Tools to Make Gears on 3+1-Axis Mill or Live Tool Lathe

[UtahTechFabLab]

Hey Y'all,

I'm looking for a good way to make various gears on a 3+1-axis mill or on a live tool lathe.

I manage the engineering fabrication lab at a state university, and our engineering students commonly need to make gears. Straight-toothed spur gears most often, but I also want to be able to make bevel gears, worm years, helical gears, internal gears... Basically anything that students can come up with.

Do you know any good attachments or tools that I could use on the machines I already have?

I should probably get a set of involute cutters. I've also been exploring the use of straight-shank broaching tools with involute-shaped inserts on the Mill and Lathe. Does anyone suggest a good brand or have tips on that? Or have any they're wanting to part with...? ; )

NOTE: I'm not interested in production speed. We often only need to make one or two of a given gear. So as cool as they would be, a skivving or broaching unit for the CNC lathe (for example) are overkill.

 

[EmGo]

Hey Y'all,

I'm looking for a good way to make various gears on a 3+1-axis mill or on a live tool lathe.

Don't even think it. This is all shit.

If you stick to spur gears, you could get a wire edm but otherwise, for a lab ? Forget it.

And involute cutters are crap. They have a place in the world but it ain't the 'accurate gear dep't'. Or even semi-accurate gear dep't.

Ten grandish would be reasonable for a start point. Otherwise, wire edm maybe. Or use one of the weirdass methods that emulate a Maag. Or even try to profile them with a tiny ball end mill, which takes forever but is okay for one part. Or find a real gear shop to do your occasional work. But kiss it off on the involute cutters and bogus hobbing attachments and the rest of that stuff.

 

[Jaxian]

Is this the kind of thing you were talking about as too expensive? They are but I think they are the only way to really get a good gear on a live tool lathe. I guess you could just use a stub arbor and a involute gear cutter but I have never tried it on a lathe.

 

[angelw]

They are but I think they are the only way to really get a good gear on a live tool lathe.

The OP is talking about a mill. He would need synchronization between the Spindle and the +1 axis if its a true 4th axis and not just an indexer and that's not going to happen.

And shite! This is the first time I've agreed with EG.

Regards,

Bill

 

[Peter from Holland]

Get the right software (Klingelnberg I heard) and a 5 axes machine
You can make everthing you can come up with Every pressure angle you want Therefore optimizing the rolling action of a gear
S shaped bevel gears and so on and so on
Big cyclo paloid gears are made that way nowadays

 

[UtahTechFabLab]

Is this the kind of thing you were talking about as too expensive? They are but I think they are the only way to really get a good gear on a live tool lathe. I guess you could just use a stub arbor and a involute gear cutter but I have never tried it on a lathe.

I'm actually in the middle of an e-mail conversation with MD Tools regarding some of their live tools. A new hobbing unit like that is probably outside of my budget, unless I get a good used one somewhere. But that said, I'm going to see what they can do as far as discounts. Most companies have some kind of programs for educational institutions.

Thank you!

 

[UtahTechFabLab]

The OP is talking about a mill. He would need synchronization between the Spindle and the +1 axis if its a true 4th axis and not just an indexer and that's not going to happen.

And shite! This is the first time I've agreed with EG.

Regards,

Bill

Hey there! So I think I misspoke... I have a 4th axis rotary table and I can do all the synchronizing I need to. It's a DMG Mori Duravertical 5100 with a DDRT Rotary Table.

Thanks!

 

[Vancbiker]

Hey there! So I think I misspoke... I have a 4th axis rotary table and I can do all the synchronizing I need to. It's a DMG Mori Duravertical 5100 with a DDRT Rotary Table.

Thanks!

Still not synchronized between the spindle and 4th axis so hobbing is not possible. Ash Gear sell single tooth cutters that one can use to make straight spur gear shaped objects on a machine like you have. Not good gears, but probably good enough.

 

[EmGo]

Not good gears, but probably good enough.

I'm not a snob like some people who are missing in action but ... no. Space cutter gears are okay if you are replacing the drive on your Peerless hacksaw and don't have any money and live 100 miles from anything better but ... not for real teeth.

And the hobbing attachments, jesus. You can buy the real thing for what they want for those, then you can cut real gears instead of wobbly slots on a shaft.

