Is GR&R (Gage R&R) required for simple attribute gauges like plug gauges?

Andrews

Involved - Posts
#1
Gauge repeatability and reproducibility study helps us find out what effect the variation in method of checking has on the inspection of a particular parameter. This variation is partly contributed by the instrument and the appraiser.

IMHO for a simple plug gauge , there can be no two results because the method of checking will not vary. So Gr&R is not required?

Correct me if I am wrong.
 

Ron Rompen

Trusted Information Resource
Trusted
#2
Andrews:

I must disagree with your previous statement. Although the 'method' of checking may not vary (using a calibrated known-size pin to check a hole diameter) the results may vary significantly.

Operator 1: Male, 18 years old, excellent physical condition, 5 days experience of inspection using Go/NoGo gauges.

Operator 2: Male, 63 years old, poor physical condition, advanced arthritis in both hands and elbows, 5 days experience of inspection using Go/NoGo gauges.

Operator 3: Male, 42 years old, average physical condition, 25 years experience of inspection using Go/No Gauges

Criteria #1: GO member must pass through part using a minimum amount of effort. Force should be used to align gage with hole, but should not be used to 'force' gage through hole. (NOTE: Parts are being sorted for rolled-over burr due to sanding operation. Hole diameter is compromised by 0.02mm maximum, distribution of defective parts is normal, actual hole diameter uses entire specification tolerance when running 'as best can be achieve'),

Criteria #2: 25,000 pcs are require to meet shipment schedule in 4 hours.

Criteria #3: only the 3 people noted above are available to sort the parts.


How many parts do YOU think will be accepted by the 3 operators? Would you expect the results to be the same?
 

Jim Wynne

Forum Moderator
Moderator
#3
Andrews said:
Gauge repeatability and reproducibility study helps us find out what effect the variation in method of checking has on the inspection of a particular parameter. This variation is partly contributed by the instrument and the appraiser.

IMHO for a simple plug gauge , there can be no two results because the method of checking will not vary. So Gr&R is not required?

Correct me if I am wrong.
Ron's example is a little extreme and assumes information that can't be known in advance, if at all, but it still serves to show that there can indeed be variation in simple attributes measurement tasks. The fact is that your statement, "...there can be no two results because the method of checking will not vary..." is nothing more than conjecture unless you have the data to back it up.
 

Caster

An Early Cover
#4
We tried and failed

JSW05 said:
Ron's example is a little extreme and assumes information that can't be known in advance, if at all, but it still serves to show that there can indeed be variation in simple attributes measurement tasks. The fact is that your statement, "...there can be no two results because the method of checking will not vary..." is nothing more than conjecture unless you have the data to back it up.
I must concur with the replies so far.

As much as we sometimes find all this stuff a pain, it is a requirement because enough problems have happened in the past that people wrote standards to make us think about preventing and repeating these same problems.

We just did some MSA on simple plug gages in simple old holes. We thought this would be just a huge waste of time. And they failed. We are still working on fixing all the problems we found.

The biggest one was training. It seems that some of our foremen think use of gages is so old hat, that they did not spend much (or any) on hands time with new people.

We also found shop floor lore had replaced the Work Instructions...my favorite person "they" said to do it this way...even though the WI clearly showed the correct way.

But, as is suggested just try it for yourself. Get the data. You may have no problems in which case you can argue to a registrar/customer from a position of strength.

However, can I suggest you do the test for fair. Do not let the person see any other previous test data. Use a veteran, and the most recent qualified hire. Make them do the parts in random order. In the location where they usually do the work. Under the same time pressure as day to day. Make sure some parts are just barely fails. As was suggested, put a burr in a few good parts. Make a few oval...if this can happen in your process. Make some parts oversize (our guys forgot to check the NOGO!)

I guarantee you will be surprised and learn a ton about your process and people.

I used to fight the requirements tooth and nail, now I try to withhold judgement till I learn a little more....I now reailize that every requirement is a problem that has happened tens of thousands of times, and we have been given an opportunity to not repeat that mistake by the ISO/TS/QS folks...
 

Atul Khandekar

Quite Involved in Discussions
#5
I fully agree with Castor. Try conducting a study, make sure you take parts that are close to limits - both inside and outside limits. It's sure to throw up a surprise or two - usually in the area of training and handling.
 
Top