New MSA Reference Manual: More Clarity, Better Organization
Here's a quick review of improvements made to the newly released Fourth Edition of the MSA Reference Manual.
– By Michael Down
Data and measurement impact all aspects of every industry. Manufacturing, assembly, retail, services, construction, labs, etc., all require review and analysis of data, followed by action that is taken on the basis of that information.
Although it is normally assumed that a measurement device is accurate, reliable, and giving high quality data, in many cases this is not only untrue, but many of the perceived problems may be due to the measurement system itself.
Changes to the MSA reference manual
Several editions of the Measurement System Analysis (MSA) Reference Manual have been released. In general, all issues that were identified by customers over the last few years are addressed in this Fourth Edition. The major areas of change are:
Additional examples to aid in the understanding of basic measurement system fundamentals;
Improvements to the organization of the manual and the table of contents;
Clarification of the relationship between calibration and MSA;
Clearer definition of the measurement decision (5-10-30);
Improved the bias and linearity sections;
Clarified and improved the attribute and the non-replicable (i.e., destructive) testing section.
These changes have helped to make the manual an asset in the understanding of measurement systems and the use of those systems to improve quality and services. The new version of the manual is now available at AIAG.
Purpose of MSA
The Measurement System Analysis (MSA) manual, one of the GM, Ford, and Chrysler reference manuals offered through the AIAG, is a reference manual wherein methodologies and principles are recommended, but not mandated. OEM's expect organizations will meet the intent of MSA in relation to measurement systems that are used in the manufacturing and assembly of supplier parts and subsystems.
Organizational improvements
One of the most important roles of a reference document is to provide a clear understanding of the subject. Improvements have been made in the usability of the MSA manual. For example, the table of contents is reformatted and structured to improve the reader's ability to find a specific subject. Some of chapters have been moved and combined to aid in the flow of material.
Relationship between Calibration and MSA
Another key improvement is the addition of information showing the relationship between the measurement system on the floor and how it tracks back to the National Measurement Institute Standards. This standardization helps to connect all who measure to a single source and reduces variation which could cause measurement errors. Measure Assurance Programs or MAPs, which keep measurement devices calibrated to the standards, are also discussed.
In general, several elements are present that can impact measurement results. For example, when taking blood pressure, the result is affected by the operator taking the measurement, the device itself, the patient, the environment, etc. The calibration method is used to verify that a measurement device can measure the characteristic within the specified design range. Normally, a "check" standard or "reference" value is used to validate the measurement device on a regular or annual basis.
5-10-30 Rule
The section of the manual discussing how to determine the minimum acceptable GRR (Gage Repeatability and Reproducibility, a statistical measure of the precision of a measurement system), has been expanded and includes clear guidelines for the decision making process. The four areas covered are:
Assembly error
Fixture error (should always be checked prior to any changes or improvements to the measurement system)
Location error
Width error (which includes the following table)
GRR Decision Comments
Under 10 percent Generally considered to be an acceptable measurement system. Recommended; especially useful when trying to sort or classify parts or when tightened process control is required.
10 percent to 30 percent May be acceptable for some applications Decision should be based upon, for example, importance of application measurement, cost of measurement device, cost of rework or repair.
Should be approved by the customer.
Over 30 percent Considered to be unacceptable Every effort should be made to improve the measurement system. This condition may be addressed by the use of an appropriate measurement strategy, for example, using the average result of several readings of the same part characteristic in order to reduce final measurement variation.
It is also pointed out that this information should not be used to establish threshold values for GRR. Each measurement situation requires review based on its own applications and merits.
Improvement to Bias and Linearity Sections
Some basic improvements have been made to the Bias and Linearity section of the manual. Calculations have been updated to use standard deviation instead of range, which was used in the past because the calculations were done manually. Since computers are used most of the time, standard deviation is a more robust value.
As a reminder:
Bias: the difference between the true value (reference value) and the observed average of measurements on the same characteristic on the same part.
Linearity: The difference of bias throughout the expected operating (measurement) range of the equipment is called linearity. Linearity can be thought of as a change of bias with respect to size. This is critical to understand in relation to linearity. Linearity can be thought of as a change of bias with respect to size.
Bias and linearity both include an evaluation or repeatability. The bias and linearity are useless if the repeatability is not acceptable.
GRR and the Target Pp
The explanation of how the target Pp (process performance index) is used has been clarified in this new release of the MSA manual. There are four different approaches to determine the process variation for measure acceptability:
Process Variation (Preferred Method) – Process variation takes on the parts in the GRR Study itself. Use when the selected sample represents the expected process variation. This would be a planned sample, instead of a random sample, to make sure the process variation is represented in the sample.
Surrogate Process Variation – When sufficient samples from the targeted process are not available, but samples can be taken from an existing process with similar variation.
Specification Tolerance – This method can be used when the measurement system is to be used to sort the process and the process has a Pp less then 1.0.
Pp (or Ppk) Target Value – When a sufficient sample to represent, the process variation is not available and an existing process variation is not available, or the new process is expected to have less variation then an existing process.
Improvements to the Attribute Section
Improvements have been made to the attribute section to help in its application, including clarification and additional examples on the use of this methodology:
Definition of cross tabs is improved, with examples of their use in the hypothesis test method.
Sample size for attribute gauging analysis has been clarified.
In the signal detection approach, the table creation was clarified, its origin identified, and step-by-step instructions with enhanced tables and graphics provided.
Improvements to the Non-Replicable section
Time was spent improving the information and locating the analysis for non-replicable measurement, or the case when multiple measurements cannot be taken on the same part or it is impossible to measure the same part or system multiple times. In addition, the index in the front has been improved and multiple gauging examples are identified.
Conclusion
The committees hope the improvements to the manual will aid in its use and application in the field.