Part Variation, Unilateral Tolerance, Destructive Testing - Gage R&R Experts pls help

[Delajo1]

Hey Guys,

I'm new to forums but could not resist registering to this and posting my question in the hope that someone can help. I have read through a few posts on destructive testing and one side tolerances but can not seem to find the answer I'm looking for.... or maybe there isn't a clear one.

I have a device that is tested in line ..... 100% of parts.

The test is preformed in two parts.... and the measurement is a combination of vacuum and time.
Example... test one can not exceed time x and test two can not exceed time y.

We are being pushed for R & R on this equipment and would like to know the best method to use and the best approach.


Question 2...

An additional complication is that the unit will not always perform identically each time it is started or run (unlike a drilled hole in a steel block that remains constant when measuring or preforming an R&R study).... therefore distinguishing between part to part variation, operator variation, machine variation, and an inheirt variation in the individual part at every time it is tested is difficult to distiguish.

What is your recommendation... any direction or advice would be greatly appreciated.
Hope to hear from someone soon.... many thanks.

 

[Miner]

Re: Gage R&R Experts...Part variation, unilateral tolerance, destructive testing

I am afraid that we will need additional information about the test equipment and what you actually measure to provide a response.

Regarding question 1) and without knowing any information about the test, I would say that you need to perform 2 separate MSA's. One for each test. Unless there is a particular reason why this is not feasible.

Regarding question 2), I have no way of advising on this until I understand the details of the test. Why is it not repeatable?

 

[Delajo1]

Re: Gage R&R Experts...Part variation, unilateral tolerance, destructive testing

I am afraid that we will need additional information about the test equipment and what you actually measure to provide a response.
Answer:
The test equipment is a volume with sensors to measure the drop in atmospheric pressure. The products purpose is to create a vacuum. The vacuum must be achieved in two stages within two specified time frames.
50% in say 2 secs and 75% in 5 secs.

Regarding question 1)
Yes we are planning to do two separate studies.
Is it possible to advise the best way to handle the specification of <2sec and <5 secs as its not a typical +/- tolerance.
We expect all results to range between 1.75-1.9secs and 4.75-4.9secs
Should we use a half way point say 1+/-0.5 for the <2 sec time.

Regarding question 2),
The pump itself that creates the vacuum when started and can by design run the test in say 1.7sec and if tested again could do it in 1.9 secs. This may be due to the position of the fan inside etc. So my question is how do we eliminate the variation in the part from impacting the results of the repeatability study on the test unit ? or should we ? Im only concerned that the person 1 may have very constant results by chance and person 2 may not ..... wouldn't this indicate in the analysis that the variation is by operator rather than the unit itself (assume it can be proven) ?

We are now at the point where have eliminated all other known issues and are planning to re-run the study ..... so any advice on the analysis of the results would be greatly appreciated. We typically perform these studies on products for dimensional measurements rather than functional so its alittle different for us.

 

[Miner]

Re: Gage R&R Experts...Part variation, unilateral tolerance, destructive testing

Q1) What is the purpose of this gage? Do you use it for inspection or for process control? If you use it for process control, the metric would be % of Part Variation and tolerance is irrelevant. If you use it for inspection, then tolerance is important. I would recommend using 0-2 and 2-5 as your tolerances. I mean, would you be upset if it passed the first test at 0.1 sec and the second test at 2.01?

Q2) What kind of results do you achieve with the current situation? That is, how big of an issue is this?

 

[JaxQC]

Re: Gage R&R Experts...Part variation, unilateral tolerance, destructive testing

Miner is correct that there is not a problem with defining your “spec” as zero to 2 & 5 respectfully.

2nd question – Don’t worry about the machine variation (1.7 1st time 1.9sec 2nd time). Actually that is the whole point of the Gage R&R to determine that. It turns out that you’re lucky that you can re-run the parts for testing (unlike say destructing welded parts where you only get one run per part). Since the Gage R&R breaks down the end value into equipment, appraiser, & part… use that to determine if it is acceptable or not. Since you are concerned about operator variation I’d use at least 3 people rather than just 2 and say 10 parts repeated 3 times. If it comes out great, no worries. If not it shows that method is important (how the one oper does / sets it up or the machine repeatability). That is why it is important to also look at the secondary charts that go with an R&R (whisker chart, appraiser run chart, interclass correlation plot, range chart), not just the one page report. They will help you define why the numbers are what they are.
Hope this helps.

