There is a great deal of science involved in measuring things, but considerably less on what the baseline dimensions and clearances should be. This situation has not changed much since I wrote about it in one of my first columns for Gear Technology. Thanks to the online archive you can read that piece yourself. If you are not able to access it now, it included a recollection of a high school classmate having the engine in his Oldsmobile 442 “blueprinted” for drag racing. As most of us drove old cars we could barely afford, the concept of taking apart a running motor just because its parts might not be of “optimum” size was very shocking. I bring this up during a discussion on non-conforming parts to remind that lots and lots of equipment performs very well using parts that may not be quite “optimum” in some ways. Amongst quality professionals the very thought of adjusting tolerances to match the parts available is heresy worthy of a bonfire. In the custom manufacturing arena, however, it is a frequent topic of discussion. Or, as some characterize it — a hostage negotiation. There are many things I refused to budge on; you simply cannot risk using “soft” parts, allowing inadequate interference fits, or sloppy bearing or seal fits. On the other hand, how much backlash is too much? Will a step key prevent the gearbox from lasting 30 years? The answer, in the words of engineering sage Ray Drago, is that “it depends.” Your job as an engineer or designer is not to just say “no” when these situations arrive, but rather to explain how the tolerance was arrived at and how a revised tolerance will affect performance. Backlash deviations were requested in every gear shop I ever set foot in. Up in my ivory tower, this puzzled me. Didn’t the machines have dials calibrated in tenths of a thousandth? Why couldn’t those clowns make it like my drawing for once? Then I started spending more time in the shop and realized that working conditions, set-ups, tooling, and cutters were far from “optimum.” In the “green” operators had to sneak up on a pre-heat treat size over very slippery balls or an unwieldy caliper. On big parts more than one shift might be involved. Or the cutting tool had to be shifted or replaced. Then the wonder that is heat treat occurs. Or the re-centering/re-boring process happens and gear grinding has clean-up issues. Eventually, you must decide whether .008" of backlash against a .004/.006" spec requires starting the “lottery” all over. Don’t be that stereotypical “boffin” who refuses to leave his or her ivory tower. There is much to learn and great value to be achieved through engaging with the real world.