Is conducting of all MSA studies required, even if not required by the customer?

[antonjav]

Re: Is conducting of all MSA studies required, even if not required by customer?

Hi Amit, for sure it is unlikey that you would be able to get an exemption in fulfilling the requirements of 7.6.1. TS 16949 does not specify any maximum frequency for doing MSA nor does it suggest one. This is because the standard only states the "What is required" portion and not the methodology for achieving the requirement.

In several engineering organizations, the maximum frequency is at times determined by the amount of variation present at the time of the MSA study. This means that if you do MSA once a year but then you find out that the GRR (or linearity and bias) results yield unfavorable results, like kappa values are close to zero or that the components of variation are larger than the part-to-part variation, then you are constrained to adjust your MSA frequency to 6 months instead.

As to which MSA study would be compulsory, again TS 16949 would not prescribe. It all depends on the measurement type that you would do an MSA on, including its intended use. For example, if you would do a MSA on a caliper, it is expected that you would do GRR, linearity, and bias studies. However, if the caliper would not measure the different sizes except for a fixed value, then you may exclude linearity in the study. You would of course include bias and GRR.

Hope that this helps.

 

[AMIT BALLAL]

Re: Is conducting of all MSA studies required, even if not required by customer?

Hi Amit, for sure it is unlikey that you would be able to get an exemption in fulfilling the requirements of 7.6.1. TS 16949 does not specify any maximum frequency for doing MSA nor does it suggest one. This is because the standard only states the "What is required" portion and not the methodology for achieving the requirement.

In several engineering organizations, the maximum frequency is at times determined by the amount of variation present at the time of the MSA study. This means that if you do MSA once a year but then you find out that the GRR (or linearity and bias) results yield unfavorable results, like kappa values are close to zero or that the components of variation are larger than the part-to-part variation, then you are constrained to adjust your MSA frequency to 6 months instead.

As to which MSA study would be compulsory, again TS 16949 would not prescribe. It all depends on the measurement type that you would do an MSA on, including its intended use. For example, if you would do a MSA on a caliper, it is expected that you would do GRR, linearity, and bias studies. However, if the caliper would not measure the different sizes except for a fixed value, then you may exclude linearity in the study. You would of course include bias and GRR.

Hope that this helps.

Thanks for help & suggestions.
I would like to ask one more question:
Should I conduct Stability studies ? And if yes, for which measurement techniques ?


Thanks in advance,
Amit

 

[antonjav]

Re: Is conducting of all MSA studies required, even if not required by customer?

Thanks for help & suggestions.
I would like to ask one more question:
Should I conduct Stability studies ? And if yes, for which measurement techniques ?


Thanks in advance,
Amit

Again, the TS standard will not require whether you shall perform a stability study or not. It would be your engineering or QA group responsible for MSA who would decide on whether stability shall be performed or not.

Of course, during audit, your organization should be able to justify why no stability study is done when it seems that one is warranted. If you can't justify a no stability study, then you might as well consider doing one - provided further that the concerned measuring or test equipment is referenced in you control plan.

Hope this helps Amit. Thanks.

 

[antonjav]

Re: Is conducting of all MSA studies required, even if not required by customer?

... on the business side, if you are so constrained with current resources and that doing a stability study would make your staff nosebleed, then you could take some risk and not do stability study at this time. The worse thing that could happen is that a minor nonconformity report (or observation only if there's no big lapses in other TS requirements) would be issued to the organization responsible; you are given time to respond and close out.

Lastly, there are other process controls in place that would capture adventurous quality excursions, yes?

Thanks.