What's new
What's new

What could cause my thread pitch to be off in a G76 cylcle?

vmipacman

Cast Iron
Joined
Nov 21, 2014
Location
Virginia, USA
So on a Hitachi Sieki, with a Fanuc 6T, threading a 4"-8UN x 1" lg
Made 4 rush parts. Checked pitch with wires and tracked down a 3-1/2" electrical conduit fitting that should have had 4"-8 threads just as a feel good. I cut my threads on the low side knowing that the thread purpose was just for positioning, not really loaded. My female conduit fitting threaded on just fine if not pretty loosely.
Got a call the mating parts don't thread on. Hang up about 3.5 threads in. My pitch is off. Gaining pitch. Checked my G76 cycle. F value = .125, rpm is 200.
Never had this problem before! Really annoyed at myself to get that call but really confused how this happened. Any secret values in fanuc or nuances that could cause this?
Thanks
 
Hi. I'm sorry to post this here. But I need to know how to post my thread... Can anyone help?
 
So on a Hitachi Sieki, with a Fanuc 6T, threading a 4"-8UN x 1" lg
Made 4 rush parts. Checked pitch with wires and tracked down a 3-1/2" electrical conduit fitting that should have had 4"-8 threads just as a feel good. I cut my threads on the low side knowing that the thread purpose was just for positioning, not really loaded. My female conduit fitting threaded on just fine if not pretty loosely.
Got a call the mating parts don't thread on. Hang up about 3.5 threads in. My pitch is off. Gaining pitch. Checked my G76 cycle. F value = .125, rpm is 200.
Never had this problem before! Really annoyed at myself to get that call but really confused how this happened. Any secret values in fanuc or nuances that could cause this?
Thanks

Before you go ripping your machine apart, I'd make sure your parts are actually right... And hopefully the
mating parts are WRONG.. It would not be the first time that has happened..

Go out to your machine right now, and no matter what material you have in right now, just do a skim
with a threading tool. Cut an 8tpi thread, doesn't have to be deep, just a scratch, and do it at
200 rpms.

Now, you have a pitch gage, right??? No.. Do you have a 1-8 tap, that works too.. No? Good
thing its a rather course pitch, calipers will work... Even go down to the hardware store and
buy a 1-8 bolt...

Check that first...

Wouldn't be the first time that a customer has said I need *This*... And they really
needed *That*.
 
I dialed in the threading cycle on the first part, then ran the last three start to finish. Thanks for the suggestions Bob! I already remade one ready to thread so I'll do that in the morning. Luckily they just shaved three threads so it works for their needs but I'm not happy. Very imbarassed still. Need to figure this out.
 
Hi. I'm sorry to post this here. But I need to know how to post my thread... Can anyone help?

Hi Saldash, welcome,

To make a new thread go to the Forum tab above and then click on whatever category you would like to post in. When that page is up there should be a tab in the upper left that says "post new thread"
 
Are you accidentally using CSS by chance? I was getting some weirdness when I used G96 (default out of CAM) then it occurred to me the threads were fine after 2-3 pitches, so the issues must be caused by the acceleration (which was still in progress by the beginning of the part due to a small lead-in distance). It's cutting at a constant diameter too, no reason to change speed anyway.
Changed to G97 and all was well. I should have known, but I was new to lathe-land and just figuring things out at the time.
 
It's pretty obvious when you use CSS while Single point Threading-----the Threads are garbage and not salvageable.

I think the mating part is bad. I can check a .125 pitch with calipers, your feel good gauge wouldn't go on if the pitch were wrong. If you made 4, they would look alike. If the part were moving in the Chuck the part wouldn't be mistaken for a good part. (I hope)

R
 
When they said "gaining pitch" maybe they meant gaining pitch diameter? Are you sure that it's not just a a taper issue?
 
Pure coincidence, but a couple years ago I made a couple dozen parts with a 4"-8 internal thread. I first made a gauge on a manual lathe for checking the female threads. Cut that to the middle of the tolerance measured with wires.

I used a bar with top notch inserts for the internal thread, cut to the upper end of the tolerance since the end use of the part would work best with fairly loose threads. I used an insert whose largest internal pitch was 8 tpi. Threaded the first hole and my gauge would only go in a couple turns and lock up. Material was 1117 and the thread was slick as could be. Checked my numbers and re-ran the thread cycle. Final pass took no metal but skimmed off most of the Dykem I'd sprayed in the thread.

Finally. for no reason other than not being able to think of anything else to try, I switched to an insert that was good down to 4 or 5 tpi. Futzed around with offsets till I was about cross-eyed as I didn't want to scrap an $80 chunk of stock. Ran the cycle, and the gauge threaded in perfectly with the little bit of slop I'd expected. Rest of the parts came out on the money with no additional spring passes, etc.

About a year later I was looking for some top notch grooving inserts and happened to notice the manufacturer had changed the max internal pitch on the insert I'd initially used to 9 tpi rather than the original 8 tpi. Nice to know, as I'd wasted about 3 hours to figure out they didn't work for 8.
 
I had this problem that was caused by slipping spindle drive belts, just a thought. I could cut 3/8" threads just fine, 1-1/8 and the belts would slip screwing up the pitch unless I took very light cuts.
 
Don't know if it would actually matter or not, but I was always told pitch should be input as six digits even if the last few are zeroes.

Any chance your stock is moving?
 
Last edited:
Pure coincidence, but a couple years ago I made a couple dozen parts with a 4"-8 internal thread. I first made a gauge on a manual lathe for checking the female threads. Cut that to the middle of the tolerance measured with wires.

