Atlas S2s splitting their own axles?

[ek2000]

Absolute Bollocks.

You've not been around here long have you 

What does this even mean? One should not expect a manufacturer to have the right spares? Atlas gives a 90 day warranty, the least of any MRR manufacturer I've seen and while I may be happy with their product, their warranty and in this case the spares leaves much to be desired, atleast for me.

[wazzou]

Absolute Bollocks.

You've not been around here long have you 

What does this even mean? One should not expect a manufacturer to have the right spares? Atlas gives a 90 day warranty, the least of any MRR manufacturer I've seen and while I may be happy with their product, their warranty and in this case the spares leaves much to be desired, atleast for me.

You'll just need to ignore it.  Many of his recent posts are sadly of a similar nature.

[mcjaco]

Well, we're on post #24, and still no other reports, firsthand or hearsay, of this particular model splitting gears. Perhaps if we're lucky, this may be a fluke? Or has anyone else experienced this defect? It would be good to know, rather than speculate...
Otto K.

Agreed. 

[up1950s]

I wonder if instead of having a tight fitting metal rod in a plastic tube , they that a splinned or otherwise shaped metal rod that slid into its female counterpart . This arrangement would not have to be so tight as to counter torque , just tight enough to stay put . Maybe LocTited , or Goo-ed .

[peteski]

I wonder if instead of having a tight fitting metal rod in a plastic tube , they that a splinned or otherwise shaped metal rod that slid into its female counterpart . This arrangement would not have to be so tight as to counter torque , just tight enough to stay put . Maybe LocTited , or Goo-ed .

Yes, some companies use that method.  Just now I'm working on a Walthers (ex Life-Like) Mallet and they use linear knurling (if that is the correct term) on the metal half-axles which are pressed into plastic axle sleeves.  I think I've seen something similar on some Bachmann models.  But majority of manufactures (Atlas, Kato, etc.) just press fit smooth half-axles into tight plastic sleeves.   I suppose this might be related to manufacturing costs: Knurling adds a step (and cost) to the manufacturing process.

[narrowminded]

I wonder if instead of having a tight fitting metal rod in a plastic tube , they that a splinned or otherwise shaped metal rod that slid into its female counterpart . This arrangement would not have to be so tight as to counter torque , just tight enough to stay put . Maybe LocTited , or Goo-ed .

Proper Delrin and a proper fit (.0003"/ .0006") won't split the material and will hold well.  Those could be considered tight tolerances but with today's machines, in reasonable shape, that isn't really too hard to accomplish.   

I've seen Bachmann axles that were .002" tight on one wheel and then another .002" on another.  I was actually surprised at that variation.  If your machine's any good at all it would actually be hard to miss a dimension that badly.  I suspect some of the newer ones have that fixed but I think they still tend to run them tighter than necessary.  Maybe they still want the tolerance for sloppy parts that still go together enough to ship.  I don't know why they continue to do that with the decades of problems that have existed around that one issue.  Of course, it doesn't pop at the plant or maybe a year or two on the shelf so they can say, "We never saw a problem".    But a few years?  Yeah, "Pop goes the weasel(y gear)".  And nobody should be surprised and certainly not after decades of the same old same old.

When I say "proper Delrin" I mean good material and then processed at the correct temperatures, the proper parameters.  I have used Delrin bar for machined gears and parts of all types and not had the kinds of repeated problems that they have had with pressed parts.  Get the numbers right!  At some point it's not pressed, it's cold headed.   The material quality is reliable purchased as bar but it can be molded properly, too.   You just have to do it right.

[Doug G.]

Thanks for that info NM. And splined axles would still require a friction fit which would be OK if it can be controlled. The problem arises when tolerances are too loose and a metal axle with splines at the high end of tolerance ends up plugged into a plastic hole at the low end of tolerance. CRACK!

Doug

[narrowminded]

Yes, some companies use that method.  Just now I'm working on a Walthers (ex Life-Like) Mallet and they use linear knurling (if that is the correct term) on the metal half-axles which are pressed into plastic axle sleeves.  I think I've seen something similar on some Bachmann models.  But majority of manufactures (Atlas, Kato, etc.) just press fit smooth half-axles into tight plastic sleeves.   I suppose this might be related to manufacturing costs: Knurling adds a step (and cost) to the manufacturing process.

Knurling is an extra step but would be performed in the same setup as the turning, requiring a tool change from the cutting tool to knurling tool, and a speed change, but that would not add a lot of time especially in a newer CNC machine with their high speed tool changers.  There are problems with that, though, in that the small axles aren't that rigid and the knurling process is a little flaky by virtue of what it is.  There is a series of peaks and valleys in the knurler and it distorts the metal, depressing at the knurler peaks and then that material flows high into the knurler's valleys.  It creates high and low spots but the high spots aren't really easy to control a tight dimension on as it is a function of the distortion and material volume, a result of the knurl position but ALSO, the original machined diameter.  Further, because the distortion is not perfectly controlled, it tends to be a one time, one way, assembly that can't be reliably disassembled and reassembled.  You may get away with it a few times but may only get the initial assembly, too.  And if you're working with steam axles where orientation will matter (quartering) and may require a few tries, you get only one.  Once the knurl has been assembled and made its impression in the bore, that's it for orientation.  And if that's not enough, there's the series of stress risers introduced at each knurl crest.  

