Test Uncertainty Ratio (TUR) for Ring Gage Calibration

[czielsdorff]

Hello all,

I have read several threads, but I could not find the specific answer I was looking for (hopefully I know what I am looking for ).

The company that I work for has a calibration system and we routinely send gages out for calibration from certified labs. Recently we have been improving our own knowledge on calibration and a question has been raised regarding the TUR for thread ring gages.

Per our calibration lab the TUR they provide on the certs is less than 1:1 (about .7:1). We have discussed this with them and they are confident that this is the best level of calibration short of sending it directly to NIST. They also feel that other labs are not stating the "true" uncertainty and therefore may be stating a better TUR value.

Does anybody have experience or insight into how this is done throughout industry? Are people commonly accepting calibration certs with a less than 1:1 TUR? Any guidance or direction would be appreciated.

I am far from a calibration expert, but I have been reading a lot of papers lately.

Thanks,
CZ

 

[Stijloor]

A Quick Bump!!

Can someone help?

Thank you very much!!

 

[BradM]

Hello CZ!

You asked a very good question.

First, it's good that you seek out more information regarding TUR. There are multiple way to calculate it, and even more ways to present it.

Consider: The lab may report their best TUR for the month, their best ever, their worst, an average, etc. Too, I tend to rely on a lab this is more forthcoming with their uncertainty budget and how they calculated it, even if it not as good as others. Unless I know how they arrived at those values, I approach it with a healthy skepticism.

Having said that, .7 to 1 sounds a bit weak. I would see what other competent labs are reporting and ask them about their uncertainty budgets.

There is not a problem... until there is a problem. If one of your ring gauges fail, you'll need to check within the range of the reported uncertainty.

But again, they may be reporting an ultra-conservative, worst case uncertainty. Which in reality, may be as good or better than other equivalent labs.

Not sure I answered anything; or if I just made the water muddier.

 

[czielsdorff]

Brad,

Thanks for the response. Through my discussions with this lab and through our work with them, I have more confidence in them than our other labs and I believe we are getting a higher level of calibration (based on their opinion of the equipment used). Unfortunately they are stating the worst-case TUR with several different factors included. While this results in a lower TUR, they feel strongly that it is the right way to calculate it.

I appluad them for sticking to their beliefs rather than inflating the TUR value. However, this puts us in a difficult position because other people do not want to accept the calibrations with that TUR value.

We are currently in the process of requesting more calibration quotes with specific verbiage regarding the TUR and hopefully this will shed some light on the situation.

The water is already muddy, so you certainly didn't make it worse.

Thanks,
CZ

 

[DietCokeofEvil]

I'm going to throw more in here regarding this- from the calibration aspect.

I remember when we had to do our uncertainty budget for adjustable thread rings- and I still feel that it cannot be done.

To simply report the uncertainty of the set plug is not correct IMO. It does not account for basics like thermal expansion, technician training etc. Thread rings is one of the few areas where technician training has a direct influence on the calibration of the gage.

When calibrating a thread plug for instance- there is a procedure that is followed. Using the right wire, pressure, setting the machine properly according to procedure, subtracting the correct constant removes much of the possibility of the technician influencing the calibration of the gage.

With a thread ring, the end result is up to the technician's ability to determine how close the gage is. Since every technician seems to have their own opinion of what that is- I've talked to technicians who firmly believe a NoGo thread ring should not go onto the NoGo set plug. Everyone has a different idea of how to quantify a "2 finger drag". Plus, how do you physically measure how far over a thread ring is from the set plug. If your set plug is 0.3945 and your thread ring is exactly 0.3945, it's not going to fit on the set plug. You have to loosen it so it fits- how do you measure that numerically? You opened it 0.0002? 0.0010?

Since the tolerance on a standard pitch diameter thread ring is 0.0002-0.0006 or so depending on the size and pitch of the ring-I don't see how holding that tolerance can really be done when it all comes down to feel.

Maybe I'm missing the big picture on this- since I measure thread rings all the time and have been doing so for a number of years- but I think that people are putting way too much stock into a tight uncertainty and TUR ratio on thread rings.

 

[czielsdorff]

Thank you for the response DietCokeofEvil, this is exactly what I am after. It sounds like you have seen labs that only report the uncertainty of the plug gage but do not include the uncertainty of the process or other factors when calibrating the ring gage. Is this correct?

What are we supposed to do? Either the lab only includes the set plug uncertainty or they include all of the different factors (more accurate) and get a poor TUR value?

This is the conundrum my company is currently facing and we still don't have a clear direction.

Thanks,
CZ

 

[DietCokeofEvil]

Since this question has been asked of us by a few customers now, I did a search of scopes on the A2LA website- of the 15 or so I looked at, 2 of them reported just the set plug uncertainty. The rest were either right around what we report, or even larger.

I'm currently checking to see what is all included in our uncertainty budget and then I was going to talk to our accrediting body to see what their response is. When I get some answers, I'll let you know.

 

[Pezikon]

Uncertainty contributors in thread ring gage calibration:
1. Uncertainty in the setting plug calibration
2. Roundness errors in the setting plug
3. Variations in 'feel' when adjusting the gage
4. Roundness errors in the ring
5. Linear pitch errors in the ring and the setting plug
Source: AMTMA "Searching for Zero" dated 2004

I've never had to do an uncertainty budget for a ring gage. I don't know how to quantify these uncertainty contributors. I'm just listing them for your info.

For what it's worth, you can also calibrate ring gages using a a lab master comparator like this one: prattandwhitney.com/Content/LabMaster_Universal_Length_Comparator.asp