Drawing 10 - Reversing Gear, Oil Pumps, Saddle
1. Smokebox Saddle    
The drawing suggests using 2mm copper or brass sheet to fabricate the smokebox saddle but I had considered making it from mild steel. However, I couldn't find a local supplier so bought a piece of 14swg brass sheet at the Midlands MEX and used that. First, I cut out the blanks I needed, remembering to cut the width of the main saddle piece to the development length to accomodate the bending of the material. The four sides of the saddle support will be riveted together before soldering the saddle in place. The crossways pieces are cut to fit exactly in the frames and the lengthways pieces are cut to the nominal dimension minus two thicknesses (0.160") and will fit between the two crossways pieces.
Next I marked out the cutaways and then stitch-drilled them, removed the waste and cleaned up the shape will rotary burrs and sanding disks. I also drilled and countersunk two rivet holes at each end for fixing the corner angle pieces to and am using 3/64" countersunk brass rivets for this. Once drilled, I clamped an angle piece flush with the end and drilled through the existing hole, countersunk the opposite side and fixed the rivet, repeating for each hole until both sides were complete. Finally, I clamped the end pieces to the sides, drilled through and riveted these to produce the box shape of the saddle support.
To make the main saddle piece I calculated the spacing between holes, as this is not given on the drawing, and marked them all out before drilling for 10BA clearance. I chose not to remove the centre section before bending because I would have ended up with a rather weird shape. I also turned up a piece of MDF to the size of the smokebox to act as a former since I don't yet have the smokebox material. I don't own bending rolls but we have some up the club so Saturday morning I took the plate and former along for the next operation. I have never used bending rolls so a collegue kindly gave me five minutes tuition and then left me to get on with it.
About ten minutes work saw the job complete and I have deliberately rolled the shape a little tight as it will be no trouble to ease it back to the exact diameter of the smokebox material but not so easy to go back and roll it through again. I'm also aware that it is normal to allow extra material at each end because of the leading and trailing flat spots but I've managed to get quite a good shape without this. After easing the circumference back to fit the template, I marked out the ends of the saddle support and then removed the waste to leave it ready for silver soldering to the main saddle piece. Here it is waitng to be clamped to the saddle prior to joining.
The components of the smokebox saddle needed to be soldered together and I had to decide how to keep the two parts together during the soldering process. I decided that the best way was to drill a hole though the saddle and fix a cross member across the saddle support with a nut and bolt. The main advantage was that it kept the mass down and I wasn't wasting all my gas heating up clamps. At this point, I undid the clamp and gave the brasswork a really good clean-up.Then I mixed up the flux and put some on the edges of the saddle support before locating the two parts together and clamping down again. Following this, I fluxed the rest of the assembly, placed my pieces of solder into the inner corners and then painted the rest of the flux over the solder. Then I assembled my hearth from my fly-ash blocks and placed the assembly inside.
I was approaching this job with much trepidation but took a deep breath and off I went, directing the flame around the sides and underneath the assembly. After a few minutes the flux had turned to a pinkish slag, bubbling away, but the solder wasn't melting and I started to give the occasional blast of heat over the top. The mild steel clamp turned cherry red but the solder still hadn't melted. At this point, I chickened out and turned off the heat. I decided to mix another load of flux, a little more watery this time, and applied this liberally around the area. A moments thought made me realise that if the solder hadn't melted, then I couldn't have got it hot enough. So this time, I took no prisoners and really blasted the flame into the area and wached the whole assembly start to glow. And then, as if by magic, the solder started to melt at the front and slowly crept along the sides and across the back until it was all gone! I kept the flame going for another five seconds or so until I was sure the solder had all melted, then turned off the gas. This photo was taken twenty seconds after removing the heat.
Once it had all cooled down I gave the assembly a good scrub in clean water but had to use an old sciber to get rid of the remaining encrusted flux and had a good hard look to ensure that a decent joint had been made. I'm pleased to say that the solder has flowed evenly throughout and appears to be quite a good joint. It now needs a really good rub-down to improve the appearance, and I have spent ten minutes or more with emery cloth trying to clean it up but I will give it a few minutes in the shot-blast cabinet later to see if that will speed up matters. It did, and here is the final result.
