|
Shay Cylinder Cocks
Nelson Riedel Nelson@NelsonsLocomotive.com
Initial: 10/18/2003 Last
Revised: 06/11/2004
Steam condenses when it
enters a cold cylinder. The cylinder cocks are
valves located at each end of each cylinder that are
opened to drain this condensate (water). The next photo (courtesy
of friend Murry Mercier) shows the cylinder cocks on Cass No
5. The escaping steam indicates that the cocks are open
--- maybe to warm the cylinders.
|
The photo on right shows the cocks on one cylinder
in more detail. All 6 are ganged together and operated by a
single lever in the cab.
Apparently cocks are not required on the 1.5"
scale shay engine. However, cocks make the engine look
authentic and steam coming out of the open cocks looks neat!
I was unable to find manufactured cocks that look
like these. The closest is Coles' straight nose cocks.
However, the handles would need to be modified to gang the six
together and the output is straight out rather than down.
The cocks are quite simple, consisting of a body and
a plug. When the cock is open there is a hole from the input
in the body , through the plug and then through the output from the
body. The cock is closed by rotating the the
plug. The challenge is to make a reasonably tight seal when
the cock is closed. The full sized cocks use a tapered plug
with a matching taper in the body. There is an adjustment to
hold the plug securely in the body.
|
 |
I checked out Kozo Hiraoka'a three books on building the Shay,
the Climax and the Heisler. He used both straight and tapered
plugs. In the last of the three books he said that he favors the straight plug because it is easier to
fabricate and more reliable.
Last year I tried to make some cocks ---- and they leaked
really bad. A few months later tried again. A second set of
failures. I was testing with125 psi air and not looking for a 100%
seal, but 98% or so was a reasonable objective ---- and the best that could
be achieved was a perceptible difference in flow when the cock was
moved from open to close. Shutting off the flow was a fantasy at
that point.
After some months of thought I concluded that I couldn't make tapered plugs that would seal since both the plug and a
reamer with identical taper had to be fabricated. The standard taper
pins were tried and found that they would seize --- the taper is too
shallow. So, it has to be straight plugs.
The next task was to find a way to make a tight seal with a
straight plug. Teflon was selected; the idea
was to surround the plug with Teflon and to make the hole for the plug undersize. The Teflon is slippery so
the plug should turn easily relative to the tightness
of fit. Also, Teflon has some give so it should maintain
pressure on the plug. Teflon can tolerate temperatures up to 400 degrees
F so temperature shouldn't be a problem.
| Models: On this, the third attempt, a number of models
were made over a a several day period. Some of these models are
shown on the right. The
handles and plugs weren't trimmed on the early models --- the
challenge was to find a way to make a good seal.
It was found that the cock would seal initially but after a
few operations would start to leak. The hole though the body
and plug had been drilled after the cock was assembled. It was
found that the hole through the plug left a burr that tore the
Teflon after a few operations.
The solution to this was to drill the hole in the the
plug in a separate operation and to smooth the the plug before pressing
it into the Teflon
sleeve. The first successful model continues to seal after
several hundred complete revolutions of the plug.
The final version is a reduced size model of the successful model using
the same size plug and Teflon sleeve with a smaller body and output
pipe.
The smallest Coles Straight Nose Cock is shown for
comparison. It uses a tapered plug and spring to keep it tight
in the body. A nice design if you can make reasonably precise repeatable
tapers. |
 |
| This shows the final version mounted on the cylinder.
The nut is 2-56 that is 5/32" across the flats (industry
standard small pattern nut)
. The modeler's 2-56 nut that is 1/8" across the flats
would look better ---- make the cock look smaller.
Also, a slightly longer discharge pipe would be closer to the
prototype.
|
|
| This shows the disassembled cock. The parts of the
body are silver soldered together. The handle and plug are
also silver soldered together. Note the Teflon sleeve pressed into
the body. |
 |
Design: The following are the particulars of the design
used.
-
The Teflon Sleeve is 3/16" OD pressed into a
0.182" (#14 drill) hole in the body that is scored to
keep the sleeve from rotating.
-
The Plug is 1/8" OD brass pressed into a
7/64" hole in the sleeve.
-
The output pipe is 3/23" OD 1/16" ID
brass tube threaded 3-48.
-
The flow hole through body and plug is 0.052"
(#55 drill)
-
The central part of the body is made of 1/4"
brass rod with a width of 0.187"
-
To insure alignment, the input and output pipe are
made as one piece and soldered into the central part of the body
and then the hole for the Teflon sleeve is drilled separating
the input and output pipe.
-
The input pipe is threaded MTP 1/8"
-
The end of the plug is threaded 2-56
-
When the plug is forced into the sleeve some of
the Teflon is squeezed out beyond the faces of the body.
-
The plug length is about 0.010" greater than
the body width.
-
The nut and the side of the lever apply a small
pressure on the ends of the sleeve that are squeezed out of the
body.
-
The lever is made of 5/32" wide 0.50"
thick stainless steel.
-
The effective lever length is
3/8"
-
The hole in the plug is drilled in a separate
fixture and all burrs on the plug smoothed before being pressed into the sleeve.
