Compressor noise reduction....

Thread view:

Enable EMail Alerts

Start New Thread

Thread rating:

kifuyu@gmail.com - 20 Mar 2006 07:28 GMT

Hello! Everyone

I have a curious things about compressor noise.

>From measurement the blade passing frequency of impeller is identified

as dominant noise source. There are many peaks of BPF in spectral graph
of noise measurement.
Overall level of noise is 99dBA, however, the summation of such a BPF
peaks is only 92dBA. I think that remaining noise level 98dBA is
broadband noise induced by aerodynamics.

Is useless the effort of reducing BPF noise ?

BPF noise levels are 10dBA higher than other spectral noise source. I
think that reducing impeller noise is very efficent way from a
practical view such as masking effect.

thanks.

Reply to this Message

Noral Stewart - 20 Mar 2006 12:46 GMT

> Hello! Everyone
>
[quoted text clipped - 3 lines]
> as dominant noise source. There are many peaks of BPF in spectral graph
> of noise measurement.

You are apparently speaking of the BPF and its harmonics.

> Overall level of noise is 99dBA, however, the summation of such a BPF
> peaks is only 92dBA. I think that remaining noise level 98dBA is
> broadband noise induced by aerodynamics.

If the BPF and harmonics do truly sum to 92 and the over is 99, then the BPF
and hamonics are not dominant.

> Is useless the effort of reducing BPF noise ?
>
> BPF noise levels are 10dBA higher than other spectral noise source.

You just said this was not the case, unless you mean that the tones are 10
dB above adjacent frequencies in a narrow band analysis.

I

> think that reducing impeller noise is very efficent way from a
> practical view such as masking effect.

Not sure what you mean here. Is your concern overall noise reduction
(probably) or masking of tones? Very different problems and results.

> thanks.

Reply to this Message

Brian Marston - 20 Mar 2006 14:56 GMT

> Hello! Everyone
> I have a curious things about compressor noise.
[quoted text clipped - 10 lines]
> practical view such as masking effect.
> thanks.

If noise control were flat dBA reductions and not functions of frequency
then we'd ignore the BPF and concentrate on the remainder (at least down
to the 92dBA level).

BPFs tend to be dominant down in the low frequencies, where noise
measures tend to be least effective. That's why compressors have
pulsation dampers. General noise control measures tend to be most
effective in the mid and high frequencies, and less effective in the low
frequencies.

Noise reduction should not be simply viewed as dB(A) reduction,
frequency band reduction can often be far more important. Once noise
reduction is divorced from the human element it becomes meaningless.

Reply to this Message

ari - 21 Mar 2006 18:41 GMT

> Hello! Everyone
>
[quoted text clipped - 14 lines]
>
> thanks.

That's quite common that 'dominant' noise components depend on the way
of evaluation (as I understood other comments). In linear noise emission
sense narrow band components might be / probably are dominant. Possibly
the applies to disturbance/annoyance as well?

According to your numbers, in A-weighted noise emission sense broadband
noise is more important, as you suggest. On the other hand the noise
comtrol measures depend on your target level.

I like to think that BPF is also due to aerodynamics. For example only
(a small) portion of it is due to blade thickness noise. Inflow and
outflow conditions (to/from impeller) effect both narrow and broadband
noise. On the other hand, if you prefer to look at BPF as volume
velocity source, I'm not sure how to separete or name the noise sources.

And finally, how "high" the BPF nad multiple components are depends on
frequency resolution. I prefer to look at BPF noise relative to total
level, i.e percentage (in case of noise emission).

BR,

ari

www.machineryacoustics.fi

Reply to this Message

Natural Science Forum / Physics / Acoustics / March 2006

Compressor noise reduction....