3/4" Scale Coupler Update

[RET]

Hi,

I see its been about a year and a half since I last added anything to this thread. I kept working on the couplers intermittently, I just haven't added anything.

Now I'll add the coupler knucle fabrication.

Knuckle Fabrication

I used Don's sketches to make the knuckle drawing (it shows an end view of the knuckle). In Anvil it is called Knuckle (see the picture below).

End View of coupler drawing

In the drawing, the green line is the surface of the knuckle, but the red line 1/16" outside the green one is the center of the 1/8" dia. ball end end mill that will cut the profile. The green circle at the right on the drawing is the location of the 3/32" dia. hole for the knuckle pin. It goes all the way through from top to bottom. The green circle on the left is on the full size couplers. It accesses the hollow interior. I haven't drilled it on the miniature ones.

Use Interrogation in Anvil to get the lengths of the radius values every five degrees (from the centre) for the center of the ball (the inner numbers on the drawing) and then use a spreadsheet program to subtract those values from 1.000 to get the depth the cutter (z co-ordinate) has to sink (outer numbers on the drawing) at every 5 degrees to create the surface profile. Write the "G" code program (knuckle) to create the surface in steel from a 1/2" dia. steel bar. Note: this program has two subroutines that determine the horizontal travel of the cutter on each 5 degree pass (one to go out and the second for the return travel). The first knuckle machined was made 1" long as a test piece and that is the one that is shown as the finished part in the picture.

Knuckle

Coupler knuckle#1a.jpg (14.86 KiB) Viewed 3594 times

In the knucle picture above, the 1" long piece has been cut to a finished overall length of 11/16" with a 3/32" step on each end to make the part that fits in the coupler a little less than 1/2" long (0.485) since that is the opening that was cut in the top & bottom coupler halves. The knuckle still has to have the locking "tail"silver soldered on and milled to shape.

Below is a picture of the initial test piece being machined. As you can see, it turned out OK.

Test knuckle being machined.

As you can see in the picture, this method of making the knuckle produces little grooves in the surface which must be removed with a file or belt sander. In this operation I use cutting oil on the knuckle blank but the coupler castings are machined dry.

Since that one was successful, a longer one was tried (7 1/2"). Because of its length, it had to be supported by a dead centre.

The x values in the subroutines had to be increased to accommodate this additional length. Since each finished knuckle is 11/16" long, I was able to get 9 finished pieces from that one bar. Adding the initial piece, that made for a total of ten.

Second coupler bar being machined showing tailstock support.

second bar in a finished state

As I said previously, I made the coupler knuckles before some of the other parts just to see if I could. In these pictures you can also see the rotary headstock that I made. It has turned out to be VERY useful.

There's more, but this is a good place to stop.

Richard Trounce.

[RET]

Hi,

Here is the next installment.
Here's what the partly finished nine looked like (no locking tails silver soldered on yet).

Partly finished knuckles

The next step was to silver solder 3/16" thick bits of steel in the middle of the coupler so that I could use CNC to cut the lock form as part of the knuckle. I used a 1/16" pin to locate the bits for soldering.

As you can see from the picture, it didn't really matter what the shape of the piece was as long as there was enough material to form the necessary lock shape.

Right two have been milled to shape

As you can also see in the picture, the two parts on the right have been contoured to their final shape by the mill. If you look closely, you can see the locking slot in the top face of those parts.

Next, I had to make the holding fixture for the knuckle so I could machine it. Like the other fixtures, especially fixture #2 which has to take large machining forces, it takes some thought.

The knuckle fixture needs to be simple, but it also needs to hold and locate the part so there is absolutely no possibility of any movement while the tail is being machined. Finally, it needs to allow for reliably locating the "G" code "Home" when setting up the fixture in the mill.

As you can see, the fixture is simple, but a little explanation is necessary. The hole in the knuckle goes over the pin and there is a recess for the wide part of the knuckle so it doesn't affect the clamping force on the main knuckle body. This way, the top clamping bar bears only on the part of the knuckle with the hole in it. The recess also prevents any possible rotation of the knuckle under the machining forces.

In the next picture, you can see the top clamp which holds the part securely and repeatably while the machining is in progress. You can also see the cutter following the path generated by the "G" code. In the picture, while the cutter follows the complete path (including the locking slot) on every pass, at this stage it hasn't gone low enough to actually start cutting the locking slot in the tail.

This is the end of this installment but there is still more to come.

Richard Trounce.

[RET]

Hi,

This is the final addition to bring everything up to date.