And to finish the job off, I loaded the assembly to my large vice on the mill and removed the central area where the pipes come through. I just need to spot through the frames to mark the boltholes and this part is finished.
2. Lubricators    
There are commercial lubricators available but I thought I would like to have a go at making my own and decided to make the tanks first. I didn't have 18swg brass in stock so chose to use 20swg instead. I started by making a hardwood former to bend the material around, making the former a little larger to suit the thinner material and preserve the outside dimensions. A pair of brass strips were cut 1.7/8" wide by 5" long. I made use of my large milling vice to get the first bend square, positioning the former level with the end of the jaws.
With the first bend accurate, I was able to finish the other sides using the bench vice, finishing with the final side overlaying the first side. Before removing the waste, I removed the former, inserted a junior hacksaw blade in the fabrication and re-inserted the former before clamping up. The waste from the two overlapping sides was then removed using a junior hacksaw, the packing blade ensuring that when the gap was closed the tank would become square.
The larger of the two offcuts was then cleaned up and used as a soldering strap at the back of the tank. A recess was milled into the hardwood former to clear the strap and pushed back into each tank section in turn to enable milling the ends square and to finished length.
A pair of base plates were cut from 14swg brass sheet, squared up and the centre section milled away by about twenty thou to create a location for the tank sections, followed by drilling the mounting holes. Then the bases and tank sections were soldered together, the clamp in the picture being used to ensure the earlier solder didn't fall apart. The bases had taken on a slight bend after soldering and each was reloaded to the mill and skimmed flat. It was only a few thou but enough to cause rocking of the assembly. To finish, the various holes were added to the tank sides ready for the pump components to be fitted.
The body of the lubricator pump should be brass but I have used a couple of offcuts from some bronze plate that I worked on recently. The bodies are 3/8" thick so I was able to machine all round to get them to size. I have added a clearance chamfer at the bottom because of the solder in the tanks. All the drilling, tapping and reaming operations were done on the big mill using co-ordinate positioning because all holes are on the same centre-line so I only had to set the "Y" axis and leave it locked all the time.
To suit my tooling, the 1/8" reamed hole has been made as 3mm and the 5/32" holes as 4mm. Threads are all 40ME or 8BA, as specified, except for one blanking plug that I drilled the wrong size hole so went to 6BA instead. The one awkward hole was the inlet channel which is at twenty-five degrees to the horizontal. For this, I used a slot drill to create a small island to start the drill, and had a piece of sacrificial material in the cross-hole so that the drill didn't wander as it crossed over.
The collection of plugs, unions and other assorted pieces were all made from brass bar. The only part that was not straightforward was the shuttle on the outlet side which required some grooves milled in the sides to allow the oil to flow. After turning to size on the lathe, the three grooves were put in using a dovetail cutter on the mill. A piece of hexagon bar was used to index the three sides as it was quicker than setting up the rotary table. Apart from the piston and the spring, these are all the parts that make up the body of the pump.
Next up was the plunger. To make the plunger and collar I started with some 5mm flat black bar intending to make the whole thing in one go. Two are required and I cut material to make both by splitting after milling. A pair of pilot holes were drilled first followed by milling away the internal form with a 1/8" slot drill. The dimensions on my sketch are the plus and minus co-ordinated from the hole centre for machining to finish size. Moving to a larger cutter for more rigidity, the outside form was produced next leaving a 1/16" thick wall all round..
Finally, the workpiece was held vertically in the vice and the waste cleared away to leave the shape shown. This was then sawn in two and then it was over to the lathe to turn the 3mm diameter using the 4-jaw chuck and a small supporting centre. However, after wrecking one of the pieces because of a poor setup, I decided to take a different approach. The leg was sawn off, a small boss milled and an M3 hole drilled and tapped into the end. Into this was screwed a length of 3mm stainless rod of the appropriate length. Even this was too restrictive and the boss was cleared away just prior to assembly.
