What does the P/t ratio (Precision/Tolerance) =RR/(USL-LSL) say about my gage study?

R

rderoeck - 2006

Can somone explain to me what the P/t ratio (Precision/Tolerance) =RR/(USL-LSL) tells me about my gage study? Is there some magic number that needs to be met?

Thanks,
Rich DeRoeck
 

Jim Wynne

Leader
Admin
rderoeck said:
Can somone explain to me what the P/t ratio (Precision/Tolerance) =RR/(USL-LSL) tells me about my gage study? Is there some magic number that needs to be met?

Thanks,
Rich DeRoeck
Welcome to the Cove:bigwave:

The precision/tolerance ration is a measure of how much of the tolerance is being consumed by gage error. The generally accepted goal is 0.1 or less (10%). Here's a basic explanation (scroll down):
http://www.itl.nist.gov/div898/handbook/glossary.htm#P
 
R

rderoeck - 2006

So......this is diferent than the % R&R requirement of <10%. Do I have this right?

Rich
 

Jim Wynne

Leader
Admin
rderoeck said:
So......this is diferent than the % R&R requirement of <10%. Do I have this right?

Rich
Well, more precisely, it's a different way of looking at it that takes the 3-sigma magnitude of R&R into consideration.
 

Miner

Forum Moderator
Leader
Admin
There are an increasing number of ways to evaluate a measurement system, and each of these is intended to answer a different question about the suitability of the system.

If your question is "Is this gauge adequate to inspect a part to this tolerance?", then P/T ratio (known as % R&R before the proliferation of indicators) answers this question. The typical standard for P/T ratio is 10% or less is ideal, between 10% to 30% may be acceptable depending on the importance of the measurement, the cost of new gauging, etc., and > 30% is not acceptable.

If your question is "Is this gauge adequate to inspect a part as part of study (e.g., Six Sigma, DOE, capability, etc.)?", then %TV (Total Variation) answers this question. The standard is the same as above. In addition, the Number of Destinct Categories is used. An ndc >/= 5 is good, between 2 & 5 may be acceptable, and < 2 is unacceptable. Wheeler uses a Discrimination Ratio, which must be >/= 4, and requires the gauge resolution to be < the StdDev of PV (Part Variation).

If you want to know where the variation is coming from, % Contribution of each component answers this question.

See AIAG's MSA manual and/or Wheeler's "Evaluating the Measurement Process" for more information.
 
R

rderoeck - 2006

Here's my issue....

Someone showed me a GRR study using JMP software. It printed out the following summary:

10.3334 % Gage R&R=100*(RR/TV)
0.16857 Precision/Tolerance Ratio=RR/(USL-LSL)

So the first one is comparing the percentage of measurement variance to process variance (as determined by the study data) while the second one is comparing the proportion of measurement error to the part tolerance.

Can you then say that the percentage of P/t is 16.857%?

Thanks a lot,
Rich DeRoeck
Supplier Quality Engineer/SPC Coordinator
Calibration Manager
 

Miner

Forum Moderator
Leader
Admin
rderoeck said:
Someone showed me a GRR study using JMP software. It printed out the following summary:

10.3334 % Gage R&R=100*(RR/TV)
0.16857 Precision/Tolerance Ratio=RR/(USL-LSL)

So the first one is comparing the percentage of measurement variance to process variance (as determined by the study data) while the second one is comparing the proportion of measurement error to the part tolerance.

Can you then say that the percentage of P/t is 16.857%?

Yes. P/T = 16.857 %, which compares the gauge error to tolerance. This falls into the marginal are, but if it is not a critical dimension, it is probably acceptable.

The 10.3334% GRR % TV (Total variation) means the gauge is acceptable for performing process studies, such as DOE or capability studies.
 
B

bryan3com

Re: What does the P/t ratio (Precision/Tolerance) =RR/(USL-LSL) say about my gage stu

About P/T ratio there are two way to control it:
6sigma/USL-LSL*100% OR 3sigma/Tolerance

For the tolerance what kind of number should I put?
ppl told me to use "80% average" be the Tolerance. Is that correct?
 

Miner

Forum Moderator
Leader
Admin
Re: What does the P/t ratio (Precision/Tolerance) =RR/(USL-LSL) say about my gage stu

Put your USL - LSL in as your tolerance. If you have no tolerance, you cannot calculate P/T Ratio.
 

KSHILENDU

Registered
Can P/T Ratio calculate in MSA of Bevel's Protractor?

If it is yes, then in one of example, RR=0.0187 & USL-LSL=2 Degree, what is P/T RATIO?
 
Top Bottom