Rainhill

stenella · Rainhill

The hour of truth has arrived - will Rocket be able to comply with the instructions at the Rainhill trials to win the prize? Preparations have taken until the last day of STWWW ...

First a little history of this particular locomotive: built in 1979 by Hornby, it needs a real Stephenson to bring it up to standards. However, when carefully prepared, it is a fine runner, and in the original configuration I have got it to haul three coaches.

Some stage 1 modifications were necessary to increase power and comfort when steaming it:

1. The gas tank. The original plastic gas tank, which has to be unscrewed for refilling, sucks majorly. I drilled out the mounting hole for the fitting in the tender to be able to accept the “internet gastank”, which was subsequently hacksawed shorter, and the internal baffles and gas pipe were removed before epoxying it back together.

The yellow barrel on the back of the locomotive was drilled with a 29mm hole saw to be able to fit the new tank into the barrel.

The silicone tubing used to connect the new tank to the burner, while gas-tight, relatively secure and making it easy to uncouple the tender from the loco, is still prone to jumping off either the tank or the copper pipe that leads to the burner, especially if the tender derails and/or uncouples itself going over a bump. This can (and did) lead to accidents. A few large fireballs later, I had made a fitting with a home-made union nut for the 4mm tubing in the original gas pipe covered in steel mesh. I cut the tube right behind the ferrule, removed the nut, annealed the ferrule and ran a drill through it, turned a new M8 union nut and an adapter piece M8-M4. The tank was further modified by drilling out the nipple to 3.3mm and tapping the resulting hole for an M4 bolt. When all this went back together I had a safe and workable gas delivery system. Since the “neck” of the tank is relatively short, part of the ribs on the bottom of the tender had to be ground off in order to be able to fit the nut and adapter at all.

The system can only be screwed together (with difficulty) while the gas tank is on the tender, as it is too large to pass through the hole. Uncoupling the locomotive from the tender is now tedious again – since the tube cannot rotate, literally the entire tender has to be screwed off the gas pipe, but this is also the case in the original design by Hornby.
2. The burner was taken apart, and all threads and o-rings lubricated. This makes the action incredibly more smooth and predictable. Regular cleaning of the burner is necessary, because the butane, no matter how pure, will still leave deposits that gunk up the jet and needle valve. No gas flow because of gunked-up burner/ronson valve in the original tank means a weak flame, no steam, and no movement!
3. Front axle. An eccentric brass sleeve was made to fit over the axle. On the wider side, two holes were drilled about 12 mm apart, through the sleeve and about a mm into the axle, and the part of the hole in the sleeve tapped for M2.5 bolts of suitable length. Another hole was then drilled in the axle on the same line as and between the previous two, with its center 4 mm away from one of the outside holes. The axle was then cut between the single hole and the pair 4mm apart. This allows to bring the track width to an actual 3.5in (89 mm) instead of the 85mm that Hornby uses.

The axle can now easily be converted to 85 mm to run on the Hornby plastic track, of which the points do not tolerate an 89mm gauge, or to 89mm, to run on outside 89mm track without constant derailing. 4mm of washers are added between the frame and wheel spacers when running at 89mm.
4. Rear axle: far less problematic. Remove wheels, add washers. Similar for the coaches. The tender is more of a problem – the wheels fit exactly between the black plastic axle blocks. 2mm was filed away from these blocks.

Removing them from the tender is somewhat problematic. I find the best way to be the application of a steady increasing outwards pressure on one of the wheels until it pops loose. First use a fine screwdriver for undoing the ends of the leaf springs and the lip on the bottom. Once you remove one axle block, the axle can be fed into the resulting hole, and slight outwards pressure on the horn block on the other side will allow the wheel to slide past it. When mounted at 89mm gauge, the wheels foul one of the leaf spring attachment points where it clamps the inside of the tender frame. About half has to be ground away for the wheel to fit, but this does not impair their function, and it is completely invisible when the tender is on the rails.
5. Removal of the reduction gearing for a more prototypical chug and somewhat higher boiler pressure. Also the losses in power associated with the very large brass bushings through the wheels and the gearing are eliminated. The eccentrics steering the valve gear cum wheel bushings were mounted in the wheels, and the part of the eccentric overlapping the wheel drilled through with a tiny drill, into the metal of the wheel. They were then pinned in place. The intermediate gear was simply removed from the frame of the loco. The gear on the axle on the inside of the wheel was kept as a spacer, not least because it has a conveniently drilled hole for a pin present on the eccentric.

