How to reduce airborne gear noise? 1

[autonoise]

I'm working on a cast aluminum gearbox that has high frequency airborne gear whine. (throughout ~500-900Hz, with a few peaks). We checked for resonances on the case, but they don't line up with noise peaks. Adding nice ribs has not helped. We can't use a sound blanket or other acoustic treatment to block the noise due to the operating environment :-(. Looking for advice on how to modify the case. Also, why don't the ribs help?

TIA
Dan

 

[Hatch]

If your gearbox is not in resonance, then ribs or stiffening of the case will do nothing except change the existing resonace frequency. What you need is additional transmission loss from the casing (ie add mass), assuming that you have done all you can (or want to) on the gearing itself. Precision machining, gear selection etc can reduce the gear noise.

Can you use a steel or cast iron case? They would provide more transmission loss than the aluminium. Other than that your only recourse is to use an external acoustic treatment. What is the environment which precludes the use of blankets? C. Hugh (

http://www.Hatch.ca)

 

[GregLocock]

Hatch is absolutely right. We've just changed teeth profiles and got 5 dB that way (at a cost).

If you are looking for a quick and dirty one-off fix go to a plumbing shop and buy some lead sheet. depending on your environment you must be able to find some sort of fibre to pack between the case and the lead sheet.

What are your actual constraints on a solution?

Incidentally, the peaks must be coming from somewhere - maybe there are internal resonances in the box.

You might want to look at magnesium for the case - it has higher damping than aluminium, tho still nowhere as good as cast iron.

Have you checked the coincident frequency for your wall thickness/material? 900 Hz sounds a bit low but that may be the problem.

If I were you and working in the auto industry I'd beat the gear supplier up. Don't accept any casing vibration greater than 1 m s-2, at tooth mesh frequency. Cheers

Greg Locock

 

[autonoise]

We are working on an automotive transmission, and there is insufficient clearance between the box & the chassis for acoustic treatment (as well as durability issues in operating environment).

Yes it seems the resonance is internal to the box...working with FEA to understand it.

Fc for the case is > 3.5 kHz.

Not all the boxes are objectionable; when we put in gears which have been sorted for good quality, in a "bad" box, it doesn't help.

Greg is your 1 m/s^2 limit for surface panels only, or for any point on the box? We measure more than that at the mounting point for the "good" boxes.

Thanks
Dan

 

[GregLocock]

Well, to be honest that 1 m/s^2 figure is our first guess - we are trying to get the gearbox supplier to accept the idea of a vibration limit. We have specified that at the transmission mounting itself, rather than all over the casing, but for an airborne sound problem you (obviously) need to specify a surface velocity for the whole surface. We do not have an internal model of the gearbox (they are a black-box item), so I haven't got the faintest what level of vibration is acceptable internally.

I'm amazed that known good gearsets don't cure bad boxes, as I would have thought that the boxes wouldn't change much. Is it the gear transmission error due to the difference in shaft spacing perhaps? Is a bad gearset in a good casing OK? What's your gear transmission error? Is it worse in a bad gearbox?

Our gearnoise problems are at similar frequencies and we have an expensive fix - a torsional damper on the front of the driveshaft. Tuning it is by guess and by god, but it does work very nicely. We also have an internal torsional damper in the driveshaft, but that's usually tuned for diff whine, not trans whine.
Cheers

Greg Locock

 

[autonoise]

We're assuming that good gears make noise because the tolerances on the supporting structures (carriers, etc.) are too big, causing poor alignment. Problem is, we haven't been able to find out which assembly parameter(s) are creating the (presumed) mis-alignment.

We can't actually measure the TE in a gearbox. The modelers can estimate the TE, but we don't have a "target".

In this case it doesn't seen like a torsional damper help the noise radiating right off the box? On that subject, how do you determine where to start when you build your 1st iteration of a damper?

Thanks

Dan

 

[GregLocock]

"In this case it doesn't seen like a torsional damper help the noise radiating right off the box?"

I'm only saying it works for us - and I'm not saying that I know why!


" On that subject, how do you determine where to start when you build your 1st iteration of a damper?"

We get our first tune for torsional dampers by

1) measuring the local modal mass at the frequency of interest, of the mounting point.

2) working out what frequency we'd need on a fixed (infinite base) mounting to give the right tune.

3) develop the damper on a fixed base, using a hammer and accelerometer

4) fit the design to the car

5) start again!

To get torsional measurements we cement a small aluminium block to the periphery of the shaft or damper. If anybody is feeling super techy they sometimes use opposed pairs of charge coupled accelerometers as a torsional accelerometer.


Cheers

Greg Locock

 

[autonoise]

Start again (and again) we shall.

Thanks to all who replied.

Dan

 

[Rob45]

Check back issues of Sound and Vibration magazine; some time in the past year they had a good case history on gearbox noise. The first thing you need to know is which gear is giving you trouble. Relate your noise frequency to gear tooth passing frequency.
It's also possible that you are experiencing bearing noise rather than gear noise. Do an internet search on "condition monitoring" to find the noise signatures of various bearing defects.