edwards air pump operating speeds

[malcolmd]

I am building a Leak Compound the 3"+5" by 3" one...

I am told that the pumps driven via a lever off the cross head become too noisy at full speed (~500rpm) and do not operate well...

So I was planning to drive the pumps via a worm reduction gearbox....

However, I read somewhere that Edwards Air Pumps only operates properly at quite high speed (presumably some adiabetic or iso-thermal cleverness)...

Is this true?
If it is, how slow is too slow?
Should I drive the airpump from the cross head and just move the feed pumps to the worm-geared calmness?

any (practical answers) would be much appreciated...

Malcolm

[barts]

I am building a Leak Compound the 3"+5" by 3" one...

I am told that the pumps driven via a lever off the cross head become too noisy at full speed (~500rpm) and do not operate well...

So I was planning to drive the pumps via a worm reduction gearbox....

However, I read somewhere that Edwards Air Pumps only operates properly at quite high speed (presumably some adiabetic or iso-thermal cleverness)...

Is this true?
If it is, how slow is too slow?
Should I drive the airpump from the cross head and just move the feed pumps to the worm-geared calmness?

any (practical answers) would be much appreciated...

Malcolm

I've never run one, but the design relies on a splash effect to help fill the chamber. Prototypical practice of 250 rpm was typical; a worm gear is likely too much, but 2:1 reduction w/ a spur gear driver sounds about right.

- Bart

[barts]

Ok; I had a little more time to look into this. From

MODERN MECHANICAL ENGINEERING
A PRACTICAL WRITTEN TREATISE BY SPECIALISTS
EDITED BY
A. H. GIBSON, M.lnst.C.E., D.Sc. M.I.Mech.E.
ALAN E. L. CHORLTON C.B.E., M.lnst.C.E., M.I.Mech.E., M.I.E.E.

page 239.

https://archive.org/details/ModernMecha ... eeringVolV

The Edwards Air-pump.

This type of pump is commonly run at a high speed. The bucket, as
shown in fig. 16, is without the valves, and as it descends vacuum
formed above the bucket is as perfect as the temperature will
allow. When nearing the bottom of the stroke the conical end of the
bucket strikes the water and gradually sets it in motion round the
curved end of the pump barrel, the velocity acquired being sufficient
to impel the water through the ports in the barrel to the top side of
the bucket. Also, as soon as these ports are uncovered on the
down-stroke, air from the condenser rushes into the barrel, because
the vacuum there is greater than that in the condenser just before the
ports are opened. Before the water has time to fall down and run back
through the ports the bucket has returned and re-covered the ports,
after which the air is compressed and discharged through the head
valves, followed by the water. The clearance space the barrel in Fig.
17.-Diagrams Air-pump from Edwaras is practically filled with water
the topof the stroke, on thereturndownstroke is very at and there
little re-expansion. Fig.17 illustrates the character of the
indicator diagrams obtained from the top end of a dry Edwards air-pump
when (a), there only a moderate was leakage air, and (b) when the
air-leakage was abnormal. The outside views in fig. 18 refer to an
Edwards triple air-pump having three cranks set at 120°and driven by a
directly-connected electric motor. Usually,however,the motor drive is
by a pinion and gear-wheel, that the motor can so run at a high
speed. The speed of rotation of the Edwards air-pump connectedto a
surface condenser commonly high as 250 r.p.m.,but if is as working as
a wet air-pumpattached a jet condenser, speed to the would
Fig. 18.-Edwards Triple Air-pump perhaps not be more than half this
value, becauseof the much greater volume of water to be discharged.

Found by googling "Edwards air pump rpm".

- Bart

[malcolmd]

Bart, that was really helpful and echos the comments here
http://www.thesteamboatingforum.net/tec ... pumps.html (saying they don't work well at low speed) - I just realized that that link is here !!!

I will have to think a little further, as I was gearing up the counter-shaft to run an alternator, and the worm box was driven from this to get the feed pump speed down...
ah well....

Ma

[malcolmd]

Does the team think that an Edwards pump will work where the condenser is below the inlet height?

All the diagrams I see are for surface condensers (and the leak was designed with one) which "drain" into the pump. but I am planning to use a keel condenser, so the pump will have to "suck" to prime itself - and I am not sure if this is an approved approach - any thoughts?
M

[barts]

If the pump is capable of pulling a reasonable vacuum, it will certainly lift the water out of the condenser a few inches. Remember that 28" of vacuum is something in excess of 20 feet of water lift. If the pump doesn't pull a vacuum because it is dry, the exhaust back pressure from the engine will force the condenser contents right into the pump .

- Bart

[malcolmd]

Yes Bart, that was my thinking too - (but then I have learnt over the years, that it's often a good idea to check that with another!!)


thanks....

[malcolmd]

However, mulling this over I am left with the following question/confusion...

According to the reference I cited above "water at 29"Hg will flash into steam at 25C..." the temperature we are hoping to return the feed water at will be above 25C, so that seems to suggest that we are never pumping water at all but water vapour, as it will not have condensed at all. Or is some kind of equilibrium reached with a 50/50 steam/water mix?

However, one can imagine that the keel condenser will be below 25C, so will we not wind up with a condenser full of cold water, while we bubble the steam/water vapour thru it....?
But if that were true it would seem to indicate that in the end we will wind up with a condenser full of really quite cold water which finally gets "pushed" into the pump, and the only reason we don't flood the engine is that the pump is below the exhaust port....

Bah! I think this steam stuff is too confusing - I guess I'll just build it and see what happens!

[steamboatjack]

Malcolm,
The Edwards air pump was designed for surface condensing ships and as such is situated at the lowest point of the system so than the condensate runs into the pump by gravity, very large engines running at 70-100 rpm have no problems with this.
I know there are Edwards pumps fitted to boats where the pump is not the lowest part, i.e. keel condensers and they still work ok? but it could be the position is causing the condenser to alternatively fill & empty?? thus causing noise, just a thought. looking at Mr Leak's drawings I would say it should be ok as its a standard Edwards design but make sure the bucket is a really good fit in the liner and that it is set up with the minimum bottom clearance that you dare, this should of course coincide with the full port opening position. Should you use a different drive system than rocking levers (these were rarely used on launch engines in the golden age) I would suggest you look at alternative pump designs, particularly the so called Dave Fogg pump, a uniflow design, or a common air pump but fitted with Cross's patent floating bucket (see funnel159 p72).
I am in the process (among many things) of re-designing the Leak engine although this has taken a back seat at the moment as the casting providers seem to have a supply problem.
Regards
Jack

[barts]

David Fogg's design is patterned after the US Navy air pump; as it is double acting the vacuum is likely to be steadier.... it also seems to run at speed pretty well which can save some complications. From what I've read, orientation is apparently not important either which gives the designer a break.

Since our boat plans include a box keel for both prop protection and 'sitting on the hard', I want to place the air pump underneath and forward of the engine. This saves space and places the pump beneath the keel condenser, which I plan to run at the top of the box keel.... Of course, servicing the pump is trickier, but if the support structure is easy to unbolt from the engine and is plumbed with flexible collapse-proof lines it should work out just fine.

- Bart