Improving the Whistle

Feldbahn whistle leaking waterWhen I ordered my Stafford from Station Road Steam in 2009 they refused to fit a whistle of any kind to the locomotive so I had to fit my own, and as you will see if you read the Stafford pages of this website getting it to sound cleanly without spluttering to life as it expelled condensed steam took me some time.  When I traded in the Stafford for the Feldbahn I had expected SRS to be offering a whistle that worked nicely and so I was quite disappointed to find that the whistle valve they fitted constantly leaked and water ran freely from the whistle down the boiler cladding.  This looked unsightly but the worst part of it was that every time you sounded the lovely chime whistle it spluttered into life imitating a decrepit fountain. You can see the water collecting around the base of the whistle in this photo.
 I lived with this for over a year while other more important tasks were tackled but finally the time arrived to solve the problem.  The root cause of the problem was simply that the commercial whistle valve fitted by SRS always leaked slightly, and resultant slow flow of steam condensed in the pipe leading to the whistle where it then dribbled out.  I had learnt from the Stafford that metal on metal valve seats rarely seal 100% against steam at boiler pressure so what was really needed was a whistle valve with a PTFE valve seat (or something similar).  As I couldn't find such a device for sale it seemed that I had two alternatives; either to fit one of the R A Barker PTFE seat whistles on the front of the cab and somehow rig up an operating cord / chain or to design and make my own whistle valve.  I chose to design my own whistle valve for the simple reason that R A Barker do not currently sell their excellent whistle valves with a chime whistle fitted.
This page explains the reasons behind my design and gives details of the whistle valve I created for the Feldbahn chime whistle.

Sectional view of the O Ring whistle valveThe starting point was to adapt my O Ring clack valve design which has proven to be leak free by adding an operating push rod to open the valve (click here to see the relevant page on the Stafford section of this website). Unfortunately although my new design worked perfectly using compressed air the first time it was tested with steam the O Ring blew off the shuttle and the whistle valve would no longer close.  With no other shut off valve available I had to endure the shriek of the whistle for about 15 minutes while I dropped the fire and got rid of the boiler pressure.  After that I redesigned the valve to use a PTFE valve without the O Ring and this seems to have solved all my problems.  Throughout the design of the new whistle valve I had the following objectives in mind:
1)  There should be a minimum of high pressure joints or seals.
2)  The valve seat should be easy to machine e.g. not at the bottom of a deep hole.
3)  The cross sectional area of the steam passageways had to be equal to (or greater than) that in the original factory fitted steam valve.
4)  The valve should fit directly to the existing steam manifold and not obstruct the use of the new R A Barker steam valves.
5)  The valve should be operated by a lever as opposed to any form of chain.
The existing dimensions of the clack valve O Ring initially set the basic dimensions of the new whistle valve and thus the valve spring, while the original whistle valve gave a minimum cross sectional area for the steam passageways of 13 sq mm.  A suitable stainless steel compression spring was easily sourced from ENTEX Springs but I then spent hours using my CAD program sketching many potential valve designs before I settled on the design shown here.

The drawing reproduced from my CAD program shows the cross section through the middle of the whistle valve.  The main body of the whistle valve is shown in light blue and the spigot on the right screws straight into the Feldbahn's steam manifold (1/4" BSP thread) and is locked in place by the nut shown in green.  Steam enters via the offset steam passage to reach the valve chamber which contains the spring (light brown) and the PTFE valve (red).  The PTFE valve seats against its valve face on the bottom of the removable threaded bush (white).  Experience of my O Ring clack valves had taught me that I could machine such a bush and main body face to be steam tight against the full boiler pressure.  The valve top cap (dark blue) is bolted onto the main valve body using M4 socket cap head screws, but being on the outlet side of the valve this joint does not need to be so steam tight as the main valve body components.  Steam exits the valve top cap via the spigot on the top left which is threaded 3/8"x32 to accept a normal 1/4" diameter pipe nipple.
 

Whistle valve body and cap for a Feldbahn or Stafford steam engine20This photo shows the whistle valve body (top) and valve cap (bottom) machined from 1 1/4" square phosphor bronze bar.  To save on metal they could probably have been manufactured using several parts silver soldered together but I had a suitable piece of bar so the parts were turned using the four jaw chuck on my lathe.

