5" gauge Britannia (a builder's diary)

[springcrocus]

With the two sections brought back together, the next job was to make the rearmost part of the foundation ring and bring the backplate to shape. The rear angle of my inner firebox is not quite correct and I need to use a 1/2" thick piece of material instead of 3/8" to fill the gap. I've made a temporary filler using two pieces of material but I shall replace this with a single piece at the appropriate time.



The holes for the stays were drilled next. The drawing calls for phoshor bronze stays, threaded 2BA and nutted in the firebox which is the old-fashioned way, using comsol to seal the threads. I've chosen to use copper rivets silver-soldered into place as per modern practice. After marking out grids at the given 7/8" spacings, all the 3/16" diameter holes were drilled with a pistol drill.



One of the advantages of drilling all the holes now is, although it's unlikely at this diameter, I can strip everything down and recover from a broken drill scenario. It's also a lot easier to deburr the backs of the holes using a ball-shaped burr in the dremel. We will solder the stays first because it's impossible to fit solder rings to the stays in the water space once the foundation ring sections are fitted.



I also needed to get the front tube plate to be a good fit and this was the perfect opportunity as it's the last time the two sub-assemblies will be separate. The front had gone a little out of round and was now gently teased back to the correct shape using the front plate as a former, then the flange of the front plate was eased outwards until a decent fit all round was obtained. I've already soldered in the bush for the regulator block and steam collector pipe using 38% silver solder.



Although the copper rivets will be used for all other stays, the front throatplate gap increases as it rises and the rivets are too short. Some 3/16" copper rod was turned into stays by parting to length and centre-popping to create a shoulder to stop the rivets falling through. To assist assembly, a decent chamfer was filed on the front.



Meanwhile, my colleague made a load of solder rings to fit over the stays. 1mm silver solder was wound onto a 4.4mm drill like a spring to create about a dozen rings per length.



A view from the back.



There are 180 or so required so this is just short of half-way.



And then we...

Good heavens, look at the time!

Steve

[ettingtonliam]

Umm - I thought it was usual practice to fit the stay rivets with the heads on the inside of the firebox, and the legs on the outside? I might be wrong, and doubtless you have your reasons.

[springcrocus]

Oct 5, 2019 4:03:38 GMT ettingtonliam said:

Umm - I thought it was usual practice to fit the stay rivets with the heads on the inside of the firebox, and the legs on the outside? I might be wrong, and doubtless you have your reasons.

Oops! Not a good choice of photograph. They are being used as placeholders only, but they are not long enough for soldering. The heads of rivets will, indeed, be on the inside.

Regards, Steve

[runner42]

Hi Steve,

great work I was suitably impressed by your use of the dummy smokebox tubeplate to re-align the tubes and flues, I was hoping that you would show the setup for drilling the holes in the outer wrapper and particularly the ones for the throatplate, which are hampered by the length of barrel.

Brian

[springcrocus]

Brian - no setup, all holes were drilled freehand using a 4.2mm pilot followed by a 4.7mm sizing drill for the rivets (they are undersize) and 4.8mm for the rod stays.



Fitting the stays to the front throat plate took quite a time. Starting with the two lowest ones, flux was wiped into both front and back holes. A solder ring was held in a pair of pincers that I had specially modified for the job, lowered into the water space and a stay pushed through from the outside, passing through the solder ring and into the inside of the firebox.



Another solder ring was placed over the outside of the rivet and more flux was then added both inside and outside. This whole routine was repeated until all the front stays were in place.



The boiler was carried through to the hearth and blocks built up around it. We've decided to keep the boiler in the inverted position with a block under the front giving a slight backwards tilt. As can be seen, the blowdown bush has been inserted and all of the existing solder joints in the vicinity fluxed over.



Blocks have been placed over the barrel to keep the heat in and in the combustion chamber to protect the tube ends from excessive heat. My colleague will play the normal burner over the outside and down into the water space, at lower than normal power and keeping the flame moving. I will be using the cyclone burner and heating the inside front of the firebox, providing the most power. This will pull the solder through from the outside.



