Worm Gear Programming

jephw · Jan 16, 2021

Good day to everyone.
Any suggestions on how we could program and machine this worm gear? Material is 4140, 2.375 diameter. For our CAD software we are using Fusion 360, but I haven't been able to make the desired toolpath. Anyone else had to deal with a similar project that can give me advice?
Our machine is a Mori Sieki NL3000y.

npinson · Jan 16, 2021

I’ll be honest, I’m going to follow this thread because this is something I’ve ran across and I am equally interested. For the OP, here is what I know. That appears to be a globoid worm shaft. There is a radius that “wraps” around the female part of the worm gear. The area where the “threads” are concave. As far as I know the threads are also concave. Someone well versed in gears will probably correct that terminology. Only way I can think that could be made is with G32 threading with a profile set from either a spreadsheet with start stop coordinates. I’ve seen YouTube videos of quick leading threads, multi start, and heavy taper done with G32. I’ve used G32 to accomplish multi lead oil grooves and thread forms, but it was done by hand and very much trial and error. The videos I have seen have used Mastercam to accomplish it. Would love to hear more about it regardless, because I’m no expert.

Sent from my iPhone using Tapatalk

npinson · Jan 16, 2021

After closer look at your picture, it looks like a single start worm gear shaft. Also, maybe not a globoid.

Sent from my iPhone using Tapatalk

EmGo · Jan 16, 2021

npinson said:
Also, maybe not a globoid.

Double-enveloping worms are bigger in diameter at the ends, skinnier in the middle. They wrap around the wormgear in one plane, while the wormgear wraps around the worm in another plane. For more information, look up "cone drive". Mr Cone invented them.

This is just a regular old worm, straightforward to cut on a lathe.

npinson · Jan 16, 2021

EmanuelGoldstein said:
Double-enveloping worms are bigger in diameter at the ends, skinnier in the middle. They wrap around the wormgear in one plane, while the wormgear wraps around the worm in another plane. This is just a regular old worm, straightforward to cut on a lathe.

I noticed that in my second post. Most double-enveloping worms I’ve seen are also multi-lead. Is that generally correct?

Sent from my iPhone using Tapatalk

EmGo · Jan 16, 2021

npinson said:
Most double-enveloping worms I’ve seen are also multi-lead. Is that generally correct?

Not really, but now you're getting past my explanatory powers

Imagine a worm drive, in mesh, from above like on a topview drawing, with the worm blank laid out horizontally and the wormgear looks like a circle. Imagine the worm blank above the wormgear towards your eyes (z axis ?) far enough to not touch.

Now sharpen the face of the wormgear by grinding it concave. It looks like a very slightly helical gear now, with a sharp top edge.

Make everything rotate together in timed relationship. Now lower the worm blank past the sharpened wormgear. The cutting teeth are travelling in a circular path through the worm, right ? So it's not actually a lead, it's an arc.

Make sense ?

Initially these were called Cone Drive after Samuel Cone, who thought them up. The cutting machines actually worked sort of like my crappy description. Then the Cone Drive Co. was bought by Michigan Tool in Detroit, who made them for decades. Michigan Tool made quite a few gear cutting machines, some pretty nice ones. I believe they went tits-up some time in the late sixties ?

npinson · Jan 16, 2021

I had my head wrapped around the concept until that [emoji16]. So can a double enveloping worm be manufactured using a lathe without a C Axis? Technically? In my opinion yes, but not easily.

Sent from my iPhone using Tapatalk

EmGo · Jan 16, 2021

npinson said:
So can a double enveloping worm be manufactured using a lathe without a C Axis? Technically?

I don't think so. Not unless you could rotate your tool around a point in x-y as it traversed in z. Integrex probably can tho.

npinson · Jan 16, 2021

So the lead does also does wrap around the pitch diameter essentially

Sent from my iPhone using Tapatalk

EmGo · Jan 16, 2021

npinson said:
So the lead does also does wrap around the pitch diameter essentially

Exactly. But I don't know if you should call it a lead, that usually means a straight line.

Whatever you do, it's confusing to describe

And don't forget, the shape of the tooth has to arc around the center of rotation of the wormgear. So the angle of the sides of the teeth relative to the axis of the worm is not static as you traverse in z (if that makes sense

).

edit: here, much better than me :

https://www.geartechnology.com/issues/1084x/loveless.pdf

Jaxian · Jan 17, 2021

I answered you over in the CAD/CAM area with a picture. You have to use the Rotary Toolpath as I recall and have the Manufacturing Extension. I wanted to see if I could cut threads with a 4th axis on my VMC. Seemed to work fine. Would need a form tool though and need to model the form correctly. That path will do the correct motion.

EmGo · Jan 17, 2021

npinson said:
Most double-enveloping worms I’ve seen are also multi-lead. Is that generally correct?

