Time For Some R & R
(Gage R & R, that is!)
Have you ever noticed that your weight at home is sometimes different that your weight at the gym or doctors’ office? This is mostly due to variation in the scales, or gages of the measurement (and only a little bit because of how big your breakfast was or how heavy your clothes are). Measurement errors are inherent in any measurement process and device. If you’re measuring parts to ensure they conform to required specifications but your gage or measurement process vary too much, you might pass parts that should fail and vice versa. If you can’t trust your measurement system, you can’t trust the data is produces. To make sure that your customer gets what they want, you need to make certain that your measurements are accurate. You do this by conducting a Gage R & R study.
Gage R & R, or Gage Repeatability and Reproducibility, focuses on the precision of a measurement system by measuring the variability induced from the system itself and comparing it to the total variability observed in order to determine the validity of the measurement system. Measurement systems contain variation from three main sources: the products themselves, the operator taking the measurements (reproducibility) and the equipment used to perform the measurement (repeatability). The purpose of conducting a Gage R & R study is to be able to distinguish the source of the variation.
Let’s break this down: Gage Repeatability is the ability of the measuring device to provide consistent results. It is the variation obtained from one gage and one operator when measuring the same part several times. Gage Reproducibility is the variation induced when different operators measure the same part using the same gage, otherwise known as operator-to-operator variation. Repeatability and reproducibility together are called “measurement error”. What you hope to see in a Gage R & R study is that most of the variation is caused by variation between the parts and very little of the variation is caused by the operators or equipment .
To conduct a basic Gage R & R study it is recommended to have 10 parts from one batch or lot, 3 appraisers (the people who measure the parts), 1 measurement tool (gage), and three measurement trials on each part by each appraiser. The results are then charted or graphed for analysis. The generally accepted variation is 10%. This indicates that if a measurement instrument exceeds 10% of the part tolerance, the measurement instrument has no ability to measure the process variation. If the variation is less than 10%, the measurement system is excellent. The study separates the variations into its sources (part-to-part variation, repeatability, and reproducibility) and helps operators determine how to correct a poor measurement system. If the study shows a high repeatability relative to reproducibility, it indicates the need for a better gage. If the reproducibility is high relative to repeatability, that indicates the need for better operator training in the use of the gage.
A common mistake people make concerning a Gage R & R study is forgetting that it is evaluating their measurement system and not their products. Gage R & R doesn’t care how awesome your products are – it only cares how precisely you measure them.
Challenges arise as processes improve and products are produced so precisely that it’s difficult to detect part variation by existing gages. More and more products are produced using advanced manufacturing techniques and tolerances are measured in microns. Given these very small tolerance ranges, seemingly small differences can have significant impact on the measurement value. As products and processes improve, measurement systems also need to improve.
The table below is an example of the required repeatability for a given tolerance range and Gage % when measuring at the micron level:
|Tolerance Range (um)||10% Gage (um)||20% Gage (um)||30% Gage (um)|
When designing automated metrology (measurement) systems for advanced manufacturing, it’s important to know if production parts are available. Be wary of pre-production parts; they may differ from actual production parts and this will impact vision and laser based metrology systems. New products going through iterative design and process changes will also affect the metrology solution. It is recommended that the system take this into consideration and not design to 100% of the specified capability of any component. A minimum of 10% safety margin should be used in all calculations. For example, in reference to the table above, a system needs to be repeatable to less than 1.9-microns with at least a 10% margin to achieve a 100-micron tolerance and 10% gage. It is also important consider the environment where the system will be operating and whether conditions could affect measurements.