GWR 94XX Tank Loco


This belongs to a fellow club member who had it professionally built many years ago. The workmanship is top notch and the platework is absolutely stunning. The chassis appears to be based on LBSC's Pansy design but the topworks are obviously different.

It's not had a run for several years but John the owner has brought it down to the track a few times recently and tried unsuccessfully to get steam up and give it a run. It has however run very well in the past. Several people have had a go with it with the same lack of success. It's possible to get the boiler pressure up to about 50psi with an electric blower but as soon as the loco's own blower is used, the pressure rapidly falls again. This could be the fault of the blower or an air leak in the smokebox etc. As usual, I took pity on him and offered to take the loco home and see if I could find the problem.

The first thing was to connect up the compressor and put some air into the boiler. I pumped it up to about 30psi and was greeted by the sound of air escaping from several sources! The most obvious leak was from the gauge glass which leaked at both the top and bottom fittings. I had helped John fit a new glass recently when we found that the gauge was reading very erratically, making it difficult to tell whether the boiler had any water in it or not! It transpired that the original glass was too long and was blocking the hole through to the boiler in the top fitting. The new glass did leak a bit but I assumed this had been sorted. Obviously not.

It turns out that the nuts for tightening the seals (slices of silicon rubber tubing) were too long and the threads were bottoming before tightening down on the seals. A quick skim in the lathe to shorten them soon sorted that out and the glass was refitted with no leaks.

Next leak was the injector steam valve which was cured by taking out the spindle and recutting the conical end in the lathe.

I noticed that a pool of water was gradually flooding the bench top and this was traced to the blowdown valve leaking. The spindle on this seemed very tight so I decided to unscrew the complete top of the valve (after draining the boiler!). This promptly sheared off at the thread that screws into the valve body! A few choice words followed! Although the body of the blowdown valve seemed to be bronze, the top fitting was made from brass and the threaded part had corroded away due to dezincification. The stainless valve spindle had also virtually seized into the brass fitting.

Fortunately, it proved easy to unscrew the blowdown body from the boiler but I was surprised to see that the valve seemed to screw straight into the boiler backhead without a proper bush?

In order to unscrew the valve I had to remove a short copper drain pipe on the bottom of the valve body and the nipple on that snapped straight off due to the same problem - dezincification of the brass.

Dezincification of a brass pipe nipple

A new top for the valve was quickly turned up and the threads in the body cleaned up as well as the threads on the spindle. The valve was then refitted, as good as new. It also sealed this time!

That seemed to cure all the leaks on the outside of the boiler except for a slight leak from the whistle valve which I wasn't too bothered about.

There was, however, air coming out of the blast pipe which meant the regulator was leaking when shut. I had noticed when John had tried to steam the loco at the club, there seemed to be steam coming from the chimney, even when the blower was turned off. The valves on this loco (as on Pansy) are underneath the cylinders so they fall off the port face when there is no pressure in the steamchest. The leak from the regulator wasn't enough to seat the valves so the air was going straight up the exhaust.

I decided to see if I could sort this leak out as it was quite serious but where was the regulator? Pansy has a dome with the regulator inside that, but the 94XX doesn't. It turns out the regulator is a horrible disc in tube type situated between the superheater header and the bush on the front tubeplate. This meant removing the superheaters to get at it. Hmmm. It turned out to be not that difficult in the end. The steam boss on the header was only held on by 4 screws and the connections to the cylinder were nut and nipple joints. It was a tight squeeze as the wet and dry headers would only just fit through the hole in the smokebox door ring but I got the lot out eventually.

The port face of the regulator just bolts between the boiler bush and the steam pipe boss and the disc has a slot in it to uncover the holes in the port face. The disc is operated by a long stainless steel rod with a square on the end and a spring to keep the disc against the port face. Very simple but, as is often stated, a devil of a job to get to seal properly.

Superheaters with disc type regulator

(Can you spot the obvious that I completely missed somehow?)

The port face and the surface of the disc were a bit scuffed and the disc had stuck onto the square on the end of the operating shaft so the spring was not actually doing it's job. The disc was removed from the shaft and the square hole eased with a needle file to stop it sticking again. The face was then lapped flat by rubbing it on fine carborundum paper on a surface plate. The port face was also lapped flat by the same process after removing the stop pin that limits the rotation of the disc.

The whole lot was then reassembled and refitted to the boiler with a bit of steam oil on things to lubricate it for now. Refitting the regulator and the superheaters was quite fun as several times the disc came off the end of the rod just as I was getting it into final position and it took a few attempts to finally get it all back! I would imagine it was originally fitted to the boiler before the boiler was fitted to the smokebox.

