The correct way to do a Linearity and Stability Study on a Critical Dimension

[Priss80]

The customer request to conduct bias, linearity and stability for the special characteristic. I conduct bias, linearity and stability study by using production part. I use the part as master and pick 3 measurement as master and periodically measure the same part and compare to the master. Is this the correct way i conduct the study? Some of the people buy a master block to conduct the study which to me is not measuring to the customer special characteristic. Use production part or master block is more appropriate? If we use master block we will add calibration cost, isn't it?
Please help i am a bit confuse

 

[t.PoN]

Re: Linearity and stability Study

Using the production part:
How do you its true value or reference value?

if you have a good source for estimating the reference value of your production part, then I think its ok to use it!

 

[Priss80]

The referance value is measured by more than one same type of calibrated measuring equipment to confirm part dimension is correct.

 

[t.PoN]

As long as you have traceability of your measurment then I can't find a reason why not!

Your uncertainity will increase but that's all

 

[t.PoN]

I just want to add this:
Make sure that you the goal of study is clearly defined!

i say this because re-reading your post I was not sure if you are having the study for your measurment method or your products!

if the goal is the bias of the measurment method then using master block or tracebale product part is ok

but if the goal is to measure the stability of your production, the you should conduct the measurment using your product part not your master block! your master block can be used to determine the bias for the method in this case

 

[Priss80]

Referring to the aiag MSA manual, only mention to take from production part. Can I understand that the manual is considering to measure the product stability instead of equipment or method?

 

[t.PoN]

The MSA is targeting the industrial world.

The measurement system is usually inetrrupted as the production process as you can see in the attached image.

but yes, the MSA usually talks about the stability of your production process (including the equipment used to manfacture the products).

It can be used as well to measure the stability of your actual measurment/test/analysis method (including the equipment used to capture the data (such as a caliper).


I think your customer is intersted in your production and not your measurment/analysis method.

 

[smkeilman]

I think you may be misinterpreting what is being asked. The bias and linearity are more associated with measurement system analysis than manufacturing process analysis. Stability could apply to either. If you're being asked for a straight MSA on this characteristic, then the only relevance to the production part is on WHICH measurement system you choose. The MSA is performed on the measurement system in a vacuum, not the production process.

 

[Priss80]

For stability study, i use production part to perform MSA, i need to identify the specification of the characteristic but what is the tolerance i should put? exm: 2.50+/-__?__
If i use master block to perform stability, what should the specification be?