Here we can see that the solder rings in the water space have melted and been drawn towards the maximum heat.



The rings on the outside have melted and been drawn through to the water space. All those pink areas are where flux was added to help protect the existing solder joins



As before, the assembly was left to cool slowly while we had lunch. In fact, this time we left all the blocks in place to prevent any rapid cooling. After lunch, the assembly was rinsed and pickled and we could see more clearly how the solder had flowed. Inside the box we had perfect penetration.



Looking into the water space, it appeared that we had successfully soldered all the rivets



Outside on the throatplate, however, we noticed that the solder on the lower left rivet had not formed a fillet. It looks almost like the ring fell off before completely melting but it may be that the clearance was too great and all the solder has flowed right through.



It will need to be redone and at the next heat-up when doing the side stays, I will try a little intense, local heat with a different nozzle and a new solder ring. It should work fine because the rivet has definitely been tinned, just not enough solder.

And I need to rig up a pulley system. Lugging over 25Kgs (plus liquid) in and out of the wash tank is no mean feat.

Steve

[springcrocus]

Just a quick update on the side stays. These have now all been soldered in and we chose to do both sides at the same time. There is no point covering this in detail as the setup was the same as last week, the only minor difference being that we were able to get heat into the waterspace more effectively from the back. A couple of pictures taken after it was all cleaned up.



They are not shown in the pictures, but a couple of copper blocks were clamped at the top corners in place of the foundation ring to ensure the outer corners of the wrapper stayed in the correct place. Without those, any movement of the wrapper would have meant that the backhead wouldn't fit and it would be extremely difficult to try and correct matters once the stays were holding everything firmly in place.



My colleague and I will get together again in two weeks time when we will fix the front tube plate. Meanwhile, I'm going to get some machining done (if I can remember how).

Steve

[springcrocus]

I have made the safety valves as per drawing but I will probably make some Gordon Smith valves as well, based on the information that Jim kindly sent me a couple of years back.

I turned the outside form first, including the diameter for the hexagon section, then drilled the 3/16" diameter hole through. To make the ball-seat and tunnel I mounted an 8mm end mill in one of the toolpost carriers, centred it off and plunged to depth. I find this easier and more accurate than messing around with cutters in the tailstock.



The top part was then bored to size and depth, lightly chamfered and tapped 1/2" x 40 tpi. Then it was parted off to length and the second one made.



Next I set up a button die and tapped a one-inch length section of 1/2" dia brass bar to be used for the adjusters. I also made a mandrel to screw the valves onto for milling the hex and for making the external threads on the valves. The tapped valves were used as the gauge.



The hex was milled using the rotary table to index round, it's only a spanner flat and half a degree or so out matters not at all.



The mandrel was then set up in the lathe and the external 1/2" x 40 thread cut. I just run the lathe at it's lowest speed, hold the die holder on it's guide, start cutting and wait till it hits the front and let go. I've put a root-diameter undercut at the back of the thread to assist seating of the valve body



To make the adjusters, I stood the threaded rod upright in a chuck on the mill and drilled all the holes using co-ordinate drilling. I made the 1/16" dia holes about 3/4" deep.



The rod was then returned to the lathe and the two adjusters parted off. The bit left over was turned down to below root diameter and three panel pins loctited into the holes. This will become the adjusting spanner, I just have to drill a small cross-hole.



I still have the plunger rod to make, then find a suitable spring and ball.

Steve

[92220]

That's a lovely bit of boiler making Steve! Keep up the good work and the photos!

Bob.

[timo]

Steve,

Always an education - nice machining.

Tim

[springcrocus]

One of my concerns regarding the fixing of the front tube plate was how to ensure getting the tubes hot enough for the solder to wet them. I decided to use the small cyclone burner because it was able to fit through the dome bush and heat the tube plate from the underside. The top burner is the one I usually use.



