After several years of trouble free performance, I've started to notice consistent issues with the performance of some of my trains at recent events. I've observed some aren't pulling cars as well as they used to, or on some layouts, they don't pull at all. In preparation for an upcoming show, I finally setup a test loop and took a closer look at my engines and rolling stock.
Up first was weeding out any bad 9V motors. For these tests, I timed motors around a loop of track. The loop has the usual curves and straight runs, but also a few switches as well, which tend to slow things down a bit. I set the regulator to about half speed and didn't touch it again until all tests were complete. Most motors, stripped of everything but a coupler, made the run in about 15 seconds, including several motors fresh out of the box. Out of 11 motors, 2 were in the mid 20's and one took a full 30 seconds.
For the next test, I loaded the motors down by having them push a single car. For most motors, this added a few seconds to the run bumping the normal time up to about 20 seconds. The 30 second unloaded motor, could not complete this test. In the end, I sat 3 motors aside for repair.
As a final test, I took one of the remaining motors and tested all of my rolling stock. Some of my older cars, the Santa Fe series in particular, have consistently required more pulling power for their weight than others so I was hoping to find something with these tests and the results did not disappoint. Lighter weight cars naturally completed the loop a bit quicker, but all cars completed the loop in between 15 and 18 seconds. That is, except for 3 Santa Fe cars which took about 25 seconds.
It turns out, wear on the wheel assemblies was slowing these cars down. On a good wheel holder, there is clearance between the wheel and the holder.
On a worn one, this clearance has disappeared allowing the wheel to contact the holder.
On a light grey holder, this is easy to spot in the form of black scuffs. The wear at these points isn't visually significant, so it could be hard to spot on a black part.
Removing some plastic from the holder to restore the clearance solved the issue for me. With this quick mod, all cars were restored to the 15 - 18 second loop time. In the case of these Santa Fe cars, that represented roughly a 30% improvement!
Another item worth mentioning is to also check idle bogeys on engines. I found wear on these as well. Rather than running a test, I preemptively fixed these too. Finally, the more recent idle bogey designs that use technic bricks and axles seem to inherently have a much higher rolling resistance than the old style using the wheel holder (part #2878).