How to measure circular runout of an internal thread?

[Mr.Chipeater]

Imagine a rectangular block with a 5/8-24 ID thread on either end with a 3/8 diameter bore between them. The print calls for 0.002 circular runout on the major, minor, and pitch diameter of each thread with the through bore as the datum.

Due to other features on the part, we fixture the block inside a lathe so we end up single pointing both the bore and the threads so in theory it should be as accurate as the year old Mori NLX its cut on. However the customer is claiming some crazy runout measurements from 4 to 10 thou out and is rejecting dozens of parts. We took a look, saw nothing wrong, shipped the same rejects back and now miraculously only a couple are getting rejected. Its an ongoing order and they keep rejecting parts from every batch for runout which is causing a lot of headaches for us. Its pretty clear to me that their method is not consistent but haven't had any luck convincing the QA guy of that. Last we spoke, he said if I think their method is bad, its on me to figure out a better method for them to use.


So how would you measure runout of the pitch diameter on an ID thread?



FYI, their method consists of inserting a gauge pin for the part to rest on, screwing a threaded plug with a hole into the thread, placing an indicator on the hole of the plug and spinning the part.

 

[plastikdreams]

Ask how they are measuring it and request to see the process.

 

[Mr.Chipeater]

Ask how they are measuring it and request to see the process.

As described to me, their method consists of inserting a gauge pin for the part to rest on, screwing a threaded plug with a hole into the thread, placing an indicator on the hole of the plug. Then they spin the part around and watch the dial indicator to get a measurement.

 

[plastikdreams]

Sounds hokey at best...I'd ask to see it in person

 

[Mr.Chipeater]

Sounds hokey at best...I'd ask to see it in person

I agree and the inconsistent results also say its hokey but there's no convincing the QA that it is without "proof of a better method"

 

[Kallam]

A threaded bolt or stud will never run concentric there is always some clearance between the threads .

 

[guythatbrews]

Their check is well crap. How are you checking it?

It is tough to check. Put the part on a tapered arbor in the 3/8 hole and put the assembly on bench centers. Confirm the arbor is running true. Mark the runout on the block all 4 sides each end. Replace arbor with threaded jo plugs runout should becthe same.

 

[guythatbrews]

Their check is well crap. How are you checking it?

It is tough to check. Put the part on a tapered arbor in the 3/8 hole and put the assembly on bench centers. Confirm the arbor is running true. Mark the runout on the block all 4 sides each end. Replace arbor with threaded jo plugs runout should becthe same.

Even better put on tapered arbor get jo plugs with a through hole big enough to pass arbor. Check TP directly on jo plugs in a bench center.

Problem is special high dollar jo plugs.

 

[Mr.Chipeater]

A threaded bolt or stud will never run concentric there is always some clearance between the threads .

That's what I said! I asked if their plug could adjust for differing pitch diameters between parts. The question was lost on them...

Even better put on tapered arbor get jo plugs with a through hole big enough to pass arbor. Check TP directly on jo plugs in a bench center.

Problem is special high dollar jo plugs.

how do jo plugs solve the problem of clearance between the mating pitch diameters allowing the center of the male thread to deviate from the female thread?

 

[Vancbiker]

Machinist reproduction material.

Reprorubber®

Reprorubber^® is a metrology-grade self-curing casting solution that sets the industry standard for self-curing casting applications.

www.reprorubber.com

Check the reproduction on a comparator

 

[guythatbrews]

Threaded jo plugs are self-centering, assuming a vee thread form. The pitchline forms a continous centering cone. They must be shouldered on the face of the part. If the face is not perpendicular to the pitch line they will tilt.

They do make jo plugs with spilt threads that spring out but I've never used any of that type.

Without a doubt your customers check is bullshit. If the 3/8 pin is an interference fit and long enough to get in a vblock it might be ok. And doesn't sag due to part weight. Can they repeat their measurement reliably? If they can't demonstrate that you can prove their way don't work.

Do you as the maker have a way to check the TP or are you assuming it's pretty close?

 

[crossthread82]

Their way of checking it is total crap. Obviously if the threaded plug screws in it has to have clearance which means its going to sit to one side when rotated...duh.

Xray or CT would be the best but also the most cost prohibitive. A transfer measurement of sorts like @Vancbiker suggested could work as well. For quick and dirty though I'd just get a gage pin that barely fit's through the minor on the thread and indicate it while spinning the part on the 0.375 bore.

I'm willing to bet if they're rejecting parts it's b/c they're screwing something into the part with a 5/8-24 thread that has an end the goes into the 0.375 bore with a very close fit. The "slop" in the threads is cause the end of the part to rub/gall on the side of the 0.375 bore. Which would be a design flaw on them since for that to work you'd need a piloting shoulder behind the thread to hold center. But as usual they're going to say it's your problem b/c it's impossible that their design sucks.

 

[Conrad Hoffman]

The major and minor should never touch anything, so no idea why the spec. Test method seems to have too many pitfalls to work. Wondering if you can get a custom thread gauge made that has a pilot to fit the center hole? Everything ground on centers. Or make one yourself on the lathe just to try the concept.