This shows the situation before:

6. Cylinders: it is imperative that you open them up and check whether the rubber o-rings that prevent steam loss along the piston rod are still in place. They perish, and are most probably the reason for many Hornby Rockets being so sluggish that they will hardly move themselves. Here the o-ring is NOT present:

7. Piston and connecting rods: Hornby used a sort of plastic clamp on the piston rod which grabs a hole in the connecting rod. When used intensively, this system has the tendency to let go in the middle of a run. Once it starts letting go, there is no more stopping it. The solution for me was to find a nail of suitable diameter, drill two 1mm holes through it to accept locking wire, and drill out the plastic clamp, so this pin can be fitted through the little end clamp and the hole in the connecting rod. The next issue is the plastic clamp actually letting go from the metal piston rod. Cleaning and application of cyanoacrylate glue will get the Rocket through a fair number of steam-ups before it lets go again. For the definitive solution, see below.

8. Rubber o-rings on the oilers instead of the fiber washers. They can now be screwed tight with finger pressure only.
9. Boiler cladding – I stuffed rockwool into the space around the boiler for insulation.

This concludes stage 1 of the modifications. It will yield you a Rocket that will pull 3 coaches in a straight line, and with the gas burner fully open reaches speeds light engine that are too fast to be safe on the plastic Hornby track. Pulling three coaches through a number of Hornby points is still somewhat problematic, because using the points implies a large number of curves, causing a lot of drag on the wheel flanges. In addition, the damn thing, when run hard, runs out of water and/or out of gas about every 3-4 minutes, even with the much larger new gas tank. Topping up the gas with the new tank only involves taking the yellow barrel off the tender to access the filler valve, but to top up water, the hot filler cap has to be laboriously screwed out. These two issues, and sprucing up Rocket to make it more fun to use, made me go to the following stage 2 modifications, which mainly involve a new boiler with some improvements.

Stay tuned ...

stenella · Rainhill 2

The new boiler:

10. Boiler: I used 54mm copper tube of 1.5mm thickness. To this I added six inserts
a) one of 10 mm with 12mm collar, tapped M8, in the position of the filler neck, which will now function as the mounting point for a steam dome.
b) one tapped 8mm with 10mm flange tapped 1/8 40 tpi for a Dream Steam filler valve, right behind the second boiler band. This will do away with burning my fingers to unscrew the filler cap between steaming sessions.
c) One blind bush (8/10mm), drilled with a 2 mm drill from the bottom inside, but not through the top. This is subsequently drilled by 1.6mm (1/16nd) from the side. Explanation follows below.
d) One bush (8/10mm) tapped with M6 x 1m, for a Regner 3 bar (40 psi) safety valve, which is small enough to fit under the cap that is provided on the boiler for this by Hornby. This turned out to be a mistake – the safety valve is actually threaded metric fine, M6 x 0.75mm. An adapter was quickly made, which still keeps the safety valve small enough to fit under the original cover.
e) A bush in the backhead (8/10 mm) threaded on the side of the outside of the boiler with an M5 bottoming tap, halfway through the bushing, leaving a plain 4mm hole for feeding the steam pipe into on the inside. Fitting this bush and the pipe was obviously a horror story – the safety valve bushing is very close to the backhead, protruding into the boiler, and it is nearly impossible to get everything to fit.
f) A bush tapped for 1/4in 40 tpi on the bottom of the boiler, to fit a Bruce engineering blowdown valve.

The boiler endcaps were cut and filed out of the same copper pipe as used for the boiler, cut open and hammered flat. Geometry teaches us that 3 boiler caps can be made of one 54mm long section of 54mm diameter pipe. A 22mm hole was drilled in each endcap for the flue. The flue was made out of 22mm copper pipe, with 7 8mm cross tubes. This was all brazed together.