The boiler was stood vertically on it's firebox and a hearth built all around to keep the heat in. The firebox was kept open to allow an occasional blast of heat to help keep the tubes hot and the top three or four inches of the barrel kept clear as this is where most of the heat would be applied.



Flux was pasted around all the tube ends and in the front plate to barrel joint, solder rings placed over all the tubes and bent rods put in the gully. The regulator/steam collector bush, which had been fixed in earlier with higher-temp solder, was blocked to try to keep the heat in.





While I kept the cyclone burner going through the dome bush, my colleague concentrated heat around the outside of the barrel. Eventually we reached the point where all the solder had melted so we placed another celcon block and on top and covered the boiler with a fire blanket to encourage slow cooling.

After cleaning up it became obvious that we had not got enough heat into the tubes because a dozen or so of them had no solder on them, most of it had flowed onto the tube plate. We made some more solder rings and set up the boiler as before, along with another load of solder in the gulley. This time I decided to use the big cyclone burner instead to get a lot more heat inside the barrel, the main disadvantage being that it wouldn't fit through the hole and had to be rested on the lip instead. This picture was taken later but shows the difference in nozzle sizes.



We also removed the bung in the regulator bush to see if this would allow a better throughput of heat. This time, it worked a treat and these last two pictures show a perfect melt.



The small mark at eleven o'clock on the outer ring is a shadow from a slight dip in the solder, not a blowhole.



I'm well pleased with our progress and with the end result. It's slow, for sure, but we're learning all the time and not messing it up.

Steve

[timo]

Steve,

Nice going.

Tim

[don9f]

Hi Steve, yes well done. I know from my own experience recently with the Jinty boiler how frustrating it can be when things don’t quite go to plan. However it is also most satisfying when having understood the problem and taken appropriate action, it all turns out well in the end.

Cheers Don

[springcrocus]

The horizontal stays at the top of the firebox are quite a dominant feature and some thought was given on fitting them accurately. The first challenge was to get the drilled and tapped holes square to each other taking into consideration the taper of the firebox. A reminder of how they are supposed to fit.



The boiler was set up with the firebox sitting on 3/8" packers and the barrel supported at the far end. With the crown level, the centreline of the holes in the girder stays was picked up with a gauge and scribed onto the outer wrapper.



The positions of the stays were then centre-popped onto the lines on each side. Over on the mill, I scribed the same height line onto an angle plate and drilled and reamed a 1/2" hole to take some bushes.



Next, I made up three drill bushes from 5/8" mild steel, turned to 1/2" diameter up to a shoulder and parted off. These were turned round and drilled 9/32", 11/32" and 3/8" respectively for a pilot drill, sizing drill and 3/8" x 40 tap.



On a spare piece of kitchen worktop, I screwed a pair of steel angles to make parallel rails and packed the boiler up as before, squaring it to the rails and clamping firmly at the rear.



Using a 3/8" dia point that I quickly turned up, the first centre-pop was located and the angle plate clamped to the rail.





Then the smaller bush was inserted and the pilot hole drilled using a pistol drill, the bush changed and drilled at tapping size, finally changing to the largest bush and tapping by hand. This was repeated for all twelve holes. After the first hole, I also drilled and tapped an M5 hole in the top of the angle plate for a locking screw to prevent bush rotation.



Finally, the stay were made from 1/4" dia copper rod, each one individually sized to the width of the firebox at each point and threaded 1/4" x 40 tpi. Here they are assembled into position. Only eleven stay-ends are on view in the finished model, one being hidden under the cleading.



This setup is quite simplistic and not super accurate but it's as good as it needs to be. The threaded holes are square to their opposite number and the stays assembled easily without causing any distortion. These will be soldered into position at the next heat-up.



The stays are a little forward of where they should be because of my earlier mistake with the combustion chamber and inner firebox but by extending the cleading forward by half the difference I'm hoping it won't notice too much.

