Trucks, Wheels & Derailing

I clearly remember buying NWSL wheels with 1.015 axle lengths. Said so on the packet even! I have even checked their website and they sell wheels with flush ends, and pinpoints with axle lengths 1.015 and 1.025. What is the point of this I hear you ask?

Well, actually that is an interesting question. I've been having problems with freight cars derailing, in particular my Proto 2000 tank cars that have Kadee AAR standard trucks, colloquially known as Bettendorfs, which is incorrect, unless the AAR standard trucks were made by Bettendorf (just showing off my freight car credentials here), and laying out my Railway Prototype Modeller agenda. Those of you who are not interested in such things you should look away now, before I go into uber geek model maker mode.

Okay, all of you who are still reading can sit back and relax while I talk you through the problem...

I was talking to Bill of Mousa Models about ordering some wheels with 1.020 axles from NWSL as a special order. I quite like NWSL wheels, because they don't wobble and therefore run true. He quite rightly asked why bother as I can just use the 1.015 axles I have, as five thou shouldn't make any real difference? I said well, I've researched another manufacturers site, Reboxx, and downloaded their PDF that recommends 1.020. We agreed to disagree, as friends can.

However, Tony Thompson reminded me of tram-lining wheels on his blog, which I can't find the link to post here. Perhaps Tony might consider putting tags on his posts to help searching for stuff? Just a suggestion.

Anyway, I thought to myself, lets test the assumption that five thou will not make any great difference.

The way I did this was by taking a Kadee AAR truck and begin measuring the axles Kadee supplied, which turned out to be 1.020. Then I started working on calculating the depth of the journal boxes and the total amount of play across the axle length when the wheels are in the journals. What I found was a range of "slop" between 0.016 to 0.021 of an inch each side, or 0.42mm to 0.54mm, which means that there is a range of movement from 0.032 to 0.042 of an inch, or 0.82mm to 1.08mm across each axle.

This is not good, as it means that the amount of slop in each axle might result in the wheels not being in line with each other. To me this seems like way too much slop.

I then wondered about what length of axles I should be using?

So I started by measuring the axles in my tank cars and found that they were not 1.015, but rather 1.010. Okay only five thou difference, so no real biggie, but those five thou's add up, and ten thou, approximately 0.25mm, would be a thing to be concerned about. At this point I just had to go through and measure all the axles of all my wagons, which took the rest of the morning.

Didn't have many 1.015 axles, in fact they all seemed to hover around 1.010, ranging from 1.007 up to 1.013. So about three thou either side of 1.010. So I thought, ah ah! I've either bought the wrong wheels, or NWSL had a packing glitch. The evidence being that I found seven axles that were 1.005, and three axles at 0.997 of an inch.

Then I went and checked the website and found that NWSL don't market Code-88 wheels with either  1.005, or 1.010 axles, which made me think further. My conclusion is that NWSL manufacturing tolerances will not meet my needs.

My best guess at the axle length I need is 1.030 (this is 1.010 from my original axle, plus 0.32, less 10 thou, which I consider to be an acceptable amount of slop, which comes to 1.032), but this is a bit of a guess, because it is dependent on the manufacturing tolerances of both Reboxx and Kadee being consistent. So fingers crossed Reboxx here I come. I will report back when the order arrives in a few weeks.