I am not a quality expert and I am starting to look at gage R and R to justify changing gages becuase I strongly believe it is not capable. I have read MSA manual and it has lots of useful information but I don't have a thorough enough understanding of what the results mean. Can anyone point me to a simpler treatment (GR&R for Dummies) :read: which will help me understand more easily and convince my boss to invest in more capable gage.

Any help appreciated.

Here is the GR&R form that we used up until a few years ago.

Please note that this only calculates Repeatability and Reproducability; per the MSA manual, you also have to consider Linearity, Stability and Bias.

However, it is something that is (fairly) simple to use, and should give you enough evidence to either proceed forward, or forget your plan.

Here is the GR&R form that we used up until a few years ago.

Please note that this only calculates Repeatability and Reproducability; per the MSA manual, you also have to consider Linearity, Stability and Bias.

However, it is something that is (fairly) simple to use, and should give you enough evidence to either proceed forward, or forget your plan.

I am sorry Ron but this file contains the 2nd edition constants that have been changed.
This file Partcertall_6.15 3rd ed constants (http://elsmar.com/Forums/attachment.php?attachmentid=3287) has the correct constants as well as other useful forms.
There are also other forms in the attachments but please check the constants
e.g K1 now = 0.8862 for 2 in the past 4.56

Hello,

I am very new to Gage R&R and have been tasked with performing a few of these experiments. Now I know what the purpose of Gage R&R is, can anyone please give me an explanation of the mathematics? I would specifically like to understand the K1 and K2 constants and how they are calculated/determined.

Even a recommendation of a book would be most helpful!

I hope you can help! Thanks

Nanci

Two books on Gage R&R are the AIAG Measurement System Analysis (MSA) book available through the AIAG website, and Evaluating the Measurement Process by Wheeler and Lyday.

If you are in automotive, it will be less painful to use the AIAG MSA book than to fight with the OEMs. But, if you have a choice, the Wheeler book is a more useful way of evaluating your gages.

I think I've recommended this before...For basics you may refer to "Concepts for R&R Studies" by Larry Barrentine. Available from ASQ Press.

Hello, thanks for the book recommendations, I have bought the Evaluating the Measurement Process book and it doesn't seem to tell me exactly what I need to know. What I need to know is quite simple, and I really hope someone can help me. I understand a bit of mathematics, and would like to know how to calculate the Equipment Variation (E.V.) and the Appraiser Variation (A.V.) I have got sets of results that I am going to work with...I have seen a few spreadsheets but would simply like to know how the calculations are made...This is especially important when I sometimes get different E.V. and A.V. Values from different spreadsheets, with the same data!
I know what the targets for these EV and AV values are, but I would just like to know what the formulas are to calculate them. I know there are some K-Constants in there somewhere - Please could somebody explain this?

Many thanks, fingers crossed for someone who can help!

Nanci

Hello, thanks for the book recommendations, I have bought the Evaluating the Measurement Process book and it doesn't seem to tell me exactly what I need to know. What I need to know is quite simple, and I really hope someone can help me. I understand a bit of mathematics, and would like to know how to calculate the Equipment Variation (E.V.) and the Appraiser Variation (A.V.) I have got sets of results that I am going to work with...I have seen a few spreadsheets but would simply like to know how the calculations are made...This is especially important when I sometimes get different E.V. and A.V. Values from different spreadsheets, with the same data!
I know what the targets for these EV and AV values are, but I would just like to know what the formulas are to calculate them. I know there are some K-Constants in there somewhere - Please could somebody explain this?

Many thanks, fingers crossed for someone who can help!

Nanci

A lot of clients use GageTrak, or GagePak software for their calibration activities. I am told both are very easy to use, and I believe both automatically calculate the math for all the common gage studies, including Gage R&R's.

A lot of clients use GageTrak, or GagePak software for their calibration activities. I am told both are very easy to use, and I believe both automatically calculate the math for all the common gage studies, including Gage R&R's.


We are now using GageTracker 6 since March '05. It also caclulates stability and linearity. Was recommended by CB auditor. I think it was approx. $1,600

Hello, thanks for the book recommendations, I have bought the Evaluating the Measurement Process book and it doesn't seem to tell me exactly what I need to know. What I need to know is quite simple, and I really hope someone can help me. I understand a bit of mathematics, and would like to know how to calculate the Equipment Variation (E.V.) and the Appraiser Variation (A.V.) I have got sets of results that I am going to work with...I have seen a few spreadsheets but would simply like to know how the calculations are made...This is especially important when I sometimes get different E.V. and A.V. Values from different spreadsheets, with the same data!
I know what the targets for these EV and AV values are, but I would just like to know what the formulas are to calculate them. I know there are some K-Constants in there somewhere - Please could somebody explain this?