It's actually a process that was more common in the past when machine tolerances weren't as easy to hit, especially in metal fits on smaller shafts where zero clearance is a press and a few tenths is way too much.  It was a low cost way to mass produce  mating metal (and plastic) parts without the extreme tolerances a true press required.  But with today's machines and  cutting tools and the natural forgiveness of the materials used in these latch ups, it really isn't as necessary because you can often get the fit you need efficiently without the dimensional forgiveness knurling affords and I feel the straight press makes (on average) a superior assembly.   

[peteski]

The worst split-axle problem Bachmann had was with their white plastic axles (I believe it was Nylon, not POM). But either way, I'm sure the causes were the same as what narrowminded described for POM.

I'm a big fan-boy of Kato, but they also have had their share of split plastic parts press-fit on metal shafts (axles and universal cups).

I'm not a big fan of knurled axles either. I have a a gut feeling that when they are installed in a plastic axle tube the raised parts of the knurl cut into the plastic possibly affecting its integrity and strength, making it more prone to cracking (just like a small crack in a car windshield can spread, or like scoring a sheet of plastic makes it much easier to snap).

[Doug G.]

Bean counter faults.



Doug

[nkalanaga]

I have heard that Nylon shrinks as it ages, which adds to the problem.  I don't know if Delrin does, but given that it's intended as an engineering material, it shouldn't.

Knurled axles would be a pain if they weren't perfectly in gauge from the factory, and we know how often new locos are in gauge...

A splined axle would resist turning, but if it was loose enough to guarantee no cracks, it wouldn't stay in gauge.

Better to go back to a metal sleeve over a plastic axle.  It can't crack, but can still be regauged if needed.

[peteski]

I have heard that Nylon shrinks as it ages, which adds to the problem.  I don't know if Delrin does, but given that it's intended as an engineering material, it shouldn't.

Knurled axles would be a pain if they weren't perfectly in gauge from the factory, and we know how often new locos are in gauge...

A splined axle would resist turning, but if it was loose enough to guarantee no cracks, it wouldn't stay in gauge.

Better to go back to a metal sleeve over a plastic axle.  It can't crack, but can still be regauged if needed.

Linear knurling I was describing is like tiny splines all around the axle. The axle can be moved in and out of the axle tube but it will not turn.

[narrowminded]

The worst split-axle problem Bachmann had was with their white plastic axles (I believe it was Nylon, not POM). But either way, I'm sure the causes were the same as what narrowminded described for POM.

Yes, but a proper fit on proper nylon could have stayed together, too.  

I don't know if you ever fooled with the Bachmann PCC trolleys (the high speed race ones) but any internet search will bring up a string of comments on the cracking gears in them and then that the ones with the black gears had fixed that problem.  Well, I had a few of the white geared ones that I ran under my lady friend's Christmas tree and I put gears (or truck assemblies) in them about once a year.  Then the black geared ones came out with great fanfare and GREAT!  No more broken gears!  Except I quickly found that wasn't true. They split too.   Maybe got half again the life but far from fixed.  That's when I finally decided to dig into it and measured a new one.   Press was WAY too tight!  I finally took a new one apart, made a collet to hold the wheels in the lathe and took about .002" to .004" off the axles, if my memory is correct.  Whatever it was, it was a LOT for that fit.  Haven't had one break since.  I suspect the white gears would have kept working fine too had they had a proper fit.  I have since found bad axle diameter tolerances on some other Bachmanns  that had a problem and am still pretty amazed how tight they had them fitted.  After all of the decades of this it baffles me why they kept doing it.

For the original problem with the Atlas, it's always possible, and one could hope, that on assembly they got the gear started crooked and just rammed it home, stressing the part and starting the failure before it ever had a chance.  Or maybe it was the first gear of the day and the mold was cold.    Or maybe....    Hopefully it's the only one, we'll never know why it failed, and who cares anyway if the replacement works!

[narrowminded]

Nylon is affected by moisture.  Delrin is a much better product for these gears and other parts.  It's better in almost all of the desirable characteristics.  

Here's a Delrin engineering and applications guidelines brochure from Dupont.  If you really are interested... you might want to make an appointment and get that checked.   Meanwhile, you can look at its characteristics and also what it says about process conditions as I have mentioned, like mold temps and such.  Around page 15, give or take, are some interesting remarks for those who care.

http://www2.dupont.com/Plastics/en_US/assets/downloads/design/DELDGe.pdf

[Atlas Paul]

Our repairman was out on Monday, so I got with him this morning about this.  This is the first report we have had of the gear splitting.  We pulled some of our spare parts and they are showing signs for splitting as well, so I will now contact the factory and see how quickly we can get some replacement parts.