All the boiler inserts were fitted through the 10mm hole for the largest bushing and brazed in place. The largest steam dome bushing was brazed last, with the collar on the outside. Next time, I should remember to braze in the bushings before I braze the endcaps in … that would make for a more robust boiler.
Brazing was done with two solder lamps. That is barely adequate for a boiler this size. And I had a lot of problems getting everything to look clean. Do not start brazing without 10% sulfuric acid, it makes the clean-up incredibly much easier. If I had to do it again I would also invest in a larger burner.
Then the boiler was decorated.
After not finding a suitably compact valve with any of the usuaI suppliers that also looked right (all you find is globe valves, I considered the Dream Steam regulator, but that only comes with a handwheel now, and still needs 1.5 turns from closed to open), I fashioned a regulator which resembles the prototypical one by buying a set of 5-way metal aquarium air valves, mentioned on this forum before. They are simple “plug with a hole” affairs from ebay.hk, which turn out to be made from 5mm nickeled brass with 3mm inside diameter. The valves can be cut off one by one, the sideways outlet threaded to screw into the M5 bushing at the boiler head, and the through-tube on the other side of the valve cleaned up to accept 3mm copper tube. Reassembly and test-fitting indicated that about 10mm extra space was necessary for everything to clear the fake boiler back plate and the cylinder support across the top back end of the boiler. A hole which would comfortably fit the bushing (10mm) was drilled in the correct spot of the black fake boiler backhead and the insulating disk contained in it. The new regulator was lapped in to reduce leakage. The regulator plug is held in with a round nut on the bottom, which loosens in use, and if the regulator is mounted in the locomotive, it is almost impossible to reach. Four flats were filed on this nut to be able to grab it with pliers, and after tightening it as far as it would go, a drop of superglue was applied to the threads and the bottom of the nut. After the handle fell out, the hole was drilled through at 2.3mm and tapped M3. A steel handle was fashioned which features M3 thread on the end.
The Mamod top-up valve was added to the boiler, the Regner safety valve and conversion bushing, and a steam dome was made from 3/4in hexagon brass bar. This steam dome was originally 25mm high and threaded M8 over the lower 8 mm, right up to the nut. This turned out not to be long enough, giving only one full turn of thread engagement into the boiler when everything was assembled. A new one was made which was 6mm higher. A blank part was left below the nut, since the nut has to sit on top of the external steam dome cover on the yellow boiler cover anyway. The inside of the steam dome was drilled with a 5.5mm drill. The copper steam pipe inside the boiler has an outside diameter of 4mm. Brazing in the last bushing into the boiler over the steam pipe severely burned the pipe, which obviously heats up much faster than the rest of the boiler. A hole burned/molten in the pipe was plugged with a drop of superglue. Not enough heat and sloppy brazing led to the steam pipe being brazed to the inside of the bushing, luckily this could still be carefully cut by judicious use of a dremel, and the threads were cleaned up with a steel center-drilled M8 bolt (to clear the steam pipe). After breaking off the steam dome on final reassembly, everything was drilled back out, and a new bushing tapped for M10 fine was brazed in. Then a new piece was soft-soldered onto the steam pipe, and another steam dome made which was drilled with 6mm on the inside and threaded M10 fine on the outside.
Finally, I purchased a 0-100 psi ½ in pressure gauge from PPS model engineering. It is the smallest one I could find, and uses 1/16th copper pipe for the connection. The pipe was soft-soldered into the mysterious blind bushing with the sideways hole mentioned higher, and routed externally to the bottom of the boiler. I put on the tiny nut and olive, and promptly melted the olive while trying to solder it in place. I fashioned a new olive (smallest thing I’ve ever made on the lathe!) and soft soldered it in place with a soldering iron instead of the torch. Lesson learned here…
When I ran out of water the first time, the soft solder melted and the pressure gauge tube sprang a leak. I cleaned everything up and brazed the copper pipe into the bushing, being very careful not to overheat the tiny copper pipe. Hardest thing I’ve ever brazed – keeping the right balance between a whopping large boiler and a teeny-tiny tube is a thin balancing act.

Everything was then mounted on the locomotive, and the banjo nuts desoldered from the original manifold. 3mm copper pipe was soldered into both sides of the regulator, and bent so as to arrive at the hole at the bottom of the cylinders. It was marked at the exact length it should be cut with the help of a spare manifold I had purchased (depth of the copper tube in the banjo nut is fixed, and obviously the center-to-center distance of the nuts on the spare manifold corresponds to the dimension between the holes in the cylinder they need to connect to). After cutting the pipe, the regulator/manifold was mounted on the boiler through the fake backhead (impossible later, the firebox prevents the regulator from being screwed into the boiler when it is in place), and the boiler thus installed in the locomotive. The banjo bolts were used to secure the banjo nuts to the cylinder, and they were soft-soldered in situ with a 35W soldering iron, to ensure correct angle and position. Everything could then just still be disassembled. It is now an extremely tight fit requiring careful juggling with the manifold and regulator below and around the cylinders and cylinder supports, but it can (just) be done.

The boiler was hydraulically tested up to 90 psi with the help of the Mamod top-up bottle and valve. The safety valve was tested also at this point.
A comparison of the old and the new boiler:

The pressure gauge was removed and the yellow lower boiler cover slipped over the pressure gauge line.

The locomotive was assembled, and a steam test up to 80 psi was performed.

After everything cooled down, the safety valve was installed, and tested under steam. With the burner at full blast and the engine stationary, the boiler pressure does not rise over 50psi.

A hole for the top-up valve was drilled in the top boiler cover and the edges filed smooth (ahem ...*after* I took this picture).