Many thanks, fingers crossed for someone who can help!

Nanci

Hi Nanci

The AIAG MSA Manual most definitely will explain these formulas.

I also have the Wheeler book and it is a much different (better?) approach.

Last year I broke down and made a study of the MSA manual, just like I was back in school. We had troubles with "poor" GR&R on World Class equipment and it fell to me to find out why.

I made notes, worked all the examples, built some spreadsheets, and kept reviewing it until I finally got it. I think I still have my notes, I would happily share them if you like, but be warned they are rough study guides for my own use and may not help much.

It took some time and effort but all became clear.

Each spreadsheet I have seen rounds off slightly differently and gives numbers just different enough to drive you crazy. It seems that you have to carry every significant fugure you can in the calculations.

Good luck, I suggest the AIAG MSA as a good book.

Thanks everyone for the help! Caster, your notes sound like they would give me exactly the information I am searching for. Please can you get them to me somehow (email/scans)?

Many Thanks
x

Can anyone give me an example of how to do a MSA study on a measurement tool with no know value. We use pyrometers to record the temperature of aluminum. We cannot control the temperature and therefore have no value to measure our temperature against. The TS standard states that we shall conduct analyse on each type of measurements referenced in the control plan. Any ideas?? :
Thanks

Can anyone give me an example of how to do a MSA study on a measurement tool with no know value. We use pyrometers to record the temperature of aluminum. We cannot control the temperature and therefore have no value to measure our temperature against. The TS standard states that we shall conduct analyse on each type of measurements referenced in the control plan. Any ideas?? :
Thanks

Why do you measure and record values you can't control, and don't care about? In other words, why is it in the control plan if you're not controlling it? What does your control plan entry say to do?

I am sorry Ron but this file contains the 2nd edition constants that have been changed.
This file has the correct constants as well as other useful forms.
There are also other forms in the attachments but please check the constants
e.g K1 now = 0.8862 for 2 in the past 4.56

can you tell me where I can find the new d constants in the appropriately updated table - all my old R&R literature uses the previous edition's. ALso, what is the rationale for such a large change in the constants? prety radical change for statistical constants ( that I thought were as unchangeable as gravity or acceleration)

can you tell me where I can find the new d constants in the appropriately updated table - all my old R&R literature uses the previous edition's. ALso, what is the rationale for such a large change in the constants? prety radical change for statistical constants ( that I thought were as unchangeable as gravity or acceleration)

This thread is 3 years old, but maybe someone might be able to help.

D2 for 2 Trials = 1.128, for 3 Trials = 1.693
D3 for 2 Trials = 0.853, for 3 Trials = .0888
D4 for 2 Trials = 3.27, for 3 Trials = 2.58

The 3rd edition made two notable changes:

The first was to change from 5.15s (standard deviations) (99%) to 6s (99.73%)
The second was to simplify the formula by omitting the multiply by 6 and divide by 6s to simply divide by s.

special procesess (painting, plasma coatings, brazing)

Few questions:
1. Can i use 10 simulated samples (simulation parts, scrap parts) instead running parts?
2. what is the procedure if i have tolerance min or max?
3. what measure must be write if: tolerance for dia is .755-.775, act dia is .7555-.7745 (his i sexample)? "big dispertion"

these are imprtant question forme to start doing this :truce:

regards!!!

What will the gage be used for, inspection or process control? The answers to your questions depend on the answer.

What will the gage be used for, inspection or process control? The answers to your questions depend on the answer.

Olny for inspection by worker and quality inspectors.

Since this gage is only used for inspection, you should use the P/T Ratio (% Tolerance) metric for assessing acceptability of the gage. This compares the gage error to tolerance. Process variation is not part of the calculation, so the dimensions of the parts are not terribly important unless the might affect how the operator may inspect them. It is good practice to select parts the use the entire range of the tolerance, but this is not mandatory.

1) As long as operators will measurethe simulated parts the same way they would measure an actual part, yes. Be careful that this is a correct asumption.

2) Do you mean a unilateral (1-sided) tolerance? You may want to use the PME (Probable Measurement Error) propsed by Donald Wheeler. The P/T Ratio has not been adapted to unilateral tolerances.

3) This sounds like Guard Banding. That is a business choice. Most do not unless the dimension is critical.