hardest plastic for joining axles?

[mmagliaro]

Center boring a 1.5mm hole in a Delrin rod is not as hard as it seems.
Start with a 1/8" rod, bore the hole by eye "close".   Press a 1.5mm axle in it, chuck the axle in
the mill, then turn down the Delrin rod around it to true it up.

It's the same approach I used on the brass hubs, my spoke soldering jig, and more.
It works really well. 

As for this whole flatted shaft/epoxy idea, here's what I do not understand.
You can position escape holes and inject glue however you want, but in the end, you have a
metal shaft inside a plastic tube, with a crescent-shaped slug of epoxy in there.   Unless you have holes or cuts
into the inside of the tube so that epoxy can get into them like "fingers", you are relying on a glue bond,
not a mechanical block.  Without cuts in inside wall of the tube, there is really nothing to keep
the crescent-shaped glue slug from spinning around inside the tube with the axle.

Like so:


Now... if I were to drill through the tube into the epoxy slug after the fact, and insert a short pin,
that might work great.    And JB is hard stuff.  I can see putting a short wire pin into it, coated in more epoxy.

[narrowminded]

I am envisioning an axle assembly with two stub axles inserted from each end with an insulating gap between them.  It is that void including the flatted axle shafts that would be filled with the epoxy in one piece, engaging both axle shafts, bridged through the insulating gap, resulting in one cast in place slug that has adopted the mechanical features of the surrounding parts.  Maybe I'm not understanding the assembly. 

[mmagliaro]

I am envisioning an axle assembly with two stub axles inserted from each end with an insulating gap between them.  It is that void including the flatted axle shafts that would be filled with the epoxy in one piece, engaging both axle shafts, bridged through the insulating gap, resulting in one cast in place slug that has adopted the mechanical features of the surrounding parts.  Maybe I'm not understanding the assembly.

No, you've got it right.  So you are assuming that the epoxy between the two metal shafts, with their flatted ends,
is where the strong bond would be, and the glue would thus not have to worry about bonding to the slippery
plastic tube?  The tube would be there to keep them in axial alignment, but otherwise, the bond between
the metal ends is what would keep them from sliding in/out, or rotating out of quarter with each other?

I think I get it now.  I suppose that might work.  With the flatted regions, there would at least be
some surface for the epoxy to stick to.

[Lemosteam]

If you are going to fill the brass tube with epoxy there is no need for the delrin insulator.  Just make two jigs one for each end that fits the outside of the brass tube cut to length, with a bore in the center to hold each flatted axle half  shaft in place, while through a third hole you inject the epoxy after clamping the jigs together..

[mmagliaro]

No, the metal axles cannot have any chance of both touching the brass tube anywhere.  That would
create a dead short across the frame halves.

[Lemosteam]

That could easily be controlled with a styrene washer in the center of the tube, which would have a short pilot diameter in the centers of both sides, both to serve as a locator and a spacer.

[peteski]

I think that making the insulating tube out of epoxy (and keeping it from oozing out)  while trying to keep the axles perfectly aligned, and the wheels perfectly quartered is not something that is really feasible.  There is just too much going on at the same time.  I think you guys are over-thinking this.

[mmagliaro]

I think that making the insulating tube out of epoxy (and keeping it from oozing out)  while trying to keep the axles perfectly aligned, and the wheels perfectly quartered is not something that is really feasible.  There is just too much going on at the same time.  I think you guys are over-thinking this.

So do I.   

I'm just pressing them in there, with a flat on the shafts, getting it quartered, and then drilling a small
hole through with an M1.0 thread for a small screw.

[nkalanaga]

I like NarrowMinded's method.  Since the axle would be solid after the epoxy set, it wouldn't matter if it was glued to the tube or not.  Even if it spun in the tube, the wheels would have to stay in gauge.

[peteski]

I like NarrowMinded's method.  Since the axle would be solid after the epoxy set, it wouldn't matter if it was glued to the tube or not.  Even if it spun in the tube, the wheels would have to stay in gauge.

