Oil check valve

The mechanical lubricator supplies oil to the cylinders. Two check valves (or none return valves)  are used, so that the steam from the steam chest is not escaping through the pump, and thus disturbing a constant oil feed to the steam engines.

With my other locos these valves are made as check valves with stainless steel balls, acting as valve. This design was used by LBSC and Martin Evans in most of their locomotive designs.
Although this works fine, it is not the best solution under all circumstances. Once in a while a piece of dirt or grid will enter the oil feed system and the valve will not close properly. When this occurs, within a few minutes steam will be 'boiling' in the oil container and a mixture of oil and condensed water will start to emerge from the lubricator. Due to lack of lubrication the locomotive will come to a halt very soon.  Cleaning out the pump, pipes and valves takes some time. After refilling and pumping fresh oil through the system, the problem is not always cured. Than the disassembling of the valves is needed.

 Maintenance on the oil check valve of my 5" gauge loco, in the running shed of  Leek  in 2015 

I talked about this occurring problem with several model engineers. There are several solutions nowadays for this problem, using modern materials that are now available.

This was a solution I've tried in 2004 on my 5" gauge 'Didcot'. The inner part of the valve is shown in the picture. It works like an old type bicycle valve. A silicone tube is used as the valve; oil pressure will lift the tube from the holes and steam pressure should close it. I found that it wasn't a success and couldn't dependent on. If the cause was the material of the silicone tube or the overall dimensions of the valve I do not know. It was later replaced for a normal ball valve, which is more reliable but not a 100% 'fail save' solution. 

A German model engineer (Ralph-Peter), whom I've met at an international steam meeting in Leek,  explained me his methode to overcome the problem. He drew up a sketch of his valve, which he used with succes on his own locomotives. This check valve does not contain a stainless steel ball, but a small spring loaded valve that has a rubber seat. He even found that during an inspection of his valves after some time in operation, that small pieces of grid were embedded in the rubber seat, without disturbing the proper working (sealing) of the valve. A longer trouble free operation of this type of valve is thus expected.

I've made a drawing of his sketch in Solidworks, suitable to my locomotive

The spring loaded valve body contains a small piece of rubber, which is pressed against a small embossed part of the valve seat. 

The main body is 10 mm in diameter and 18 mm long. The valve seat and end cap are screwed in with Mf 8 mm thread.

For turning the embossed part on the valve seat a special tool was ground with the aid of small diamond cutting blade in a high speed Proxxon grinder. 

This tool cuts the valve seat in one operation.

The main body of the valve is positioned on the top of the steam chest; a small feed pipe will deliver the oil droplets straight into the live steam pipe. 

It looks a bit big; but it will be inside the steam engine cover afterwards and out of sight all together. 

A small piece of round 5 mm Viton-rubber was inserted (press fit) in the valve and cut  al little bit over size with a knife. Afterwards the rubber was grindend  square (flat) on the lathe to the correct length with the Proxxon grinder. Because the rubber is sitting on the valve seat, a smooth and flat surface is needed.

 The components of the valve: on the left the end cap with stainless steel spring, in the middle the valve with the small rubber insert (on the photo black, but later replaced by Viton which I could get from a friend), and on the right the valve seat with the small bulge that functions as the valve seat. The body has a small extension pipe, so that the oil is fed directly in the main live steam pipe.

Viton was used for the valve seat.

Because the type of my "Viton" O-ring was not exactly known, a test was done by putting the valve in cooking water for some time.

Afterwards I could see that the seat withstand the heat without much problems. The spring loaded closing force was perhaps a bit high.

The valve spring was cut to length, so that closing force was about 200 grams;  firmly closed, but not over stressing the rubber valve seat.  The valve is easy to be opened by the oil pressure from the lubricator.

I hope this type of valve will solve the problem of leaky valves; I'll keep you posted!

This week I've also made an in-line version of this valve for my 5" gauge locomotive, so I can test it very soon in running conditions when the loco is in steam.

(By now, June 2016, I can tell you that it works just fine on my 5" loco)

It is in principle the same valve, but the end cover is now drilled with a 2 mm drill and a union and nut is made.

The valve has three small groves, so the oil has an unobstructed way passing the valve.

Reconnecting the oil pipe.  (It was disconnected, so the oil could be blown out with compressed air, before silver soldering the ferrule to the pipe)

The valve in-line in the main oil feed pipe.