Measurement System Analysis (MSA) is one of the five tools of quality management, why it is applied, and what statistical characteristics it has?
What is Measurement System Analysis (MSA)?
A measurement system is a collection of instruments or measuring tools, standards, operations, fixtures, software, personnel, environment, and assumptions used for quantitative measurement or qualitative evaluation of the measured characteristics. The whole process used to obtain the measurement results is called the measurement process or measurement system.
Measurement System Analysis refers to the category of variation analysis, that is, to analyze the size of the variation brought by the measurement system relative to the total variation of the process, so as to ensure that the main variation of the process comes from the process itself, not from the measurement system and that the measurement system capability can meet the process requirements. Measurement system analysis is aimed at the stability and accuracy of the whole measurement system. It needs to analyze the position variation and width variation of the measurement system. The position variation includes the bias, stability, and linearity of the measurement system. The width variation includes the repeatability and reproducibility of the measurement system.
When to Do MSA
- The newly produced products have large product variation
- When new equipment is introduced
- When the measuring operation is replaced with a new person
- The analysis frequency of vulnerable instruments needs to be noted
Purposes of Measurement System Analysis
Understand the measurement process, determine the total error in the measurement process, and evaluate the adequacy of the measurement system used in production and process control. MSA promotes understanding and improvement (reducing variation). We often analyze the process status, and process capability and monitor the process changes based on the measured data of the process processing parts How to ensure that the analysis results are correct? We must guarantee from two aspects:
1) Ensure the accuracy/quality of the measurement data, and use the measurement system analysis (MSA) method to evaluate the measurement system obtaining the measurement data.
2) Ensure that appropriate data analysis methods are used, such as SPC tools, experimental design, analysis of variance, regression analysis, etc. MSA uses mathematical statistics and charts to analyze the resolution and error of the measurement system, to evaluate whether the resolution and error of the measurement system are appropriate for the measured parameters, and to determine the main components of the measurement system error.
MSA Statistical Properties
- The measurement system must be under statistical control, which means that the variation in the measurement system can only be caused by common reasons rather than special reasons.
- The variation of the measuring system must be smaller than that of the manufacturing process.
- The variation should be less than the tolerance zone.
- The measurement accuracy shall be higher than the higher one of the process variation and tolerance zone. Generally speaking, the measurement accuracy is one-tenth of the higher one of the process variation and tolerance zone.
- The statistical properties of the measurement system may change with the change of the measured items. If so, the maximum variation of the measuring system shall be less than the smaller of the process variation and the tolerance.
MSA Evaluation Stages
The evaluation of measurement system analysis is generally divided into two stages:
Phase I: verify whether the measurement system meets the requirements of its design specifications.
1) Determine whether the measurement system has the required statistical characteristics, which must be carried out before use.
2) Find out which environmental factors have a significant impact on the measurement system, such as temperature, humidity, etc., to determine the space and environment for its use.
1) The purpose is to verify that once a measurement system is considered feasible, it should continue to have appropriate statistical characteristics.
2) Mainly determine the R&R of measuring tools. It is usually used as a part of calibration, maintenance, and measurement.