But the piece of epoxy which makes up the key part (the crescent shape filling the flat in the axle) is very small - just a tiny sliver of epoxy - to me it seems that it will shear very easily.  We are taking about something like 0.005" X 0.010" cross-section of epoxy. That will not have much shear strength.

[mmagliaro]

That's what I think.   It is too much to expect of that little bit of epoxy, and it makes a little slug of epoxy
have to act like part of the axle.  Also remember that on one of these, it won't be a plain
plastic tube.  It will be a tube with a molded gear on it, in which case it does matter if the axles spin inside
the tube, even if they spin together.

After considering all these options, I still like the idea of flatting the shafts, and then filling up that area with
epoxy, but drilling small holes in the tube for a pin or a small threaded screw right into that epoxy.  Now,
the hardened slug of epoxy just has to act as a physical dam between the flat and the screw and nothing has to
depend on being an adhesive.   I can push it all together, quarter it, then drill and tap the hole.

[nkalanaga]

If the epoxy is between the two flats, it can't very well shear, because the torque will be trying to compress one side while the other is in tension.  Epoxy won't compress, so there's nowhere for the other side to open into.

Now, if it was between two flat ends, yes, shear stress would probably break it fairly easily.

[peteski]

The way I understand Max's design is that the half-axles do not overlap (just like in any standard Kato wheelset (passenger or locos like GS4 or FEF). 

[mmagliaro]

Correct... It is a standard half-axle design where two half axles just press into a plastic tube and
don't touchin the middle so they don't short.

This is done all the time on engine, and it usually just relies on the axles pressing in tightly enough that they
don't slip, which I don't really trust.  Even with epoxy in there, there is nothing to keep the two
axles from slipping inside the plastic tube unless that epoxy actually adheres to the inside of the tube,
and I don't want to rely on that.  And unless that little slug of epoxy makes a super-strong bond, which
I highly doubt in these small surface areas, the two half axles could still shear apart and slip independently
of each other.

Like this:

[narrowminded]

That picture is very much what I envision in what I described.  All of these details could be tested very easily by mocking up that joint.  You could consider keeping the axle spacing at .020" rather than .005".  You could also file the flats at a reducing angle towards the round section, or notching the flatted section root transition, further keying the joint, not depending on a bond but mechanical interference.  And if the injection method was used, the access holes would finalize as cast in place, zero clearance pins where they pass through the sleeve's access holes, all one piece without requiring pretty outrageous tolerances in a part that will be difficult to machine at best and near impossible to machine with conventional machine tools at the tolerances required.

And further options utilizing the same basic approach still exist if that isn't enough. Two flats could be filed opposite each other on each axle, four total, and the opposing axle flats assembled somewhat aligned (tolerance loose as the part will be cast in place at whatever final shape results after quartering and gapping) and then the flatted sections will be full dimension across the axle gap. Each key will bridge one to the next at full cross section and the whole thing girdled and held in alignment by the sleeve.  I think this general conceptual approach to the problem has a very good chance of success whereas the machining of miniature pins and miniature threads engaging plastic seems inherently fraught with material strength and physical space problems before it's even started, and then imposing tolerance requirements that could only be described as cruel and unusual punishment to the machinist tasked with that job.   Testing the concept as well as refining the process before to attempt assembly on the final parts would be imperative but also easy to do.

Being a newer guy around here with minimal history I hope these posts aren't interpreted as being argumentative or carrying any motivation other than the spirit in which they are offered, to be helpful where I think I can be helpful.  I still admire and marvel at  the kinds of work this community performs and only offer suggestions where I have at least some experience or feel I can contribute something drawing on my experience, 30 years of special machinery design and manufacturing.  This is one of those rare times.  I think the basic idea is solid and has a reasonable chance of successful execution with basic tools.  A few trial parts should quickly prove the validity or not as well as resolve the execution methods.  Good luck whichever method you choose, sincerely.   It's already an amazing piece of work!