It's just a case of "no", at least if this is really a university shop with (I would hope) some standards.

If there is a real need to cut teeth then there's a few ways that are not ridiculously expensive (but they still aren't cheap) but just because someone thinks "it'd be nice if we could" ? Not a good plan.

 

[guythatbrews]

Also remember gears cut with involute cutters are not involute gears. They are a poor approximation of involute gears. They don't roll well and are as loud as a jet at anything but low speed.

Back in my gear shop days my sucker partner was constantly making stuff free gratis for college kids. You need to find a guy like that.

Bevel gears, even straight, are a whole different, higher level.

Bottom line, good gears is tough without the know how and proper equipment. Sorry to say but you may be on a fools errand.

 

[JSL_MFG]

This is easy... been there and done that on this exact machine. The DDRT rotary has really good indexing accuracy. You'll be fine.

Make a blank, mount it to an arbor so you have enough clearance, set zero to rotational center, slot through the center of the tooth, rough & finish with a ball endmill, and bob will be your uncle.

You could rotary surface these, but honestly, you're probably better off programming a simple parallel stepover running in your X-axis on the mill. Subroutine that around your part based on the # of teeth and shaboom, gears.

You can get away with a lot if you're able to use ball endmills and rotary indexing. Full rotary surfacing is also nice, but it's usually more complex than it needs to be. You're only making 1 offs, just don't overthink it, when it doubt, surface it.

 

[IceCzar]

Get the right software (Klingelnberg I heard) and a 5 axes machine
You can make everthing you can come up with Every pressure angle you want Therefore optimizing the rolling action of a gear
S shaped bevel gears and so on and so on
Big cyclo paloid gears are made that way nowadays

Or you could find a University Alumni that would get Peter's Mikron 102 (which I covet) for you

Hobbing machine Mikron 102-03 | Veltman Machines

veltmanmachines.nl

as mentioned gears aren't easy, even slow and prototype
My primary aspiration is to make high precision gear bearings
there are a number of old and ingenious ways to make good gears but as to educational value that's debatable
it is interesting to note though involute standards and master gears cannot be made by cnc

I'd imagine you have access to this

https://journals.sagepub.com/doi/abs/10.1177/0954405412469554?journalCode=pibb

(everyone else would need to act like it's September 19th)

 

[UtahTechFabLab]

Or you could find a University Alumni that would get Peter's Mikron 102 (which I covet) for you

Hobbing machine Mikron 102-03 | Veltman Machines

veltmanmachines.nl

as mentioned gears aren't easy, even slow and prototype
My primary aspiration is to make high precision gear bearings
there are a number of old and ingenious ways to make good gears but as to educational value that's debatable
it is interesting to note though involute standards and master gears cannot be made by cnc

I'd imagine you have access to this

https://journals.sagepub.com/doi/abs/10.1177/0954405412469554?journalCode=pibb

(everyone else would need to act like it's September 19th)

Oh my that Mikron looks like such a pleasure! What a beautiful old machine! Maybe one day...

But thanks for your comments!

 

[UtahTechFabLab]

This is easy... been there and done that on this exact machine. The DDRT rotary has really good indexing accuracy. You'll be fine.

Make a blank, mount it to an arbor so you have enough clearance, set zero to rotational center, slot through the center of the tooth, rough & finish with a ball endmill, and bob will be your uncle.

You could rotary surface these, but honestly, you're probably better off programming a simple parallel stepover running in your X-axis on the mill. Subroutine that around your part based on the # of teeth and shaboom, gears.

You can get away with a lot if you're able to use ball endmills and rotary indexing. Full rotary surfacing is also nice, but it's usually more complex than it needs to be. You're only making 1 offs, just don't overthink it, when it doubt, surface it.

Hey thank you! I wasn't sure how well we'd be able to do using the 4th axis, but you've given me some hope!

 

[EmGo]

it is interesting to note though involute standards and master gears cannot be made by cnc

That's definitely not correct. I doubt that anyone making master gears runs mechanical gear grinders anymore, and definitely mechanical involute testers are a relic of a bygone age. They still work fine but if a new purchasing kid from a big customer walked in and saw you using one, they'd probably shit a brick.

