Having injured my back just before Christmas I was relegated to 'light duties' in the workshop. Wishing to find something suitable to do, I found the partly made-up body of a GWR type brake valve I had started some seven years ago. The first valve I made had been fitted to 'THE DEAN' and was slightly different. Anyway over the Christmas and New Year break I completed the second valve.
Note... Express engines on the GWR did not have steam brakes. They were therefore fitted with a much simpler type of valve.
The full size valve operates as follows.
The setter valve was very complicated and ensured that the loco steam brake was applied if say the guard destroyed the train pipe vacuum to apply the brakes. The steam brake was controlled by vacuum.
My miniature valve is a simplified version of the fullsize but it does everything the fullsize valve does apart from the setter valve function. Steam is admitted to the rear cover in which is located a rotating circular flat valve. The valve has one hole which in the 'off' position admits steam direct to the vacuum ejector and in the 'on' position admits steam to the steam brake cylinder. A blind slot in the valve connects the steam brake port to atmosphere in the 'off' position and in the 'on' position connects the ejector port to atmosphere to destroy the train pipe vacuum and so apply the brakes.
| A | Air to train pipe to destroy vacuum |
| B | 1. Steam to steam brake |
| " | 2. Steam release from steam brake |
| C | Air entry |
| D | Steam entry from manifold (Rear cover) |
| E | Steam to ejector |
The slots in the front face were milled by hand using a hand-held Dremel engraver fitted with a small dental burr. There are 16 slots instead of the scale 21 because 16 slots are much easier to do freehand! It took about 45 minutes. It is obviously important that the valve doesn't leak. The valve was lapped to the body with Brasso, turning the valve only in the position that it moves in service. The two simulated studs and bolts are 12BA Hex with 10BA threads. I have to be in a good mood to turn up 10BA bolts in steel! As I didn't have a 12BA box spanner, one had to be made by casehardening the end of a piece of hex rod and 'bashing it' into the end of a piece of brass rod pre-drilled to the A/F size held in the lathe chuck.
Unlike the fullsize valve, where the ejector is incorporated in the back of the valve, the ejector is a separate item.. As I had finished the brake valve before the New Year I decided to make the ejector as well, which took about four hours. It is a mixture of the well-known Brian Hughes / Reeves design and the Martin Evans' design for STRATFORD. The steam cone is drilled No74 and the exhaust cone No53.
I connected up the ejector to my compressor and an ex-BR Vacuum gauge. At 15psi the vacuum created is 5in Hg, at 30psi - 10in Hg and at 60psi the vacuum is 15in Hg.
To complete the assembly I made a vacuum limit valve to limit the train pipe vacuum to 10in Hg as per the Hughes/Reeves design. This was actually tricky to make. I turned the cylinder parts out of 1.3/8" diam brass bar and so the dimensions are smaller than the Hughes/Reeves design. In the cylinder is a rubber diaphram moved by the train pipe vacuum. The tricky part is to get the diaphram to move freely and without friction otherwise the sensitivity is lost. On testing I found it sensitive to 1in Hg - which isn't much until you realize that this is a 10% error.
Eventually the whole lot was really to be tested and I am pleased to report that it works extremely well on air. Some of you will have seen a demonstration at the January meeting. In use the ejector should use only a small amount of steam which (like the fullsize valve) can be controlled to an extent by the brake handle position. I will will also fit a steam stop valve on the ejector supply to save steam when not pulling vacuum fitted stock.