Drawing 15 - Valves, Ejectors and Pipework
(Marked as drg No. 21 at bottom right)
1. Smokebox Ejector    
The drawings offer a somewhat sparse version of an ejector but we have to start somewhere. However, scaling from the drawing the 9/16" width is actually 1/2" which is par for the course with these rubbish drawings. I made a block to the dimensions given from an offcut of 1" dia brass bar just to get the ball rolling. I decided to try and make something a little closer to prototype and printed this picture from a blow-up of a Nigel Fraser Ker picture. I've milled out the basic shape and drilled and tapped holes to suit based purely on the visual aspect of the ejector.
Contrary to the drawing, the pipe flange at the bottom of the ejector is triangular, not square, and was made from some brass hexagon material. After turning on the lathe, the triangular shape was milled and the picture to the left shows how easy it is to make a triangle from hex bar - just index round two flats at a time. This was then returned to the lathe and parted off. The flange that connects the smokebox pipe to the ejector body was made in similar fashion, although the square section was created in the vertical mode instead. After finding the centre, it's simply a case of winding round at the same x-y reading, just plus or minus to suit. The central hole is for a 6BA screw to clamp it to the body. The flange holes are 10BA.
After this, it was a case of nibble a little off here, a little off there and generally keep removing material to leave a similar shape to the photo. There are no dimensions, everything just done visually. Once the various screws were in it started to look more the part. The bit to the side was made next but, not knowing anything about ejectors, I've no idea of what to call it! The front end was turned first, taken to the mill for the flats and flange holes to be machined, then returned to the lathe for parting off. The M6 thread is to screw the next piece of pipe to.
To finish the ejector, I bought an 8mm brass elbow from Screwfix for 2.09 (part no 78141 if anyone else wants to get one), cut the threads off and linished the smokebox shape on one end, flat on the other. A short length of 1/4" bar made the joining pipe, a piece of hex material for the left-hand nut and the crenellated nut was made from some 7/16" dia bar. A 1/16" dia end mill formed the crenells, of which I made eight by indexing round forty five degrees at a time. The base of the ejector was shaped on the linisher and finished using emery cloth on the smokebox to get the radius correct.
The mounting holes were drilled 10BA clear and the smokebox drilled and tapped to suit. All the parts are just press-fitted together and there is also a plug wedged into the other part of the elbow, tapped 6BA to allow a screw to come though from the smokebox to fix it. The various sections of pipe will be cut later, possibly substituting solid bar depending on how easy they bend. Meanwhile, here is the assembly finished, although I will probably spend some idle moments rounding a few more edges and filing away some more material before it finally gets painted.
2. Exhaust Injector    
I've looked around for a casting for the exhaust injector but there only seems to be one on offer, a white-metal casting for about 38 but it looks a bit squashed to me. So I've sculpted one out of some old brass offcuts. The first part to make was the main body and for this I used a lump of brass that my father cast in sand nearly fifty years ago. There were loads of inclusions and it was no good for anything else. The middle section of the injector was turned first, finishing at about half-inch diameter, using a radius form tool to help mimic the actual casting. The big hole is one of the sand inclusions and will be filled with stopper before painting.
Then it was turned round in the chuck, the major diameter turned to just over 1.1/8" and a 3/4" dia spigot turned that will be the pipe flange. Next, it was over to the mill and reduce the sides to about 9/16", leaving the flange 3/32" wide.
Top and bottom were also milled to their respective sizes. None of this is measured beyond rule-accuracy, just machined until it looked about right. The basic shape is starting to appear. The photo I am using throughout is a grey-scale version of one from the Nigel Fraser Ker site, which Nigel has kindly given me permission to use.
A 3/16" end mill was used to clear the underneath area behind the flange. Then the first of the angles was machined on the top, just aligned by eye.
The photo shows a small flat area before the slope of the other angle, so this was machined next. This front slope appears to be a bit steeper.
There is also some sort of pad on the top so this was machined next. The three square pipe flanges were made next. The drawing shows them as all the same size but the one on the right is smaller so I've made two from 1/2" square brass and one from 3/8" square. There is a small locating spigot on each to assist assembly.
Because all these parts are cosmetic only, the flanges were faced to twice the thickness to simulate both top and bottom flanges at once. Then the were drilled for the dummy 10BA nuts and bolts. The next part made was the thinner barrel-section at the front, a piece of 7/16" diameter brass faced flat at the joining point and with a 6BA tapped hole at the front. Into both this and the main body, spot-faced holes were machined to locate the pipe flanges.
