Failure to Measure Conformity of Product

[QATN11]

I am faced with a similar issue, but not necessarily with just mill tolerances. One customer has drawings for parts dimensioned in fractions. Stated tolerance is usually +/- 1/32 and sometimes +/- 1/16. Within the same drawing, other dimensions are to two decimal places and other times to three decimal places.

It has been my understanding that the dimension rules the precision of the measurement device. i.e. for a dimension with a 1/16" tolerance, it is permissible to use a calibrated tape or rule incremented in sixteenths of an inch. If the dimension is written as a decimal fraction with a 2 place decimal, the appropriate device must have precision to read at least to the third decimal place and standard rounding rules apply and so on to the stated decimal precision.

My issue is the customer's Incoming QA Inspector reads every length dimension to 3 decimal places regardless of the stated units and tolerances. This includes dimensions indicated a "REF" which should have no implied tolerance. Mill tolerances are generally governed by the appropriate ASTM specification which almost never go to the third decimal place.

My position has been measure to the drawing or redraw and restate the tolerances needed, then I can tell you if the tolerances can be met.

I don't want to win the battle and lose the customer, not do I want to add cost chasing an unnecessarily tight tolerance.

 

[Jim Wynne]

I am faced with a similar issue, but not necessarily with just mill tolerances. One customer has drawings for parts dimensioned in fractions. Stated tolerance is usually +/- 1/32 and sometimes +/- 1/16. Within the same drawing, other dimensions are to two decimal places and other times to three decimal places.

It has been my understanding that the dimension rules the precision of the measurement device. i.e. for a dimension with a 1/16" tolerance, it is permissible to use a calibrated tape or rule incremented in sixteenths of an inch. If the dimension is written as a decimal fraction with a 2 place decimal, the appropriate device must have precision to read at least to the third decimal place and standard rounding rules apply and so on to the stated decimal precision.

My issue is the customer's Incoming QA Inspector reads every length dimension to 3 decimal places regardless of the stated units and tolerances. This includes dimensions indicated a "REF" which should have no implied tolerance. Mill tolerances are generally governed by the appropriate ASTM specification which almost never go to the third decimal place.

My position has been measure to the drawing or redraw and restate the tolerances needed, then I can tell you if the tolerances can be met.

I don't want to win the battle and lose the customer, not do I want to add cost chasing an unnecessarily tight tolerance.

Which is greater, 1/16 or or .0625? They are equal, and regardless of the method used to express the value, they will always be equal. By the same token, .0626 is greater than .0625, regardless of how the values are expressed. If you have a specification that says, for example. 1/16" max, and you or your customer measures it at .0626, the tolerance has been exceeded. The option of any kind of rounding must be explicitly stated, and in the absence of an explicit statement it must be assumed that if the absolute value of the tolerance has been exceeded, a nonconformity exists.

Now, having said that, how you deal with these things is a different thing, and we have to hope that in the end, pragmatism and fitness for use will win the day. If you have a dogmatic customer who only cares about whether one number is bigger than another and not about whether the thing in question is usable, you're probably going to lose the argument.

 

[Rickser]

Not to be sarcastic, but I would think the Customer QA Insp needs retraining, especially if he/she is rejecting items based on his/her own interpretation of specifications. If that is the case, I would have an eyeball to eyeball discussion with my customer management and come to a resolution. However, sometimes you may have to "lose" the customer if they become unreasonable. We just did and most of the people there were/are personal friends of many of us. It was hard but they just refused to negotiate and took the attitude that we had to make them happy. Sorry about that.