Friday, September 16, 2022

Tolerance Stack Up

 Tempting as it might be to save this title for a commentary on modern day culture, I decided to use it for this piece with a more traditional understanding of the term.  As it pertains to motorcycle mechanics, the term is commonly understood to refer to the manufacturing tolerances of parts that are used together and how those plus/minus tolerances can affect overall clearances and fit.

When dealing with just two parts, it’s a very simple concept to understand.  Say a shaft is designed to have an outer diameter of 1.000” with a manufacturing tolerance of +/- .0005”, and the bushing it rides in is designed to have an inner diameter of 1.002” inner diameter with a manufacturing tolerance of +/- .0005”. That means if the both parts are spot on spec, the combo will have .002” clearance.  If the shaft is at the low end of the tolerance (.9995”) and the bushing is at the high end (1.0025) you will end up with .003” clearance. On the other hand, if the reverse is true (1.0005” on the shaft and 1.0015” on the bushing, then all of sudden you only have .001” clearance. 

Obviously when dealing with multiple items that are part of the same assembly, it can get interesting in a hurry.  Consider the “squish” clearance (distance between the ledge on the cylinder head and the piston deck at TDC) on an Evo or Twin Cam.  Each part that contributes to that distance has its own manufacturing tolerance which affects the final assembled clearance.  Here are some of the measurements that come into play: The distance from the centerline of the crankcase main bearing to the surface where the cylinder mounts, the cylinder length, the gasket thicknesses, the center-to-center length of the connecting rods, the distance from the centerline of the piston pin to the upper surface of the top ring land, and even the centerline distance from the main shafts to the centerline of the crankpin (which also determines the stroke, BTW).  If all those tolerances stack up one way, the squish distance would be smaller than if they stack up the other direction. And that is why those who build performance engines take the time to measure and adjust that final distance to make the squish clearance as efficient as practical.

But what brought this issue to mind and caused me to write this piece, is much more mundane, and in fact a bit humorous.  The other day I was putting the finishing touches on a cone Shovelhead lower end.  I had already installed the new oil pump, checking that it turned smoothly. I had selected the pinon gear which provided just the right amount of lash between it and the cam gear to provide quiet operation (not too loose and not too tight). I had test fit both the cam and breather gear for proper end play before the final installation of the cam cover with Loctite and a torque wrench applied to the screws. Satisfied with my work for one day, I would get back to it in the morning.

But the next morning, when I went to the next step, that of installing the tappet blocks and lifters, much to my surprise, the crank would no longer turn! Where in the world had I messed up?  The cam still had end play so the next thing that came to mind was that perhaps the cam cover gasket had compressed more than expected, taking up the breather gear end play that I thought that I had established.  Off came the cam cover again. Nope, …still locked up.  Now this would have really been a puzzler except that right about then I remembered that I did one the last thing after installing the cam cover the previous evening.  I installed the timing hole plug. 

The brand-new S&S crankcases came with a nice zinc plated timing plug that would easily pass for an oem part. But the customer had a nice new shiny chrome plated Colony brand timing plug, so of course that’s the one I installed.  Sure enough, the chrome plug was just enough longer that it contacted the flywheel and kept it from turning.  But before we condemn Colony, let’s think again about this whole tolerance stack up issue. This particular motor has a set of original Harley flywheels, …you know the kind with the wide and heavy left side wheel. So, the fact is, the Colony timing plug would not have touched thinner S&S flywheels when used in S&S cases, and neither would it have touched stock flywheels in stock crankcases. And of course, the S&S timing plug doesn’t touch the stock flywheels when installed in the S&S case. But the combination of parts from two different sources with their differing tolerances, turned out to be just enough to make them incompatible.



And that brings to mind a similar situation I ran into about 30 years ago, also while building a Shovelhead engine.  The parts list included Delkron Shovelhead cases and a set of the generic aftermarket lifter blocks that nearly all the major parts distributers sold (and still do to this day). One of the lifter blocks, whether front or rear I do not recall, would not physically fit into the crankcase.  In an effort to determine where the fault was, I took a stock lifter block and it dropped right in to the Delkron case.  But then, before counting the aftermarket lifter block out, I tried it in a stock Harley case.  Sure enough it dropped right in.  Try as I might though, there was no way that aftermarket lifter block would go into the aftermarket Delkron case without one or the other having some metal removed!

Maybe those two examples don’t quite fit the technical definition of a tolerance stack up, but they give some idea of the adventures that await those who source their parts from multiple manufacturers in their efforts to keep old Harleys on the road.