If somebody has had success lapping the slider portion of K-1 Triples, please PM me. We have had good luck with other repairs but trying to lap the sliding bit has resulted in just making it worse. Any advice appreciated!
Clyde Putman
Friends of the Cumbres and Toltec Scenic Railroad

Bernie Watts at Backshop has lapped a lot of flat valve surfaces in compressor heads - I'd get with him about adapting that techique to the flat valves you need to make smooth.

Are we talking a sliding type port made of brass? I have been working (slowly!) on rebuilding m-3 feed valves with the same arrangement. Best I can tell these parts need to be near mirror finish and very flat to work right. I switched to a jeweler's lapping compound and have gotten much better results, but haven't had time to put them back into the test rig to test the seal. From what I can tell the regular grits used in a machine shop are gravel to these valves.

First take a small triangular file and relieve the corners of the surface to be lapped in the body so the block won't ride up. (Figure File cuts)

You need a lapping block that fits the body. Make your lap from cast iron and diagonally score the face. Figure out how to hold on to it. I machined a spherical seat into the top, hard soldered a ball bearing that fit the seat to a worn out screw driver and put a 90 degree bend in the screw driver. (Figure Driver and lap)

Using a fine lapping compound lap the body seat with the cast iron lap. Then lap the slide to the cast iron lap, then lap the slide to the body and repeat until all three surfaces are flat. Finish with an extra fine compound.

Clean everything scrupulously. Good luck!

CCdW

Don't forget that if the surface of your plate is not flat neither will your part be. We use rotating valve plates at work for leak detection equipment. These have to maintain a seal at 30 RPM without any type of gasket. Imagine a K type locomotive brake section roatating at that speed and maintaining a perfect seal. We measure vacuum decay in 10" of water vacuum. Your body can barely produce and maintain this small amount of suction.


The specifications for lapping these plates are measured in light bands and when finished are a near mirror surface. Unfortunatly I can't tell you more about the process as we send the work out to a tool maker shop. It is not cheap and we can tell when a new employee is being trained as rejects go up.

Lapping is a lost art but there are industries that still do it. Another use that comes to mind is granite surface plates for precision measurement. We would have a company come in once a year to do ours as the plates were about 4'x6' and 10" thick. Not easy to move. It would take them a few hours with a smaller cast iron plate used to do the lapping by hand of course. Along the way he would stop periodically and inspect his work with a light band meter. Find any high spots and keep lapping when they were done it the plate was guaranteed to be within .00005" for another year.

Part of the point of having the cast iron lapping block is to have the lapped surfaces in sets of 3. If you have 3 surfaces and lap them together in succession, they will all end up perfectly flat to an extreme amount of accuracy. Surface finish will still be dependent on the type of lapping grit used.

Some WABCO literature also recommends using a piece of chamois to rub powdered graphite into the surfaces after lapping is completed.

Also, use a non-embedding lapping material (such as garnet) to lap brass parts.

Three surfaces being the magic number. If only two are used you will end up with one convex and one concave surface. This is the method used to grind the spherical surfaces on telescope mirrors prior to their being parabolized.
Keith

I've had some success with lapping various slide valves and will share a few thoughts.

My first thought and caution is that if you are not tooled up to do this and are not in for
a steep learning curve, send this valve out to multi-service, PABCo or another qualified
shop that will still work on these older valves. Lapping slide valves sucks.

Now, some terminology: The rectangluar block of brass with the various ports milled in
its surface is the slide valve proper. The mating fixed surface in the body of the valve
itself (which is in fact part of a somewhat removable bushing, but that's another story) is
known as the slide valve seat.

Any attempt to lap these two surfaces to each other by merely spreading lapping compound on
them and rubbing them to and fro will end up making things worse. In effect, you are just replicating
the process of wear which lead to the valve being out of lap in the first place. To lap a surface
and make it both locally smooth (surface finish) and globally planar (flatness) requires a constantly
changing motion so that the abrasive particles don't get stuck in a rut.

