CNC Threading issue that took me a while to catch. Feed ramp, end of the thread

[laminar-flow]

We have been making some parts that have a short 20 pitch thread. The last run was almost fine, but the gauge was just a bit tighter at the end of the thread. I didn't think anything of it but on the next run, I just had to go and mess with the program to make it faster. To try to improve the finish of the thread, I increased the spindle and the thread was way worse. It was acting like taper as the gauge did't even get to the bottom of the thread. And then I realized what was happening. The feed was too fast and had to ramp slow down as it approached the end of the thread, thus changing the pitch. Is there any way to predict this so we can run the fastest RPM and still not affect the pitch? Haas HPCL lathe. Fusion 360 CAM

 

[CORONA VIRUS]

We have been making some parts that have a short 20 pitch thread. The last run was almost fine, but the gauge was just a bit tighter at the end of the thread. I didn't think anything of it but on the next run, I just had to go and mess with the program to make it faster. To try to improve the finish of the thread, I increased the spindle and the thread was way worse. It was acting like taper as the gauge did't even get to the bottom of the thread. And then I realized what was happening. The feed was too fast and had to ramp slow down as it approached the end of the thread, thus changing the pitch. Is there any way to predict this so we can run the fastest RPM and still not affect the pitch? Haas HPCL lathe. Fusion 360 CAM

A while back I came up with a general rule. Keep the inch per minute feed at or below 60ipm. Most lathes do not have a problem at this feed.

 

[Ox]

I started a job recently with 8TPI @ 1500 rpm, and that dog didn't hunt.
1250 was still a bit too fast.
Running fine at 1100 now.


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

Think Snow Eh!
Ox

 

[angelw]

Is there any way to predict this so we can run the fastest RPM and still not affect the pitch?

Hello laminar-flow,
An erroneous Thread Lead exists at the start and end of the thread due to acceleration and deceleration respectively. The error at the start of the thread can be circumvented completely (if sufficient standoff is available) by starting further away from the workpiece. The following two algorithms result in the distance that contains the erroneous Thread Lead.

To calculate the distance of thread lead error due to acceleration:

TED = 0.002 x RPM x L

where

TED = Thread Error Distance

RPM = Spindle Revs

L = Lead of Thread


To calculate the distance of thread lead error due to deceleration:

TED = 0.00055 x RPM x L

If you know the amount of Thread Error Distance that's tolerable, you can transpose the above algorithm to give a result in terms of maximum spindle revs.


RPM = TED / 0.00055 / L

Regards,

Bill

 

[Bill in PA]

I am following this thread. Does having chamfering turned on or off affect the number of erroneous threads?

Bill

 

[wmpy]

Hello laminar-flow,
An erroneous Thread Lead exists at the start and end of the thread due to acceleration and deceleration respectively. The error at the start of the thread can be circumvented completely (if sufficient standoff is available) by starting further away from the workpiece. The following two algorithms result in the distance that contains the erroneous Thread Lead.

To calculate the distance of thread lead error due to acceleration:

TED = 0.002 x RPM x L

where

TED = Thread Error Distance

RPM = Spindle Revs

L = Lead of Thread


To calculate the distance of thread lead error due to deceleration:

TED = 0.00055 x RPM x L

If you know the amount of Thread Error Distance that's tolerable, you can transpose the above algorithm to give a result in terms of maximum spindle revs.


RPM = TED / 0.00055 / L

Regards,

Bill

Are these formulas only applicable to the OPs machine? It seems to me that the ability of a machine to accelerate/decelerate would vary by builder/ control/ age.

 

[angelw]

Are these formulas only applicable to the OPs machine? It seems to me that the ability of a machine to accelerate/decelerate would vary by builder/ control/ age.

Hello wmpy,
It does vary, but not enough that you would notice. The constants used in both algorithms are correct commencing with the Red Cap Fanuc Servos and later controls don't vary much from that. I use these constants for all controls, with later controls being covered by the marginally larger distance result than would actually be required. It's a better proposition than stabbing in the dark by incrementally dropping the revs down until you get one that works for you.


Regards,

Bill

 

[Nmbmxer]

A lot of manufactures have tables and formulas in the programming manual to calculate the ramp up and down distances. The same thing can get you on the start of thread if your work is small and center is large, you don't have enough run up before engaging the work.

 

[laminar-flow]

Well thanks for that information. I've never had an issue with lead in and only an issue with the end this time. Thanks.

 

[sinha]

I am following this thread. Does having chamfering turned on or off affect the number of erroneous threads?

Bill

I do not think so.
The acceleration / deceleration for the axial motor should not get affected by the motion of the radial motor.