Steve

[David]

Oct 29, 2019 19:19:33 GMT springcrocus said:

This setup is quite simplistic and not super accurate

Are you joking? Given you're working with a huge lump of soft copper getting repeatedly heated I'd say it's way more than accurate enough. But then your boiler doesn't have banana tubes and too little crownspace like mine does so maybe you're onto something ;)

It still seems way above and beyond anything I've heard of down here when it comes to boiler work.

I think you've probably built that design of boiler as well as it's possible to do, given the various design features that make it more difficult than necessary.

[springcrocus]

David: I was really referring to the kitchen worktop, angle irons, wood screws and pistol drill. Not items usually associated with precision engineering. But if they do the job....

I've now soldered the gauge glass bushes to the backhead with high temperature silver solder using the fixture I made earlier. I also soldered the lower clack bushes...



... or so I thought.



A closer inspection of the lower clack bushes revealed that the solder had failed to wet them although it had flowed all around the back plate. A gentle tap with the hammer and they promptly fell out.

Undeterred, I cleaned up the bushes on the lathe, cleaned out the holes in the backhead, fluxed up and had another go. Same result! I don't know what sort of bronze I have made them from but the rest of the material has been suitably marked up as rubbish and some fresh ones made from a new piece of phosphor bronze.



Today, the new bushes were soldered in and this time all is well.



The three anchor blocks for the floor support brackets were also made and soldered at the same time, still using the high-temp solder.



This is now ready for the final stage of the build. Getting close now.

Steve

[JonL]

That fixture for the water gauge bushes is a great idea.

[delaplume]

Quote}--- "I don't know what sort of bronze I have made them from but the rest of the material has been suitably marked up as rubbish and some fresh ones made from a new piece of phosphor bronze."

Aluminium Bronze ??.......Just a guess on my part...I have never knowingly tried to Silver solder the stuff as such but for some reason your flux certainly didn't have any effect....When machined did they have a sort of Golden look to them ??

It would be interesting to find out just what it is....Have you any left over that you could send to one of our "Experts" for analysis ??

[springcrocus]

Alan: I'm pretty sure it's ali-bronze, I had some in the factory for a regular job twenty-odd years ago and this will be an offcut of it.

Our Friday session in the workshop saw the girder stays soldered to the crown, along with the bolt heads, and the horizontal stays soldered into position. This was really just a caulking operation but 55% silver solder was used throughout. The girders didn't really need soldering as the screws do the work but we decided on a "belt and braces" approach anyway.



Everybody seems to agree that the longitudinal stays in most boilers are superfluous but, since I'm neither willing nor able to try and prove this with calculations, then the stays get fitted as per drawing. One of the stays in the Britannia boiler is hollow and used as a blower stay but Adam (he with the silver tongue) Cro has persuased me to reconsider using this stay because it adds an unsightly carbunkle to the backhead that is not on the prototype.



The offending lump protrudes here under the gauges. The other stay ends are hidden under the false backhead.



Adam suggested running a pipe down the outside in the rectangular trunking on the nearside of the loco as is done in full size but I've decided to try a compromise which I will probably live to regret.

Having the blower pipe inside the boiler has the advantage of reducing condesation in the pipe and, therefore, no spurt of water from the chimney when the blower is first applied. To try and get the best of both worlds, I have made a blind end cap for the blower stay with a 5/32" hole in the side, the head being 7/32" long which is the space available behind the false backhead. The one on the right is the connector shown on the drawing.



This screws into the backhead, although it could just as easily be straight-turned the same as any other bush, and the 5/32" copper pipe shaped to fit onto the blower valve as normal and soldered into the bush. The pipe (when I get some) will exit from the side of the false backhead and, hopefully, be quite unobtrusive whilst the rest of the controls can now be modelled more like the prototype.



And if it fails to work then I can always follow Adam's suggestion and use an exterior pipe.

Steve

[Cro]

Looks great Steve, like the solution for the blower sorry for the extra work but if it works out you can thank me when I come to see her running!

Keep up the great work.

Adam

[ettingtonliam]

Just curious but in the 2 lines of text above the photo, what was 'the Didcot tragedy'?