Thread milling form error

[MazatrolMatrix]

Hi,

Apparently thread milling with a tool bigger than approx 70% of a thread will introduce a form error on the thread profile, near the root of it from what I understand. On an internal thread that is.

Can someone explain to me why this error occurs and is the error relevant at all in standard applications?

Thought it was a little bit interesting.

Thanks.

 

[Philabuster]

You talking about ID or OD threads?

 

[cngbrick]

It helps to mentally visualize these things if you push the parameters a bit. For instance, in the case on an external thread you can imagine that a very coarse thread pitch such as 1 thread per diameter cut with a large diameter (let's say 120 in) thread milling tool will only have the required form at a plane running through the axis of both tools but at any point before or after the milling tool will gouge material. As you gradually reduce the extreme dimensions to normal levels the gouging will reduce but is still there relative to ideal form.

You can run the same thought experiment with internal threads if you assume a thread mill with a diameter of root diameter + thread depth.

RT

 

[MazatrolMatrix]

You talking about ID or OD threads?

Sorry, I edited my post now. For internal threads that is.

 

[sinha]

The thread has a helix angle, but the cutter does not have. At any point of time, it cuts in a plane which is perpendicular to the axis. Ideally, it should cut along the helix (which is not possible). Therefore, there will always be some gouging. The larger the cutter diameter, the more the engagement length, hence the more would be the gauging.

 

[memphisjed]

The thread has a helix angle, but the cutter does not have. At any point of time, it cuts in a plane which is perpendicular to the axis. Ideally, it should cut along the helix (which is not possible). Therefore, there will always be some gouging.

If you tilt the head and use rotary you can cheat the gouge out of cheating the threads.
I guess you could do similar on a 5 axis.
Possible- not easy.

 

[mhajicek]

If you tilt the head and use rotary you can cheat the gouge out of cheating the threads.
I guess you could do similar on a 5 axis.
Possible- not easy.

Shank would hit.

 

[cngbrick]

Going back to the 70% figure: there is always form error but the 70% is probably a reasonable rule of thumb regarding common thread pitches and permissible form error. The form error increases with pitch so a fine thread is more geometrically accurate than a coarse thread.

RT

 

[EmGo]

I guess you could do similar on a 5 axis.
Possible- not easy.

That'd be a great demo for a 5-axis video or at a show ... thread mills and single-v thread grinders do tilt the cutter. Real thread mills, I mean

@haj, interference would depend on the thread you were cutting, yes ?

 

[mhajicek]

@haj, interference would depend on the thread you were cutting, yes ?

A female thread would have to be very shallow to allow tilt without shanking.

 

[EmGo]

A female thread would have to be very shallow to allow tilt without shanking.

You're too used to tiny Thread milling is even more useful for doing something like a 2" pipe thread etc etc ... put a single tooth cutter on a longish 1/2" shank in there and away ya go. Would look impressive at a show.

 

[mhajicek]

You're too used to tiny Thread milling is even more useful for doing something like a 2" pipe thread etc etc ... put a single tooth cutter on a longish 1/2" shank in there and away ya go. Would look impressive at a show.

If you're doing a 2" pipe thread with a 1/2" shank, you don't have enough form error to worry about anyway.

 

[sinha]

If both internal and external threads are made with the same cutter, the form error would match.

Edit: The statement is not correct. See the comments below.

 

[mhajicek]

If both internal and external threads are made with the same cutter, the form error would match.

Not really. I've actually modelled the form error for tapered multi-lead threads in bone plates. You get a lot more form error on internal threads, since the thread is wrapping around the tool, and it's worst when the cutter size approaches the thread size.

 

[sinha]

Not really. I've actually modelled the form error for tapered multi-lead threads in bone plates. You get a lot more form error on internal threads, since the thread is wrapping around the tool, and it's worst when the cutter size approaches the thread size.

I stand corrected. Thank you.
Since the engagement length is more on the internal thread, it will have more form error.

 

[EmGo]

If you're doing a 2" pipe thread with a 1/2" shank, you don't have enough form error to worry about anyway.

Actually, this whole thing is not exactly correct. You get the same "form error" with any cutter, no matter the size. It's partly like single-point threading on a lathe, if your tool is not 60* and normal to the helix, then the "form" (aka angle) will be off. This is the same for any diameter thread cutter in a three-axis mill.

The other part of the error is, the cutter teeth rotate in a plane, compared to the helix of the thread, so that's not exactly "form error" either, it's more interference between a plane and a helix. With larger cutters there's more clipping/interference/metal removal on the flanks, because the arcs of the part and the plane of the cutter interfere more.

So, if you took a single-row cutter and put it normally into a hole, yes you will get error even if it's way under this 70% number. If you five-axed it in, then you've eliminated the way the cutter v and the thread v are not normal to each other. That's why thread mills and thread grinders tip their wheels.

I would say that "form error" actually applies to the way *all* thread mills are not at the correct angle, while "cutter interference" is more appropriate for the way the plane of the cutter interferes with the helix of the thread, in larger cutters.

But nobody listens to me anyhow, so carry on

 

[mhajicek]

I would say that "form error" actually applies to the way *all* thread mills are not at the correct angle, while "cutter interference" is more appropriate for the way the plane of the cutter interferes with the helix of the thread, in larger cutters.

But nobody listens to me anyhow, so carry on

I think you're trying to redefine terms. The cutter interference results in form error on the part, which is what everyone cares about.

 

[EmGo]

I think you're trying to redefine terms. The cutter interference results in form error on the part, which is what everyone cares about.

I guess what I'm trying to figure out is, the term is sort of misleading. It's not that you get "form error" when the cutter is too big : you have form error always because the cutter is not normal to the helix.

What happens is that as the cutter gets bigger, the interference gets worse, until it's not acceptable. So to me that should be different words, it's describing a different thing.

But you're right that in the end it means the thread is crappy.

(Have a bit of interest in the subject because of worms. There is a lot of literature on 'form error' in those, due to different methods of cutting. Same thing applies to threading but they are so much smaller you can get away with a lot more.)

 

[mhajicek]

you have form error always because the cutter is not normal to the helix.

This is true, but if it's small, it can be accounted for by changing the flank angle of the cutter. I've also never found a satisfactory answer to this question: Is the ideal 60 degree thread form measured along the thread itself (at the pitch angle), or as a section along the axis of the screw, as if turned with a flat 60 degree insert? Most will measure with an optical comparator normal to the axis of the screw, which I'm pretty sure is wrong, since the screw will shadow itself due to the pitch angle; which is why you can't injection mold a perfect thread in a two piece mold.

 

[EmGo]

I've also never found a satisfactory answer to this question: Is the ideal 60 degree thread form measured along the thread itself (at the pitch angle), or as a section along the axis of the screw, as if turned with a flat 60 degree insert?

I'm pretty sure it's really supposed to be measured normal to the helix ... if you want to drive yourself crazy, there's a whole chapter about this in Analytical Mechanics of Gears by Earle Buckingham. He goes into the differences you get by turning, milling, thread milling, grinding, and whatever other method you can come up with. Whirling ?

It's oriented to larger pitches such as worms and acme threads and multi-start threads but the principles apply.

Warning : you better love math. And buy fresh batteries for the graphing calculator.