Basic Text MSA (Measurement Systems Analysis) Procedure

1.0 Purpose: To describe a step by step process for the evaluation of measurement systems.

2.0 References:
2.1 Measurement Systems Analysis (MSA) Manual Second edition
2.2 ASQ Automotive Statistical Process Control Manual 1986
2.3 Production Part Approval Process Manual First Edition
2.4 14-021-001 Corrective action procedures for quality control
2.5 02-000-001 PPAP procedure
2.6 QA-16 MSA data sheet

3.0 Definitions:
3.1 Bias-The difference between the observed value and the reference value.
3.2 Repeatability- the variation in measurements obtained with one measurement instrument when used several times by one appraiser while measuring the identical characteristic on the same part.
3.3 Reproducibility- the variation in the average of the measurements made by different appraisers using the same measuring instrument when measuring the identical characteristic on the same part
3.4 Stability-the total variation in the measurements obtained with a measurement system on the same parts when measuring a single characteristic over an extended time period.
3.5 Linearity- the difference in the bias values through the expected operating range of the gage.
3.6 Discrimination(resolution) - the ability of the measurement system to detect and indicate small changes in the characteristic.
3.7 Gage- generic term for all Measuring and Test Equipment

4.0 Responsibility
4.1 Process and Design engineers identify what equipment needs to be evaluated. All gages in the control plan are evaluated using this procedure and the MSA.
4.2 The quality control manager is responsible for choosing trained individuals, evaluation the results, choosing the samples, and providing an environment consistent with MSA requirements. In addition, the quality control manager is responsible for the overall integrity of the study.

5.0 Safety/Environmental Instructions:
5.1 Safety glasses and shoes must be worn at all times during testing.
5.2 Obtain MSDS sheet for disposal, storage, and safety instuctions for any chemicals used to perform the MSA

6.0 Equipment/Tooling:
6.1 All reference standards must be NIST Traceable
6.2 Equipment as described in customer control plan

7.0 Procedure:
7.1 There shall be procedures for the use of all gages in the study
7.2 There shall be documented training records to ensure that all appraisers are trained on the proper use of the gages in the study.

7.3 Ensure the gage has the correct discrimination. The resolution must be at least one-tenth of the six sigma value of the tolerance.
7.3.1 For example if the tolerance is .0031” to .0033” then the resolution of the gage must be capable of reading in increments of .000002”.
7.3.2 Use consistent units of measure. Avoid conversions (i.e. Inches to millimeters)
7.4 Ensure the stability of the measurement system, taking into account:
7.4.1 The environment including:
7.4.1.1 Is the measuring system influenced by temperature/humidity fluctuations(i.e. expansion coefficients for gage blocks/calipers)?
7.4.2 Does the system need to be cleaned on a regular basis?
7.4.3 Is the system subject to wear over time?
7.4.4 Does the system need a “warm-up” period (for electronic gages)?
7.4.5 Consult owner’s manual or procedure for additional sources that could affect stability.
7.5 AISI will conduct long term stability studies at the discretion of the quality assurance manager.
7.5.1 The QA Manger will consider:
7.5.1.1 Cost of the study
7.5.1.2 Customer requirements
7.5.1.3 Time to perform the study
7.5.1.4 The availability of other gages
7.6 Determine the bias.
7.6.1 Measure one sample on the NIST reference instrument.
7.6.2 Measure the same sample 10 times using the gage being evaluated.
7.6.3 Calculate the average of the 10 readings.
7.6.4 Bias = observed average - Reference Value. This means that the observed measurements will be on the average +/- the bias of the reference value.

7.7 Convert the Bias to a percentage of the tolerance by multiplying 100 and dividing by the tolerance. If the bias is large look for:
7.7.1 Error in the master
7.7.2 Worn Components
7.7.3 Instrument made to the wrong dimension
7.7.4 Instrument measuring the wrong characteristic
7.7.5 Instrument not calibrated
7.7.6 Instrument used improperly

7.7 Determine the linearity
7.7.1 determine the range of use of the measurement system.
7.7.1.2 For example a scale may have the capability of taking measurements from 0 to 100 pounds, but Armstrong only uses it from 50 to 70 pounds. Linearity must be determined in the range of 50 to 70 pounds.
7.7.2 perform 7.6.1 - 7.6.2 exempt repeat the process at the low end of the range, mid range and upper range.
7.7.3 Use formula in 7.6.4 but substitute Linearity for bias and repeat for all three measurements.

