By using 3D printing, I should be able to save time in my experiments with Percy. That in turn means I will be able to learn more in the same length of time. Indeed, I should be able to run so many experiments, that I can afford to change one variable at a time.
The first variable is 3D printing, itself. I printed one wheelset to test the supports, which worked fine. However, that first wheelset would bind twice per revolution. The axle was a couple of thousandths of an inch wider than designed, most likely due to the accuracy of my Mars printer, and the surprisingly tight tolerance of the original axles.
After spending a number of minutes filing the axle, it rotated okay, but before I had it running sweetly, I decided to change approaches. I still had to print the second wheelset, so I simply redesigned the axle as an oval, and re-printed the first wheelset.
The axles still bound a little, but I switched front for rear, and then only the rear had the problem. This time the solution was a very few swipes of the file. Then the chassis coasted smoothly when pushed down the test track.
The next step was to test with the connecting rods. To keep the rods in place, I cut a yogurt lid into clips that fit into grooves in the crank pins. With the rods on, the chassis still coasted smoothly down the test track in exactly the way that my engines never have. I’m confident I have a basis for experimentation.
Pembroke:87 
These are good looking experiments. We have a Percy around this house too, though of the Brio track sort. Never graduated the little railroaders here to electric trains – I guess I was not the ideal role model!
I’m having a little trouble connecting the dots on how the experiments will help your work on 622. The other day you mentioned the drivers need re-doing, so I am wondering whether some or all of the following are contributing to issues:
– are they concentric with the axles?
– are they perpendicular horizontally and vertically as between wheel face and axle centre line?
– do the drivers vary in diameter?
– is there variation in the throw distance from axle centre line to the pins/screws used to hold the drive and connecting rods?
– are the pins/screws used to hold the rods perpendicular to the wheel face? i.e. concentric with the axle both vertically and horizontally?
– are the axles are horizontal/level & perpendicular to the rails?
I’m not sure how important tiny variations of these elements might be, but they strike me as relevant.
Or is it something altogether different?
Rob
Hi Rob, thanks for your questions. All of the above are (or were) as good as I could make them at the time. I really believe quartering is the biggest issue, but I’m hoping to understand where there is wiggle room through the experiments with Percy. For example, Percy’s connecting rods are much looser than my engines’. How much does that matter? Also, Percy’s crank pins are more than double the diameter of 622’s; what effect does that play? Basically, Percy will help me get a “feel” for what works.
As they say, the finest line in model engineering is the one that separates running clearance from slop. I need to an intuitive feel for that line.