Gage R&R - Applied threshold for the %EV and %AV values?

Hi everybody,

Is there someone for helping me ?

I would like to know if someone knows if there is an applied threshold for the value of %EV (Equipment Variation) and %AV (Estimated Variation) in the GRR calculation. How can we explain these two indicators?

Thanks in advance.

Rach.

Star18 said:

Hi everybody,

Is there someone for helping me ?

I would like to know if someone knows if there is an applied threshold for the value of %EV (Equipment Variation) and %AV (Estimated Variation) in the GRR calculation. How can we explain these two indicators?

Thanks in advance.

Rach.

Click to expand...

"AV" stands for Appraiser variation, not estimated variation. There are no thresholds; the information that comes out of a GR&R study should be used to help target improvement efforts. If EV is disproportionately high it might mean that the gage needs calibration or repair, or that the gage is inappropriate for the application. IF AV is high, it might mean that operators are using the gage improperly (or inconsistently) and might benefit from training.

Jim Wynne said:

"AV" stands for Appraiser variation, not estimated variation. There are no thresholds; the information that comes out of a GR&R study should be used to help target improvement efforts. If EV is disproportionately high it might mean that the gage needs calibration or repair, or that the gage is inappropriate for the application. IF AV is high, it might mean that operators are using the gage improperly (or inconsistently) and might benefit from training.

Click to expand...

I also should have mentioned that "AV" is also referred to as Reproducibility and "EV" as Repeatability.

Star18 said:

Hi everybody,

Is there someone for helping me ?

I would like to know if someone knows if there is an applied threshold for the value of %EV (Equipment Variation) and %AV (Estimated Variation) in the GRR calculation. How can we explain these two indicators?

Thanks in advance.

Rach.

Click to expand...

While there is no specific threshold value for EV and AV individually, there are threshold values for the combined value %GRR or P/T ratio.

These thresholds are: < 10% ideal, between 10% and 30% acceptable, > 30% unacceptable.

The purpose of breaking it out to AV and EV is to identify the major source of the variation to allow you isolate the cause and reduce the variation.

%ev & %av!

Hi everybody,

I understand what you want to mean but my question is about the meaning of a value of %EV and %AV according to the result of %GRR.
You will find as an attached file a form of MSA application. Could you help me please ?

Rach.

I am not completely clear on your question. Can you be more specific?

I did review the attached file, and can offer some advice on the results. The %GRR = 80% and the ndc = 1.05 both indicate that this gage is unacceptable for use in process control (SPC) or for use in any type of statistical evaluation such as a DOE.

There may be several explanations for this. The first question that must be answered before going any further is: Is the part variation representative of the actual process variation? If it is not representative, you must repeat the study using parts that exhibit the same variation as the process.

If the part variation is representative of the process variation, you then look at the components of variation. The high %EV = 80% tells you that the equipment itself is a likely place to begin looking.

Check the validity of the 10.010 measurement made by Appraiser A, Part 1, Trial 2. Even changed to 10.0001, the equipment does not have sufficient resolution to "see" process variation. This gives you the low ndc value. If the parts are representative of the actual process variation, you will need to invest in a gage with greater resolution.

Before you do this, what are you using the gage for? If you are looking at SPC, do you have an economic case for it. That is, will the benefit of SPC justify the new gage? Are you using the gage for inspection to tolerance? If you are, then calculate the P/T ratio for the gage. This may be more acceptable than the %GRR.