On Thu, 03 May 2018 11:18:01 -0700, tabbypurr wrote:

On Thursday, 3 May 2018 18:07:49 UTC+1, Johnny B Good wrote:

The other option of running the generator inside a soundproof box
brings
a whole new set of problems such as providing the necessary
ventilation, usually using a large fan or two to provide forced
ventilation. Not only does the power consumption of a large fan reduce
the net power available,
the ventilation provision (fans and associated vents) often severely
compromises the soundproofing benefit of the enclosure which ideally
should be fabricated from brick or concrete panels with sound absorbers
installed within (but not applied as panels completely covering the
walls to act as counterproductive thermal insulation).

I saw a lot of such ill conceived attempts at enclosing generators in
such housings, intended to trap the noise emissions, on youtube and the
use of lightweight materials was the predominant error in their design.
There's a lot to be said for the KISS approach to this problem (aka,
intelligent use of sound deflector panels). In short, all these youtube
videos of "Generator dog houses" were, in essence, best regarded as
"How Not To" videos for all the help they were. :-(

The principle is known & simple enough. Provide a sepentine path for the
sound, ie at inlet & outlet. I would have though the exhaust would be
capable of driving ventilation.

I'm rather surprised that the youtubers hadn't at least done some
homework on the basic principles of noise suppression before spending on
their ill fated follies (noise suppressionwise - the other aspects of
protecting them from the elements and possible theft wasn't so much in
question).

Not so much serpentine as a zig-zag on the intake side to reflect the
sound back into the main chamber for multiple 'visits' to the sound
absorbent blocks or curtains within.

As you say, the exhaust *is* put to good use to accelerate the slower
moving engine cooling air that is exhausted in a vent that surrounds the
exhaust outlet (as per the Honda and the Parkside gensets) using the same
principle as the one used by turbofan jet engines on commercial airliners
to improve thrust efficiency. In this case, the slightly warmed engine
cooling exhaust is also put to use in cooling the exhaust muffler box,
allowing the use of such snug fitting plastic housings in this 'suitcase'
class of generator.

With that in mind, I think I'd design my brick "Doghouse" not only to
have ample room within for sound absorbers (curtains or blocks) but also
to allow for generator size upgrading.

Rather than have a relatively small exhaust aperture in a 'back
wall' (just bigger than the rear exhaust vent of the genset), I'd extend
the back out for another meter or two and then mount a sonically
transparent tube just a little larger in diameter than the rear vent of
the machine, around which I'd fill the gap with suitable sound absorbing
fibre, terminating the end of the tube to the end of the doghouse
extension with a rigid panel.

Leaving the inner side of this packing held in place with an
acoustically transparent panel would contribute to the sound absorbing
materials used within (possibly to the extent that it could even suffice
all on its own).

I think I spotted a reference to the engine rev range between eco idle
and maximum load somewhere on the 'net', possibly relating to the earlier
PGI 1200 A1 model (it isn't mentioned in the specification section of the
printed user guide), where ISTR figures of 3900 and 4300 rpm which struck
me as a rather conservative ratio between eco idling revs and flat out. I
think the figures for the Honda eu1000i were 3000 and 4300 rpm
respectively, still a conservative ratio imo.

The reason I'm interested in engine rev figures is that it decides the
lowest fundamental frequency of engine noise (65Hz and 50Hz respectively
for the Parkside and Honda gensets) and half those frequencies in the
case of single cylinder 4 stroke engine exhaust sound pressure waves[1].

To be fair, the higher the rpms, the higher the frequency, which eases
the problem of absorbing the sound energy at these shorter wavelengths.
Also, of course, the lower torque demand means lower mean cylinder
pressures making for a quieter exhaust note at partial load. Still, since
we ideally need a quarter wave resonant structure within which to place
sound absorbent material for maximum effect, it's a vital consideration
when it comes to sizing our rigid enclosure.

A quarter wavelength in air at 20 deg C for 65Hz works out to 1.32m (at
0 deg C it's 1.27m - the speed of sound in air goes up with temperature).
However, a quarter wavelength sound absorbing resonant cavity needs to be
within a room with a dimension of at least half a wavelength in order to
function properly. If we're going to size up a rigid enclosure to act as
a resonant absorber it needs to have a dimension of a minimum of half a
wavelength for any sound absorbers to work effectively.

In this case, that works out to 2.64m or 104 inches (nearly nine feet!
8' 8"). I think, if I was going to embark on a brick 'doghouse' for an
emergency genset, I'd go for a 4 1/2 by 4 1/2 by 9 foot tunnel split by
an acoustically transparent panel with an exhaust vent sized hole linking
to an acoustically transparent tube of the same cross-sectional area
leading out to a back wall vent with acoustic absorbent material filling
the surrounding space in the exhaust half of my 'doghouse'.

The front door would be a reasonably substantial panel sealed against
the front with rubber gasketting suitably vented with zig-zagged trunking
into the interior space designed to reflect any noise back towards the
exhaust end of the enclosure where they'd be dissipated into the exhaust
vent sound absorbing lining.

Even with zig-zagged trunking, the necessary vents will compromise the
noise containment performance to some extent but there are limits as to
how long a 'serpentine' path you can use before the issue of restricted
air flow starts to rear its ugly head. In any case, there's little point
in making the front door noise leakage any more than 2 or 3dB quieter
than the exhaust vent leakage.

As to whether I embark on such a major DIY project (and Gawd knows, I
could do with the exercise!) is another thing altogether. I might be more
inclined if I had more than a vague notion that my plan was theoretically
sound (forgive the pun). It looks like a little more research is needed
before I embark on what could turn out to be an unwanted folly.


[1] The pulses of exhaust gas going into the muffler from the exhaust
port approximate to half cycles of twice that frequency so although the
fundamental is half that of the vibration frequency due to the engine
revs, a significant portion of the exhaust energy is carried in the 2nd
harmonic which the muffler attenuates whilst magnifying the fundamental
which probably explains the barely perceptible improvements made by
replacing a generator's weedy muffler with that from a motorcar designed
for a four, or more, cylinder car engine where the 'tuning' factor is
less critical than for a single or twin cylinder 4 stroke engine.

--
Johnny B Good