 

[aproddutoor]

Re: Gage R&R Experts...Part variation, unilateral tolerance, destructive testing

Hi,

I want to do a Gage R%R on a part which has unilateral tolerance and it is destructive testing with 4 critical engineering dimensions on it.

So how do I collect my samples for destructive testing?
The same gages are used for both inspection and process control. I know how critical it is for the parts selected to represent the entire process spread. But it is really difficult to get parts for 4 critical dimensions throughout the process spread and I won't know until I measure that I can select that part for my study?

This process hasn't been easy for us in the past.

Someone please help me find a solution

Thanks

 

[Miner]

Re: Gage R&R Experts...Part variation, unilateral tolerance, destructive testing

There are several approaches for destructive MSA, but I need to understand the process, parts and characteristics to advise you.

Also, do you actually use the gage for process control or for inspection? If you only use it for inspection, the selection of parts is not important. If you use it for process control, it is important. However, you do have another option. It will involve some manual calculation, but you can substitute the process variation obtained from a capability study, or from SPC, for the process variation of the ten parts. This will actually give you a higher confidence in the result.

 

[aproddutoor]

Re: Gage R&R Experts...Part variation, unilateral tolerance, destructive testing

There are several approaches for destructive MSA, but I need to understand the process, parts and characteristics to advise you.

The wire is wounded on a spool and the end of the wire is stripped. The stripped wire is then welded to a ring. It has four critical engineering dimensions. To measure the dimensions, you have to remove the wire from the spool and you can't wind it back onto the spool. But once out of the spool, the same part can be measured several times by different people. That's the process.

Also, do you actually use the gage for process control or for inspection? If you only use it for inspection, the selection of parts is not important. If you use it for process control, it is important. However, you do have another option. It will involve some manual calculation, but you can substitute the process variation obtained from a capability study, or from SPC, for the process variation of the ten parts. This will actually give you a higher confidence in the result.

We use the gage for Process control and inspection. So the selection of parts is important.

How can I do the manual collection for the ten parts using capability study?


Let me know if you want to know something else

 

[Miner]

Re: Gage R&R Experts...Part variation, unilateral tolerance, destructive testing

The wire is wounded on a spool and the end of the wire is stripped. The stripped wire is then welded to a ring. It has four critical engineering dimensions. To measure the dimensions, you have to remove the wire from the spool and you can't wind it back onto the spool. But once out of the spool, the same part can be measured several times by different people. That's the process.

Okay, you are in luck. While collecting samples is destructive to the product, it is not a destructive MSA. A destructive MSA occurs when the act of measuring the part a single time destroys it. This is not the case here. The standard MSA approach will work although it does destroy your parts. I hope they are not too costly.

We use the gage for Process control and inspection. So the selection of parts is important.

How can I do the manual collection for the ten parts using capability study?


Let me know if you want to know something else

If you have Minitab, see Intro to Measurement System Analysis (MSA) of Continuous Data – Part 5a: R&R, Part Selection

If you need to do this manually, do the following:

% Study Variation = 100 * [StdDevR&R / StdDevTotal Variation]

StdDevTotal Variation = SQRT[StdDevR&R^2 + StdDevPart Variation^2]

Manually substitute the StdDev from a capability study for StdDevPart Variation

 

[aproddutoor]

Re: Gage R&R Experts...Part variation, unilateral tolerance, destructive testing

If you need to do this manually, do the following:

% Study Variation = 100 * [StdDevR&R / StdDevTotal Variation]

StdDevTotal Variation = SQRT[StdDevR&R^2 + StdDevPart Variation^2]

Manually substitute the StdDev from a capability study for StdDevPart Variation

As I said before, the same gages are used for Process control and Inspection.
I'm certain that I can't get parts which represent the entire process spread? So, How should I collect my samples?

Is it ok to collect those 10 parts without representing the entire process spread?

If so, does the standard deviation of part variation from the process capability study and the 10 parts that I select be from the same lot?

Thanks