ANOVA Gauge R&R

ANOVA Gauge R&R (or ANOVA Gauge Repeatability & Reproducibility) is a Measurement Systems Analysis technique which uses Analysis of Variance (ANOVA) random effects model to assess a measurement system.

The evaluation of a measurement system is not limited to gauges (or gages) but to all types of measuring instruments and test methods.

Purpose

ANOVA Gauge R&R measures the amount of variability induced in measurements that comes from the measurement system itself and compares it to the total variability observed to determine the viability of the measurement system. There are several components affecting a measurement system including:

* Measuring instruments, the gauge or instrument itself and all mounting blocks, supports, fixtures, load cells etc. The machine ease of use, sloppiness among mating parts, "zero" blocks are examples of sources of variation in the measurement system;

* Operators (people), the ability and/or discipline of a person to follow the written or verbal instructions.

* Test methods, how to setup your devices, how to fixture your parts, how to record the data, etc.

* Specification, the measurement is reported against a specification or a reference value. The range of the specification does not affect the measurement, but is an important factor affecting the viability of the measurement system.

* Parts (what is being measured), some items are easier to measure than others. A measurement system may be good for measuring steel block length but not for measuring rubber pieces.

There are two important aspects on a Gauge R&R:

* Repeatability, the ability of the device to provide consistent results. It is a measure of the variability induced by the system if the same operator measured the same part using the same device repeatedly.

* Reproducibility, the variability induced by the operators. It is the variation induced when different operators (or different laboratories) measure the same part.

It is important to understand the difference between Accuracy and precision in order to understand the purpose of Gauge R&R. Gauge R&R only address how precise a measurement system is.

Anova Gauge R&R is an important tool within the Six Sigma methodology, and is also a requirement for Production Part Approval Process‎ (PPAP) documentation.

How to perform a Gauge R&R

The Gauge R&R is performed by measuring parts using the established measurement system. The goal is to capture as many sources of measurement variation as possible, so they can all be assessed and addressed. Please note that the purpose is not to "pass". A small variation reported on a GRR may be because an important source of error was missed during the study.

In order to capture reproducibility errors, multiple operators are needed. Some (ASTM E691 Standard Practice for Conducting an Interlaboratory Study to Determine the Precision of a Test Method) call for at least ten operators (or laboratories) but others use only 2 or 3 to measure the same parts.

In order to capture repeatability errors, the same part is usually measured several times per operator.

In order to capture interactions of operators with parts (e.g. one part may be more difficult to measure than other), usually between 5 and 10 parts are measured.

There is not a universal criteria for minimum requirements for the GRR matrix, being up to the Quality Engineer to assess risks depending on how critical the measurement is and how costly they are. The 10x2x2 (10 parts, 2 operators, 2 repetitions) is common, although it has very few degrees of freedom for the operator component.

Common Misconceptions about GRR

* Need only one GRR per family of gauges. It is usual to say "There is an acceptable GRR for this caliper". This statement is false, as a GRR is for the measurement system, which includes the part, specification, operator and method. As an example, measuring a steel block with a caliper may be achieved with a good precision, but the same caliper may not be suitable to measure soft rubber parts that may deform while it is being measured.

* The GRR will not pass using parts, so it has to be done with standard weights and blocks. The GRR done in this way will assess the precision while measuring standard weights. The device might not be suitable to measure that specific type of parts. If the part "changes" while being measured, this has to be counted as a measurement system error.

* Need to report on PPAP documentation GRR results for everything that is measured. This is not necessarily a requirement. The Quality Engineer usually makes an educated assessment. If the characteristic is critical to safety, a valid GRR is required. Instead, if there is enough understanding that some particular part is easy to measure with acceptable precision, a formal GRR is not required. Customers may ask for additional GRRs during PPAP reviews. Knowing that a GRR is not good and still uses the measurement system does not make sense. This is like using bent calipers to get measurements, you get a number but it does not mean anything.

*Performing a GRR is very expensive. In order to perform a GRR usually a number of parts (sometimes between 5 to 10) is required to be measured by at least 3 operators (some suggest ten or more) 2 to 3 times. So the measurement costs are the ones associated with those additional measurements. For simple devices this may not be very costly, and the results is a known measurement error that can be used to assess all measurements subsequent to that. The costs can be higher for destructive testing.

* GRRs must be within 10% to pass. There are AIAG guidelines for GRR errors relative to the specification, and what to report on a PPAP process. The final call is between the supplier and customer, and it is a function of the critically of the characteristic and the assessed measurement error. GRR is a tool that helps making this assessment, but it does not gives you the answer.

External References

* GRR Definition: http://www.siliconfareast.com/grr.htm
* [http://www.bentham-open.org/pages/content.php?TOIMEJ/2008/00000001/00000001/29TOIMEJ.SGM GRR and Adjusting Acceptance Limits]

ee also

* Measurement uncertainty