Gage R&R for Hardness Testers - Heat Treating Plant

D

dbzman

Were are a Heat Treating plant and we use a number of Hrdness Testers in the Plant.

Is it possible to perform GR&R on Hardness Testers?

Hardness Testers make indentations on the part that can not be repeated exactly.

The hardness can vary from spot to spot on the part (although not by much).

Thoughts?

Opinions?

Thanks!:bonk:
 
N

noboxwine

Sure ya can

I audited a Heat Treat facility a few moons back and they did indeed conduct an R & R for a Versitron Rockwell C tester on hardened tool steel. They treated 10 peices of flat stock and conducted the usual 3 x 3 x 10 R & R. Nothing that I saw was unusual about it. It satisfied me an an auditor and most importatly proved to them whether or not the measurement system was sound. No reason why this is any different from other measurement techniques. Hope this helps

:D
 

howste

Thaumaturge
Trusted Information Resource
I suppose that as long as you understand that material variation is also included in the equation, then it can be done. If your results show that the R&R of the system is acceptable with that variation included, then it certainly would be OK without.

BTW, most R&R studies have material variation included anyway. Example: a sample of wire being measured with a micrometer will have variations in both diameter and roundness along its length.
 
A

Al Dyer

Look at it as a benefit! If you have 10 heat treated samples each appraisor can measure on a slightly different part of the sample.

Also think that unless your going through 17025, the auditor might not even know what you are talking about.

The key is to identify, tag, keep lot numbers and the samples for future reference, this way if there is a problem you can identify it and prove your point.

Al...
 
A

Atul Khandekar

Two thoughts:

1. If the variation in hardness from spot to spot is not very significant, you could consider them to be homogeneous samples and use Nested ANOVA method which is recommended for non-replicable testing.

2. Is it possible to get a 'standard' or 'certified' block of known hardness? e.g. NIST's Standard Reference Material catalog. (https://www.nist.gov/srm). This is likely bo be an expensive proposition.
 
D

dbzman

Gage R&R for Hardness testers

Hello everyone!

Does anyone know if you can perform R&R studes on hardness testers?

If so, how is this done?


Thanks!


Mike W (Confused & abused)

:bonk:
 
N

noboxwine

Yep

You sure can. It's like any other non-destructive measuring device.

I always did a 3 X 3 X 10 R & R.

3 people measuring 10 parts 3 different times. Plug the numbers into the sheet and thats its. Here's a groovy ready-to-use Excel sheet attached. Let us know if you need anything else and have a day



:cool:
 

Attachments

  • gage r & r.xls
    76 KB · Views: 505
D

dbzman

Question

Question.....

I was told that you could not do a R&R study on hardness testers because it is destructive and hard to repeat.

You can't test in the same spot as the other person.

Thanks for you help!

Mike W
 
N

noboxwine

yeah--you cant penetrate the diamond or ball in exactly the same spot every time..... just the same way you cant meaure a diameter with calipers in exactly the same spot every time.... there is some variation in every measuring technique, but no more in hardness testing than any other application. i have done it with certified calibration blocks, sheet metal parts, hardned tool steel, etc...and had successful studies every time. give me a report on it !:bigwave:
 
D

dbzman

GAGE R&R - Look what I got

We finished our study but the numbers are not what we expected.
We used 10 parts, two operators, and two trials.

We got an R&R of 85.58!

We then took a secound study, added it to the first, and got an average.

When we used this number we got an R&R of 82.55

What have we done that is wrong? We know that we have the proper hardness testers for the part. All of the readings were within specification.

Should I do this with ANOVA instead(because of the nature of the study)?

If so, is there a spread sheet that I could use for ANOVA available?

Thanks!

Mike W :bonk: :bonk:
 
Top Bottom