I've spent most of the last few days in the garden making the most of the dryish weather but I've managed to find some time in the evenings for injectors. I've had a play with one of the 8oz ones with not a lot of improvement but this one of the two I made did not work very well to begin with. By tweaking with the steam cone I managed to get it to pump and lift ok but it still will not run dry and drips badly no matter what I do.
As way of a change I made a new set of cones for the commercial injector that my brother bought to convert it into an 8oz one instead of the 20oz that it was. These were made to the same dimensions as mine as regards cone sizes and tapers but with the outer diameter increased to 3/16" to suit the bore of the existing body and the length increased to cater for the longer body. The injector worked quite well first time (a vast improvement on the original cones!) but again, it would not run dry despite tweaking. I abandoned it for the time being while I did some more gardening as a bit of light relief but had another go tonight. I had another look at Derek Brown's articles and drawings of the cones and noticed that the angle of his delivery cone for the 8oz was 6°. On my drawings I had got it at 9° and had made my cones to my drawing. I had assumed that only the 4oz delivery cone was 6° and the 8oz and 12oz were both 9. Wrong! The drawings in ME are actually very difficult to read and some of the dimensions are very blurred, including some of the cone angles. Some of the other articles suggest that the cone angles are not that critical but just in case I made a new delivery cone with a 6° angle. This was a big improvement and by reducing the gap between the delivery cone and combining cone it now runs virtually dry with only a very slow drip from the overflow. This can be stopped by careful adjustment of the water valve. Actually, like the 4oz, it can be made to run dry by turning the water off and back on quickly. Measuring the output gives a rate of 8oz/minute - spot on!
Obviously the problem was due to the incorrect angle and this could well be the problem with my two 8oz injectors. I'll make a new delivery cone for the one that I've been testing and try again. The other is still on Helen but that too has the same fault.
I have done very little on the injector front for months now, being pre-occupied with painting Helen etc.
The injectors I fitted to Helen continue to be very temperimental, so much so that I have tended to use the handpump and axle pumps and forget the injectors.
The other day though, a friend at the club lent me a new Gordon Chiverton 12oz injector for me to try so I gave it a go. I tried it on the test rig and it worked extremely well, starting at 90psi and running dry until the pressure fell to about 60psi. It carried on working down to 30psi but with water coming from the overflow. By reducing the water inlet it could be made to run dry again.
I then fitted it to Helen but it did not work anywhere near as well. It would pump with no problems but I couldn't get it to run dry which points to a problem with the loco pipework etc rather than the injector. After a bit of thought I suddenly thought that there may be a problem with the clack valves on the boiler. Maybe there was insufficient lift on the ball and the output from the injector was being restricted, hence some of the output was coming from the overflow. I took the top off the clacks and measured the lift of the balls and noticed it was only about 1/32". To increase the lift I recessed the inside of the clack tops and tried the injectors again. The Chiverton injector worked much better and I could actually get it to run dry with all the water going into the boiler. The increased lift made little or no difference on the homemade 8oz injector but obviously the clacks need more lift for injectors than for pumps.
The Chiverton injector puts water in very fast which is useful but the steam consumption is high, dropping the boiler pressure rapidly. 12oz is much to big an injector for a 2½" gauge boiler.
Once Helen is completely finished, I am hoping to get back to the injector saga and carry out some more experiments.