Another air test showed the regulator to be sealing ok now but on opening the regulator there was a lot of air still coming out of the chimney. There was now sufficient pressure to seat the valves but still the air was coming from somewhere. Investigation showed that although there was a little bit of air coming out the blast pipe (probably the valves were dry and so not sealing perfectly), there was a hell of a draft coming from the back of the smokebox. Using a plastic tube as a stethoscope, I tracked this down to what I thought was the steam pipe flange on the regulator. I wasn't entirely convinced that was where it was coming from as it seemed to be coming from one of the superheater flues but decided it had to be the flange. I had used high temperature silicon flange sealer on the joints and I assumed it hadn't sealed properly.

So, out it all came again! Then I immediately saw the problem. There was a 1/16" hole in one of the superheater elements! This is glaringly obvious in the above photos of the regulator but I completely missed it at the time!

Why didn't I see this before!!

The hole is perfectly round and could have almost been drilled. The rest of the elements seem sound so I wonder if this could have been a fault in the tube?

I didn't fancy making a new superheater so decided to try and repair it. I simply cut down a 1/16" diameter copper rivet to leave just a short shank, thoroughly cleaned the area around the hole, and silver soldered it in. It looks sound so should last for a bit yet. If the superheater goes again, I think it will be a new stainless steel replacement!

Hole repaired with rivet silver soldered in

The whole lot was refitted again (with several attempts required, as before!) and this time no leaks! Yippee!

Another problem that had become apparent when I had to remove the superheaters etc. was that the whole chimney assembly was wobbing about on top of the smokebox! The bolts were all very loose and the chimney/petticote pipe was not sealed to the smokebox. Between that and the leaking regulator and superheater, it's no wonder the fire would not burn using the steam blower. It would be impossible to get a decent vacuum in the smokebox.

After finally refitting the regulator and superheaters, the chimney was put back with some silicon gasket goo and the bolts properly tightened up. This was fun as it uses bolts with nuts inside the smokebox and getting the nuts back on at the rear of the chimney involved much fiddling about to try and get the nuts in the right place and then hold them there while the bolts were fitted. It took ages and severely tried my patience!

It seemed a good idea now to see how effective the blower was with all the 'improvements' so I pumped the boiler up to 50psi and opened the blower. There was a good blast up the chimney and the smoke and flames from a lighted taper held near the open firehole door were sucked into the boiler with no problems. This was with the open grate so I figured that there must be a fair vacuum in the firebox. I think all should be well now in the steaming front.

John had mentioned that the handpump was not very effective so I decided to have a look at this. It's mounted in the left hand tank and operated through a big slot in the top and I think this is the one thing that really spoils an otherwise excellent loco. In fact, I think this position of the pump spoils all tank locos although it is easy to put them there. I will always try to fit them into the rear bunker if possible.


I assumed that the pump had been fitted after the tank was completed so it must be possible to get it out again! By taking off the tank stay over the top of the firebox and then removing the piece of thin angle alongside the firebox, you can remove a rectangular section of the tank top at the side of the slot and the pump can be wangled out through the opening. The pump is held onto a platform inside the tank by 4 screws which you can just get to, and the nut on the delivery pipe can be undone with a spanner held at an angle through the slot. Not easy, but not impossible.

The two rear fixing screws were already loose, in fact, it turns out they had sheared off completely so the pump was only held by the front ones. I wonder if the pump had been removed before and the rear screws had snapped off when someone had tried to undo them? Anyway, I got the pump out and it's a fabricated job of 1/2" bore. The ram seems to be some sort of plastic with an O ring seal and had a groove worn on the bottom edge at the valve box end. This had been caused by the end of one of the bolts used during the fabrication process being left sticking up into the bore! The offending bit was removed using a fine file and some carborundum paper wrapped around a piece of bar, and the ram refitted with a new O ring.

Both valves were leaking so the seats were cleaned up and new balls fitted and seated in the usual way.

The problem now was to drill out the broken screws in the pump mounting plate and then retap them for new screws. Fortunately, I found a very long 6BA tap and an equally long tapping size drill that my father got years ago and this job was quickly accomplished.

Refitting the pump took longer than getting it out as I had a job to get the union nut back on that connects the delivery pipe to the pump. It didn't quite line up with the pump thread and took a lot of fiddling to eventually get it on square, especially in such a confined space (cue the gynaecology jokes!).

