Do sound waves caused by a vibrating machines correlate to the vibration waves of the machine itself?

Question

I'm a computer engineer, so please excuse my lack of mechanical knowledge :).

We have a machine that has large electrical motors causing massive amounts of vibrations, the machine also consists of many frames and large plates. Due to this vibration a part of the machine is generating sound waves that make it unbearable for humans to come close to it, let alone talk with each other. They have added support to the machine, and although this did make a difference to some of the vibrations, the ones causing the unbearable noise is still present.

So I came up with a solution. I want to take a high quality mic and measure the sound, from there on I want do do a frequency analysis to determine the frequency of the noise. Then I will measure the vibration of individual parts of the machine and find the part who's vibration frequency matches that of the sound. Then we can hopefully isolate the sound by adding extra reinforcement and vibration damping to that specific part or parts.

My question is, will the vibration of the machines frequency match that of the sound, and will this be a viable way to solve the problem?

Edit 1: I can't really share any more technical information about the project, because of NDA.

Edit 2: Just to clarify, I will not be making design decisions on how to improve the sound once the problem is located, we have enough talented mechanical engineers to do that, my solutions (adding reinforcements) were only some assumptions made by me.

Answer 1

Yes they do.

If a large piece of rotating machinery is creating unbearable amounts of noise, it is because its rotating parts are unbalanced, which causes it to shake back and forth as it rotates- and because it is rigidly attached to its support frame, it sets the frame into vibration as well and this makes the noise worse.

Note here that by clamping the machine more rigidly to the frame, the coupling between the machine and the frame is enhanced and the radiated noise can actually get worse in this case!

Instead, you need to decouple the vibrating machine by using flexible mounts between it and its support frame so the frame doesn't vibrate as the machine vibrates. the engine mounts used in cars work this way; they are blocks of rubber the allow the engine to shake without making the rest of the car's frame shake.

As pointed out above, electric motors are very well-balanced at the factory and produce only a mild humming noise as they spin, plus a whooshing sound from the cooling fan. The noise is coming instead from the machinery that the motor is connected to.

You say the noise is unbearable, but please tell us this: is the sound an unbearable, high-pitched scream like a jet plane taking off? or is it an unbearable, moaning groan like a saw blade cutting through a piece of wood? or an overwhelming, rumbling roar? or a percussive, high-speed bang-bang-bang like a machine gun? or a series of big explosions going off one after another that shake the floor? you don't need a precision microphone to determine this: for superloud noises you ears are good enough.

If you can describe the noise, we can offer concrete guidance to you! I'll sign an NDA if you want ;-)

Answer 2

The noise is from the attachments whose natural frequency is close to the natural frequency of your machin/s. they resonate and the amplitude of vibration exponentially increases for them.

You can try to disconnect the elements temporarily from the machine/s one by one and find out which one is the culprit!

Then you can try to add test ballast to that individual part to change its natural frequency so that it doesn't resonate with your machine.

Answer 3

"My question is, will the vibration of the machines frequency match that of the sound, and will this be a viable way to solve the problem?"

Yes, they will. One way of finding which part of an engine is causing a particular noise is called acoustic intensity via structural vibration. In this we measure the vibration of the panel and do some maths and that gives the strength (intensity) of the local sound. The topic begins, but does not end, with the idea of a piston in a semi infinite baffle.

Alternatively you can measure the acoustic intensity directly, using an intensity probe, but curiously this tends to be more problematic than the structural approach.

Either way you will be going fairly deep into Noise and Vibration analysis, you'll need to hire someone.

Having measured it then you have to figure out how to solve it... again experience is useful, but it might be as simple as building a huge concrete box around it.

Good luck (and next time, consider noise reduction when designing the system).