A few other drilled and tapped holes were made in the body where bits stick out or are bolted on, the six bolt holes in the main flange drilled and the area behind this flange hacksawed and file out to allow nuts to fit the space. Then the front barrel was silver-soldered to the main body. The flanges have been attacked with a junior hacksaw to simulate the join. The smaller flanges in the picture are not yet fixed, and also on view is a lump of 3/16" brass plate shaped to suit the overflow valve at the front. Once I was happy with the shape, I drilled and tapped a pair of 10BA holes and one 6BA hole for the piston and lever parts.
One intricate little part made was the fork in the centre, made from 1/8" square brass, turned and threaded 10BA, then cross-drilled with a 0.8mm drill to take a 1/32" rivet as a hinge pin. For the size of my kit, these are seriously small sizes. This is the little valve completed. the tiny stirrup on the left is a piece of ten thou shim screed onto a 10BA stud. The lever is 20 swg brass filed to shape.
The next couple of pictures show the valve fitted to the barrel with a 6BA screw. There is a blind tapped hole in the back of the body and a hole through the barrel. Two of the three flanges have also been soft-soldered to the body.
The platform for the exhaust injector was made from 16 swg brass and shaped to suit the angle of the ashpan. A pair of holes were drilled, the idea being to use a pair of 8BA screws with spacers. Also in the picture is a piece of brass, about 3/16" square, that screws to the back of the assembly to simulate the boiler steam inlet. It's a good job I managed to get the two flanges in line. Drilling the two 8BA fixing holes in the bottom of the barrel.
A slim triangle of brass was made to form the web in the top of the injector and the platform had it's cutaway made for the overflow pipe. The small piece of brass angle and the flat offcut bolted at each end are the mounting feet. The little brass rectangle is the overflow flange and the turne piece of 1/4" hex is the exhaust steam valve nut on top of the injector.
This is a view from the back showing how the two parts are screwed on. None of this will be seen once all is assembled. And this is the upper web soldered into place, a bit scruffy but needs filing down somewhat. I haven't made the small front web, that's a detail too far.
With all the screws in place, and ready to fit to the ashpan, it became obvious that the platform was never going to hold the injector in the correct position. The designer had made a complete pig's ear of placement, as seen when comparing the drawing to the photo. The injector sits below the ashpan, not above the baseline as drawn, so a new bracket had to be made. This entailed a bracket folded in the opposite direction with fixing screws above the injector. This is the new platform with an acute-angled bend.
This is the injector temporarily mounted on the ashpan. It needs tweaking slightly but is pretty-much in the correct position. The pipes, once made, will be a light push fit into the recesses except for the exhaust steam pipe which will have it's own flange and be bolted on. On the right, the large pipe flange has been fitted and the overflow pipe soldered into place. A coat of paint makes all the difference.
3. Washdown Valve    
I've also made the dummy washdown valve from more bits of brass offcuts and this one was very straightforward. A piece of 10 swg for the backplate, some 3/8" x 1/2" for the body and some 16 swg brass for the top flanges. A short length of 1/8" dia had some flats milled onto it to make the spindle. The body had a 1/4" dia recess put in with a slot drill and a piece of bent 1/4" dia pipe soldered into it. The thread was formed on 5/16" dia bar with a spigot which was fitted into the pipe and soldered. The rest was a few drilled and tapped holes and some screws.
4. Duplex Valve (used as Blower Stop Valve)  
The duplex valve on the driver's side is not needed and is used instead as the blower valve. The drawing shows a rather box-like construction with only a passing resemblance to the prototype. I have tried to construct something a little more realistic-looking while still making it a functional device. A bronze offcut was used to make the body rather than brass as this will carry live steam. After milling up a basic block it was loaded to the independant 4-jaw chuck for machining of the steam entry point which is offset in both planes. A 3/32" diameter hole was drilled straight through and the front turned to create a stem with a 1/4" x 40 tpi male thread. Over on the mill, the three mounting holes were drilled 3/32" dia for 8BA clearance, the tipped on edge and the holes for the valve and the steam outlet drilled, also 3/32" diameter.
These holes were then opened up to 7/32" diameter and tapped 1/4" x 40 tpi. A 3/16" end mill was run between the holes and round the outer edges to create the raised bosses. The external circular form of the bosses was created with a home-made tool set in the boring head. This is a piece of 12mm dia mild steel drilled and reamed 3/16" to take ground-up broken centre drills. By setting the tool up to the outer size of the boss, it is easy to create the external circle. The inset shows the form tool next to a standard boring tool.
The boring head is positioned over the hole centre, all adjustment being done directly on the boring head. This shows the first one finished. The workpiece was then laid over onto it's side and the DRO centred on the middle of the three mounting holes. I need to bring the valve to the right thickness leaving a raised centre section and need to determine the amount of material to remove. Feeler gauges were used to find the distance from the boss to the packing piece.