It is very easy to lap the slide valve but difficult to lap the seat, since you can't put the seat upside
down on a lapping plate and swish it around. Therefore, you need to make a cast iron block which
will be known as the "slide valve lap". As a rule of thumb, this lap must be about 2 inches longer
than the slide valve *seat* you are trying to lap, because if it is shorter or the same length, your
reciprocating motion will lap down the middle of the seat more than the ends.

Here is another important point: you'll notice that the slide valve seat has two longitudinal grooves,
one at each side. These resemble the gutters of a bowling lane. The slide valve is generally made
slightly narrower than or equal to the inside-inside dimension (the solid land of the seat). You want to
make your lap *wider* than the inside-inside dimension, so that it covers the entire width of the seat,
but not so wide that it binds against the side walls of the seat. A good rule of thumb is to caliper the
width between the raised walls and take off about 1/64" (or approx 15 thou) for the width of the lap block.
Verify that this dimension will cover the gutters. If the valve is worn so badly that the gutters are gone
and/or are not at least 1/64" deep, you'll need to re-groove them with a narrow jeweler's saw blade.
The gutters are essential for the operation of the valve. Without them, the slide valve will lift off the
seat if it wanders sideways.

There are two general approaches: lapping with compound or grinding with a succession of sandpaper
grades. If you are going with the former technique, as the other fellow suggested, you need to mill
grooves at approximately 45 degrees to the axis of the lap, about 1/16" deep and about 1/2" apart.
The spacing depends somewhat on the width and length of the lap, with smaller parts generally demanding
finer-pitched grooves. The purpose of the grooves is to give a place for the lapping compound to go.
If you are "lapping" (really truing) with abrasive paper, you can make the lap perfectly flat.

Regardless of method, be sure to lap both the slide valve and the lap to the same reference surface,
either a master lapping plate or a smooth surface plate with abrasive paper on top. A surface grinder
is handy to get the lap fairly smooth and square and will cut down on the lapping time. I have generally
preferred the abrasive paper approach with slide valves because it avoids having to re-true the lap
as frequently, and more importantly, avoids the messy cleanup of trying to get the lapping cpmpound
out of all the little ports and passages under the slide valve seat.

Then, as the other fellow suggests, you'll need a way of moving the lap back and forth while exerting
downward pressure. The pressure needs to come at the center of the lap, otherwise you will tend
to rock it and not true the seat evenly. I have tried the method he illustrated with a hook but for
heavily worn seats, I have resorted to a stiff steel bar maybe 3/8" thick and 1" wide, with the
spherical projection on it to engage the lap. This lets me bear down on both ends of the bar.

Depending on the nature of wear, it can take several hours to get the seat true. Back in the day,
there were automatic machines and tooling to do this.

I would also add that you don't want to be doing this without a test rack of some sort. IIRC
C&T does own a 3-T rack. Make sure you clean the valve thoroughly and lubricate it
before putting on the rack. It is simply not possible to test the valve accurately by mounting it
on the car or engine that it came from, besides the enormous inconvenience of doing so!

I'm not a machinist, but I do make telescope optics. Again, three surfaces is the key for a flat. Against brass, you likely can use a piece of 1/4-inch or thicker green (float/window) glass as one of the surfaces. Use pitch to attach a wood block to the back of the glass if you need to hold on to it. Ideally you not only rotate between the three pieces, but also flip them so the glass is on the bottom some of the time. Go slow, and don't repeat motion more than a half dozen times in a row, but change directions, stroke, etc to introduce variability. Heat is bad, go slow.

Once everything is cleaned up and if you get everything aligned just right, with a non-continuous source of light above your head (like a old fashion fluorescent bulb or a mono-color LED light), the glass on top, pressing down on one side with your fingers, looking down through the glass you may start to see band/rings of light. This is an interference pattern the forms as the light rays bounce between the surfaces. Each band represents a wave length of light. If the bands are straight, you are flat. If they curve more than the distance between the bands they are more than one wave length off flat. Expect them to curve quickly at the edges. The edge is the most difficult to get flat. Ideally, you test with a flat of known quality, likely not needed in this case.

Bob

I wonder if a Foucault test on a triple valve would be of value?
(Tongue firmly in cheek!)