I made the eccentric from phosphor bronze although the drawing calls for mild steel. I think it will wear better but may be wrong, easy to replace if neccessary. There are different ways to produce eccentrics including independent four-jaw chucks but I chose to use an offset bush because it is repeatable without clocking up the workpiece. First I turned the main boss, holding the 9/32" dimension to within a thou for a good fit in the bush. I also drilled and reamed the hole 3mm, which I'm using instead of 1/8" dia. The parent material is 5/8" diameter because the workpiece will clean up with about ten thou to spare. Next I made the offset bush from a piece of 3/4" diameter mild steel, turning the O/D to 11/16" dia and leaving a shoulder for butting up against the chuck jaw, then reversing and facing to length. The offset hole was made on the mill by holding in a table-mounted chuck, clocking true then offsetting by the given throw dimension of 3/32". The bush was drilled and reamed 9/32". I also made a small brass plug to fit in the other end of the bush to ensure it didn't distort when clamped up in the lathe chuck.
The eccentrics were then loaded to the chuck and pulled up really tight, and I mean REALLY tight, otherwise the intermittent cut could have moved the workpiece round in the bush and ruin it. All turning here was finished in the one visit, using the compound slide to control lengths. To drill the locking screw hole, I made a simple drill jig from some hex mild steel as I find this easier than trying to set up fiddly components in vee-blocks and vices etc. I've used M3 rather than 8BA because I have a load of M3 grub screws in stock.
The drive shaft was made using 3mm stainless steel rather than 1/8" dia material solely because it was in stock. The various bushes are as per drawing, except for the hole size, of course, but the rear supporting bush has been made to solder in from the outside, rather than the inside, because of the strap at the rear of the tank. The other variation I made was the turned and threaded part of the shaft. My 8BA button die is in very poor condition and not up to producing a decent thread on stainless but I have M2.5 available in both HSS split die and spiral-point HSS tap, my favourite combination of threading tools. As a long-time machinist in industry, I really don't like carbon-steel dies or hand taps and avoid them where possible. The part called "adjusting bush" with its locknut is straightforward turning and threading with tap or die, as appropriate, but I'm not sure what it's meant to adjust - it bolts to the tank and is simply a support. The priming handle has been made from 3/4" dia brass bar. The stem was turned, drilled and tapped first and the handle created on the mill. A filing button was used to guide the shaping near the stem. The picture shows the components loosely assembled in the tank.
To make the ratchet, I started by skimming down to 7/16" dia a short length of 1/2" dia bar which I aquired from a local factory and is tough as old boots. It's BS S154 which is equivalent to EN25. I also parted through part-way so that I didn't have an intermittent cut after the teeth were formed. The bar was then transferred to the mill and the teeth cut with a dovetail cutter. I have chosen to make 36 teeth because that is ten degrees, and two turns of the handle on my rotary table and anyway, I don't have a dividing head, I went round three times before I was happy with the depth and then returned the bar to the lathe drilled the hole and finished parting them off.
I've chosen to make the lubricator operating levers from 16swg brass because I have a box full of brass offcuts to use up. They would just about come out from 1/2" wide strip but the holes would need to be offset quite accurately to make it work so I used some 5/8" approx pieces. Although they are left- and right-hand, they are symetrical so I made them as a stacked pair, starting with the holes. These were co-ordinate drilled using the DRO on the mill.
I then set up one of my universal milling fixtures and drilled a matching pair of holes for clamping the work down with. The two levers were bolted down using 8BA bolts and with small washers underneath for some cutter clearance. It only needed a few thou. I then roughed out the maximum dimensions at the head, tail and spring-mount positions. Once I had those dimensions correct, I then needed to skew the jig over to form the tapered shape. I don't have any slip gauges but an appropriate sized drill was good enough to provide the correct offset.
I didn't want to unbolt the work from the fixture so had to move the jig in the other direction also, just using a combination of smaller parallel and drill to get the same amount of movement on the other side. Here is the final side just finished. The rest was just filing and linishing, a filing button being used at the head end and freehand at the bottom. Four pivots were needed and these were made next on the lathe using 3/16" dia mild steel, simple turning and threading, and then it was onto the pawls. I have made these from 1/16" brass as a temporary item to check "proof of principle" and to create a pattern. They will be remade when I get some ground flat stock.
