Axle Pump, Donkey Pump, Blower, Blast & Superheated Steam

[jma1009]

I wont go into the theory and practice of fullsize exhaust injectors.

The reason they are not used in miniature is firstly the added complication in making them, plus they need to be left on to work (long runs with the regulator left open, which we dont have in miniature), plus the rather insurmountable problem of separating out oil from the exhaust steam.

I agree that a donkey pump needs keeping hot and the steam pipe to it should be short and direct, or if the donkey pump is mounted on the side of the smokebox the steam feed should go through a hollow longitudinal stay in the boiler in the same way that a blower feed usually does.

Cheers,
Julian

[joanlluch]

Back to topic.

At some point before investigating further the subject I believed that double piston (dual) donkey pumps should be more reliable than single ones. However after looking into the way they work I don't see the point.

I learned that miniature dual pumps just have a complex set of internal passages that feed steam in sequence to both pistons. Such as this: When piston 1 is at the bottom, steam is feed to the top of piston 2, when piston 2 is at the bottom, steam is feed to the bottom of piston 1, when piston 1 is at the top then steam is feed to the bottom of piston 2, finally when piston 2 is at the top steam is feed to the top of piston 1, the sequence repeats. So after all it's just a trick to make pistons move that way, but there's no crankshaft or anything that would mechanically link both pistons.

I would have assumed that two pistons phased at 90 degrees would be used to make sure the pump would never stall, just the same as regular pistons of a locomotive. Maybe full size pumps work like this?

(1) So, in the view of the above, what's the point of a dual steam pump, specially in miniature (other than appearance) ?

(2) Another question. Why are donkey pumps vertically designed?. Is there any compelling reason why a steam pump could not be designed to have horizontal cylinder and water pump?

(3) Finally, I am considering designing my own pump, so I have another question. What's the usual proportion of bore areas between the steam cylinder and the water cylinder?. The pump will have to push water into the boiler at the same pressure as the steam feed. So I understand that, provided there's no loses, the area on the water cylinder needs to be just slightly smaller than the steam cylinder. However, i'm sure there must be some empirical figures based on what works best according to practical experience. Any suggestions?

Thanks in advance for any input.

[steam4ian]

Joan

In duplex (double piston) pumps the pistons are not phased at 90 degree as there is no mechanical interconnection. One piston drives the valve gear for the other. Piston A has to get near the end of its stroke to drive the valve for piston B to the steam in position. Piston B is thus driven away for top or bottom thus reversing the valve on piston A so it can be driven in the reverse direction. Such pumps are fascination to watch.

In a single cylinder pump there is a shuttle valve coupled to a small piston whose valve is driven by lost motion off the piston rod. The valve for the main piston is incorporated in the shuttle. In model form the shuttle and associated steam passages get to be very small compromising reliability.

Single cylinder pumps are generally mounted vertically to conserve real estate, important in say a ship's engine room.

My loco and some others at my club have single cylinder pumps with the valve operated by springs and a trip mechanism moved by a striker coupled to the piston rod. The striker from the piston rod moves along the valve spindle compressing springs at alternate end of the spindle. The spindle is restrained against moving due to the spring pressure by a latch bar. The striker form the piston also hits cams on the latch bar releasing the latch allowing the spindle to move to the alternate position. Such pumps are proving very reliable and mine starts ticking away before the pressure gauge has come off the bottom stop. For reliability I do keep it just ticking way when running the loco; it is supplemented by an axle pump and an injector.

Ian

[uuu]

A late friend, a lifelong marine engineer, made several single cylinder pumps for his boat. He could never quite get them right. They would not start up every time without a whack. Perhaps the dual piston ones are easier to make reliable. Although, as we've heard, you can get the single ones to work.

Wilf

[joanlluch]

Hi Ian, Thanks for your explanation. In the description of your pump there's some wording that's hard for me to translate (not your fault, of course, but my English skills), but I think I get the idea. Maybe when you have a chance, you can post a couple of PHOTOS. :-).

From my own thought I came to the conclusion that for a single cylinder pump to work reliably you must fit some kind of mechanism for the valve that acts as a 'switch' with only two stable positions. I.e.. when the piston nears the end of travel it makes the valve to 'switch' to the alternative position and so on. In digital electronics there's a kind of circuit called 'latch' or 'flip-flop' that keeps a particular state between two possible ones based on short living triggers. This is what I understand the valves on your pumps do.

My idea is to have a classic sliding valve that acts like a 'latch' which is 'switched' by the piston end of travel in each direction. The two stable positions of the valve can be maintained through springing (same as a wall mounted electrical switch). Alternatively, I can think of a trip mechanism based on two opposite faced magnets in a way that the valve wants to stick to one or the other depending on which one is near. This would avoid any actual springs and may result in a reliable enough system. One design constraint on using magnets is that they loose strength even at moderately high temperatures.

There's a question that remains to be unanswered. What's the relative area of the water cylinder respect the steam cylinder? For instance, is it enough to provide just 10% less area in the water cylinder than the steam cylinder, or do we need less to make it work?

[weary]

Joan,

Re: Relative area steam cylinder to water (pump) cylinder.

This is the relevant text from LBSC, The Live Steam Book, 1950, page 68. -->


'On paper' a 5/16 inch steam piston will easily overcome the back-thrust of a 1/4 inch water piston, and such a pump would feed perfectly if made perfectly; but as there is not a wonderful lot of 'spare power' floating around with such small bores, it is not advisable for a novice or anyone whose work is not fully 'up to scratch', to make the pump bores too big. The combined back-thrust and friction may 'stop the clock' in a manner of speaking and she won't pump at all. Anyway I think a 7/32 inch water piston might be chanced with safety, considering the average amateur's skill.


For clarity - The dimensions above refer to cylinder bore or diameter and refer to a duplex pump.

The imperial dimensions reflect the period in which the book was written.
5/16 inch = 7.95mm
1/4 inch = 6.35mm
7/32 inch = 5.55mm


So, LBSC is recommending a 50% (approx.) reduction in piston surface area of water pump relative to steam cylinder piston surface for security of operation. Even the 5/16 inch steam piston to 1/4 inch pump diameter involves an approximate 35% area reduction steam piston to water pump piston.

I have not built a 'donkey pump' myself, so cannot comment. However the above text may give you a starting point, or provoke pertinent contributions from those with direct experience.


Regards & best of luck

Phil

[joanlluch]

Thanks Phil, that's a very useful contribution. Oh, and thanks for having taken the time to put it in mm.

[Cro]

Joan,

I do have drawings for both single and double acting pumps if you would like me to photograph the drawings for you.

I know both to be working designs with the double acting version being used on our large Halton Tank club loco. Having recently been overhauled it works a charm but before that it needed some help to get started where it was worn.

Adam

[joanlluch]

Sept 2, 2016 7:33:59 GMT Cro said:

Joan,

I do have drawings for both single and double acting pumps if you would like me to photograph the drawings for you.

I know both to be working designs with the double acting version being used on our large Halton Tank club loco. Having recently been overhauled it works a charm but before that it needed some help to get started where it was worn.

Adam

Yes please, do you have my email address? I'll send you a PM just in case.

[Nigel Bennett]

I recall LBSC writing about exhaust steam injectors. He said he'd done it - but as Julian remarks earlier, it's getting rid of the oil in the exhaust steam, and not putting it back into the boiler - that is the problem.