"Harold and Susan Vordos" wrote in
. net:

While it may be a bit of a stretch for you to come to terms with what
I'm about to say, when I was grinding I could measure with a tenths
mic and hit the size, *reliably*, within .000050". We had a
Sheffield shadowgraph at our disposal, which verified our readings.
Probably not very impressive in today's world, but a Sheffield was
cutting edge when I was in the shop.

While it _can_ be done, I fail to agree that it can be done reliably over
the long haul. You may be able to do it repeatably, but the day you are
out on vacation, Bob your coworker most likely won't be able to repeat
your success.

By today's quality standards, a gauge must have a resolution a minimum of
10x finer than the tolerance you are measuring. It _must_ long-form R & R
(Range & Repeatability) at less than 20% of your tolerance for general
work, and most all precision component customers require R & R of 10% or
less.
All of our work falls into the latter 10% category.

A long form R & R requires the range and repeatability of the gauge to be
measured, including the human factor, cumulatively all the way back to
the NIST standard.

It's a fairly easy task for a +/-0.001" tolerance, it's an entirely
different story for +/-0.00005" tolerance.

--
Anthony

You can't 'idiot proof' anything....every time you try, they just make
better idiots.

Remove sp to reply via email