Keith

Nope, Ronchi followed by Ross-Null or nuthin'. :-)

Well, I've never lapped a triple valve, so maybe I'm missing something but I do have twenty years of tool making experience, much of it requiring generating flat and parallel surfaces...

Let's not get too hung up on the three surface deal. Yes, when one has no access to other equipment, lapping three surfaces to each other will generate flat surfaces, but it's a slow go. Much faster if you can introduce a pre-made reference flat into the mix, such as a cast iron lapping plate, or just 3M wet-or-dry paper on a granite flat. Small Chinese surface plates are so cheap these days that everyone should have one, and brass cuts fast with wet-or-dry paper.

Likewise the cast iron lap used on the valve body seat can be repeatedly reconditioned on a surface grinder, rather than laboriously lapping it against a third surface to remove the wear imparted by the valve body.

Let's put the effort into lapping the valve and seat flat, not trying to flatten the laps by lapping.

.

I second the sentiment. Although this work process on air valves has historically been called "lapping"
there are some important differences with respect to the "textbook" definition of a lapping process:

True lapping must be done with a lap that is softer than the target material, so the lapping compound
embeds in the lap. Lapping a hole that must be a precision fit to a matching pin in a fixture, for
example, would generally done with a copper lap against the steel of the fixture. On air brake work,
the lap is cast iron and the lapped surface is brass, which violates this principle.

Lapping falls into a series of machining processes going from milling to grinding to lapping to polishing,
in decreasing order of metal removal rate and surface roughness index. What is actually needed for
practical air brake work is more towards the grinding side of the spectrum and certainly not polishing.

An air valve surface which has a mirror finish may actually perform worse than one which has a dull,
frosted appearance, if the former is not flat. The fluid viscosity of air and the actual leakage rates
which are acceptable in practical air brake work do not call out a mirror finish. The device will perform
fine with a lesser finish. Flatness and particularly little high spots are the biggest issue. Even small
scratches will have no measurable effect. Large scratches (that you can feel with your fingernail)
that lead directly to a port may cause leakage.

Optical flats and monochromatic light sources are great in the lab, or in optics work, but they are too
high of a standard for this work and they will be very, very difficult to apply to a slide valve seat. You
will have great difficulty getting the surfaces sufficiently clean and oil-free to the point where you'd be
able to get the flat down. You'd also have to make an optical flat that is ground to the precise dimension
so that it fits into the slide valve chamber.

A toolroom grade granite or steel surface plate makes an excellent reference surface for lapping the
slide valve and the lap. As I said before, I prefer abrasive paper because it does not wear down
the reference surface. You'll generally get the brass slide valve done with one sheet of coarse
wet-n-dry SiC abrasive paper, about a 200 grit, followed by one sheet of 400-600 grit. Finish with
a 1000-1200 grit if mirror finish makes your happy. Skip the coarse pass if the valve is only slightly
out of true and there isn't that much metal to remove.

Doing the seat is much more tedious. The hardest part is holding the paper to the lap. I've experiemented
with a number of approaches. Sticky tape can work -- anchor the strip of paper to the top side of
the lap, pull it around fairly taut and then stick it down on the other end. If you are using cutting oil,
which is helpful when working at the coarser grades, it quickly impedes the tape's stickiness. Another
approach is to mill a couple of slots on the top of the lap, get two little strips of metal, drill and tap
so the strip can be dropped into the slot and held down with a thumbscrew. You'll also need a precise
paper trimmer to make strips of paper that are exactly as wide as the lap: too wide and they'll curl
and bind on the sides, too narrow and they may not cover the entire seat surface.

The paper will load up much faster when working the seat this way vs working the valve against the
surface plate. I often pull the lap out every 10 strokes or so and blow down the paper and the seat
with an air nozzle.

There is an investment in time making the lap. Protect it. Don't ever stamp the lap to ID it. Light
scribing is acceptable, or paint marker on the top. Keep it in a cardboard box when not in use.
Record the critical dimensions and the valve or valves that it is intended for and also write this
information on the box.

Always interested in other's experiences with this tedious task.