The feeler gauges were then placed on top of the boss, the 1/4" diameter end mill rested on top and the "Z" axis readout set to zero. This is the height of the tool for the final pass. The end mill was then run round three sides using the hole co-ordinates as the reference positions. Finally, two blind 6BA holes were tapped on the far side of the block for a pair of dummy pipes to be fitted to. All the rest was handwork with needle files. Once it is painted it should bear a passing resemblance to the full-size valve, even though one of the outlets is missing. It certainly looks better than what was on the drawing.
5. Injector Stop Valves    
There are two injector steam valves on Britannia and I have made a pair from some bronze offcuts that I had. There will be a pair in here somewhere although I have another piece in the scrap bin and will make an extra pair at the same time. Rather than waste this material, I set up each operation with brass bar first, then followed with the bronze. The first thing to make was a tool to form the neck radius of the valve and for this, I chose to use the back end of a 1/2" diameter end mill. The drawing is wrong (again) where it specifies 1/2" radius, it should be 1/4" radius (1/2" diameter).
As can be seen, the tool gets mounted upside down in the toolpost holder, the angle being enough to give a bit of clearance without losing too much of the form. The holder is set high in the toolpost and because I want to finish at 1/4" diameter, I set the tool just a bit less than 1/8" above the centreline of the work.
This is another shot showing the tool in action. It's also a good example of why setting tools exactly on the centreline of the work is a nonsense. I've never used a tool-height setting gauge in my life, always just eyeballed the height and moved it if it wasn't cutting quite right. So here is the fully-formed curve of the valve neck ready to be threaded.
To get the outer form of the globe I used a HSS tool already in the toolbox. The diameter is nearer 7/16" than 1/2" but it's close enough for this. The 1/4" x 40 tpi thread has already been cut with a button die. Before they were parted off at an inch long, a centre dill with a ground-back pilot was used to create the front form for the nipple. Here are the first two, one with a nut and nipple attached.
The next job was to create the recess for the branch of the valve and they were held in the milling vice as shown, it not being worth setting up a fancy holding solution to get around the differing diameters problem. Although it can't be seen, there is a nut screwed to the front to protect the thread from crushing. To turn the opposite end, they were screwed into a tapped mandrel and the procedures repeated. To release them from the mandrel, a piece of emery cloth was wrapped round the valve and pliers used to grip. The recess for the branch came in handy here.
The last job done today was to part off 1/4" lengths of 1/4" diameter brass and silver-solder them into the branch recesses. A turn of 0.5mm silver solder wire was wrapped around each branch and the parts held with some sacrificial clips. The threads were coated in correction fluid for protection. This picture was taken immediately after soldering. And here is one alongside the drawing. These will go in the pickle overnight and I will machine the rest of the valve tomorrow.
Once the globe valves were removed from the pickle and washed, the next job was to machine the spindle branch. To ensure this was running true, I used the drill chuck to align and load them to the independant 4-jaw chuck. It might have been smarter to use 5/16" diameter material and skim it down but it's what it is. With the branch axially aligned to the lathe centreline, they were then clocked for minimal runout.
It was important to be able to control the depth of the drill and not penetrate through the base of the valve so I set up my spare digital caliper in the contraption I cobbled up a few years ago. I made sure that the 3/32" pilot drill didn't exceed 5/8" deep, leaving about 1/16" wall thickness at the bottom. This was followed with a 3.3mm drill that I resharpened as a counterbore, then tapped 5/32" x 40 tpi although I could have used M4 if I had thought about it in time. The 1/4" x 40 tpi outside thread was cut last. Setting the parts to drill the angled steam ports was achieved by holding them, tilted over, in a small machine vice using the counterbore drill to help set it true in the other plane. Not easy to describe, nor to photograph, unfortunately.
This vice was then held in the main milling vice and the hole drilled using a PCB drill. Alignment was done purely by eye; as long as the drill meets the drilled spindle hole then all is well. The opposite end was done by turning the vice over. The draughtsman doesn't show a spindle, he had obviously lost interest by this time. Design your own seems to be the order of the day and I have made mine from 5/32" diameter stainless steel. I have designed my spindle to act onto a 1/8" diameter ball bearing and to be captive in the valve. Length is also an unknown until the pipework is in place and my best guess is about 3.1/4" max. The front of the spindle was reduced to 0.115" diameter by 3/16" long, then threaded 5/32" x 40 tpi.
The main part of the spindle was reduced to 1/8" diameter by holding on the front turned portion and supporting the far end in a hollow live centre. The middle section grew a few thou, as expected, and was abraded out with emery cloth. I left about 3/32" length of thread.The final picture shows the component parts of one valve and the other assembled. Once the final overall length is known, a square will be milled on the end for a handwheel. I also need to make a pair of gland nuts to hold the spindle captive, the one shown is a standard pipe nut with oversize hole. This type of valve is directional, steam flowing right-to-left as drawn, and I have embossed arrows on the valves to ensure correct orientation.
6. Next Item...