CNC lathe gear hobbing

[Garwood]

I'm shopping for a gear hobber, but in searching around for manual hobbers I am running across these: https://www.ebay.com/itm/2242499829...585M7Ude33hoalQqUS17UzP/83|tkp:Bk9SR-zFtJSJYQ

This VDI40 hobbing head looks super wimpy. However, I have a VDI60 machine with C axis. I'm wondering if a VDI60 hobbing head (if anyone makes such a critter) would be a substitute for a real gear hobber for 10DP helical gears?

I would imagine there's compromises like trying to get a big fat hobbing head anywhere close to a 16" chuck holding 4" material.

Anyone used a hobbing head on a CNC lathe?

 

[wmpy]

I haven't used one. Do you need to sync the rotation of the live tooling to the rotation of the spindle? I don't think all live tool lathes can do that. Like if the cutter is under load and slows down, will that mess up your teeth if the main spindle doesn't slow accordingly?

 

[implmex]

Hi Garwood:
There is a guy who used to post regularly on here named Zahnrad Kopf.
He is a gearmaker by trade and had only unkind things to say about trying to hob decent quality gears with a hobbing head on a CNC lathe.

His complaint was multi faceted, but as I recall, he complained of both the rigidity of these units and of their accuracy.
I don't recall him explaining what it was that made them intrinsically inaccurate, but he seemed pretty adamant that you could not make a tight tolerance gear with one.

I trust his judgement!
Here's a link to his website if you want to reach out to him:

American Machine & Gear Works

American Machine & Gear Works, Surgical, Medical, Dental, & Industrial Machining, Rapid Prototyping, & Gear Making

www.thegearmaker.com

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com

 

[DouglasJRizzo]

I would imagine there's compromises like trying to get a big fat hobbing head anywhere close to a 16" chuck holding 4" material.

Anyone used a hobbing head on a CNC lathe?

When I worked for Doosan we had a demo in the showroom doing this on smaller gears.
It worked well enough.
However, I also grew up in a gear house, and for what you're thinking of, it's far better for a Barber-Colman, Fellowes, G&E, etc to do the job than on a CNC with an attachment.
You're talking about a pretty steep cut and a large PD. I'd much rather hob or shape than try it in an expensive CNC lathe with a $15k attachment.

 

[Dan from Oakland]

Save yourself a lot of pain and just buy a used old gear hobber, unless you need really high quality gears, then buy a nice gear hobber. Those lathe mounted hob heads might get you teeth cut on fine pitch so- so quality gears, but not a 10DP gear unless its plastic or possibly aluminum.

 

[Richard King]

I sent you a PM - guys I know who rebuild and sell gear machines. Zahnrad Kopf. is a member here too (also a student like you).

 

[sinha]

Zahnrad Kopf has also reviewed my book on Macro B ...

 

[Zahnrad Kopf]

Hi Garwood:
There is a guy who used to post regularly on here named Zahnrad Kopf.
He is a gearmaker by trade and had only unkind things to say about trying to hob decent quality gears with a hobbing head on a CNC lathe.

His complaint was multi faceted, but as I recall, he complained of both the rigidity of these units and of their accuracy. I don't recall him explaining what it was that made them intrinsically inaccurate, but he seemed pretty adamant that you could not make a tight tolerance gear with one.

I trust his judgement!
Here's a link to his website if you want to reach out to him:

American Machine & Gear Works

American Machine & Gear Works, Surgical, Medical, Dental, & Industrial Machining, Rapid Prototyping, & Gear Making

www.thegearmaker.com

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com

Thank you for the kind words, Marcus. As always, they are appreciated. (:>)

I *do* have some experience using the attachments and software on some turning centers. I have been extensively involved in evaluating and choosing some machines, based upon their performance, as well. And, as luck would have it, have been discussing this very topic with several of the machinery OEMs for the last 6 months while preparing to purchase a machine for the shop.

The bottom line is that they still basically suck. For all of the same reasons that I have written about here on this forum (and some others, in the past. Basically, it just boils down to rigidity and electronics being the biggest culprits. Mostly, it's rigidity. After those two reasons, it starts getting into applications and ease of use specifics.

