Type 2 and Type 3 (commonly referred to as GRR study) are repeat measurements on real test objects. The known variables repeatability and comparative precision are determined in the standard (many other parameters are still possible).
The difference between Type 3 and Type 2 is that “user influence” is excluded and therefore only repeat measurements are performed. But when can this be ruled out? An expert statement on the topic is as follows: If a robot removes the components from production and inserts them into the measuring equipment, then one can assume a method 3. If a tester touches the component, it must be proven that the influence of the user does not matter! It is therefore the responsibility of the quality management to make the decision in consultation with the customers.
Accordingly, the following information is required in the standard:
The specification limits, the resolution and the actual value of the normal, as well as the repeated measurements
Specification limits:
These are the specification limits of the characteristic. However, the MSA also allows other references, such as process variation or part variation. If the strategy is designed for this, then the respective references must be specified.
Resolution:
The resolution is the smallest readable change in the display value. With a classic caliper with digital display, it is the smallest digit that can change. For measuring machines whose values are calculated from “point clouds”, there is a discussion as to whether the artificial resolution (which only results from the reduction in output) should be taken into account.
Measured values:
During data acquisition, care should be taken to ensure that the reference part does not remain in the measuring device but is removed from the measuring device and reinserted before each measurement.
For this reason, a preferred measuring sequence should be the following: First the first operator measures all components, then the second, the third. This is followed by the first operator, etc.
This forces the removal of each component after each measurement.
Sample taking is also important. As in this example, the components should not all be in the middle of the tolerance:
This is because it would not be possible to make a statement about the ranges in the direction of the specification limits.
As far as possible, the parts used should cover the tolerance range and may, of course, exceed it. This is because it is not the quality of a test specimen that is assessed, but rather how precisely and repeatedly the value can be measured in a comparable manner.
This “Part variation” becomes important later on when the NDC is evaluated, or the part dispersion is the reference variable.
Basic Analysis
Main analyze is done with Form 3 and different charts. Form 3 is different depending on the MSA-Version and the calculation method, the activated requirements in the line where the "Smiley" is shown.
