Natural Science Forum / Physics / Acoustics / July 2006

Dear Christos

I have measured a number of generators for projects where I had an
existing installed Gen of similar type and needed the data for a new
install. I think that I used the ISO standards that you mentioned.
Engineering Noise Control by David Bies and Colin Hansen also has a
good chapter on SWL estimation by various methods.

In any case from your description I think that you will have a lot of
difficulty getting an accurate SWL due to the closeness of the
enclosure walls especially if the internal walls had a low absortivity.

If you could install (temporarily) lots of insulation on the internal
walls of the enclosure you may be reduce the reverb level enough so
that you only have the near field  to measure ie close as you can
practically get to a free field considering the confines of the space.
Measure in the present condition then install insulation until the
measured noise level no longer drops, this should get you close to a
free field as you can in the small space.

You could then measure at 1 metre and extimate SWL = 10 x Log (surface
area of envelope at 1m from surface of Gen) + Lp (avg of measured
locations) then add a healthy + 3 dB or so for the new design.

Hmmmm sticking my neck out here. It would be good to see what some
others say as I also just a humble consultant.

Tom

This question seems rather confused. The question jumps from "sound
pressure levels at 1 metre" to "Sound Power". Which is being measured ?

Is it 1 metre from the outside of the generator enclosure or 1 metre
from the generator which is inside the enclosure? If outside the
enclosure, is this a compliance requirement for hearing protection, or
if inside is this for noise reduction?

If "1 metre from the outside of the enclosure" which is near a wall then
the reason for the measurement needs to be rethought. If for hearing
protection then measure where people are likely to be. If people are
likely be in the 1.4m gap then measure there. The reverberation will be
a legitimate part of the measured sound pressure level.

What is the purpose of the measurement ? and don't get tangled up in
measurements simply for measurements sake.

If it's sound power you're after then there are several old techniques
to take these measurement that would ruffle the purists but still give
reasonably accurate results.

Please clarify ?

    The classical method, giving perhaps "engineering accuracy" is to apply
a reference sound source. Here, the sound pressure level is measured
anywhere convenient in the enclosure, preferably more than 1 m from the
walls. Then the generator is shut off and a reference sound source is
placed inside the enclosure by the generator, again preferably 1m from
the walls, and the sound pressure level is again measured. From the
calibration of the reference sound source (RSS), one then can compute
the sound power with some confidence.

    Overall, ISO (and other countries) have written a number of standards
that claim to measure sound power. I find the titles to be confusing and
perhaps misleading.

    The ISO overall set is:

    The determination of the environmental factor K2 is imbedded somewhere
in these standards.

    It is my opinion that the RSS method gives you the best opportunity for
precision.

    The ASTM E1124 method is the next best thing:

    There is an American ASTM method (ASTM E1124) called the "Two Surface
Method" that was developed specifically for coal pulverizes, where the
method was tested with a calibrated reference sound source and found to
be acceptable. The degree of accuracy was not clearly stated, but it was
believed to be adequate for the purpose intended, to compare one
equipment against another, and to give a useful estimate of the true
sound power. My inspection of these precision statements leads me to
believe that it is precise certainly better than +-5 dB, and mode likely
+-3 dB; certainly over the mid frequencies. The 5 dB limit might be
applicable below 100 Hz; all my opinion.

    The inner surface is anywhere from o.15m to o.3m from the noise source
surfaces. The outer surface is further, where a consistently measurable
drop of sound pressure level occurs. This difference, if consistent,
could be as little as o.1 dB,, but larger values are preferred. (It
seems like one or two dB would be ideal.) A two channel sound level
meter is used so that for any given radial line location, both
measurements are made simultaneously.  Many positions about the tested
equipment must be measured. It is also allowed to sweep he surfaces of a
rectangular volume.

    The area of the inner and outer surfaces must be determined. The
surfaces can be nearly conformal (shape following the shape of the
tested equipment). Computation of sound power uses the inner and outer
SPL's and areas.

    Re your citations:

 >      BS EN ISO 3744:1995  Acoustics. Determination of sound power
levels of

    3744 allows the reference sound source ("substitution") method, and
provides means to determine K2

    3747 is aimed specifically at the RSS substitution method.

    11204 is only concerned about the SPL (not sound power) that is found
at the operator or bystander location. Ut is intended for health
purposes and not for sound power determination.

    For obtaining RSS's, see our web pages, RSS section (RSS300):

        www.campanellaacoustics.com

            Angelo Campanella

    I re-read your post, and there is a different interpretation of your
question:

    I believe you have asked about an enclosed source operating in a much
larger building hall.

    In that case it is more likely that you should use the ISO 11204 method
since you desire SPL for personnel located between the unit enclosure
and a wall only 1.4m away.

    It is certainly true that if personnel are there, and the wall reflects
sound, that the SPL there shall be higher than for other locations where
there is no nearby wall.

    It implies that the generator should not be near that wall, r that
personnel should not be allowed to be between the generator enclosure
and that wall unless they wear hearing protection.

    3744 and 3737 have little or no meaning in the health context.

    (WSYWYG): "What you see is what you get".

    The SPL found there is not to be disputed. If it is above a maximum
(e.g.  85 dBA), then hearing protection may be required. Certainly, a
sign indicating "High Noise Level" must be placed there.

    On the other hand

1- If you want to measure the sound power that is being sent into the
large building hall where this enclosed generator is located, then
perhaps 3744 is needed. Here, if all noise producing equipment can be
shut down (e.g. run the test in early morning hours when all noise
producing equipment can be shut off), then the reference sound source
"substitution method" can be used.

If this cannot be done, then

2- The absolute method where the reverberation time of the hall can be
measured, can be applied.

But if that also cannot be done, then the only alternative is:

3- The ASTM 1124 "Two Surface Method" is your remaining alternative.

4- (The "Intensity method" is also available, but this acoustical test
equipment is much more expensive and more difficult to use.)

    Angelo Campanella

As already mentioned, first you must clarify to what purpose the meaurement
shall be taken.

If you want to evaluate the hearing risk of an operator working in this
environment, you must measure the A-weighted sound pressuren level in this
place, because it is the sound pressure level to which the hearing of man
responds. Compare the measured value with the respective standardized noise
limit, for instance an equivalent continuous sound pressure level of 85 dB
(A) as the limit of a daily exposure time of 8 hours valid in my home
country Germany and many other countries.

If, however, you are interested in learning the noise emission of your
generator, you will have to apply one of the standardized procedures
mentioned by Angelo Campanella. In my optinion, you can skip the methods of
sound power measurement based on sound pressure, because they fail in
critical situations such as short distances and mixed sound fields (neither
free field nor true reverberant field) where they yield only limited
precision. You should concentrate on the methods of sound power measurement
using sound intensity, i.e. the ISO 9614 series. It might be a little more
expensive - you need a sound intensity probe, a normal sound level meter is
insufficient. On the other hand, this is the only method that will give you
reliable results in situation where all the methods based on sound pressure
yield vague approximations at best.

Heinz