Oh, reading this over again, maybe I misunderstood you .... did you mean multi-start ? If so, I guess it's just as common as any wormgear setup, but isn't inherent in the design, just depends on the ratio your want.

If you did mean multi-lead, then I got it right, the "lead" is really an arc, so it's more of a continuously-changing than multi lead.

But there are also real multi-lead worms (but not wormgears) that give you an increasing tooth thickness if you move the worm along its axis. That's because the lead along one side of the tooth is less than along the other side of the tooth.

G&E for one called that "duallead", it's a fairly common way to make the backlash adjustable. Just moving the worm deeper into mesh (radially) is the cheap way, but often you get a step from wear in either or both parts, so that's not as good as teeth that get thicker if you move them along the worm axis. If you just move the worm deeper but there's a step, it seems like you got rid of the lash but a few hours later the step is worn away and ya gots nothing. Duallead doesn't have that problem.

Sorry if I mistook your meaning.

npinson · Jan 17, 2021

EmanuelGoldstein said:
Oh, reading this over again, maybe I misunderstood you .... did you mean multi-start ? If so, I guess it's just as common as any wormgear setup, but isn't inherent in the design, just depends on the ratio your want.

If you did mean multi-lead, then I got it right, the "lead" is really an arc, so it's more of a continuously-changing than multi lead.

But there are also real multi-lead worms (but not wormgears) that give you an increasing tooth thickness if you move the worm along its axis. That's because the lead along one side of the tooth is less than along the other side of the tooth.

G&E for one called that "duallead", it's a fairly common way to make the backlash adjustable. Just moving the worm deeper into mesh (radially) is the cheap way, but often you get a step from wear in either or both parts, so that's not as good as teeth that get thicker if you move them along the worm axis. If you just move the worm deeper but there's a step, it seems like you got rid of the lash but a few hours later the step is worn away and ya gots nothing.

Sorry if I mistook your meaning.

Multi start would be more correct

Sent from my iPhone using Tapatalk

EmGo · Jan 17, 2021

npinson said:
Multi start would be more correct

In that case, I'd say "it depends"

Single start is usually considered more accurate than multi-start, but the ratios are more extreme. They use cone drives for a few different purposes - one is, they can transmit more power because of larger wear surfaces. So in that case, maybe multi-start because they'd rather have 3:1 than 50:1.

They also use double-enveloping for accuracy and no-backlash situations, such as instruments or perhaps a precision dividing head. Those uses would be inclined to go with single start.

Different horses for different jobs.

npinson · Jan 17, 2021

In my Marvel saw, it has a developing worm. Certainly a multi start. Bronze lower gear was roasted by no attention, and no gear lube.

Sent from my iPhone using Tapatalk

EmGo · Jan 17, 2021

npinson said:
In my Marvel saw, it has a developing worm. Certainly a multi start. Bronze lower gear was roasted by no attention, and no gear lube.

Ouch. I bet they got a pretty penny for a replacement, too. Yes, wormgears have a lot of sliding, so they want oil. And if they don't get it, they can be real grumpy

edit: found a pretty good illustration

But for the original question, no need for any of this, a normal worm is just like a thread. (Okay not just like but close enough for this discussion.)

jephw · Jan 17, 2021

Yes, its just a single start. To me it just seems like an easy thread with simple angles, and easy measurements. Drawing it up was no problem. Just the programming I'm struggling with.

jephw · Jan 17, 2021

EmanuelGoldstein said:
Ouch. I bet they got a pretty penny for a replacement, too. Yes, wormgears have a lot of sliding, so they want oil. And if they don't get it, they can be real grumpy

edit: found a pretty good illustration

But for the original question, no need for any of this, a normal worm is just like a thread. (Okay not just like but close enough for this discussion.)

Yep, basically as simple as that. Thought it would be an easy project, not the case it seems.

Dan from Oakland · Jan 18, 2021

Just program it like any other thread. You may have to grind a tool to the correct form if you can't locate an insert. No different than an acme thread although the pressure angle may be different. I've made special tool blocks that fit into my turret slots with different lead angles to tip the tool for side clearance. HSS blanks work very well for this purpose, although for a single lead worm its easy enough to grind the side clearance into the tool. Depending on the pressure angle, you can use a canned threading cycle if the infeed angle matches.

sinha · Jan 18, 2021

Infeed angle may not match. In such a case, the nearest smaller angle can be used in the canned cycle.

Worm Gear Programming

Aluminum

Attachments

Aluminum

Aluminum

Diamond

Aluminum

Diamond

Aluminum

Diamond

Aluminum

Diamond

Stainless

Diamond

Aluminum

Diamond

Aluminum

Diamond

Aluminum

Aluminum

Titanium

Titanium

Similar threads