Bringing this topic back from the dead...
After finishing a weathering job on an old roundhouse reefer I was dismayed to discover that even my best locos could only pull a few cars up my curving 2.5% grade with this car in the train. The car really doesn't roll freely but I didn't realize exactly how big of an impact that would have. I looked for info and came across this thread from waaay back.
Inspired by the arc tester I put one together with a 900mm radius and printed it out:
https://www.thingiverse.com/thing:4171719After mounting up a piece of Atlas code 55 flex I set about testing trucks!
There were real measurable differences in the trucks I was testing but the ones from the car that didn't roll seemed to test OK. I was very confused but decided to try putting the entire car on the track. What a difference. The car in question would only change directions ONCE and come to stop at the dead center of the test rig. Other similar length cars would test at 4+ direction changes with my new Intermountain 50' box getting 6 direction changes consistently.
There is something going on with those trucks under load!
Another neat feature of testing the whole car is that it doesn't seem to matter how much the car weighs. I took a floor weight from an old car and set it on top of a few of the other cars I was testing and it did not make any difference to the number of direction changes. There are 2 reference marks on the ramp and if you line up the center of the car (center of mass) with the mark you get repeatable results.
This will let me standardize my fleet on some number of minimum direction changes to make it on the layout. Im currently thinking 4+ will be a good starting point.
I think I may draw up a 'test car' platform that will just set onto a pair of trucks (build the kingpin into the platform floor) that can be weighted to some middle ground weight and use that as a test standard.