Everything was assembled, at which point it became evident that the top screws securing the front and back plate to the boiler would never fit, mainly because of the clamp-nuts that Hornby employed over the lips of the top boiler cover. There was simply not enough room. A solution was crafted by turning 3mm nuts round, so they would just fit into the holes in the lips of the top boiler cover, and then soft-soldering these in place. 3mm bolts were cut to the correct length and everything was again reassembled.
The little ends of the loco are plastic, and have a tendency to come off the piston rod. In order to remedy this, new blocks were filed from solid brass, the piston rods were threaded M3 so the little end block could screw on, and the pin which was installed earlier through the plastic little end blocks was now transferred to the brass block.

The blow-down valve was installed on the bottom of the boiler.

A water level sight glass is still missing. A sight glass is unfortunately impossible, because there is simply no location where it could be placed. Originally, Rocket has one under the left cylinder, but there are simply no sight glasses that are small enough (capillary action …), and it would make the assembly of the locomotive impossible anyway.
Also fashioned from a piece of scrap brass and wood was a tool for opening and closing the (very hot) regulator.
Since the higher flow rate of the new gas tank provided a mixture that was too rich when the valve was completely open, the air holes in the burner were carefully enlarged with a 4mm drill bit. Be careful – the burner is brass, and sharp drill bits have a tendency to bite into the material and pull themselves along. It is best to grind a different angle on a dedicated 4mm bit for drilling brass, so it won’t grab, and you will be able to control the depth of the drill well. Drilling too far can damage the jet, as this interferes with the diameter of the hole. The burner pipe itself was drilled out to 6mm, and the diffuser retained as original. I have melted parts of the aluminum backhead, but the situation has now stabilized.

Performance was now lackluster. Two more modifications were made to make Rocket a consistent performer. Boiler lagging was added again, as it was omitted when first reassembling the loco.

After careful listening to the burner, it turned out that while it was fine with the loco stationary, the flame would flutter when steam from the cylinders injected into the stack caused backdraft with the loco running.

For this reason, new exhaust pipes were made which are turned up into the stack at the ends. Brass nozzles were made that fit into the end of the exhaust pipes with a 1.5mm hole in them, to serve as a primitive blast pipe. This really turns out to work very well. Cracking open the throttle lightly will leak enough steam through the (leaky) spool valves into the exhaust to really cause a substantial increase in draft.

When the loco is running, the gas valve can be turned all the way open.

Now ... I suppose we have to try all of this, won't we?

MrMamod

I'm afraid i am not a train person but i have been viewing all sections during the STWWW and this is one great post as there looks to me one hell of a lot of thought and work gone into this.

Sandman

This thread is simply amazing.

All the mods done here, are all things that should have been done by the manufacturer.

That is how the Rocket should have been built.

Marvelous stuff.

stenella · Rainhill 3

So this is it then ... despite the very variable weather large crowds have turned up as the fire is stoked, steam is raised, and the regulator is opened for the inaugural trial run to Manchester!

First class coaches only, as the home-built second class coach has broken two wheel bearing blocks due to a derailing incident.

This is the fifth attempt at videoing the new and improved Rocket ... I had it break once, there was no memory card in the camera the second time, the camera batteries ran out the third time, and the fourth time, by the time I had the steam up it was too dark to film!

However, today everything went more or less according to plan.

In the first video, the loco is a little slow, since not everything is thoroughly warmed up yet.

The next run is all around the track with just one short stop to raise some more steam after the incline.

The last video is the celebratory passenger run, which unfortunately was only partially recorded because the camera batteries ran out. I then had to re-record it off the computer screen because the video file was corrupt at the end and I couldn't copy it. The quality is thus obviously not what it could be. After a bit of a difficult start, this time the train completed a non-stop run around the entire track, without running out of anything!

I hope you all enjoyed it, I certainly did. Thanks for everyone's great posts! Special hats off to David, who posted the very first run of his pristine Hornby Rocket!

Cris

Sandman

Awesome videos, but I can't get the first one mate.

stenella

I'm working on it Sandman ...but youtube is being a bit pants atm.

Les

I did enjoy it a lot, you must be very pleased with it now after all your modifications.

stenella

The first video should be fixed now!

Sandman

Excellent.

Video goes fine now.

Great to see the plume of steam as it climbs the gradient.

Swift Fox

Really enjoyed reading through all of this and you have done a great job with all the modifications. It's great to see the fruits of your labour running out on the track, well done

kno3

Very nice upgrade.

Ross

Very nice work! great improvements!

Titan

I am very impressed!! well done!

MrDuck

as good as the Rocket gets and that is already a fine model but you made it an impressive runner Thumbs up!

Steve_S

And that... is how to do it. Very impressive work.... well done!

tmuir

That has to be one of the most heavily modified Hornby Rockets around, and I'm glas someone has finaly managed to make one do some real work.
I do love the look of them though, and despite their short coming would still love to own one.

bessytractor

that really is quite amazing!

Rainhill

Rainhill 2

Rainhill 3