End result - No one will put their name on the line guaranteeing anything past a Q7. And even *that* pithy level will come with a stack of requirements that be observed. (specific materials, specific pitch, specific brand <partner> attachment, specific set up, specific moon phase of the year, etc... )

A good, quality tool holder will cost $15K - $25K, and that alone doesn't allow one into the dance. The machine has to have the <cough> "option" <cough> enabled. There's another $10K - $20K. Minimum buy-in at $25K to be allowed to make mediocre gears? No thank you. You can purchase a clapped out ancient Hobber and have it fully rebuilt for pennies on that particular dollar. And when you're done it will run circles around the turning center, all day, every day, 365 days a year.

Now, if you're just making mediocre gears and shafts all day, every day, then it might pay off in the long view. (like parallel spline shafts that have other features along the shaft) But for the money and effort, we will always have good, old, solid, gear driven gear machinery around here.

We are a machine shop. Not just a gear shop. Soon, one of our turning centers will have the ability to do this. And even then, the least quality-demanding gears that we make (SouthBend Lathe replacement gears) will *STILL* be made on real, actual gear making machinery. Because even a brand new, modern turning center with the most recent tooling and options cannot achieve the quality that we require while making even those.

 

[cnctoolcat]

It was my thinking most (if not all) cnc lathes that hob do so by essentially "form milling".

Whereas actual gear machines cut using the involute process---entirely different than just plowing the cutter straight along the part, as you would with a form cutter on a milling machine.

For involute cutting, each cutter tooth follows the cross-section profile of the finished part tooth, with subsequent cutter teeth advancing along this profile of the part tooth. So, if you consider a cross-section of the part, the cutter teeth are working their way along the finished profile of the part's teeth: "down the valley and up the ridge, one tooth and one nibble at a time."

Thus the synchronization of the machine itself ,along with the shape of the cutter teeth, work together to do an involute cut.

 

[Zahnrad Kopf]

It was my thinking most (if not all) cnc lathes that hob do so by essentially "form milling".

Whereas actual gear machines cut using the involute process---entirely different than just plowing the cutter straight along the part, as you would with a form cutter on a milling machine.

For involute cutting, each cutter tooth follows the cross-section profile of the finished part tooth, with subsequent cutter teeth advancing along this profile of the part tooth. So, if you consider a cross-section of the part, the cutter teeth are working their way along the finished profile of the part's teeth, "down the valley and back up the ridge."

Thus the synchronization of the machine itself ,along with the shape of the cutter teeth, work together to do an involute cut.

Hey Cat. Long time, no write. (:>)

Nope. It's actually Hobbing, using an actual Hob, with timed relationship maintained between tool and work.**

** - One of the contributors to at least some minor error in what ultimately relates to accuracy is the fact that they are simply employing "electronic gearing", AKA - "servo slaving". This has some limitations with regard to encoders, computers, and the (literal) flexibility of belts on pulleys, allowing one point of some minor error to creep in, eventually. This, as opposed to even some "cheap" CNC Hobbers that still utilize Worm Wheels to accomplish some of the reductions, and maintain (at the least) some rigidity and accuracy of timing.

While we were at IMTS (and shopping machines), I brought the bank approval letter with me as proof of no BS offer. I told every OEM we spoke with that I would commit and sign for a machine on the spot if they would commit to Q10 Involutes in 4140 Pre-Hard repeatably, in writing. Not a single one accepted. (:>)

 

[Mud]

I saw a practical application a while back at the PRI show. Cam blanks were being made complete on a Mazak Mill-turn of some sort. The lobes and journals were grooved and turned, the cam lobes were milled with live tools, and the oil pump gear was hobbed with live tools on the machine. Cam drive gear attachment holes were put in one end. I suspect in that application the gear was "good enough" and the done in one aspect made it worth doing, for someone
This was next to the machine that was dedicated to making complete Nascar cranks from solid round stock with a conversational control.

