Fixture Gage R&R - What is Required? Fixture part of the measurement system?

R&R What is required?

  • Required

    Votes: 3 75.0%
  • Not required

    Votes: 1 25.0%

  • Total voters
    4
K

kschroeder

I recently started to evaluate the requirements of QS9000 in relation to R&R studies. If Fixtures are used to "hold" or locate a part in order to obtain a reading(for a destructive test or get a indicator reading),would that Fixture be part of the measurement system that would require a R&R study?:smokin:
 
Last edited by a moderator:

Marc

Fully vaccinated are you?
Leader
If you are deriving your measurements from a part in a fixture, yes. You have to do the R&R on the combined fixture with the operators.

Let's say, for example, you have a fixture that has 6 gages attached for 6 different dimensions on a part. Each measurement gets the treatment with fixtured parts.
 
D

D.Scott

Although I agree with Marc - if the fixture relates to the measurement (i.e. a load cell for torque/tension) it would constitute part of the measurement system, I disagree that all fixtures need to be included in the R&R.

Say for example a bolt had to be held in place while a torque measurement was taken. The torque measurement device is what needs the R&R. The bolt could be held in a variety of ways and as long as it is securely held, it wouldn't matter what held it. It could be a bench vise or a fancy specialty block.

Obviously, something has to be used in the R&R to hold the part. It would stand to reason that it would be the same thing you use in actual practice but there are many instances where the actual fixture is not available for use in a MSA study. In such an instance, IMHO it would be acceptable to "replicate" the fixture.

Dave
 

Marc

Fully vaccinated are you?
Leader
As I have unhappily found over the years, where you use a fixture to measure, it is typically for reason beyond holding it which you can use a vice for (in the case cited). I'm not saying one cannot do certain studies on the individual instruments which make up the gage - that's a given in high rel shops. The system as a whole is important, though.

In the simple situation described, you would probably get away with it because of the simpicity - but as a QS assessor I would want studies on parts in the holding fixture. As soon as you increase complexity slightly over the simple torgue example you're asking for trouble if you dont. Let's say in the torque test described you are testing to a certain 4" bolt. The fixture may have a built in height that determines how much of the shaft is exposed during the test. What I'm trying to point out is more often than not a fixture has an effect on a measurement and sometimes it's dramatic.

A good example is a spring tension test holding fixture. In cases where you have small springs (such as those which return brake shoes to their inward 'standby' position after breaking the car on the 'old' foundation {drum} brakes) if the 'hook' that goes around the spring end is rounded you will get a different reading than if it is squared off. If the spring is tested a bit off center, the reading will differ. If the 'hook' that goes through the spring end is wider (if square) or the diameter is larger (if it is round) than a hook on another 'fixture' it will read differently. In one case I was involved in, both the manufacturer and we had identical spring test machines. The only way we could get the same readings is when we shipped the holding fixture with the springs to be tested which made things interesting because the springs were made in Germany and shipped here.

I point this out only to ensure you understand that a fixture ALONE can affect the reading to such a very large degree that while you want to do certain studies on the individual instrument (or instruments) attached to, or used in conjunction with, the fixture, discounting the fixture and not doing your study with the fixture could come back to bite you.

One size does not fit all. Make sure you understand how the system as a whole functions in your specific case. Spring engineers and automatic transmission part engineers will tell you that there is seldom an acceptable alturnate to using the fixture except in very basic cases.

All this said, fixture design its self is often very important. In the case of vibration testing - back in the 1980's - I was involved in test fixture design for holding CCAs (circuit card assemblies) during vibration and thermal shock testing. The test fixtures had to be designed with their resonant frequency taken into cosideration along with the resonant frequency of the CCAs. We used magnesium for vibration fixtures because other metals do not have the damping effect (most metals tend to 'ring' like a bell) magnesium has. This is basically it (from a 1988 Mac picture):

fixture.gif

Since we were montoring vibration characteristics and testing for effects to the CCAs themselves, the fixture was one of the most important parts of the 'system'. We even did a finite element analysis on the fixture to ensure we well understood its characteristics.

