I've reached the stage with Helen where I need to sort out a couple of suitable injectors to fit and I thought it might be a good idea to make a separate section on the subject.
I think most people probably shy away from producing their own injectors as making them is seen as something of a black art and beyond mere mortals! They do require care and precision in manufacture but they are really very simple devices. The difficult part I think is producing the special taper reamers for the cones and assembling the bits in exactly the right positions.
Finished injectors in the larger sizes are readily available (at a price!) but smaller sizes suited to 2½" gauge locos seem hard to come by. The smallest commercial size seems to be about 10oz per minute (½ pint per minute) which is really too big for my requirements, but I have been told that Gordon Chiverton can provide smaller sizes to order. The main problem is that the smaller the injector, the less reliable they seem to be. This is probably the reason why most ME suppliers won't sell anything less than 10oz/min.
I've been studying the excellent series of articles on making injectors by D A G Brown in Model Engineer Volume 184 and intend to base my own on his design which in turn is developed from earlier work by Laurie Lawrence. The articles give designs for 8oz and 4oz injectors which should be about right for Helen, the intention being to fit one of each. The articles also describe how to calculate the water consumption of any loco based on cylinder size and number, wheel diameter, speed, cut off etc. I've written an Excel spreadsheet to easily work all this out and a copy can be found here if anyone wants to use it. A rough guide to the size of injector required is twice the water consumption of the loco i.e. if the water consumption is 5oz/min then fit a 10oz/min injector. Note that the calculations are based on the use of saturated steam so presumably the consumption of a superheated loco will be slightly less.
Using the spreadsheet gives a water consumption for Helen of 2oz/min at 6mph and 35% cut off so in theory the 4oz injector should be able to handle it if it proves to be reliable enough and could probably be left running most of the time. We'll see!
I've made a start by making the jig described in the articles for soldering the various bits of the injector bodies together. I intend to use the smaller of the two bodies described as this will be more in keeping with the size of the loco so the jig has been made to suit this.
Injector body assembly jig
The jig is basically just a frame with a central pin onto which the main body fits and various screws to position and hold the bosses for the water inlet, overflow, and the valve chamber whilst they are all silver soldered together.
Have spent the last two days turning up the injector bodies and the various bosses from 1/4" brass bar. I've actually made enough bits to make three complete bodies in case one doesn't turn out very well! Brass is quite adequate for this purpose and the cones will also be made from this material. Cheap and easy to work!
The most fiddly bit is actually the little cover that connects the outlet in the side of the valve chamber to the body which involves drilling a length of 7/32" dia. brass 3/16" and then scalloping both ends at 45° with a slot drill to fit the body and the valve chamber. The body was drilled for the various holes in the vertical slide using the leadscrews to position the holes accurately. The various parts were then assembled in the jig and silver soldered together using the minimum of solder.
One body assembled in the jig and parts for two more
First body after soldering and cleaning up. Body is 7/8" long
All three bodies seem to have turned out quite well so I'll probably use one for an injector for my brother's Rob Roy. Next job is to ream the bodies to the finished size for the cones and finish the valve seating in the valve chamber.
The valve seats were reamed 1/8" diameter (for 5/32" dia. stainless balls) using a guide which screws into the valve chamber. This ensures that the holes are reamed truly vertical and finish up round instead of slightly elliptical. The balls were seated using the method described in the articles i.e. the ball is pressed onto the seat by screwing in the reaming guide to force it down rather than the usual biff with the hammer. This seems to work quite well and all three valves sealed first time. The bodies were reamed to 5/32" internal diameter by hand as there was very little metal to remove.
Next job was making the reamers and I was not looking forward to this bit! Three tapered reamers are required of 6, 9, and 13 degrees and one to bell out the ends of the cones. These are made from silver steel which is then hardened and tempered. The tapered reamers were cut by setting the topslide at the correct angle and taking fine cuts with a very sharp tool and plenty of cutting fluid. All three went very well despite the very small diameter required at the pointed end (about 0.015" !). After turning the tapers were given a rub with a diamond file to get the surface as smooth as possible. A rough finish on the reamers will leave a rough surface on the inside of the cones and possibly affect performance of the finished injectors (assuming they work at all!)
