Martin Evans 5 inch gauge ‘Firefly’ Injector – issues

[Keith ChartersParticipant @keithcharters25446]

Hi, Second post of the day relating to our club 2-6-2 Prairie Tank to Martin Evans ‘Firefly’ design (in 5 inch gauge). I’m in the process of refurbishing the loco and, as it came to the club without one, I have fitted an injector (actually tried 3!) as per the drawings. The injector is plumbed to the same feed as the hand pump via a combined T and clack valve and that all seems to work as it should. However, I have now tried 3 different injectors (a no 4 and two number 2’s) and they simply will not pick up! The only thing I’ve noticed is that when the injector water valve is opened, the flow is little more than a a steady dripping and, when the steam valve is initially opened, the water flow increases significantly. I’ve disconnected the water feed from the injector itself and with the valve open water flows freely from the bunker tank! It’s been suggested to me that the injector should be on its own dedicated boiler feed with a single clack valve but, as this is how it was designed (and presumably other models of Martin’s design are running perfectly happily with this arrangement!), before I undertake a major re-plumbing of the loco, can anybody suggest what else may be causing the issues? Would this loco be better equipped with no. 3 (18oz) injector).

TIA, Keith.

[Clive Brown 1Participant @clivebrown1]

I’m no injector expert but D A G Brown’s book “Miniature Injectors Inside and Out” is about as good as it gets for making, fitting and fault-finding. Has a fellow club member got a copy?

[Nigel Graham 2Participant @nigelgraham2]

The water should run easily from the overflow so I wonder either the annular gap between the steam-cone and combining-cone entry is too small or (more likely perhaps) the overflow valve is not opening freely.

I am not sure if sharing the clack-valve with the pump would normally make much difference but injectors do prefer as free a delivery path as possible, because in any fluid system each sharp bend or valve adds a small pressure or flow loss.. You could experiment with a temporary pipe from the injector only.

As Clive says, D.A.G. Brown’s book is well worth perusing.

[John PurdyParticipant @johnpurdy78347]

Keith

I have an injector on my 5″ gauge 0-4-0 tank and it feeds via a tee into the hand pump clack like yours.  I have had no problems with that method of feeding it into the boiler at any time (unless of course the water in the tank gets hot!).

John

[noel shelleyParticipant @noelshelley55608]

A good free flow of COLD water is vital. Check that the ball overflow IS stainless, I had this issue with a commercial No 4 new injector that stood for a while. What steam pressure do you have above 60 psi ? Check that there is no blockage or scaling . Is there an air lock in the feed. No 4 should be fine for that loco. Noel.

[Nigel Graham 2Participant @nigelgraham2]

Air-lock in the feed…

Possibly but Keith says the water flows freely when disconnected from the injector. The draw created by the steam ought clear bubbles in time.

.

He has tried no less than three injectors with the same result so the fault is somewhere in the plumbing.

 

It is possible to have a tiny air-leak on the suction side, insufficient to let tank water out but allowing air in when the steam is turned on and acting as an ejector. One sign is the device injecting more or less, but chirping like an excited guinea-pig; if worse, the performance is poor with much spitting and spluttering at the overflow.

On tender locomotives the common culprit is a feed-bag consisting of a simple hose pushed onto a plain stub of pipe. On a tank locomotive it could be from a slightly leaky pipe union.

 

Ironically the steam pressure can be too high for the injector! Mr. Brown deals with this in his book, but I would not like to try the modifications he describes: a bit too much like watch-making for me. The water inlet annulus area, a function of the steam-cone setting, is quite critical; but of three injectors I would expect at least one to be right for the boiler.

I have noticed that injectors can be a bit awkward to start at full working-pressure, but once running will carry on until the pressure is way below the device’s starting minimum.

They can demand very fine control of the steam and water, as they rely on close relationships between steam and water volumes. That on my own club’s loco will not pick up until the water-valve has been turned almost off!

.

I mentioned previously, too many sharp corners in the delivery side. This may be so on the inlet side, too. Each sharp elbow impedes the flow, and although it may be fine when the injector is disconnected it can affect the flow when the steam is trying to drag the water in at high speed.

[Derek DroverParticipant @derekdrover32802]

Something you havent mentioned, the three injectors you’ve tried, are these ones which you know work (have operated them on other loco’s)?

Have the injectors all been recently soaked in pickle?

Note that the smaller the injector, the more problematic they tend to be.. going up in size requires larger diameter pipework for unrestricted flow rates.

In my experience with tank engines, you have to be very careful with the water temperature, as soon as it starts warming up the injectors wont pickup, so have you tried running the injector on an external water source (it’ll tell you if the probelm is on the feed side)?

[noel shelleyParticipant @noelshelley55608]

To see if the fault is the injector or the plumbing set up, try running it on a test boiler or separate steam supply with a COLD water feed. The feed water pipe work should be 3/16″or bigger. The flat end fitting MUST be flat on the injector or air may leak in. On tender engines it is possible for a sliver of the inside of the hose to restrict the flow if the end of the pipe is sharp. If you are near me I have a test boiler set up that uses a No 4 and will happily try your injector for you, or all 3. Noel.

[duncan webster 1Participant @duncanwebster1]

To amplify some previous comments, restriction on the feed line can give odd symptoms. Chap I knew had a feed pump which worked fine when going slowly, but poorly when going fast. He thought it was valve seating problems, but it turned out to be cavitation in the feed line. Slightly bigger pipe with gentle bends rather than elbows cured it. Once a pump or injector has ‘got hold of’ the water a bit of pressure loss is manageable if not desirable

[noel shelleyParticipant @noelshelley55608]

Duncan has a very valid point, cavitation would really spoil injectors performance. I’m in the process of reworking the hand pump and water feeds for the 9F tender to use 1/4″ pipe – even this is to small but has to be practical. Noel.

[Author]

Posts