Natural Science Forum / Physics / Acoustics / November 2005

There is BEM software for assisting this kind of problem, see
http://www.scientific-computing.info/smk/horn/paper.pdf
For general work on the BEM see
http://www.boundary-element-method.com

If I understand correctly, you believe you can achieve 115 dBA at one meter
from the exhaust, and can put the exhaust on the opposite side of the unit
where it is directed away from your operator and the unit acts as a barrier.
You get some from extra distance, and more from the directivity/barrier
effect.  However, the directivity/barrier effect is limited by atmospheric
scattering.  My expectation is that you would still be a few dB above the 85
dBA goal.  This is considering only the exhaust noise.  Do not forget about
the block radiated and cooling fan noises that will radiate out the air
intakes and discharges.

    That approach runs out of feasibility because in order to get the
barrier effect you desire, that barrier for which the exhaust opening
"would be on the other side" should need to be quite large, like the
whole side of a house or garage.

    My experience with quieting APU sound is that a large expansion chamber
(aka "muffler"), perhaps another 50" cube, or a small sized oil drum (I
used an empty 15 gallon oil drum) should be used as the proxy sound
attenuator into which the exhaust sound is fed into with a flexible
exhaust pipe. I have heard of Army units in the field doing exactly
that; employing an empty oil drum as a super-muffler. It has worked
great for me at home. Line the interior of that expansion chamber with a
few inches thick fiberglass, metal wool, or perforated metal. Some was
feed the tailpipe from the center or certainly away from the any inner
surface. But the volume in cubic feet is the critical parameter in
quelling exhaust sound.

    There exists a possibility to use active noise cancellation in your
tailpipe outlet. These have been under experimental development by the
auto industry for some years. The early-on problems were the life
expectancy of the loudspeaker driver in that unfriendly (hot and moist)
tailpipe environment, and the nonlinear behavior of gasses at high sound
pressure amplitudes. Those programs soon concluded that a conventional
muffler, as you already have identified, was needed to precede that
tailpipe ANC device.

    Certainly yes in the vicinity of the engine/generator, provided that
the cooling fan and engine combustion air intake are designed,
configured and treated to reduce their noise emissions as well, possibly
with ANC devices deployed on those air inlet/outlet ducts as well.

    In acoustics, mathemaical models are best thought of in an historic
role rather than as a prediction tool. That is, there are so many
variables in acoustics, that we are best advised to perform essentially
empirical tests whose results can be understood(!) with ad-hoc
mathematical models, rather than to create mathematical models,
expecting nature to follow them to our pleasure.

    Just my humble opinions........

        Angelo Campanella