Yesterday Mr. Copeland asked if it were possible to make an 'eductor'
or 'venturi pump', similar to a steam engine feed water injector to
create vacuum, using pressurized air as the motivating fluid.  The
answer is yes; the operating principle is the same.

As any fluid flows down a tapered path, the conical 'inlet section'
of the venturi, its mass flow remains the same along the total length.
But its velocity must increase, to cram that same flow through the
narrower end of the cone.  According to Bernoulli's principle, the
energy needed to accelerate the fluid to this higher velocity is
'robbed' from the fluid's initial pressure energy content.  The result
is, that at the narrowest point of the cone, the pressure is much lower
than at the inlet.  One can poke a hole at that point, and use it as
a vacuum pump.

Durward Center told me that a few player organs did use an eductor in
a side stream from the blower, to make the small amount of vacuum
needed to work the roll reader in the spool box.  However, it was much
more common to furnish an ordinary player-piano box pump for that
purpose.

Many commercial firms, including Schutte & Koerting and Coppus Blower,
make commercial vacuum systems using venturis.  However, these are
expensive, and usually use high pressure (often 75 PSI) air to operate
them.

The practical difficulty for organ service is the very low supply
pressure available from the organ wind.  This means a very large
ejector, using a lot of air.

Assume, jut for talking purposes, that one wanted to supply vacuum to
run an Aeolian 176-note spoolbox.  At a maximum of 16 tracks playing,
with 0.043" (#57) bleeds, about 1.8 cubic feet per minute (CFM) would
be used.  But the vacuum also operates the roll centering logic and
other accessories, so to allow for that, as well as the inevitable
leaks and seepage, assume the need for about 5 CFM of vacuum flow.

If the roll motor also uses vacuum (some do), we are talking an
additional 3 to 5 CFM.

The following web site has a free "Java" that can make rough,
preliminary venturi calculations:
  http://flow.netfirms.com/venturiflow/calculator.htm

Running a calculation for our model;

1. Assume the chest pressure is 4 inches water column (inWC, common
for Aeolian), and one could tap into the blower line upstream of the
regulators, to get a pressure of 5 inWC at max flow: p1 inlet pressure =
406 inWC absolute.

2. A vacuum required in the spoolbox of -8 in: p2 neck pressure =
393 inWC absolute.

3. One has to run at least ten times the required vacuum flow through
the eductor, to avoid the incoming low pressure air entering the hole
in the neck, from diluting the total flow and spoiling the eductor's
behavior. So the final calculated flow rate should end up about 50 CFM,
or 100 CFM if the roll motor is included

4. Change the C (discharge coefficient; essentially 'efficiency') entry
from the theoretical 0.985 to a more practical 0.6.

By trial and error, one determines that a venturi with a 3" diameter
inlet and a 1-1/16" diameter neck would fill the bill, using 53.3 CFM
of motivating air, for the spoolbox only.  With the roll motor, a 3-1/2"
x 1-1/2" eductor, using 107 CFM would be needed.

Such an eductor would probably have to be about three feet long, and
would need to be rather accurately made, to work as predicted.  But
more important, it would use as much wind as two or three additional
ranks.

And such air actuated eductors are very noisy, due to the high
velocities (144 ft./sec at the throat, in our model).  Large Coppus
Blowers, used in industry to purge hazardous gasses from areas being
worked on, are almost intolerably loud.  One would have to locate the
eductor remotely, in the blower room.

I would think it easier, and certainly cheaper at the end of the day,
to get a small suction unit from Player Piano Co.  It could be turned
way down, and located remotely, connected by some cheap 1-1/4" PVC
drainpipe, to work the spoolbox of a player pipe organ.  That's the
way I did it.

Richard Vance