Checking Threads - Discrepancy in threads that are machined In-House vs. Outsourced

[KyloDen]

I have recently had an issue pop up dealing with thread gages, specifically thread rings. We machine components with external NPSM and BSP threads in house as well as have outside vendors do some to alleviate our work load. The problem is a discrepancy in the threads that are machined in house vs. on the outside.

Internally, we have master setting plugs for all of our Go and NoGo thread rings. We also use these rings to check parts incoming from outside suppliers. We are running into issues at receiving inspection where a supplier's parts do not fit our thread rings. However, I have had one supplier in particular come in to discuss the discrepancy, and he brought along his thread rings that he uses for inspection. The fit the parts perfectly. He also brought along copies of his certifications from an accredited calibration lab showing the his rings fall within the allowable limits of the particular thread size. The certs they had showed that their thread rings were set close to the high limit of the tolerance, but still within tolerance. I don't have a number to equate to the rings we use in house, so I can't compare apples to apples, but would suspect that we have our rings set internally toward the lower end of the tolerance range.

Our purchasing department is going crazy (and I agree) because we cannot reject parts that fit a gage that has been confirmed as being the correct size, but also do not pass our internal checks. The only solution I can see (at least for the near future until I can do some more research into this) is to have the supplier send us their ring gages and have them set to our master setting plugs to make sure there is agreement between the gages. However, I also realize we are not in the business of setting gages/being a gage lab.

As I see it, if I do it for one supplier, then I need to do it for all my suppliers who machine threads for us, which is something I really don't want to end up doing. Just wondering if anyone else has run into a similar problem and how they may have resolved it.

 

[Eredhel]

Re: Checking Threads

If your company doesn't mind doing it I would send out yours and your supplier's and find exactly how different they are from each other. You might find a difference between them that highlights something.

 

[Eredhel]

Re: Checking Threads - Discrepancy in threads that are machined In-House vs. Outsourc

I forgot to mention. If you have a comparator and some Repro rubber you can check the ring gages yourself and get some good info. If you're talking plug gages have you wired them to see where they are?

 

[KyloDen]

Re: Checking Threads

If your company doesn't mind doing it I would send out yours and your supplier's and find exactly how different they are from each other. You might find a difference between them that highlights something.

I had actually suggested doing that and was met with "why would we need to do that, we know how to set our rings" from the head of the metrology lab, so I'm not sure how much more I can more that stone without going over people's heads. It's almost like talking to someone from the accounting dept.

 

[KyloDen]

Re: Checking Threads - Discrepancy in threads that are machined In-House vs. Outsourc

I forgot to mention. If you have a comparator and some Repro rubber you can check the ring gages yourself and get some good info. If you're talking plug gages have you wired them to see where they are?

I actually did have the plugs wired. They did fall within the allowable range, though I do recall them trending toward one end of the range (I forget the exact measurement without having the report in front of me).

 

[BoardGuy]

Re: Checking Threads - Discrepancy in threads that are machined In-House vs. Outsourc

Some thoughts come to mind based on same experience. Here a few questions to ask and comments:

a) Have your set plugs been calibrated by an outside lab what do the certifications show?
b) What thread class is specified on drawing and are both your and your supplier’s gages meeting it? If unspecified that could be an issue.
c) What tread standard is in play here? If one gage meets a British thread standard and the other a U.S. thread standard the gages have deference in thread fit even if they are the same thread call out.
d) Your gage could have barreled threads meaning the center of the thread form is bigger like this () or a drunken thread form.

How we solved this was to send both our and our suppliers gages out to gage manufactures lab who analyzed them. In our case our gages were found unacceptable because they had drunken threads.

 

[ncwalker]

Re: Checking Threads - Discrepancy in threads that are machined In-House vs. Outsourc

Also, if we are talking large threads, it could be temperature variation between you and the supplier. You should send your gages to his floor for a little while and see what comes of it.

 

[KyloDen]

Re: Checking Threads - Discrepancy in threads that are machined In-House vs. Outsourc

Some thoughts come to mind based on same experience. Here a few questions to ask and comments:

a) Have your set plugs been calibrated by an outside lab what do the certifications show?
b) What thread class is specified on drawing and are both your and your supplier’s gages meeting it? If unspecified that could be an issue.
c) What tread standard is in play here? If one gage meets a British thread standard and the other a U.S. thread standard the gages have deference in thread fit even if they are the same thread call out.
d) Your gage could have barreled threads meaning the center of the thread form is bigger like this () or a drunken thread form.

How we solved this was to send both our and our suppliers gages out to gage manufactures lab who analyzed them. In our case our gages were found unacceptable because they had drunken threads.

Yes, we do send the master plugs out on an annual basis. The last cert from August of 2016 shows that it (the setting plug in question) is riding right at the low end limit. This particular one is a 1¼-11½ NPSM, but it would also apply to other NPSM and BSPP threads.

Also, I just found out that our internal thread rings are adjustable, while a couple of our suppliers have solid rings. Is there a preferable style of gage to use? Or should an adjustable ring and solid ring be interchangeable when it comes to making the accept/reject decision. I haven't researched it yet, but my initital thought is that it shouldn't matter.

 

[BoardGuy]

Re: Checking Threads - Discrepancy in threads that are machined In-House vs. Outsourc

Yes, we do send the master plugs out on an annual basis. The last cert from August of 2016 shows that it (the setting plug in question) is riding right at the low end limit. This particular one is a 1¼-11½ NPSM, but it would also apply to other NPSM and BSPP threads.

Also, I just found out that our internal thread rings are adjustable, while a couple of our suppliers have solid rings. Is there a preferable style of gage to use? Or should an adjustable ring and solid ring be interchangeable when it comes to making the accept/reject decision. I haven't researched it yet, but my initital thought is that it shouldn't matter.

It always has been my preference to use adjustable.

I am assuming the associated drawing for this thread is calling out the requirement to meet ANSI B 1.20.1. If not, gaging by suppliers could be based on British Standard Pipe (PSP) or one of the ISO variants. BSP is not always the same thread requirements as ANSI

Assuming that you have thread specification and suppliers gages meet the same then it down to an issue of the thread form which will require verification of all gaging used (yours and suppliers) to determine why there is a problem.

 

[Gkokal]

Re: Checking Threads - Discrepancy in threads that are machined In-House vs. Outsourc

I use a MasterScanner in our calibration lab to check both thread gages and thread rings. One of the biggest problems that i have seen pop up in measuring thread gaging is the flank angles. If you use the three wire method, then the angles of the flanks can sit the wires either higher or lower in the thread, and give you a different pitch diameter.

We had a lot of problems correlating our supermic/ULM numbers to the MasterScanner until we realized that the PDs that give us problems between the two methods are always correlated to the gages that have bad flank angles.

As far as the adjustable vs solid ring discussion goes, i lean more towards solid. If you really think about it, a split ring is more tri lobed than cylindrical. The southern style gages are oval. So even though you can effectively set them to a plug, your contact points on the part will be smaller, and it will wear quicker than full contact.