I also have a slight concern. If the tolerance is the total band width within which deviation is permisable, why would you take the absolute maximum deviation and multiply it by two. If you were at +0.6 and -0.1 on a 1.0 profile your actual band width would be 0.7 correct? Back to what Jeff is saying wouldn't this method also show it to be acceptable when in fact the positive deviation exceeds the allowable band width by 0.1. If you multiply it it will show it to fail, yet by more than it actually does. I still think you are better off reporting it as the maximum and minimum deviation. Have we found the correct answer or have we just managed to confuse each other into submission. In the same token I absolutely agree that the correct amount of information depends on whom the report is for.

Best regards,
Terry W. Prather