Valve Body.
The input spigot was the first part to be machined and then threaded 1/4" BSP.  Since the spigot is well offset from the centre of the bar this operation has the most out of balance mass spinning on the lathe, but by using a low spindle speed with shallow cuts at a low rate I managed it without resorting to any counterbalance weights.  The part then needed to be offset again in the chuck to drill the inlet steam port.
The face opposite the inlet spigot was then faced back to create the correct width of the valve body.
Turning the part again in the chuck allowed the circular lower body (at the back in the photo) of the valve to be turned before further repositioning in the chuck allowed the remaining two sides to be faced back.  Obviously all of these operations had to end up with the various faces correctly positioned with reference to the centre line of the inlet spigot.
Finally the top face of the valve body was faced off at +1mm from the designed body thickness, the various diameters were drilled and tapped, and only then was the top face finally faced to the correct thickness and the set back area turned to create the raised boss that would locate the valve cap.
The four tapped holes that retain the cap to the body would be drilled later using the cap as a drilling guide.

Valve Cap.
This was machined in a similar fashion to the valve body, spigot first with its 3/8"x32 thread and outlet port and pipe nipple seat.  Then the sides were faced followed by the turning of the boss on the face hidden from view in the photo.  Finally the mating face of the valve cap was faced and bored to create the main cavity.  The larger recessed diameter was machined to be a close fit over the raised boss on the valve body, and the spindle hole was drilled through from the cavity side to ensure concentricity.  This hole needs to be a close sliding fit for the valve operating rod so that the rod can slide easily, but with the minimum of clearance to reduce the amount of steam that can escape when the whistle valve is operated.
A simple jig was then made that located over the boss on the top of the valve cap to allow the four holes to be drilled through the cap at the tapping drill size for the M4 bolts.  The cap was then clamped to the valve body (ensuring that it was correctly aligned) and the holes frilled down into the valve body.
The body holes could then be tapped M4 while the cap holes were opened out to clear the M4 stainless steel socket cap screws that would hold the cap to the body.

Making these two parts took the best part of a day of lathe work.
 
The 1/4" BSP locknut was machined from a scrap piece of phosphor bronze hexagon bar, remembering to relieve the hexagonal section on the mating face so that the points of the hexagon would not scratch paint off the steam manifold as the nut was tightened.  You can just see the relieved section in the photo.

The removable threaded bush that has the valve seat on its lower face was also machined from phosphor bronze hexagon.  This part looks easy to make but it needs care to ensure that the valve face is really flat and without burrs around the steam passage hole.  Basically drill the steam passage before facing the valve seat.  The mating face between the bush and valve body also needs to be really smooth with the thread correctly relieved to ensure that the bush screws fully home into the valve body where it will create a steam tight seal.  Some Copper Ease grease applied sparingly to the thread will help with the seal and prevent the bush seizing into the body.  You can just make out the residue of the Copper Ease along the edges of the bush hexagons in this photo.

The half diameter holes in the top face of the bush are there to provide additional steam passage space between the bush and the valve cap.  These were drilled through the hexagon bar before the face was turned to create the correct thickness.
 
This photo shows the PTFE valve and the operating push rod.  PTFE isn't my favourite material to turn but the part didn't give me any real problems.  A 3mm diameter hole for the operating rod was drilled 3mm deep into the seating face before that face was finally faced off by another millimetre.  As before, making the part this way results in less chance of burrs on the valve seating face than if the hole was drilled after the face had been turned.  The spigot on the left of the photo was then formed using the parting tool.  It needs to be a slightly loose fit inside the coil spring.  Finally the valve was parted off and the finished valve face was inspected using a magnifying glass.  As I couldn't see any surface scratches or imperfections I decided that it would probably be good enough to provide a steam tight seal when the valve was closed.

The operating rod was simply a length of 3mm diameter stainless steel with its ends faced off.  As it needs to fit between the closed PTFE valve and the valve lever with only a small amount of clearance I found it easier to make this part after everything else had been completed.
 
Leak free whistle valve fitted to a Feldbahn steam engine.Here is the finished whistle valve fitted to my Feldbahn.  The operating lever was cut from 3mm thick mild steel and has a varnished wooden knob on the end.  Note how the lever extends down below the pivot on the left to act as a stop to prevent the lever from lifting up further than necessary.  The valve operating rod described in the section above ends up with about 0.5mm of clearance between it and the operating lever when the lever is fully "up".

The lever pivot bracket was silver soldered from 3mm thick brass and the 4ba pivot bolt is threaded into the pivot bracket.  On the hidden side, a 4ba Nyloc nut secures the lever and with careful adjustment stops the lever from flopping about sideways to keep it centred over the operating rod.

So far the whistle valve has been working well and no water has been seen weeping out of the whistle so I hope that the leaky whistle valve problem has been solved.  I have deliberately avoided putting the CAD drawings on the website because so far the valve is an experimental part, but if you would like a copy of the CAD drawings then please contact me via the link at the top of the page.