The final reamer for making the bellmouths on the cones was cut with a tool with the cutting edge ground to 0.1" radius. One end was turned down to a very small diameter for machining the entry to the delivery cones and the other left much larger for machining the large ends of the cones.
The cutting parts of the reamers have to be machined or filed down to very slightly over half the thickness of the taper (as for home made D bits) and the articles suggest not to do this all in one go before hardening and tempering as this can lead to distortion. Instead I roughly filed away about 2/3 of the required amount before hardening and tempering and finished the job afterwards. I was a bit unsure of how to remove the remaining material and was originally going to file it away with a diamond file. I then had the idea of grinding it away by holding the reamers in the toolpost and mounting a grinding wheel in the lathe chuck. This worked very well and by using the cross-slide to feed the reamer against the side of the grinding wheel it was possible to do the job very accurately. The thickness of the taper was measured with a micrometer and left at 0.001" more than the diameter as suggested in the articles. Whether they will cut remains to be seen! The finished reamers look very delicate and easily damaged!
The four finished reamers and two bodies
Well, I've managed to make a set of cones today for the 8oz injector. I didn't take any photos this time but I'll try and get some when I make the next set. The scariest part of the job is drilling the holes as we are talking about No. 78 drills here! I had a problem at first centre drilling the bar to start the drill off as even a No.1 centre drill is really much too big. In the end I ground down the point to leave a very small cutting edge which did the trick and enabled a very small centre to be drilled. The ideal tool for drilling these tiny holes is a sensitive drill attachment for the tailstock but of course I don't have one! The drills are much too small to be gripped in my tailstock chuck so I used a small pin chuck to hold the drill. This was in turn chucked in the tailstock chuck. I started the drill off by feeding in very carefully with the tailstock feed but once it was started I slackened off the tailstock chuck just enough so that the pin chuck was loose enough to move. I then fed the drill in by finger pressure on the pin chuck which gave a lot more 'feel' to the job. This also made it very easy to withdraw the drill at regular intervals to clear the flutes - very important when drilling deep holes with small drills.
The reamers worked very well and gave nice clean tapers in the bores of the cones. The middle combining cone is made in two parts and a set of depthing tools was made to press the two parts into the body to the correct position. The steam cone and delivery cones have a shoulder on the end and they are just pushed in as far as they will go. After fitting the cones everything looks more or less in the right place with the gaps between the cones appearing to be correct.
I intend to build a small boiler for testing items such as injectors but I don't know when that will be so I'll probably test this batch of injectors on the loco itself.
Today I made the rest of the cones for all three bodies and I now have two 8oz injectors and one 4oz. Maybe! The 4oz cones were a little more fiddly as the holes are quite a bit smaller. The delivery cone for the 4oz is slightly larger than it should be as my smallest drill is a number 80 which is about 0.001" too big. Hopefully this won't matter too much. In theory the output will be increased slightly as this is dependant on the diameter of the delivery cone but the other cones have been left the correct size so I'm not sure what the overall effect will be.
As promised, a few photos of the cones being made:
Drilling the cone blank - drill held in pin chuck and then fed in by hand after starting hole using tailstock feed
Opening out hole using taper reamer
Parting off first half of combining cone
Finished two part combining cone with depthing tool
Pressing in first half of combining cone using the depthing tool
Delivery cone with No 78 drill (0.016") for size
Three completed injectors ready for testing
It occured to me today that I have the old boiler for the 2½" Flying Scotsman lying around and there's no reason why I can't use that as a test boiler. I haven't tested it yet but I think it's sound if not very pretty. It just needs a hydraulic test and some fittings making for it. That will be much quicker and easier than making a boiler from scratch.
Return to ME home pageNext Page