:argue:
What is Measurement System Analysis?
how can MSA philosphy can be benifical for manufacturing companies?
how can they start working on MSA?
What are different tools for MSA?:argue:


:thanx:
Ahsan

:argue:
What is Measurement System Analysis?
how can MSA philosphy can be benifical for manufacturing companies?
how can they start working on MSA?
What are different tools for MSA?:argue:


:thanx:
Ahsan

Hey Ahsan,

A lot of questions at the same time.
Most of them have already been discussed here at the cove.
Please use the search function and you will find a lot of info on the topic.
Maybe you better start to scroll down and look for the "similar threads" at the bottom of this page.
Some basic explanation:
In a manufacturing process you can see variation.
This variation is mainly due to the process itself but some of it is due to the measurement system.
MSA is a study to find how much variation is in the measurement system.
The study should be done in order to remove or avoid the variation or at least take it into consideration ( if it is impossible to remove )

In a manufacturing process you sometimes react to the variation that has been detected and if you don't know the contribution from the measurement system into the (visible) variation you may choose a wrong reaction plan.

The tools for attribute as well as variable studies have been discussed here as well - you will certainly find them in your search here.

Hope this helps, and if you have more specific questions after your search please let's hear it.

Best regards,

Antoine

MSA has to do with establishing that the guages you are using (verniers, micrometers and so on) are able to discriminate enough to be able to accurately measure what it is you are trying to measure.

It is a requirement of TS16949 automotive standards and the basic principle is sound particularly if you are manufacturing to very precise tolerances.

Otherwise it is not needed.

Lots of other things to worry about.

MSA is basically the methodology by which your fixtures, gages, and other variables that can influence measurment results are evaluated. MSA includes Gage R&R, Stability Studies, Linearity & Bias evaluation as well as Correlation Techniques.

A lot of people think that MSA is limited to Gage R&R which in and of itself has many iterations as a means to an end ; % of tolerance, % of Study variation, etc. etc. - Miner can best explain the statistical particulars and their relevance. There is a lot of planning and understanding of what the data is telling you in a good MSA program / evaluation. Remember, there are so many tools available, see#2 below - make sure you don't create a "head-spinner" for yourself and go overboard (I have seen this happen - 10 studies later, you arrive at the same conclusion that you would have 4 hrs. ago).

My advice is this - 1st, get the AIAG MSA "Handbook" if you can, it basically reviews all of the methodologies that I described above. 2nd, make yourself a flowchart or diagram that basically describes "Where are we going, and how are we getting there".

In other words, arrive at the conclusion of what studies or techniques you are going to apply in what situation. Of course, sometimes you have to wing it and go with what is appropriate, but, there are so many activities that you can plan ahead to incorporate prevention in your calibration system that it needs to be clearly organized and defined, or it can become a mess quickly. :agree1:

MSA has to do with establishing that the guages you are using (verniers, micrometers and so on) are able to discriminate enough to be able to accurately measure what it is you are trying to measure.

It is a requirement of TS16949 automotive standards and the basic principle is sound particularly if you are manufacturing to very precise tolerances.

Otherwise it is not needed.

Lots of other things to worry about.

Well, not exactly - it's more than simply discrimination.:yes: There are many factors that influence measurement and the precision of the gauge is one. What about the temp/humidity? The various operators, what's being measured, and so on.........

Plus, it's important not just for precision (close tolerance) stuff unless you've studied the results and found the results of measurement aren't important. True, ISO/TS 16949 requires it and ISO 9001 doesn't overtly state you must do the studies of the MSA, but that doesn't justify why someone should ignore it. That would be like saying don't do an FMEA or SPC or an 8D becasue it's not in ISO 9001! They are all useful, effective tools to benefit the organization.

Pudge's post contains good information........:agree1:

All quality systems require a gauge of some sort to qualify the product to specification. This gauge could be visual, attribute or variable gauge.

Measuring system is more than just the gauge. It includes the gauge, Operator, application of the gauge on the product and also the environment.

So how do we know that we have the best or optimum measuring system for its application. We should apply knowledge from MSA 3rd edition and perform a study on the particular gauge or its family (same type of gauges).

A gauge study, either variable or attribute, is more than calibration. Do we all read the gauge the same? Do we read the gauge the same if we measured the same part 3 times? What percentage of the process or specification ranges does the gauge take? Is it 10%, 20% or more. I have seen variable gauges that take over 100% of the tolerance. People were using that gauge to make final decisions on a part.

It really doesn't matter what type of industry one is in, a comprehensive gauge programme is mandatory with gauge studies on all families of gauges reflected in the Control Plan.

I hope this helps.

Layman terms to answer this question is we must identify and seperate measurement variation from process variation in the observed variation that we get daily.

Accuracy and precision are the two major measurement errors in the measurement variation.

From there we can breakdown to the other sources as per below

Accuaracy - bias, linearity, stability
Precision - repeatability and reproducibility

For GRR, the sandard acceptable criteria we used will be 30% and below which we compare to process variation or engineering specification, below 10% is the ideal case.

Common sense, the lower variation is better and we don't realy depend on the mutiply factor which we cover the variation distribution in 99% or 99.7% by using 5.15 or 6 in the calculation for GRR. These all are just an estimation. :bigwave:

Dear Andy
you are right that iso 9001 does not say for PFMEA, MSA and use of SPC tools but i think you shoud include them in your organizations Quality Management System.

Thanks
Ahsan

I'm happy you agree............:agree1:

I'm not able to understand how will you perform Gauge R&R in case Appraiser has no affect on measurement system?

I have 16 test equipments, 10 parts & each part is measured 3 times.

In this entire setup, although Appraiser are involved, but since the equipments are electronic equipment & the appraiser has to just put the part in the equipment & the equipment measures it, hence the appraiser has no role to play no matter how many appraisers do the same thing.

Please help
Please guide
Anurodh.

I'm not able to understand how will you perform Gauge R&R in case Appraiser has no affect on measurement system?

I have 16 test equipments, 10 parts & each part is measured 3 times.

In this entire setup, although Appraiser are involved, but since the equipments are electronic equipment & the appraiser has to just put the part in the equipment & the equipment measures it, hence the appraiser has no role to play no matter how many appraisers do the same thing.

Please help
Please guide
Anurodh.

Anurodh,

You've asked essentially the same question now in five different places. Why not stick to one thread, such as this one (http://elsmar.com/Forums/showthread.php?t=27982)? With a little patience I'm sure your questions will be answered.