A little 'Alfred' maintenance, then a nice run today at Urmston.

While running 'Alfred' at Urmston club track we'd noticed the right hand injector was reluctant to pick up. Injectors are magic devices that use steam at boiler pressure to inject more water into the boiler, against that boiler pressure! That sounds impossible, but they use cones to vary the pressure and velocity of the steam and water in the device, and also liberate energy from the steam by condensing it into water, releasing the latent heat. 

Injectors can use live steam (taken from the boiler) or exhaust steam (steam exhausted from the locomotive's cylinders). Full size main line locomotives usually have both types, but each of 'Alfred's injectors use live steam. Exhaust steam injectors are really only appropriate on locomotives that run considerable distances at reasonable speeds, ensuring a reliable supply of exhaust steam.

The steam and water supplies are controlled by driver-operated valves in the cab. Injectors have a steam cone, a combining cone, and a delivery cone as shown in the diagram below:

This is the right hand injector, outside the frames and under 'Alfred's cab. The left hand injector is in the same position on the other side of the locomotive. The steam feed (controlled by a driver-operated valve on the footplate) connects to the right hand end of the injector, the feed water from the tender (also controlled by a driver-operated valve) enters the injector from underneath, and the outlet to the boiler is the pipe connected to the left hand end of the injector. The water overflow pipe is to the left of the feed water input pipe.

When the driver (or fireman on a multi-crew engine) decides the water in the gauge glass is getting a bit low he will operate one of the injectors to put more water into the boiler. This is the most important job on the locomotive and takes precedence over anything else. If the water level is allowed to get too low, the top of the firebox crown inside the boiler is exposed, and with the roaring fire in the firebox beneath it, and no water over it to conduct away the heat, the crown will melt and collapse. With the boiler at a working pressure of 90 pounds per square inch (psi) on 'Alfred', or perhaps 250 psi on a large locomotive that will result in a boiler explosion. The energy released when a boiler full of water at perhaps 180 degrees centigrade instantly boils to steam as the pressure reduces to atmospheric is devastating and often fatal.

Locomotives always have more than one method of filling the boiler when it's at working pressure, usually multiple injectors. Alfred has two injectors, plus a mechanical hand operated feed pump mounted in the tender.

To operate an injector, first the feed water is turned on, and can be seen pouring out of the water overflow pipe on the injector. Next, the steam supply to the injector is turned on, and the water valve turned down until the injector 'picks up' Thus is confirmed by water ceasing to dribble from the overflow, and a satisfying gurgling singing sound as the water is injected into the boiler. When the gauge glass level is back to where it should be, the steam supply is turned off, followed by the water supply. Never the other way round - if you run steam through an injector with no water supply it will get hot. And hot injectors can't condense steam, so won't pick up. If you do it inadvertently, you have to run water through it until it cools down.

On 'Alfred' the right hand injector has always been more reluctant than the left hand one to 'pick up', but latterly was refusing to work at all, simply spitting steam or water from the injector overflow. One cause can be the cones (which, on these miniature injectors, have tiny orifices) can get partly blocked by scale deposited on them in use. The feed water comes through fine gauze filters in the tender so blockage by particles suspended in the water is unlikely. The cure for scaled-up injectors is to remove the unit and 'pickle' it in citric acid or vinegar which will dissolve the scale without harming the metal of the cones. So last week, off came 'Alfred's right hand injector.

The disconnected pipes once the injector was removed. The glass in the foreground isn't a whisky and soda, it's pickling solution with the injector in it (an initial 20 minutes in citric acid solution, then a rinse with water, followed by an overnight bath in vinegar).

Here's the injector getting a water rinse before refitting to 'Alfred'. The delivery cone has come out of one end on the device and can be seen in the water, along with a small collar which fits on the steam inlet end of the injector.

A closer look at the injector connections; steam input on the right, water input in the middle, feed water out to the boiler on the left.

When the water leaves the injector for the boiler, it passes through a non-return valve called a 'clack'. 'Alfred's clack valves comprise a small metal ball held against a seat in the valve by boiler pressure. When the injector forces water into the boiler it unseats the clack ball because the injector output pressure is higher than boiler pressure, allowing the water to feed into the boiler past the ball. Once the injector is turned off, boiler pressure seats the ball again preventing the boiler from blowing back through the clack and the injector overflow, emptying the boiler!

Sometimes a particle of boiler scale will find its way into the clack and prevent the ball from seating. This can often be rectified by turning on the injector again to flush through the clack, and / or banging the body of the clack valve with a wooden tool handle (not a hammer! We don't want to damage the clack!).

