TS 16949 - AIAG's MSA Manual - Thread Inspection Gage - Attribute Gage R&R

[theace]

Hello all!

I´m fairly new to TS and MSA. We are trying to do MSA for thread inspection with gage (go-nogo).I downloaded Excel charts some of you have made for attribute test from this forum.

I read attribute test from MSA manual 3rd edition and it differs little from those dowloaded ones. In MSA manual they use this kappa measure and use 50 samples and 3 trials and so on. Do you have to do this test exactly according to MSA manual in order to satisfy auditors or is it enough if you do it like in these excel programs?

Regards to all

Marko

 

[D.Scott]

Welcome to the Cove Marko.

I believe the Excel programs found here would be satisfactory. The requirement is to do an MSA, but doesn't require a certain one is followed. As long as the study is statistically sound I don't think an auditor would have a problem with it. There are many commercial and privately written programs for MSA and I doubt that any are exactly alike.

Dave

 

[howste]

TS 16949 says in 7.6.1:

Statistical studies shall be conducted to analyse the variation present in the results of each type of measuring and test equipment system. This requirement shall apply to measurement systems referenced in the control plan. The analytical methods and acceptance criteria used shall conform to those in customer reference manuals on measurement systems analysis. Other analytical methods and acceptance criteria may be used if approved by the customer.

To me, that says you should follow the customer's manual unless the customer approves an alternate method. An auditor would have to be pretty thorough to notice the difference though...

 

[Atul Khandekar]

Welcome to the Cove, Marko!

Dave is right that there is no 'strict' requirement of 50 parts - more the merrier, though . IMO, it depends on what you want to achieve from the study.

If it's a CrossTab study, where we are trying to determine the interrator agreement, you should select parts where all operators agree, and should also include parts near spec limits, ie in the zone of possible disagreement.

If you are using the Signal Detection method, where we need to identify and quantify the width of the "Gray zone" of the gage and %R&R, you need to select parts that cover the entire spectrum:
1. Unacceptable parts on both sides of the spec limits (the '-' zone on BOTH sides - rejected by all apraisers),
2. Parts from the 'Gray Zone' (near the spec limits where appraisers are likely to have some disagreement) and
3. Parts within specs (accepted by all appraisers)

IMO, You can start your study with maybe even 20 parts and go on adding parts till you get satisfactory coverage. I have seen cases where customers recommended a minimum of 20 or 30 parts - you can also ask your customer.

For Analytical method however, the MSA manual is much more prescriptive in terms of number of parts and method of parts selection.