I noticed this issue long ago too - yes, a 3.2mm (=8 ldu) bar does fit rather well into axleholes!
But axle arm thickness is not the only culprit, even with 4.59 thick arm, it fits a bar with a 6.5 ldu diameter (instead of 8)! This is easily explained because real life axlehole teeth are rather blunt! Actually this is the major contributor, as thickness increase adds only 0.85ldu diameter increase in center hole, while we need a 2.35 ldu increase.
Max Martin Richter Wrote:When I take the axle primitive and inline it, is see that each side of the cross is 4 LDU wide. That means 1.6 mm IRL.I agree with your measurements... This means that we could have divided "axle circle" in exactly 4, this gives a very close value of 6*2*sin(22.5°)=4.5922 - this would have so much simplified life, and removed all those T-junctions around axleholes.
But when I measure a random choosen axle, I get 1.8 mm (4.5 LDU) for this wide.
But axle arm thickness is not the only culprit, even with 4.59 thick arm, it fits a bar with a 6.5 ldu diameter (instead of 8)! This is easily explained because real life axlehole teeth are rather blunt! Actually this is the major contributor, as thickness increase adds only 0.85ldu diameter increase in center hole, while we need a 2.35 ldu increase.
Quote:Conclusion our axles and everything else concerning it is about 0.5 LDU too small.Updating all axle primitives to increased thickness would be feasible but would most probably leave LOTS of misfits here and there. Unsharpening teeth would be an easier route, I don't know how realistic it would look, but probably that could be done with minimal parts overhaul.
What to do now?
Update all parts or let it as it is?