The function of the clack can be tested while the injector is off the locomotive and loco is not in steam. By attaching a plastic tube to the pipe that's normally connected to the injector output one can blow through the clack, but on sucking the ball should seat preventing any further sucking. When I tried this on 'Alfred' I couldn't blow through the clack, so decided to have a look at that valve.

'Alfred' has three clack valves; one for each injector and one for the hand feed pump. They are mounted on the back of the boiler and look like small vertical brass cylinders with a hexagon nut on top which is the lid. The lid can be unscrewed to gain access to the ball and its seat. The right hand clack is clearly visible in the picture above, just in front of the fire hole door. The left hand clack can also be seen in the background. The clack for the hand pump is the same as these two but is mounted lower down on the boiler back plate, beneath the footplate floor.

On unscrewing the lid of the right hand clack and shining a torch down it, the ball could be seen. It was stuck to its seat which explained why I couldn't blow through the clack and why the injector would not pick up. I nudged the ball off its seat with a spring hook, took it out and cleaned it and the seat with vinegar, put the lid back, and did the blow and suck test again. This time it worked just fine!

The injector back in place on 'Alfred'

Today I took 'Alfred' to Urmston club again (Malc had a prior engagement at the motorcycle show at the NEC, so it was just me today). Once I had steam pressure up I tried both injectors. The left hand one worked as it normally does, and the right hand one picked up OK. It does require the water to be turned on fully and on turning on the water it dribbles, rather than flows from the injector overflow as the left hand one does, so for some reason there seems still some sort of partial obstruction in the injector, though on turning on the steam the water flows faster before the injector picks up. Might have to do the citric acid thing again or replace the injector... But at least it now picks up and delivers water to the boiler.

I had a great day running 'Alfred' round the inner track for many circuits at a time, only stopping to top up the tender water tank or the coal supply in the tender. I think driving him is becoming 'automatic' now and there is no longer that frantic feeling that there's a lot going on and not enough time to attend to everything.

I also found that 'Alfred' is happy to run 'notched back' to the maximum his indented reverser lever in the cab allows. The reverser has 7 positions; full forward, one notch up in forward, 2 notches up in forward, mid gear, two notches up in reverse, one notch up in reverse, and full reverse. In 'full forward' (or reverse) the cylinders receive steam at boiler pressure for pretty much their entire stroke. As the driver 'notches up' the admission of steam into the cylinders is cut off by the valves before full stroke, cutoff being earlier the higher the 'notch'. This is analogous to changing up through the gears in a car as speed increases. When a locomotive is notched up, the steam expands in the cylinder after inlet valve cutoff, reducing its pressure as it releases energy, so a notched-up locomotive is far more efficient than a loco in full forward gear (or reverse - it works just as well either way) as it is using expansion of steam in the cylinders to do work. However, a Locomotive will not start from a stand in high cutoff any more than a car will happily start from a stand in top gear, hence the 'reverser' lever. And heavy loads or steep climbs may require cutoff to be reduced to prevent the loco stalling.

This explains why a locomotive on pulling away from a standing start 'chuffs' or even 'barks' quite loudly, whereas one at speed 'purrs'.

Mid gear admits no steam to the cylinders - neutral in car terms.

It would be interesting to know what 'Alfred's cut off is in percent terms at 'two notches back from full' (one notch away from mid gear). He certainly ran very well in 'top notch' today, and it noticeably reduced the water and coal consumption from running in full forward. Interestingly, he would not run at such a high cutoff on rollers when I steamed him at home, tending to run rough and oscillate. The momentum of 'real travel' provides the 'flywheel effect' that allows smooth and efficient running at high cutoffs.

Jason Lau's pictures of me preparing 'Alfred' on my own in the rain this morning at Urmston. The weather cleared up by lunchtime.

'Alfred' back in the garage again tonight.

There's still a bit of experimenting to do, especially with firing rates to find the optimum sized fire - not too thick which restricts airflow through it which leads to an inefficient burn and risk of clinkering, and not too thin which means it'll burn through quickly with a risk of going out if you don't catch it in time.

The club boiler inspector completed the boiler test certificate today, so 'Alfred's paperwork is up to date. The boiler hydraulic test (pressurising the boiler to double-working-pressure with water and checking for no leaks) is due every four years, and the steam test once a year (checking the safety valves lift at maximum allowed boiler pressure and will relieve excess pressure even with the blower fully on, and that there are at least two working methods of getting water into the boiler). Both were successfully completed on 'Alfred' in October, just before I bought him, but completion of the certificate itself lagged behind.



.

.