Need 3 appraisers for GR&R (Gage R&R) of Automatic Test Measurements?

[Calvin Chen]

I have some questions regarding to calibration practices conforming to TS16949:

1. GR&R: When doing the measurement, our operators only need to put the product in a loader. The measurement machine will automatically load the product, detecting the position and place it to a fixed place, then start to do the test by the set program. Since the measurement is automatically taken, can we just do one assessment report to explain there is no significant appraiser's variation and afterwards have only one apprasier for GR&R by estimating GR&R=EV^2?
If the answer is yes, should we do the reports by different products which are tested by same measurement tools and methods?

2. Process Tool Gauge calibration: On some of our process tools, there are gauges which is used for monitoring the tool conditions, not product quality characteristics. Say, gas flow, pressure, temperature, etc. Should we do calibration on these gauges? Does TS16949 include the process tool gauge in the MSA requirements?

Any input will be highlighted appreciated. Thank you.

 

[Rob Nix]

You are right on track, Calvin. Do the standard GR&R study (3 people, 3 reps, 10 parts) to verify the non-existent reproducibility error (if indeed the operator does NOT influence the results), then simply check for repeatability (EV). The way I would do it is this: one part (product) 30 times; 5.15 std devs divided by the tolerance.

CAUTION: Look at the individual data in a trend line chart. Sometimes the dynamics of the test/inspection causes some trending, "ski sloping", or mean shifts, e.g. part or fixture heating up and expanding, oil viscosity changes, etc. I always find more meaningful data using more repetitions and fewer parts.

Then I would do the same with another product and compare the standard deviations.

Regarding the gages for process conditions, by all means, check the accuracy and repeatability of each. You may also do a correlation analysis (scatter diagram) of each process condition with the product measurement results. It is more work up front, but may save you a lot of grief during later problem solving, IMHO.

 

[D.Scott]

Hi Calvin, welcome to the Cove. We have some great Gage Gurus here but until they have a chance to jump on this for you I will try to help out.

IMHO you would not need three operators if there is no possibility of mis-placement of the part of any other variable that would give different results. So my answer to the first question would be yes.

As for different products, I would do R&R for the equipment to establish your uncetainty budget and qualify the measurement system itself. If you are submitting PPAPs for specific parts, you will need to do an R&R for each product you submit. The part variance plays a big role in suitability of the equipment for the measurement being made and therefore should be shown through the R&R.

As for process tool gages, my decision is based on whether or not the gage effects the quality of the product. For example, a temperature gage for an oven on a process where a specific heat is required would need calibration. The same would apply to timers, conveyor speeds etc. if they were required to be within a specific tollerance. I wouldn't calibrate a gage that was used soley to indicate it was turned on. I would add those gages to a PM schedule to verify they function. Rule of thumb - if it affects quality, calibrate it.

Hope this helps.

Dave

 

[Howard Atkins]

I did some work on this subject and had a lot of help and input from my friend, coleauge and knowlegable Cover Atul Khandekar .

This is something I wrote to explain what I did:

The testing machine is automatic and there is no affect of the appraiser on the results.​

The machine takes the part and positions it in place for the test.​

All variation is thus dependent on the machine (EV)​

The method used is that quoted in the MSA handbook and attributed to Duncan, A.J., Quality Control and Industrial Statistics.​

Selecting a calculation method​

By John Raffaldi and Steven Ramsier, Ph.D.

Typically, GR&R studies require two or three appraisers, and between five and 10 trials to establish the EV and AV measurement error components. However, if an automated test system lacks appraiser variation, the instantaneous GR&R calculation method that uses the within-subgroup standard deviation calculations can be used. The instantaneous method uses only one appraiser and provides information only on the within-system, EV component. If fewer than 10 measurements are taken, correction factors must be considered. These factors listed in the table, c4 Values, are taken from Acheson Johnston Duncan's book, "Quality Control and Industrial Statistics." Otherwise, use a value of c4 = 1 as the correction factor for more than 10 measurements.
For the instantaneous method, the calculations are as follows

1. Calculate the standard deviation from the trials. If using multiple parts and multiple trials, calculate the standard deviation for each set of individual part measurements, and then take the average of the standard deviations just calculated. The result is the average standard deviation
2. Divide the standard deviation by the correction factor, c4, if using fewer than 10 trials. The number of observations in the sample, n, is the number of trials. This information is shown in the table, c4 Values.
3. Multiply the number from step 2 by the number of sigma to be used, usually 5.15 or 6. <O
4. Divide the number from step 3 by the tolerance, which is the upper specification limit (USL) minus lower specification limit (LSL), and multiply by 100 to obtain the GR&R as a percentage of the tolerance. <O
The formula is: % GR&R = ((5.15 * σ/ c4) / (USL - LSL))*100 or % GR&R = ((6 * σ / c4) / (USL - LSL))*100 where sigma equals the standard deviation for the trials or one part or the average standard deviation from multiple part or trial measurements.


In the case Calvin is discussing the the machine places and performs the testing, if the operator places the part and the machine tests then this is not valid.

Atul produced a freeware program to calculate according to this method. You can download at https://www.symphonytech.com/freeware.htm

I trust this helps.