What is MSA ?
Data analysis can be largely affected by measurement system error. Measuring instruments, inspection techniques of technicians and fixturing all make up a measuring system. Any of these components can introduce variation in the measuring system. This variation in the measuring system is then reflected as part of total process variation in the SPC analysis and negatively affects key quality statistics like Cp, Cpk, Pp, Ppk, etc.
Every process contains inherent variation. The sources of variation can be related to any of the inputs from the mechanics of the process itself to raw material or personnel consistency. However, it is possible that the variation in the measurement process is a significant contributor to the overall variation of a process or is causing special causes of variation. This variation is analyzed during the MSA process. MSA is an essential step in any quality control application.
There are multiple type of variation:
- Repeatability – within an operator or piece of equipment
- Reproducibility – operator to operator or attribute gage to attribute gage
- Stability – accuracy over time
- Linearity- accuracy throughout the measurement range
- Bias – Off-set from true value
During Gage R&R studies the repeatability and reproducibility are analyzed.
Does a measuring system report the same result when the same master is “blindly” measured several times by the same person? Can one operator be consistent in measuring the same part/characteristic using the same gage? Repeatability is also referred to as equipment variation, because it often reflects on equipment design or condition.
Does a measuring system report the same results when two or three operators measure the same part/ characteristic “blindly” several times? Between-operator reproducibility is the other dimension of variation in a measuring process. Gage reproducibility is also referred to as appraiser variation, because it most significantly reflects on operators’ consistency in using the measurement system. Solutions to problems of reproducibility often involve operator training. Gage R&R studies are usually made up of two components: equipment variation and appraiser variation. These components serve to focus any remedial attention on repairing or improving the gage itself (EV-equipment variation) or training operators to use the gage correctly or more consistently (AV-appraiser variation).
Either total tolerance (USL-LSL) or total process variation (6 sigma from long range SPC data) or total study variation (6 sigma of data collected in the gage study only) can be used for calculating the proportion used in measurement error.
Guidelines for acceptance of gage repeatability and reproducibility (%R&R):
Under 10% error: measurement is acceptable
10% to 30% error: measurement system may be acceptable depending on application importance
cost of the gage, cost of repairs or improvement and existing process capability.
Over 30% error: measurement system needs improvement (depending on process capability);
make every effort to identify problem and correct them.
(Source AIAG – MSA Manual)