Fabrication Process: The fabrication
process is as important as the design -------- the process must
yield good results using the relatively low tolerance workshop
machines. Since the
setup is a significant part of the chore, some spare cocks were
made for future replacements. Also, the jigs were kept to
allow easy future setup.
| Input & Output Ports: The ports are turned from
3/16" hex stock. The first part was done in the lathe----
outside turned, inside drilled and end threaded. One
set of ports was turned on each end of the ~ 12" hex rod.
The next step was to drill the hole for the output tube as shown on
the right. That hole was then tapped 3-48 and the rod mounted in
the lathe again and the ports cut off with a parting tool. |
 |
| This shows the dimensions on the finished ports. None of
the dimensions are critical except of course the holes shouldn't be drilled too
deep and run through the other side of the piece. |
|
| Plugs: The plugs were made from 1/8" brass rod. One
end was turned down and threaded 2-56 and the other end turned down
to 3/32" The only critical dimension is the 0.197" length of the middle part. The 1/16" part
will be
soldered into the lever and trimmed after the soldering.
The threaded part will be filed down to the length that looks best
after assembly. |
 |
| The first step was to turn the end to be threaded and thread with
2-56 die. The rod is held in the lathe collet chuck
and the die holder slides on a rod held in the tailstock chuck. |
 |
| A fixture was made to hold the threaded end of the plug.
The lever end of the plug was then turned to 3/32" with the
length of the middle section 0.0197" Once the stops were
set correctly, it took only a few minutes to turn the ends of 9
plugs. |
 |
| The lever was made from 5/32" wide, 0.050" thick stainless
steel. The distance between the holes is 3/8".
The plug was soldered into one hole and a brass tube in the other
hole. The tube was made from 3/32" rod with the about
1/8" at the end turned down to
5/64" and drilled #49. The photo on right shows the
plug just after the soldering.
The lever was shaped with a file before the 3/32" rod was cut
off. |
 |
| After the 3/32" rod was cut from the handle, the hole in the
little tube was tapped 0-80.
The plug-handle combination was then mounted in a fixture to drill the
#55 hole through the plug. The fixture is similar to the cock
body except it has a 1/8" hole to match the plug diameter
and no Teflon sleeve. The fixture has a hole for a
0-80 screw into the end of the plug lever to hold it pointing up at a 45 degree
angle. |
 |
| After the hole was drilled through the plug, the burr at each end of
the hole was removed by hand turning a 5/64" drill into opening
of the hole. The plug was then smoothed with 400
grit Emory cloth (cloth wrapped around plug and held firmly while
lever rotated. ) This was repeated a half dozen times to get a bright
smooth finish. From the looks of plug at the right, it could use
some more smoothing.
|
 |
| A tapered assembly tool was made from a 1" length of 1/8"
diameter brass rod by drilling a 1/4" deep # 44 hole in one end
and filing the other end to a point forming a taper. The
tool is shown on the right between the plug and the body.
The body was placed over a plate with a 1/8" hole and
the assembly tool lubricated with a little petroleum jelly and
then inserted through the Teflon sleeve into the hole. The
threaded end of the plug was inserted onto the tapered assembly tool
and the plug and tool pushed through the sleeve. The
tapered tool eases the entry of the plug into the sleeve to avoid
cutting the sleeve.
The plug was rotated 20 or 30 times to loosen it up a
little. |
|
| The cock was then tested with 125 psi compressed air using fixture
on right. The end was submerged in water to verify no
leaks.
One additional test was done --- several of the cocks were
submerged in boiling water and there was no noticeable effect --- they
continued to seal. Also a cock with a small leak ( about 4
bubbles per second) was submerged in the boiling water and the leak
stopped within a few tens of seconds. When it was placed in cold
water it started to leak again within a few 10s of seconds. This
was repeated a half dozen times, sometimes rotating the pug a couple
revolutions between submerging. Same results. Suspect that the
Teflon swells when hot making a better seal.
The cocks haven't been tested on steam yet but expect good results
based the boiling water tests.
|
 |
| Several of the cocks leaked slightly so the plug was removed and
smoothed again with the Emory cloth and a new Teflon sleeve
installed. That fixed it every time.
The photo on right shows a finished plug with the an output
tube.
|
 |
| This was my first use of the MTP 1/8-56 die. When I used
the die I found the resulting threads too big to thread into
elbows. The die as manufactured is not adjustable. I
cut a slit in one side using a Dremel cutoff disk and then used it in
a die holder where the retaining screws squeeze the die slightly
making a smaller OD thread. That fixed the problem.
Care was taken when taping the holes in the cylinder to not run the
tapered tap in too far making the hole too big. |
 |
The photo above shows the six cocks mounted in the
cylinders. The shaft linking the three cylinders is 3/32"
stainless steel. The levers on the shaft are identical to the
levers on the cocks. The little bearing blocks are made from
1/8" thick brass and retained by one 2-25 screw per
bearing. Note that I've decided not to use the tubes on the
discharge side --- at least for now.
Update: The next photo shows the view from the rear of the
engine. This is an updated photo that shows the cock operating
lever. It was decided to leave the operating lever outside the
cab rather than trying to develop some sort of a linkage. Some
of the prototypes used a similar lever located inside the cab, in a
location that is hard to reach for the engineer of scale
model. This outside lever tucks in front of the cab side
when the cocks are off and is turned nearly horizontal when the cocks
are opened. (The shaft and little levers haven't been
cleaned after the silver soldering. The flux residue will be
removed and the shaft polished when the engine is disassembled for painting.)
| 