3. Blast Pipe    
I started the blastpipe by getting a pair of 15mm slow bends from the plumbers merchant and cut away the two outsides, as shown to the left. I then set up the chuck on the mill table in an attempt to machine the edges down to the halfway point. The Hoffman roller is being used as packing but the setup was nowhere near rigid enough.
Instead, I set the parts up in a vee-block and used a 1/2" diameter bar to hold them in place. A more complicated setup but one that held the work well during machining. A 3/16" end mill is being used, taking ten thou per pass to lessen the chance of something going wrong.
The two parts were then fluxed, brought together and wired and finally soldered using silver-bearing soft solder. The two flanges were turned from 1.1/2" diameter brass bar. Moving to the mill, the flats were first milled away followed by drilling the 4BA clearance holes for fixing to the frames
To create the lozenge shape, they were loaded on two pins, as shown to the left, and the final shaping done freehand. The assembly has to fit between the frames with fifteen thou of gasket material on each side so the pipework component was milled to suit. After finding the centre and zero-ing the DRO, cuts were taken at plus and minus until the correct width was obtained. For me, this was about 52 thou on one side and 14 thou on the other and leaves the vertical section dead central between the frames.
The blastpipe nozzle, as drawn, cannot be made as a single component. I have made it in two parts, starting with the lower body, turning the O/D and screwcutting the internal 3/4" x 32tpi mounting thread. This was then reversed in the chuck and the 3/4" diameter section turned, leaving a 1/8" wide flange at the base. I also turned a small mounting spigot on the flange to locate the upper part. This upper ring was made from the same 1.1/2" diameter material and was machined to fit on the body.
The two parts of the body are ready for soldering together, and the externally-threaded collar has been made from 3/4" brass bar and bored to suit the soldered tee-piece. This was then squeezed onto the pipe in the vice - the pipe is not pefectly round - using a clamp to support the ends of the Tee, otherwise the pipework could split. The various parts were then aligned in the frames, with a thirty-thou shim to simulate the packing,and soldered together. The various holes were drilled in the blast pipe collar and this was also soldered together.
4. Blastpipe Spacer    
The blastpipe spacer was made from 1.1/2" diameter brass, facing and turning the 1.1/16" diameter on the first operation, and drilling an 11/16" hole before parting off a little over-length. The billet was then reversed in the chuck, faced to length and the two bores formed. To create the radius to fit the bottom of the smokebox, I made a mandrel from a length of 7/8" diameter mild steel bar to mount the spacer on.
I turned a spigot on one end about 1/4" long and included an M10 tapped hole and by making the bar about 6" long, there was no need to balance the chuck. The M10 bolt was tightened with a box spanner and faced off before mounting in the four-jaw as shown on the left. The head was rested on the outer diameter of the chuck and a 3/32" spacer used at the other end to square the bar to the bedway. The only thing of note was the need to lock the cross-slide before each ten-thou pass to stop the slide being dragged in by the cut. Shown on the right is the finished article, which fits quite well in the base of the smokebox..
5. Blast Nozzle    
The blastpipe nozzle, as drawn, cannot be made as a single component. I have made it in two parts, starting with the lower body, turning the O/D and screwcutting the internal 3/4" x 32tpi mounting thread. Because it goes tight to the back, I kept the half-nut engaged and turned the chuck by hand. A thread this fine only needs five or six passes. All the bores were done at this stage also. This was then reversed in the chuck and the 3/4" diameter section turned, leaving a 1/8" wide flange at the base. I also turned a small mounting spigot on the flange to locate the upper part. This upper ring was made from the same 1.1/2" diameter material and was machined to fit on the body.

The two parts of the body are ready for soldering together but this is the only picture I have of the two parts before joining, the blastpipe and threaded collar have been seen before. Finally, the component was loaded back in the lathe and the top angle produced with the compound slide set round to about twenty degrees. The ring of thirty-thou blower holes were done on the mill using the six-hole solution from the Zeus book. The position of the hole for the steam entry fitting to the blower ring will be marked once I've assembled all the components together. If I do it now, Sod's law determines that the hole will be in the worst position possible.

6. Next Item...