Read below...
(Regnirps) wrote in message ...
There is really only one problem with your cancellation idea. To work in all
directions the sounds have to come form the same point in space. If they are
not the same, all you can do is local cancellation or an interference pattern
with some areas silent and some twice as loud.

The sounds would have to come from the same point in space with the
degree of coincidence being compared to the wavelength of the harmonic
you intend to cancel. In the case of the 660lb thrust Argus motor on
the V-1 with its 47Hz pulse rate, that would be a wavelength of about
24 ft, so with 2 of them 1 ft apart, you'd have a worst case of error
of 15 degrees from ideal. Stacked in an over under arrangement like
that rocket Long-EZ at Oshkosh you'd have 0 degree error from ideal in
a circle drawn in the plane of the wings, and that 15 degree error
from ideal directly above and below.

The harmonics would suffer worse due to their shorter wave lengths.

And yes I'd agree with you, there will be harmonics that are
reinforced in some aspects.

Just take two points and start drawing circles around each. The intersections
will form lines of interference.

On .1" graph paper, make the points .1" apart and draw circles 2.4" in
radius.

The PDE's are harder than pulse jets since the detonation wave is supersonic.
You exceed what you might call the ellasticity of the air and I don't know if
linear theory (like Fourier) will apply. There is probably a dramatic
discontinuity and Fourier requires at least piece wise continuity. But I have
not looked at shock physics in a long time. Sombody on RAM would know, like
Marry Shafer.

I've wondered about air being non-linear in certain circumstances.
But out at a distance the air molecules would have to move as normal.

I had the opportunity to play with some signals in Matlab. I captured
the sound from one of the pulse jet videos from the New Zealand cruise
missle guy. One of the problems is that the signal was so loud when
recorded, it clipped/saturated which of course brings the harmonics up
even more. But proceeded anyay. Applied a time delay (roughly 180
degrees) and summed the signals. Compared the spectra from the 2
sounds. The fundamental was supressed about 10dB. I didn't even
halve the signal before adding (like 2 full size engines). Their is a
wierd 1/2 fundamental I'm seeing. It looks like the engine is firing
in pulse pairs from spectra and time domain inspection.

Regards