This is 2023, unfortunately. If it doesn't have a keyboard, it can't be any good.

 

[IceCzar]

"The roller–rail generating mechanism is widely used in the measuring system of HIA and UMG. German National Physical Laboratory (PTB), National Metrology Institute of Japan (NMIJ) and highprecision gear research room (HGR) of Dalian University of Technology (DLUT, China) adopt this generating mechanism to form an ideal involute to test the accuracy of involute artefacts and master gears."

"The generating accuracy of the workpiece is mainly from the driving accuracy of the rotary stepping motor and the linear stepping motor, the synchronous accuracy of the two movements and the stability of electronic signals. Considering the more error sources and the disadvantages (drift, hysteresis and vulnerable to interfere) of the electronic signals, the processing accuracy of the involute is relatively low by using electronic generating method. Generally, the gear involute processing accuracy can reach to grade 4-2. According to the current processing and CNC technology, the electronic generating method cannot meet the processing requirement of HIA and UMG."

"total profile deviation is not more than 0.6 µm and the profile form deviation is less than 0.4 µm in 105 mm generating length of the specimen"

from the link above
September 19th is of course the Internationally recognized Talk like a Pirate Day

 

[Kees]

This is a proper way to make decent gears:

Nice students-project to make the program.

 

[EmGo]

German National Physical Laboratory (PTB), National Metrology Institute of Japan (NMIJ) and highprecision gear research room (HGR) of Dalian University of Technology (DLUT, China) adopt this generating mechanism to form an ideal involute to test the accuracy of involute artefacts and master gears."

Generally, the gear involute processing accuracy can reach to grade 4-2. According to the current processing and CNC technology, the electronic generating method cannot meet the processing requirement ...

Notice the source ? If you check their pants you'll see those guys have friction sores all over their dicks. Academics write all these stupid papers which the people who actually make gears ignore. Even AGMA has gone that way, it's sad. Used to be people in industry writing about these subjects.

Zeiss, M&M, Hofler, those guys, the ones that actually do this work ? If they print this stuff on soft paper they can use it to wipe their butts. Laser should be better than inkjet, I guess

^ "Invented by Sandvik Cormorant" give me a break. Same system as used by Maag and Sunderland for the past 100 years or more. And Gleason two-tools, for that matter. In fact the Gleason 14 also used a rotating disk-type cutter except it could also do crowning. Don't know when they came out with the 14 but it was before 1950.

 

[IceCzar]

As I primarily work for Universities and Academics and having grown up a few miles from NIST
UMG (ultra-precision master gears) and HIA (high grade involute artifacts) aren't made with CNC
you can direct you're invective to the Institution of Mechanical Engineers (who published that) or whatever educational institution strikes your fancy

the operative word you should have paid attention to was standards
(like as in the National Institute of Standards and Technology)

 

[EmGo]

As I primarily work for Universities and Academics and having grown up a few miles from NIST
UMG (ultra-precision master gears) and HIA (high grade involute artifacts) aren't made with CNC
you can direct you're invective to the Institution of Mechanical Engineers (who published that) or whatever educational institution strikes your fancy

the operative word you should have paid attention to was standards
(like as in the National Institute of Standards and Technology)

Kind of laughing but I won't ask to check your pecker

So tell us, Mr Educator, what are master gears used for in the real world ? Functional checks on double-flank rolling testers, yes ? On parts that are like agma 9 or 10 at the most ? Anything up to 15-ish is no problem on any brand name gear grinder today and yup, all the current ones are nc.

"Master involute" ? Like, you don't have a big red sore yet ? An involute is a mathematical equation. What are "master involutes" used in ? 1950 Fellows testers ? It's been a while since anyone needed one of those. Current, real-world, useful gear testers are all specialized cmm's which, while they aren't exactly cnc are the same basic idea.

Your super-special "master gear" and "master involute" serve no realworld purpose.

Oh wait, I'm wrong. It does give a certain group of people a pretty nice living, the ones who aren't still enthralled with novikov-wildhaber uselessness or writing long involved papers about fea on things no one with any sense would try anyhow.

Once upon a time AGMA and the rest were Darle Dudley. Now it's some group of overaged geeks who spend their efforts showing each other how cool they are.

May as well be a sociologist.