# Tolerance Width for a Measurement System Adequacy

Six Sigma – iSixSigma Forums General Forums Methodology Tolerance Width for a Measurement System Adequacy

Viewing 4 posts - 1 through 4 (of 4 total)
• Author
Posts
• #54356

Hector
Guest

In order to perform a Test-Retest Study or a GageR&R you need to apply the Tolerance, which is defined as “the difference between the Upper Specification Limit and the Lower Specification Limit”.

I have doubts when the precision of the measurement system is with 1 decimal, for example: [95.0 , 105.0]

Would the Tolerance be 10 or 100 since potential values are 98.2, 99.4, 102.1 …?

0
#195029

Chris Seider
Participant

The customer must have upper and lower specs that you consider when making the process/product. Customers don’t define their needs depending on the gage you might have or we all would just use the simplest devices/techniques and make the supplier’s life easier. Tolerance is upper spec minus lower spec.

The point of looking at %tolerance is to find how much the total variation of the gauge is to the width of the specification range and you typically want that ratio no more than 30% with the ideal being less than 10%.

0
#195030

MBBinWI
Participant

Hector – Given your specs being 95.0 and 105.0 that would make your tolerance 10.0

As @cseider identifies, the tolerance does not have anything to do with the precision of your measurement device. That said, however, it is often possible to improve an incapable measurement system with a more precise tool. That is, if the process of measurement is stable and under control. Good luck.

0
#195069

Kicab
Guest

To expand on Chris’ point of <30% marginal and <10% great, you want your precision to be <10% of the tolerance. Since the tolerance is 10 as MBBinWI noted, then a precision of 1 could be sufficient. Yours is 0.1 so that's even better.

0
Viewing 4 posts - 1 through 4 (of 4 total)

You must be logged in to reply to this topic.