chart Tmoose.....I might be able to back into some (ballpark) displacements with some other charts I have that predict human response to vibration with that ASHRAE info.

I would answer your questions....but I think that is going down a rabbit hole away from what I need to do here. I know what the guy wants. (I've dealt with him before.) And he's not going to spend the money for a lot of extra measurements. I need to disprove his theory with something like that chart (or anything else anyone can think of).

Thanks again.

 

[WARose]

By the way Tmoose, in that chart, that dashed line with the "T" next to it.....is that the threshold of perception? (Didn't know if it was explained on another page or not.)

 

[WARose]

Here is an idea: if I have a noise level reading (in dB; with 20 micro pascals as the reference base), could the calculated sound pressure level be the amplitude in a sinusoidal forcing function? (I assume the frequency in such a function would be the machine’s frequency.) This assumes the panel absorbs every last bit of it.

Thoughts?

 

[Tmoose]

"And he's not going to spend the money for a lot of extra measurements."

"some level of vibration I am measuring."

What do your vibration measurements look like? 2-500 or 1000 Hz. FFT, spectra, signatures, profiles, not words please.
Floors, ceilings, walls ?

You mentioned you favor "unbalance forces" and structure born vibration. Does the " adjacent area" (room?" vibration frequency match 1X rotating speed of the noisy equipment? Is that what you meant by unbalance forces ?

Unless the frequency(s) of the sound in the equipment room can be shown to match the frequency(s) of the adjacent area vibration (which I think you may have) in question, any argument for or against is simply specious.

Who has to pay to "fix it" if the vibration is caused by sound, vs structure borne vibration ?

 

[WARose]

What do your vibration measurements look like? 2-500 or 1000 Hz. FFT, spectra, signatures, profiles, not words please.

The device I am using measures peak (or RMS) displacements, acceleration, and velocity. I don't have a time-history, response spectra, etc. I know the frequency from calculations I've done of the previously mentioned data and the equipment drawings. (And to answer your question, it is in the 2-500 Hz range.)

You mentioned you favor "unbalance forces" and structure born vibration. Does the " adjacent area" (room?" vibration frequency match 1X rotating speed of the noisy equipment? Is that what you meant by unbalance forces?

Yes....but the unbalanced force I refer to was a proposed model to mimic (in a FEA analysis) the vibration I am seeing/measuring. The panel is modeled.....the only issue is: what is the forcing function? Can it be described by the calculated sound pressure level (as the amplitude; with the equipment frequency being the forcing frequency)? Or is a simple sinusoidal forcing function inappropriate? (I don't care about inaccurate as long as it's conservative in terms of: it gives me a max displacement possible.)

Who has to pay to "fix it" if the vibration is caused by sound, vs structure borne vibration?

They do. (If they address this at all.) I had nothing to do with it. They put this thing in without talking to anyone externally (except for the equipment vendor).

 

[Strong]

"And he's not going to spend the money for a lot of extra measurements."

From the last comment, I understand that the problem is with a Panel (perhaps instrument panel) that has unacceptable vibrations. If this is the case, then the panel vibrations can be from:

1) Structural transmitted vibrations from nearby machine -- most likely
2) Structural vibrations from machine in adjacent room -- possible
3) Vibrations produced by high sound pressure levels in room -- not likely
4) Vibrations produced by high sound pressure levels from machine in adjacent room (airborne or structure borne transmission) -- far less likely

One simple test would be to shut off one machine at a time and measure panel vibrations and nearby sound pressure level. More advanced measurements would include spectrum (FFT) of sound and vibrations. I doubt that there is an easy way to calculate a solution (instead of measurements) to your problem, unless your customer can be easily accommodated!

Walt

 

[WARose]

Vibrations produced by high sound pressure levels from machine in adjacent room (airborne or structure borne transmission) -- far less likely

I understand that.....that's the whole point of this exercise. I want to slam the door on this guy's theory. (And this thing isn't going to be shut down anytime soon.)

So.....THE question to ponder here is: Can a forcing function (caused by sound) in a time-history/dynamic model be (conservatively) approximated by a simple sinusoidal forcing function with the calculated sound pressure level as the amplitude and the equipment frequency being the forcing frequency?

And if the answer is "no".....a suggestion for something else is welcome. I realize it's probably overkill since some sound is reflected and so on. But it's a conservative answer I am after.

That's it.....no interest right now in the device I am using, possible measurements to take, the wall type, etc, etc. The focus needs to be 100% on the question in bold. If anyone doesn't know, that's fine.....I will keep looking in the meantime.....and possibly report back what I find.

 

[Strong]

Perhaps these references will help with your analytical approach:
Search: sound induced panel vibrations

https://www.bing.com/search?q=sound...HvtekIhXtaxL09PkI1bSeVZjJX9ztj18LOSYc*cZyZXG9

Acoustically induced vibration of, and sound radiation from, beams inside an enclosure
From <

https://www.sciencedirect.com/science/article/pii/0022460X76907379>

Simulating low frequency sound transmission through walls and windows by a two-way coupled fluid structure interaction model
From <

https://www.sciencedirect.com/science/article/pii/S0022460X17301207>

Relationship between sound radiation from sound-induced and force-excited vibration: Analysis using an infinite elastic plate model
From <

https://asa.scitation.org/doi/10.1121/1.4955308>

The last reference may be of most use.
Typically sound-structure interaction is by sound pressure and vibration velocity (not displacement). The coupling between sound and structure is determined by the sound field (wave) and the material size and properties.

Walt

 

[WARose]

Thanks Strong. (Your last link is to something that isn't available by the way.)

 

[GregLocock]

I think you might get in the ballpark with the assumption that it is just a pressure pushing on a panel. But it is a big ballpark, 30 dB or so.

Cheers

Greg Locock


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

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

 

[WARose]

Thanks Greg.

 

[GregLocock]

It'll be a high estimate, the highest radiation efficincy I can remember is 3 dB. at low frequencies it is very low, then peaks and says at more or less 0 dB

Cheers

Greg Locock


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

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

 

[WARose]

High estimate is what I am after. I need a number that I can say: it can't be any higher than that.

Thanks again.