How can you justify using a more accurate Pin Gage class?

QE1993

Involved In Discussions
We were having many rejects by a customer on a Ø.37535 ± .00025 ID.

We had been using Class X Gage Pins, which have a total tolerance of .00004. Following the 10:1 Gage tolerance rule; this class of Gage pins is just not accurate enough to catch the tolerance.

I told my calibration tech. to order Class XX pins (total tolerance of .00002). She mistakenly ordered Class XXX GP's (total tolerance of .00001). These GP's have a lead time of 5-6 weeks and the customer is expecting Gage R&R's using the new GP's way before that.

I figured, hey, this isn't ideal because we could be finding good parts to be bad, but at least there's no risk of calling bad parts good (due to the gage being more accurate than necessary).

The customer has concerns about this. They came back with essentially this: "this will increase scrap rates, which will then lead to using wrong measurement techniques to confirms parts which may result in errors."

I understand that it is not ideal, but I am really not concerned about the .00001 difference between the two classes of GP's increasing our scrap rate.

Does anyone have any insight on how you can justify using a slightly more accurate gage than necessary?
 

Michael_M

Trusted Information Resource
Justification to using a tighter tolerance device is easy, it meets the minimum requirement. I sometimes measure a ±.010 with a digital micrometer (.00005 reading).

I would be more concerned with using gage pins to check that tight of tolerance. Gage pins are essentially just a go/no-go test. You do not know if the hole is round, where within the tolerance band is the hole actually running.

Most holes that I personally deal with (that fall within this range) have some out of roundness and the gage pin will not show what the true size is. We have had to switch over to air gages because parts were failing at the customer.
 

Ron Rompen

Trusted Information Resource
Although there is a potential of rejecting good parts, you are much more likely to 'catch' a part which is marginal, when using a gauge that is more accurate than required. I have never EVER heard of a customer complaining about a gauge being 'too good' (unless of course I am asking them to pay for it).

On the other hand, I would question your decision to us gauge pins to check a diameter with such a tight tolerance. There is (in my opinion) far too much potential for error (particularly in rejecting a 'good' part because the pin isn't perfectly perpendicular to the hole). Have you considered using other methods?
 

QE1993

Involved In Discussions
Thanks for your inputs! I was thrown back by his concern. I guess the only thing I can think of is that as their supplier, it indirectly has an adverse affect on them if our scrap rates are high, so I can see his concern; however, it seemed like an odd place for his concern to manifest.

As for other methods, we initially proposed using the vision system on our control plans; however, the customer and I could not get a good correlation going between our two CMMs. So here we are, using GPs. I, too, am concerned about their use, but there's been so much back and forth already that I am ready to throw in the towel and say "whatever the customer wants." We ordered them a set of the same GP's too use, as well.
 

dwperron

Trusted Information Resource
Maybe its because of too much snow and too little coffee, but I am having a hard time getting my head around the idea that using more accurate gauging will cause more false rejections.

Can someone provide an explanation?
 

QE1993

Involved In Discussions
Here's what I think (and keep in mind, I'm no expect):

In the example Michael M provided (i.e. using a .00005 mic for a ±.010 product tolerance). If the nominal is 1, the USL is 1.010. The mic could potentially read out 1.01001. This part would be considered "bad," even though it might actually be "good" for what it is intended. If you were using a Caliper (which would be accurate enough ), it might read out 1.0100, which would be in-spec. In such a case, using a more accurate gage unnecessarily could find "good" parts "bad."

PS. I have good and bad in quotes because IMO it's all relative to the intended function of that part/feature. The feature might very well be 1.01001 (and therefore, OOT), but for what the wide open tolerance that it was given, it could still be considered acceptable.

As for my PIN GAGE situation, I'm with you. I just don't see how using a .37562 pin gage vs using a .37561 gage to capture my USL will result in a drastically higher scrap rate due to false rejections.
 
Last edited:

Jim Wynne

Leader
Admin
There's a pretty simple answer to this question, I think. Use whatever the customer is using to check the parts. If this means agreeing on pin gage sizes and buying duplicate sets, do it.
 

QE1993

Involved In Discussions
Yes, in the long-term we will certainly do what the customer wants; however, I am in need of those pin gages to inspect/sort contained lots and WIPs that we have currently. Also, the customer requires a GRR using the pin gages that is due way before I can actually get a hold of the Class XX GPs that the customer wants. The lead time on them is like 4-6 weeks + calibration time.
 

Jim Wynne

Leader
Admin
Yes, in the long-term we will certainly do what the customer wants; however, I am in need of those pin gages to inspect/sort contained lots and WIPs that we have currently. Also, the customer requires a GRR using the pin gages that is due way before I can actually get a hold of the Class XX GPs that the customer wants. The lead time on them is like 4-6 weeks + calibration time.
It appears that your customer, in the short term at least, is asking you to defy the laws of physics. You should, if you haven't already, explain the short-term constraints to your customer, as well as the long-term solution.
 

QE1993

Involved In Discussions
Already done! I think this thread was created more out of just bewilderment that he had even brought that up as "point of concern," as he called it. Like I said, of course I understand that our internal scrap rate could have an effect on the customer and it's in everyone's benefit to reduce rejects, but just the way he got so philosophical about it and claimed that it could prove that our measurement techniques weren't up to par, like he was just reaching to make some kind of comment, made me scratch my head like Am I missing something here?

Thanks for your input!
 
Top Bottom