I vaguely recall a general convention used for dimensional inspection that specified that inspection equipment "accuracy" must exceed the dimension's tolerance zone by a factor of 4, at minimum. For example, if the tolerance of part dimension is +-.004, then the equipment used to inspect the attribute must be accurate to at least +-.001, if not tighter. Or, was it 4x the "resolution"?

Does this make sense? Where did this originate? Would this be an acceptable substitute to use when R&R work hasn't been done?

Any input would be appreciated.

- TK :confused:

It is in fact a TAR (Test Accuracy Ratio) per ANSI/NCSL Z540-1-1994 and before that per MIL-STD-45662A. You are correct in your interpretation of how to figure it.....

If you use the TAR for uncertainties however, you must use .25 of the rated accuracy of the device at the specific reading, as opposed to an uncertainty number which may well be less.

Hope this helps.

Hershal