Gage Recall Analysis - Threaded Plug Gage has failed Calibration

[gianni]

I have a consultant who is telling me that after a threaded plug gage has failed calibration that the gage is still acceptable if it is within 10% of the gage tolerance. He is citing from ANSI B1.2 but I do not have a copy of the standard. Is this accurate?? He has not shown me the proof but he insists he is very familiar with the standard. Help!

 

[Jen Kirley]

I don't know what your consultant is talking about. Thread Check.com (I am not affiliated) quotes ANSI/ASME B1.2-1983:

...the ?NoGo (Hi) functional diameter is acceptable when the NoGo (Hi) thread plug gage applied to the product internal thread does not pass over the thread more than three complete turns.?

Does your specification give a tolerance? Van Keuren (also no affiliate, see page 13) cites the industry practice

The normal practice is that the gage tolerance should not exceed 10% of the product tolerance. Gages are furnished in four classes of tolerance, XX, X, Y and Z....Standard practices indicate that 10% of your product tolerance is divided between the Go/Not Go gages. Go gages are plus tolerance and not go gages are minus tolerance.

Please note that section of Van Keuren's catalog was referring to plug gages.

 

[gianni]

Thank you for your reply. So if I understand this correctly, for example, we use standard nuts and the tolerance is ? 0.0041". 10% of this is 0.00041". The gage fails calibration using the X tolerance of 0.003" but falls within the 10% criteria. Can this gage still be considered acceptable then? Or, would I reclassify the gage to a Y tolerance for example?

 

[Pezikon]

That sounds about right, assuming you meant 0.0003" and not 0.003".
nut tolerance: ? 0.0041"
10% of nut tolerance: ? 0.00041"
X class thread plug gage tolerance: +0.0003" (tighter than 10%)

If, for example, the certificate of calibration reports that the thread plug gage was measured to be +0.0004, that would be out of tolerance by class X spec but in tolerance per the 10% rule. As you said, you could reclassify the thread plug as a class Y since the thread plug gage meets class Y spec and not class X spec. But then the class Y spec may be too loose (ex: 0.00045") and not as tight as your process tolerance of 0.00041. This would only be an issue if you sent the thread plug in for calibration with instructions to calibrate the thread plug to class Y spec. You could potentially get a thread plug back measuring 0.00044, which meets class Y spec, but not your process spec. When you glance at the calibration cert and see PASS, you might assume the thread plug is good when it's not. What you could do is instruct your local cal lab to "limit cal" the thread plug to a specific tolerance. Give them your 10% tolerance. Or just replace it and move on with your life.

 

[Jen Kirley]

Thank you for your reply. So if I understand this correctly, for example, we use standard nuts and the tolerance is ? 0.0041". 10% of this is 0.00041". The gage fails calibration using the X tolerance of 0.003" but falls within the 10% criteria. Can this gage still be considered acceptable then? Or, would I reclassify the gage to a Y tolerance for example?

Thread plug gages are go/no go by design. If you want to verify product for MRB or recall purposes, a more sensitive measuring method should be used, and its resolution (measuring increment) should be 10 times that of your tolerance. That is, if your tolerance is in thousandths the gauge should read in ten-thousandths. I say this knowing I run the risk of telling you something you already know, but I want to be clear.

Since thread plug gauges are tapered I would not trust myself to perform an in-house calibration without specialized equipment. Replacing a thread plug gauge may be less expensive than getting it calibrated by an outside service.

 

[lukasz]

I guess the new thread gage has to be calibrated anyway. All equipment needs to be reliable. Sometimes new gages are faulty, too.


Ad rem:
OK or NOK side is out of spec? In which direction (in plus, so there's more material as it should or is it out of spec in minus)?

 

[gianni]

I think I may not have worded my question correctly and we have a misunderstanding. I understand I can just replace the gage, which is not at issue. I have to prove that I did not ship any suspect product using a known failed gage. However what I am after is, is there some type of allowance that would indicate "well the part tolerance is 10x larger than the gage tolerance, so even though the gage is out by x amount, the gage is still acceptable for the tolerances we are using for our product." I was told that if the ACTUAL result falls out by less than 10% of the gage tolerance then the gage is still acceptable for use and/or acceptable, as far as assuring no suspect product was shipped. I have gages that were on the plus (from lack of use) and on the minus side (due to wear). The minus side gages need to be replaced but again I need to show that there is no suspect product that was accidentally accepted as a result. Is there a "rule of thumb" or "industry standard" or noted in ASME B1.6 (which I do not have)? Thanks!

 

[lukasz]

gianni

if
OK side is to big or
NOK side is to small

you are in the safe area. This is why I asked about that. I'm a gagemaker and have some skills here.

If none of the cases above happen, there can be a problem. But from my experience, all cars are full of parts out of specs. Is the part you manufacture safety-related? How much out of spec the gage is? What's the thread tolerance? Did you measure the thread (not gaged) in any way?

 

[stefanhg]

HI gianni,

ANSI/ASME standards do not prescribe wear tolerance as to allow the pitch diameter of the plug gauge to be smaller than the respective tolerance limit of the product.

ISO 1502 for thread gauges for metric thread specifies following wear tolerances:
Example:
Product #12-32 UNEF - 2B, PD tolerance 0.0041,
max. PD = 0.1998, min. PD = 0.1957
GO thread plug gauge min. PD = 0.19544 worn
NOT GO thread plug gauge min. PD = 0.19960 worn

Similar case with standard ISO 15872 for thread gauges for UNJ threads:
GO thread plug gauge min. PD = 0.19534 worn
NOT GO thread plug gauge min. PD = 0.19945 worn

 

[Jen Kirley]

It is late so my answer might not be profound. But since a plug gage is designed to only perform quick checks with, when there is doubt the measurement for MRB analysis should be taken with another, more sensitive instrument and acceptability should be determined based on specification tolerance and/or customer permission. It really is that simple, or it should be.

As for calibration of a plug gage: compare the cost to that of purchasing a new one.