Gage Linearity and Bias Questions

[eturkan]

Hi Gage R&R experts..

I could not get an answer in reply to another post.I did a Type 1 study as a prep to Anova.

I have 4 questions.

1. For Bias we must use the real parts and what if they change.

I need to do a study on OD of ptfe o-rings using flat micrometer and a tapered aide. Master value of OD using vision systems increases 0.02% which caused a higher bias. If I use post test master value, bias drops to 4%.


2.For stability gage blocks or real parts? What is the common interval for measurements for a flat micrometer that has 6 months calibration interval with no problem in past?

3.Linearity: Gage Blocks or real parts?

AIAG MSA: Due to process variation, select 5 parts that covers the entire operating range of the gage. What is process variation? Potential of the gage being used in other processes or variation within a process?


4. Type II Minitab Results
I know that we use %TOL if our CPK is over 2.5 but do we still use it at times when it is higher than %GRR?

Thanks

 

[breitman]

I sure wish someone would reply; I seem to be in the same boat.

 

[Miner]

Hi Gage R&R experts..

I could not get an answer in reply to another post.I did a Type 1 study as a prep to Anova.

I have 4 questions.

To which post are you referring? Provide a link to it and I will try to answer.

1. For Bias we must use the real parts and what if they change.

I need to do a study on OD of ptfe o-rings using flat micrometer and a tapered aide. Master value of OD using vision systems increases 0.02% which caused a higher bias. If I use post test master value, bias drops to 4%.

Not 100% certain I follow this. Do you mean that the vision system, which is considered the TRUE value against which the micrometer is compared is detecting a Pre/Post test change in part size? If so, is this size change due to normal part shrinkage or due to a compression set from the micrometer measurement process?

You must use real parts because parts can interact with the gauge creating a bias that you would not see if measuring a gage block. For example, measuring the OD of a compressible part would tend to be biased low while measuring a gage block should have little bias.

2.For stability gage blocks or real parts? What is the common interval for measurements for a flat micrometer that has 6 months calibration interval with no problem in past?

Stability studies can use gage blocks, but the measurements must be taken in actual use conditions. See my blog on Stability.

3.Linearity: Gage Blocks or real parts?

AIAG MSA: Due to process variation, select 5 parts that covers the entire operating range of the gage. What is process variation? Potential of the gage being used in other processes or variation within a process?

I may need to reread this section, but it used to allow a standard or a part.
See my blog on Linearity.

4. Type II Minitab Results
I know that we use %TOL if our CPK is over 2.5 but do we still use it at times when it is higher than %GRR?

Thanks

The use of %Tol has nothing to do with Cpk. It has to do with the end use of the gauge. Use %Tol if you use the gauge for inspection, and %SV/%PV/ndc if you use the gauge for process control or statistical studies. See my blog on R&R.

 

[eturkan]

Your blog is the best application based guide I've seen so far..I appreciate your answer.

Master value change in OD
I believe it is caused since we are using taper aide to stabilize the ID/OD.Operator pushes the o-ring till they see no light but even small forces are causing OD increase since it is repeated 30 times.
%Tol
I believe I?ve read cpk 2.5 for TOL% usage on AIAG MSA book . Since the process is so stable, we don?t get enough process variation and get a high GRR%. In this case %TOL is lower


Bias must be done by the operators but Linearity can be done by an engineer??

 

[Miner]

For the master value change, try plotting the value against the order in which the part is measured. If you see a relationship, you might be able to treat it as a covariate. This compensates for it in the analysis and removes the effect.

Can you cite the page or section in which you found the %Tol vs. Cpk guidance? If you can, I'll check into it.

The MSA 4th edition manual specifies that the linearity study be done by the operators that normally use the gauge. The rationale is that linearity is the same as bias extended over the entire working range.

 

[eturkan]

Hi Miner..


It was Ford Gage R&R training.It is the best Gage R&R training I've seen so far but does not completely answer when/ when not to use TOL%





When do we use Gauge R&R % Tolerance?
When our process is so capable we can’t measure differences in the parts.
When do we use Gauge R&R % Tolerance?
[FONT=&quot]• [/FONT]When the manufacturing process is stable, in control, and
[FONT=&quot]• [/FONT]When the parts pulled represent the entire range of expected variation, and
[FONT=&quot]• [/FONT]When the process is highly capable (e.g.: Ppk >= 2.5).


When a process is highly capable and percent tolerance is used instead of percent study the criteria to have 5 distinct categories and 50% of the parts outside the blind spot of the gauge are not applicable.


​