 

[guythatbrews]

Threaded hole location gages. Also called jo plugs. These are solid. They have to shoulder on the part to centralize to the hole. Check out the FAQ section.

Home -Centerline Location Gages

Depend on Spence Industries, Inc. for various JO-PLUG® centerline location gages and other industry-approved tools.

jo-plug.com

These are split so they compress.

Thread Hole Location Gages | Vermont Gage

These aren't bolts. They are precision gauges. Neither site has any hard numbers on location accuracy and repeatability. Call these folks and ask about them. Maybe report back with what you learn.

The major and minor should never touch anything, so no idea why the spec. Test method seems to have too many pitfalls to work. Wondering if you can get a custom thread gauge made that has a pilot to fit the center hole? Everything ground on centers. Or make one yourself on the lathe just to try the concept.

Exactly so. Major and minor is designed to be clearance unless this is some special thread. How is this even called out on the drawing? Is it a verbal note?

Checking the minor diameter of an internal thread and assuming it is concentric with the pitch diameter is taking stuff for granted indeed.

 

[Mr.Chipeater]

Exactly so. Major and minor is designed to be clearance unless this is some special thread. How is this even called out on the drawing? Is it a verbal note?

Checking the minor diameter of an internal thread and assuming it is concentric with the pitch diameter is taking stuff for granted indeed.

Its got a standard thread spec with the circular runout callout under it. The callout is defined the same as my highly technical replication.

I agree about only checking minor. However if the threads are single point cut in the same setup as the datum bore on a lathe, all on center, I fail to see how there could be the 0.004-0.010 of runout that they are claiming. Taper, sure. pitch dia or size changes, sure. But none of that would impact runout. If I take the same tools and cut 2 diameters on the machine to easily measure runout, I get <0.0002 every time.

If anyone has suggestions on potential causes of that much runout, I'm all ears but I know the part isn't moving between tools nor are there any chips to get in the way.

Without a doubt your customers check is bullshit. If the 3/8 pin is an interference fit and long enough to get in a vblock it might be ok. And doesn't sag due to part weight. Can they repeat their measurement reliably? If they can't demonstrate that you can prove their way don't work.

They cannot. They got wildly different results from measuring the same "rejects" twice. I brought this to their attention but QA guy goes "its the best method we have, so we're going to keep using it"

The major and minor should never touch anything, so no idea why the spec. Test method seems to have too many pitfalls to work. Wondering if you can get a custom thread gauge made that has a pilot to fit the center hole? Everything ground on centers. Or make one yourself on the lathe just to try the concept.

I did create one and it threads in just fine. I actually made several with increasing pilot diameters but the customer was unmoved. They said "that not how its assembled so its irrelevant." They then explained that the part is assembled by screwing in 2 different pieces into each end, then a rod is inserted through one piece, through the center bore, and out the other end.

 

[guythatbrews]

They cannot [repeat their inspection relts]. They got wildly different results from measuring the same "rejects" twice. I brought this to their attention but QA guy goes "its the best method we have, so we're going to keep using it"

Unless you can get around this idiot you are effed no matter what you do. Can you make an end run around him and get someone to listen?

If you are single pointing and taking great care, concentric and parallel, you've got to be close.

 

[johansen]

Threaded jo plug but hollow instead of a pin extending out. Make one yourself and test it out.

Make the hole the same size as the 3/8 hole.

Slip a gauge pin .001" less diameter down the jo plug and it should go into the bore without much resistance

 

[Seesecurity]

Who made the threaded plug and has it been calibrated?

 

[Mechanola]

Obviously an incomprehension of that thread’s purpose. It’s a fastening thread. If they want to use it as a thread for moving parts (longitudinally), teach them that the tolerances on major, pitch, and minor diameter don’t correspond to their idea. One moving thread is the square or flat thread.

 

[Conrad Hoffman]

Its got a standard thread spec with the circular runout callout under it. The callout is defined the same as my highly technical replication.

I agree about only checking minor. However if the threads are single point cut in the same setup as the datum bore on a lathe, all on center, I fail to see how there could be the 0.004-0.010 of runout that they are claiming. Taper, sure. pitch dia or size changes, sure. But none of that would impact runout. If I take the same tools and cut 2 diameters on the machine to easily measure runout, I get <0.0002 every time.

If anyone has suggestions on potential causes of that much runout, I'm all ears but I know the part isn't moving between tools nor are there any chips to get in the way.


They cannot. They got wildly different results from measuring the same "rejects" twice. I brought this to their attention but QA guy goes "its the best method we have, so we're going to keep using it"



I did create one and it threads in just fine. I actually made several with increasing pilot diameters but the customer was unmoved. They said "that not how its assembled so its irrelevant." They then explained that the part is assembled by screwing in 2 different pieces into each end, then a rod is inserted through one piece, through the center bore, and out the other end.

IMO, that means they really don't have a good understanding. Since your gauge uses the same alignment hole, the one in the middle of the part, it really does tell them what they want to know.