I used a bar with top notch inserts for the internal thread, cut to the upper end of the tolerance since the end use of the part would work best with fairly loose threads. I used an insert whose largest internal pitch was 8 tpi. Threaded the first hole and my gauge would only go in a couple turns and lock up. Material was 1117 and the thread was slick as could be. Checked my numbers and re-ran the thread cycle. Final pass took no metal but skimmed off most of the Dykem I'd sprayed in the thread.

Finally. for no reason other than not being able to think of anything else to try, I switched to an insert that was good down to 4 or 5 tpi. Futzed around with offsets till I was about cross-eyed as I didn't want to scrap an $80 chunk of stock. Ran the cycle, and the gauge threaded in perfectly with the little bit of slop I'd expected. Rest of the parts came out on the money with no additional spring passes, etc.

About a year later I was looking for some top notch grooving inserts and happened to notice the manufacturer had changed the max internal pitch on the insert I'd initially used to 9 tpi rather than the original 8 tpi. Nice to know, as I'd wasted about 3 hours to figure out they didn't work for 8.


I am getting lost in this story....

You said that you threaded the internal threads to the upper limit, but you was using your previous part as the gauge.
Since the gauge wasn't fitting, how did you determine that you were at "the upper end of the limit"?

I'm guessing simply by X position - determined by touch-off?

When you went with the bigger insert, you had to touch that one off as well? *

And then you ran it to the same final X position in your code - and all was well?

Am I following?

Ifso - the difference between the inserts is the nose rad.
That smaller insert was good "up to" 8 TPI, but it was good "down to" what ??? 18, maybe 24 TPI?
It would have been rated down to the finest pitch that the tip rad is rated for.
Max is mere overall width.

The difference is that the bigger insert would have had a bigger tip rad, and was only rated "down to" 13 ??? TPI?

So when your cheat sheet is telling you that your thread depth is "x" for this thread, it is also ass_u_ming a certain tip rad, and if you have one much sharper, you would need to go deeper - which is what you did with the bigger insert. You may have only went "x" deep from your touch-off ref point, but to theoretical sharp - you were much deeper.


Just guessing if this is the case, and if not - I would like to understand the issue.
I just cut an 8.628 - 8 internal (AN-44 bearing nut) "class III" a cpl weeks ago and was like "How in the world doo I qualify THAT! without $1600 and 4 weeks for plug gauges?" (Or maybe a "GageMaker" setup)

So - like you - I made a plug to about means of spec per wires. (I HATE thread wires!) Then made my part - sneaking up on what it takes to let my plug spin in with a slight wiggle. But I had to go .01 if not .02 oversized to get to that point, and I concluded that it was my NT3R insert that was the problem.

I would have liked to have used a much larger tool, but didn't realize until it was time to make the part that it even had a thread in it, and with all the tools around, I couldn't really find a bigger bar/insert to make the part on hand, so I run-what-I-brung. ... and again - I think that an insert made specifically for 8TPI would have had a much larger tip rad. Prolly enough to blame .01 on diameter to anyhow.


Now, what's more interesting is - I wonder how they used that part in application? I've never seen a "bearing nut" thread actually "IN" a part before. Seems like an odd application.

???



* Getting a bigger threading insert offset close enough to re-run would be a major chore, and would almost certainly widen the thread a bit jist from slight mis-alignment. So that would have helped as well.



-------------------

Think Snow Eh!
Ox
 
I'll 2nd the additional places on the F call
First hand chasing down a similar problem with a 5/8-11"
customer said "use this nut as a gauge".
I ran 10 pc's of the 100 and none of them worked on the "other" nuts, but did fit the nut he gave me.
Corrected to 4 places from 2 and every part worked after I went back to my original offsets
 
I'll 2nd the additional places on the F call
First hand chasing down a similar problem with a 5/8-11"
customer said "use this nut as a gauge".
I ran 10 pc's of the 100 and none of them worked on the "other" nuts, but did fit the nut he gave me.
Corrected to 4 places from 2 and every part worked after I went back to my original offsets

I would be Milling Magnesium or Ti with a 1/4" Endmill, with no conveyor, and throwing lit matches at the Machine if that happened to me. That is unacceptable, Ive never even heard of that. I would flip out!!!

R
 
Original coded value may have been .090 or .091
machine is a 10t control with par set at 4 places so .090909090909 would have been useless
But .0909 worked
after 1st piece was tuned in remainder of run acceptable.
Fortunately stock was 6" long piece of 1" square leadloy so eating the first 10 wasn't a loss.
I kept 1 around as a reminder
I had called the guy who ran the lathe before I bought it and "4 places not rounded" was his quick answer....with a chuckle

Sad part for me was I had attempted several jobs before that without success because of thread fit errors that I couldn't figure out and failed

Now I actually look forward to a threaded part, evan internals
I also clarify with conditions what gauging will be used whether its an overqualified $$$$$ Go/No-go or a hot dipped nut from HD
 
4 TPI =F.25
4 TPI =F.2500

All I'm saying is that is not cause for error.

Rounding to .091 from .0909 is .0001 difference, and .090 is .0009 error, no wonder.

R
 
Don't know if it would actually matter or not, but I was always told pitch should be input as six digits even if the last few are zeroes.

Any chance your stock is moving?

It wont matter. Trailing zeroes need not be specified.

Most machines are set to use IS-B (Increment System B) which accepts three digits after decimal in mm mode and four digits after decimal in inch mode, for axis distances (e.g., X10.123 in mm mode and X10.1234 in inch mode can be programmed). Extra digits, if any, are truncated, i.e., ignored by the machine. IS-C accepts one extra decimal digit, compared to IS-B, i.e., four in mm mode and five in inch mode. However, irrespective of the increment system being used on the machine, the lead of a thread can be specified with up to four decimal digits in mm mode and six decimal digits in inch mode, like what is allowed for feedrate in feed per revolution mode.
 








 
Back
Top