7.8 If using an automated gage, or user influence is not a factor the reproducibility does not need to be performed.

7.9 Samples are chosen that represent the entire operating rage of the gage in the study.
7.9.1 Each appraiser measures the same part of the same sample. Where the measurement is supposed to be taking place is indicated on the sample to ensure all operators are measuring the same part.
7.9 Determine gage repeatability and reproducibility using three appraisers and record information on gage R&R worksheet.
7.9.1 If Gage R&R greater than 30% the fill out an NCR in accordance with 14-021-001
7.9.2 If Gage R&R is between 11-30% fill out NCR
7.9.3 If Gage R&R is 10% or less the gage is acceptable

7.10 If repeatability is large compared to reproducibility look for:
7.10.1 The gage may need maintenance
7.10.2 The gage may be redesigned to be more rigid
7.10.3 The clamping or location of the gage needs to be improved

7.11 If reproducibility is large as compared to repeatability look for:
7.11.1 Operator training
7.11.2 Accuracy of procedures
7.11.3 Calibrations on the gage are not clear

8.0 Statistical Techniques/calculations:
8.1 See procedure 11-000-010 for selection and interpretation of control charts
8.2 Acceptance criteria for gage R&R is taken from ASQ Automotive Statistical Process Control Manual, and Production Part Approval Process Manual.

9.0 Records

9.1 All information is recorded on QA-16. MSA studies are kept in the PPAP file in accordance with procedure 02-000-001.

Here is the MSA procedure that my company uses sans the header and other cosmetics.

[Marc]

Thanks for posting it. I certainly appreciate it! I'm sure others will/do as well!

First I want to thank everyone for the positive comments on the attached procedure. There are many times that the company you are working for takes your hardwork for granted, and it's nice to get a pat on the back from my peers once in a while to validate my effort. Since many people indicated that they are using this procedure to model their own MSA procedure feel like I must inform people as I make changes to it. In the most recent revision I have changed the "purpose" section to indicate that this procedure is only used for variable gages, it does not include attribute gages since my company does not use any. Section 7.3 was changed to update the calculation of resolution. Instead of the resolution being one tenth of the tolerance, the MSA manul recommends that resoultion be one-tenth of the six sigma value of the control limits. Once again thanks for the support!!!!!

[Marc]

Pat Pat Pat!

Thanks! You are appreciated here!

[kite005]

Thank you Christian I will probably use this as a basis for my own when I get to that point in my lab. Thank you again.

[gszekely]

Thanks Christian, verry useful. My coments:
1)
4.2 The quality control manager is responsible for choosing trained individuals,
7.2 There shall be documented training records to ensure that all appraisers are trained on the proper use of the gages in the study
I use to include in MSa process one professional technician, or myself, and at least two shop floor operators. Also is important for me that those involved, at least the shop floor operators, not to know that they perform a study. Usually I gave the task, to them as a daily measurement. The part indentification is not known, by them, and the measurements are done randomly.
2)
7.7.1.2 For example a scale may have the capability of taking measurements from 0 to 100 pounds, but Armstrong only uses it from 50 to 70 pounds. Linearity must be determined in the range of 50 to 70 pounds
7.9 Samples are chosen that represent the entire operating rage of the gage in the study.
While you determine linearity for a restricted range, you want to include in the analyses samples from the entire operating range of the gage.
Is this OK ?
3)
7.4.1.1 Is the measuring system influenced by temperature/humidity fluctuations(i.e. expansion coefficients for gage blocks/calipers)?
If yes, where, can I find further instructions ?
4)
7.9 Determine gage repeatability and reproducibility using three appraisers and record information on gage R&R worksheet.
How many samples to be used, or it is defined separatels for every gage, project, etc ? How ?
5)
7.9.1 Each appraiser measures the same part of the same sample. Where the measurement is supposed to be taking place is indicated on the sample to ensure all operators are measuring the same part.
I would use characteristics, instead of part.
BTW: if my equipment is calibrated, is it necsary to perform, linearity and stability study, every time I perform a study ?
Thanks for the procedure, I don't have a written one.
Gyorgy

[bmccabe]

Thank you, Christian.
I'm a little reluctant to post the one I wrote.
Some folks here might view that as an unauthorized disclosure of proprietary materials. I'm glad you did!

The procedure I wrote is 6 pages, and mostly in the form of a flow chart.
It is more specific to automated inspection, but in general follows yours.

May I ask a couple questions?

In 7.8 you suggest an operator has no influence on automated measurement equipment.
I agree with you, but not everyone does.
Q ---How do you prove this to an auditor (or doubter)?

In 7.9, you detail a procedure for a GR&R.
Q ---Why include this in the MSA procedure?

The GR&R assumes operator variability exists, and most certainly will attribute a substantial portion of the total variability in this index.
Q ---How do you interpret the reproducibility index for automated inspection?

Q ---Do you make any alterations to the GR&R procedure to accommodate automated inspection equipment?
I have a thread on this subject. Please check it out.
http://elsmar.com/Forums/showthread.php?t=14511


r/b