Summary: Never under estimate the importance of the effects of a test or holding fixture as a part of a system as a whole. And if you do not understand the significance of saying you are measuring something in a 'restrained' configuration (as is often done with very flexible parts), you may need to learn more about the role of fixturing in measurement and test of parts.

Just my 2 cents :thedeal:
 
E

energy

Can't change my vote?

While this is a Poll thread, one has to explain the reason for doing so. I voted "not required". In the example of using a vise to secure a bolt for torque testing, there was no thought to where the bolt was being held. If you were simulating the test to prove that it doesn't snap when torqued, the fixture would have to be a stack up similar to actual use. I guess that makes it part of the test. I can't take back my vote.
But, if I held a bracket in a vise while torque testing a brazed clinch nut, I wouldn't say it is required to be included. Unless I had to simulate torque breakage in actual use such vertical, horizontal or upside down positioning. So, it would appear that most "holding" fixtures would have to be considered part of the process. Not having exposure to QS, are they really that finicky? Or, is this an area subject to educated opinions of the assessor? I mean, it makes sense if actual operation is being used as the standard. Do you really have to argue the non relevance of your holding fixtures that you feel are not important to this study?
:confused: :smokin:
 

Marc

Fully vaccinated are you?
Leader
If you haven't done the QS dance, you haven't experienced 'pickiness'. Comparatively, ISO is a walk in the part. Yup - QS auditors can be, and typically are, that 'picky'. :thedeal:

Even in production. When I worked with one client in QS, they had an automated welding process. The holding fixtures measured nothing. None the less, because of their impact on the manufacturing process (weld location) was critical (and the parts were big), every fixture had to be removed every 24 hours and literally calibrated. They were in the calibration system just as measurement gages were.

QS-9000 brings new meaning to the word 'picky'. But then again, I do admit that in this case it was a necessary, appropriate approach with consideration to the part, the process as a whole, and the end product. Tracking the process proved it was cheaper to calibrate each fixture daily than to tolerate the scrap levels without daily calibration. They had 3 complete sets of holding fixtures for that line - over 80 fixtures.
 
K

kschroeder

I guess it would help to give more information for my particular situation:rolleyes: I realize that in some situations the fixture may determine the outcome of the measurement being performed or the quality of product produced,I'm just not sure in my situation it does. Let me explain.We produce alot of different powdered metal products,each of which may have a separate holder used to destructive test a part.The part basicly is just supported by the holder.(Example:supports a hub,nib,flange).Parts are compression tested using a UTM. The UTM is calibrated (certified by a outside lab )and the holding fixture is calibrated. The load is applied to the part until failure.Problem#1 How do I do a part specific R&R(for PPAP)?Problem #2 How do I do a gage R&R (the holder)? Problem #3 Do I have to? We are a QS9000 company and it has not come up but I wonder if its because it hasn't been found.:) Thanks for your previous insights!:cool:
 
D

D.Scott

We are QS-9000 and have similar problems. If you are worried solely for PPAP reasons, the customer can waive any requirement of PPAP they like and it takes precident over the QS-9000 requirements. If your customer will accept your R&R and not require a part specific one, you won't need to address the problem.

In the instance you mention, it sounds as though the test cannot be done without a specific calibrated holder. This is just what Marc was pointing out and I agree with him - you need to include the holder in the R&R for the UTM.

The situation I was trying to point out was one in which it made no difference how the part was held. The holder would be independant of the equipment.

I assume you are doing the test in an "in-house-lab" to verify your own work, or testing incoming product for suitability. If this is the case and you are QS-9000 - YES, you would have to do a gage R&R on the equipment. In the case you describe, the R&R should include the holder with the UTM. There would be no logic in requireing a seperate R&R on the holder itself as it is intrinsic to the complete test.

That is all of course - IMHO.

Dave
 
Top Bottom