Here is a different view of the CNC fixture that holds the knuckle while the tail is being machined.

The little button was added to give more support to the tail as it is being machined. If you look closely, it is just possible to see the recessed pocket in the plate for the wide part of the knuckle.


Coupler assembly operations

Here is a view of some partially made locking bars. They are cut from 1/8" x 1/4" flat bar with the sides filed half round. They have to be a loose slide fit in the bottom half of the coupler body since it is only gravity that locks the mechanism. The same thing applies to the fit of the knuckle tail in the coupler body.

Its hard to see, but the upper bar has the same dimensions and has been machined with 1/16" radius grooves on the sides. This is necessary because I found that the finished knuckles were just a little bit short, so the end of this bar is silver soldered onto the coupler knuckle, then it is sawn down the middle to add just the right amount to the end of the knuckle. A file is used to finish the shape.

The tail and the locking bar are both steel, so the assembly should be quite strong. The new weak points become the top & bottom "ears" of the coupler body.

I still have four knuckles to make to complete the fourteen coupler bodies. When I do those, I will start with a 5/8" dia. steel bar and slightly offset it in the headstock. This way, the knuckle should be long enough so that nothing has to be added to the finished piece.

The next view shows an almost complete coupler assembly. The body is dark because of the oxidation from silver soldering.

The lifting bar still has to be cut off and flattened so the clevis can be added. Then it will be complete.


Almost finished coupler with the "draft gear" rod added.

Since even this stronger design can still be broken under the right conditions, some "give" must be designed into the system. This is done by silver soldering a 3/16" dia. rod into the body of the coupler. This rod is threaded #10-32 NF on the end for 2 lock nuts and two pieces of rubber tubing are placed on the rod when it is assembled. The car body has angle iron brackets in the ends to support the coupler and to go in between the two pieces of tubing. I prefer to use tubing instead of springs to take up the "shunting shock" or any other jolts that may occur.

This design can be seen in the following picture.

Coupler installation on flat car (top and bottom) showing the mounting details.

A few of the couplers are finished, but most still need a final fitting and the last four knuckles are still to be made. This brings the thread up to where I am now. The narration and pictures give a pretty complete record of what I have done. I wouldn't attempt this without CNC and even with it, its still a lot of work, but when you wind up with interchangeable parts that all fit, it gives a great feeling of accomplishment.

Richard Trounce

[Harold_V]

Well done, Richard. Excellent results!
A little time in a glass bead cabinet would really make them look great!

H

[RET]

Hi Harold,

Thank you for your kind words. They are very much appreciated.

Like most contributors on this website, I do things for my own satisfaction, but it is still nice to have the approval of others.

Richard Trounce.

[RET]

Hi Harold,

I made a mistake with the last addition and had to redo it so you can see that it appears twice, the first time without pictures (6:22 pm. version). Could you please remove the "no picture" version and leave the second one that has pictures? Thank you.

Richard Trounce.

Done!

H

[NP317]

Richard:
Nice work!
I've been appreciating the extent you go to in order to produce interchangeable parts.
LOTS of work!
Thanks for sharing.
~RN

[RET]

Hi RN,

When you have the proper equipment, it isn't really much harder to produce interchangeable parts than it is to custom fit each set of parts, one to another. You just have to make accurate drawings in CAD that include all the critical dimensions, then use digital readouts and CNC to produce the parts. Its just a slightly different way of looking at things.

With the equipment I have, I can work to tenths if I need to, but in most things, a thousandth of an inch is plenty good enough. It is just a case of seeing what's important and what doesn't matter in each part you make.

On another subject, there is one thing I would like to mention.

You can see the riding car in the picture doesn't have any wheels yet, but the 1/4" dia. axles are already there. Its a Harry Allin design (which doesn't mean anything to most people but he was a TSME member from the 30's to the 60's). I borrowed it from Harry's son Bob and made a drawing of it.

While I had it, on one occasion I put a 4 ft. piece of portable track on the living room floor, put the car on the track and climbed on. I just wanted to see how easily it rolled and I expected to have to push it along. It started to roll on its own on what was supposed to be a level floor! As it turns out, there are needle bearings inside a hollow axle and the needle bearings run on the 1/4" axles you can see. The needle bearings are countersunk inside the axle tube so they are in the same plane as the wheels.

I already have the bearings and I intend to use the same construction when I make the wheels sometime in the future. For what its worth, some of you might want to use Harry's idea.

I will steal good ideas anywhere I find them, but I'll tell everyone where I got them. That's only fair.

Richard Trounce.