 

[Dan from Oakland]

I saw a practical application a while back at the PRI show. Cam blanks were being made complete on a Mazak Mill-turn of some sort. The lobes and journals were grooved and turned, the cam lobes were milled with live tools, and the oil pump gear was hobbed with live tools on the machine. Cam drive gear attachment holes were put in one end. I suspect in that application the gear was "good enough" and the done in one aspect made it worth doing, for someone
This was next to the machine that was dedicated to making complete Nascar cranks from solid round stock with a conversational control.

Distributor/oil pump drive gears are crossed axis helical gears, and as such only exhibit line contact between mating gear teeth and transmit minimal power. Involute profile and minor spacing errors don't have much effect on the gear mesh so that is why hobbing on the mill turn is "good enough".

 

[newcomer]

Thank you for the kind words, Marcus. As always, they are appreciated. (:>)

I *do* have some experience using the attachments and software on some turning centers. I have been extensively involved in evaluating and choosing some machines, based upon their performance, as well. And, as luck would have it, have been discussing this very topic with several of the machinery OEMs for the last 6 months while preparing to purchase a machine for the shop.

The bottom line is that they still basically suck. For all of the same reasons that I have written about here on this forum (and some others, in the past. Basically, it just boils down to rigidity and electronics being the biggest culprits. Mostly, it's rigidity. After those two reasons, it starts getting into applications and ease of use specifics.

End result - No one will put their name on the line guaranteeing anything past a Q7. And even *that* pithy level will come with a stack of requirements that be observed. (specific materials, specific pitch, specific brand <partner> attachment, specific set up, specific moon phase of the year, etc... )

A good, quality tool holder will cost $15K - $25K, and that alone doesn't allow one into the dance. The machine has to have the <cough> "option" <cough> enabled. There's another $10K - $20K. Minimum buy-in at $25K to be allowed to make mediocre gears? No thank you. You can purchase a clapped out ancient Hobber and have it fully rebuilt for pennies on that particular dollar. And when you're done it will run circles around the turning center, all day, every day, 365 days a year.

Now, if you're just making mediocre gears and shafts all day, every day, then it might pay off in the long view. (like parallel spline shafts that have other features along the shaft) But for the money and effort, we will always have good, old, solid, gear driven gear machinery around here.

We are a machine shop. Not just a gear shop. Soon, one of our turning centers will have the ability to do this. And even then, the least quality-demanding gears that we make (SouthBend Lathe replacement gears) will *STILL* be made on real, actual gear making machinery. Because even a brand new, modern turning center with the most recent tooling and options cannot achieve the quality that we require while making even those.

Wow I have spent 6 hours staying up all night reading all about the pitfalls of cnc-ing a gear on a lathe and has come to a conclusion that a live hob attachment is not the way to go. For months I've been contemplating this idea. This forum truly is a gem of information.

On a side note, would shaping with a single tooth rack on a C axis lathe be a viable option in terms of economics, quality and marketability(ie good enuf as a commercial product)? Although I must say, the idea of pumping thousands of strokes into the Z screw and ways for a SINGLE part gives me the shivers... Then there's shock load on the spindle bearings, the repetitive on/off of the chuck brakes, and so on.

I did use my c axis lathe to mill the tooth ends for a set of sliding mesh gears and it turned out perfect! I think they are called nose chamfering?

 

[EmGo]

I did use my c axis lathe to mill the tooth ends for a set of sliding mesh gears and it turned out perfect! I think they are called nose chamfering?

Normally just called pointing, at least in the US and England. If you make them round it's called -- wait for it ! rounding. Both Cross and Hey made machines to do that.

Your teeth look a little shaky, I hear Garwood has a hobber that's prime for an nc conversion

 

[newcomer]

Normally just called pointing, at least in the US and England. If you make them round it's called -- wait for it ! rounding. Both Cross and Hey made machines to do that.

Your teeth look a little shaky, I hear Garwood has a hobber that's prime for an nc conversion

The gears were pre hardened and hobbed in a local shop. Not very satisfied with the level of precision.

 

[EmGo]

The gears were pre hardened and hobbed in a local shop. Not very satisfied with the level of precision.

Didn't want to hurt your feelings but would have to agree with you. Put your location in the profile, maybe someone close can help you.