After refitting the pump, the tanks were part filled with water and the pump seems to work fine. A steam test will prove that, one way or the other.

At this point I decided to try running the chassis on air again so the wheels were lifted off the bench using wooden blocks under the buffer beams and the boiler pumped up to 50psi. The wheels rotated very jerkily but seemed to stick at a certain point and the loco would not run continuously. It ran a bit better in reverse but still not continuously. The motion seemed very stiff and eventually I decided to remove the front cylinder covers, which are very easy to get at, and have a look at the bores. Well, they were covered in something akin to black sticky tar. No wonder the wheels would not go around very easily!

Cylinder covers removed

The 'gunge' came off easily after attacking it with a brush and paraffin and the bores seemed in perfect condition. A smear of oil on them and the wheels went around much easier!

The covers were replaced after annointing with flange sealant and another air test tried. This time the loco ran but still not very well. The movement was still very jerky and it would only run in full gear.

At this point, my brother arrived and noticed that the mechanical lubricator was not going around. It's the usual ratchet type but the travel was not sufficient to drive it one tooth of the ratchet. He then proceeded to waggle the drive linkage back and forth and I said " It shouldn't move backwards and forwards like that!" The lubricator is driven by a rod attached to one of the valve rods and, looking underneath, it was obvious that the valve rod itself had about 5/32" 'slop' in it, so the valve gear on that side of the engine has 5/32" lost motion before the valve moves. No wonder the loco won't run properly! The valve timing will be way out on that side.

A quick investigation showed that the valve gear on that side has developed a lot of play in the pins, die block, etc. and needs quite a bit of work to put it right. It's a pig to get at anything without either removing the boiler or turning the loco upside down or on it's side. I don't want to do anything like that for fear of damaging the paint, or worse, so have started work by supporting the chassis at both ends leaving a gap in the middle so that I can work from underneath. What it really needs is a proper revolving building stand but that's another project for the future when work resumes on Simply Longer.


A closer inspection of the valve gear showed that the rocker arm that drives the valve rod was loose on the shaft, the die block was very loose in the expansion link slot, and the pin that the die block pivots on and connects it to the rocker arm was loose in the arm itself.

A lot of time was then spent dismantling the valve gear which was by no means easy working from underneath but I managed it eventually!

First job was to make a new die block and I decided I may as well replace both of them seeing as the other was also a bit loose . There was play on the pins that the die blocks pivot on so new pins were made as well. I had to saw a slot across the pin that was loose in the rocker arm so that I could use a screwdriver to hold the pin to get the retaining nut off, otherwise it just went round and round!

The original design uses a shouldered pin to retain the die block with a washer to space the die block away from the rocker arm. I changed this by machining a flange on the die block to do the same job and fitted a plain pin which has a larger bearing surface.

Old die blocks and pins

The new die blocks were machined on the rotary table as a pair (using my usual method) and made from bearing grade PEEK.

Blank for die blocks set up on rotary table

Milling to size

After milling, the PEEK blank was cut into two and the blocks trimmed and filed to size. New pivot pins were machined from 3/16" dia. silver steel.

New die blocks fitted to expansion links

New pin with an old one

Next job was the loose rocker arm. The rocker arm that drives the valve is fastened to the shaft with an Allen headed bolt that screws right through the arm and the shaft. At some time the bolt had been over tightened and the end had snapped off in the arm. The subsequent movement had then worn the threads in the shaft and the arm had become loose. The remaining piece of the bolt was still holding onto the shaft and I had a devil of a job to remove the arm from the shaft after the rest of the bolt was unscrewed. It needed a bit of brute force to prise the arm off and I finished up snapping off the end of the shaft, which is fairly weak.

The offending rocker and shaft

It wasn't a big job to drill out the old shaft and fit a new one so this wasn't a problem. I managed to remove the piece of broken bolt from the arm by drilling it through and tapping a small allen key into the drilled hole. This enabled me to unscrew it quite easily. The arm was them fitted temporarily to the shaft and the shaft drilled and tapped for the securing bolt.

Rockers with new shaft

Reading through the construction series for Pansy it appears that valve setting is carried out by altering the position of the loose rocker on the shaft and then pinning it. There is no other form of adjustment so I presume that is how the valves on this loco were originally set. This does not seem very good to me and I intend to alter the valve spindles so that the valves can be adjusted by screwing the valve crosshead up and down the spindle, finally locking it with a nut. For this reason, the new rocker shaft was drilled and tapped with the two arms set at 180 degrees and valve adjustment will be dealt with later.

Previous Page Next Page