PDA

View Full Version : Gear hobbing on lathe: worm gears, spur gears, and helical gears



PStechPaul
02-06-2015, 12:42 AM
I thought I'd start a new thread on this because some of the content is in a (thankfully) closed thread, but I wanted to discuss some of the ideas that I came up with based on the images of a worm gear being "free hobbed" in a lathe.

The question was raised about taking a full depth of cut using the hob, which was made using a rather long threaded rod with cutting flutes milled into it, and tapered over most of the distance. I observed that you may be able to adjust the DOC by starting with the lathe carriage far toward the tailstock where the hob is tapered. Once the cutting process has started and the gear blank has enough teeth to start turning, you can advance the work toward the chuck where the hob is fatter, until it is taking the full DOC.

There is a thread on MEM (http://www.modelenginemaker.com/index.php?topic=4650.0) about making helical gears, although I think they are more properly worms and worm gears when one has a helix angle of 10 degrees and the other is 80 degrees.

I found some videos of "free hobbing" that use a tap in one case and a specially made threaded hob in another case. But since the cutting tool is actually a screw thread with a helix angle, it cuts the gear with the same helix angle. It might be possible to cut a proper helical gear by moving the work away from the center axis of the spindle and cutter, to make the cuts all the way to the gear faces.

Also, I think it may be possible to make regular spur gears by making a hob with concentric grooves and lands with the proper tooth profile, and then spinning it in the chuck as the work is moved toward the middle of the hob. The gear blank should first get small nicks as it is moved toward the center of the hob, and it will rotate and cut more teeth as it is moved. If the hob is longer than the circumference of the gear, it will make a full set of nicks along the edge corresponding to where the teeth will be. Then the cross-slide can be advanced and the process repeated to cut further along the face of the gear. For the final cut it may be best to move the cross-slide the entire width of each tooth, and then proceed with moving the carriage by the distance of one tooth pitch to do the next one. If the hob is short, the gear blank can be moved away from the hob and rotated so that the teeth already cut will partially engage the hob, and it will then cut another portion of the teeth.

One more thing, using this method, the same gear tooth profile on the hob may work for any diameter gear. The hob would have a profile that matches a rack, and as the gear diameter (and number of teeth) is reduced, the tooth profile will automatically change to the sort of bulb shape with undercut. A similar process might be accomplished using a gear rack machined with cutting edges and clearance so that it could be used in a shaper. The gear and the rack would need to move synchronously so that each pass would be in the proper position for each tooth, and it would need to repeat at intervals small enough to produce an adequate finish. It might even be possible to position the cutting tool on the gear and use a sort of free hobbing method, as the gear would rotate as the rack moved tangentially on the teeth.

Here is a video of free hobbing for a worm gear, such as I want to make for the screw thread dial on my lathe:


https://www.youtube.com/watch?feature=player_embedded&v=-umcQtrn1CQ

dp
02-06-2015, 01:36 AM
I avoided a lot of work on the hob by purchasing a replacement lead screw from Grizzly for my lathe which I then used as the cutter. I think the hob in the video would benefit from more relief than the narrow gap. Most of the hob's circumference is rubbing against the gear.

Here's what I think is a good design and ultimately what I did.

http://www.helicron.net/workshop/gearcutting/gear_cutter/

PStechPaul
02-06-2015, 03:17 AM
Excellent! That is the sort of information I need, and I think I can make a similar hob if and when I need to make gears.

My immediate (or at least maybe short-term) project will be a 32 tooth worm gear to replace the 64 tooth presently supplied. I'll probably make a separate thread for that when I get into it. The convex periphery of the worm gear appears to be a much larger radius than the 9/32" for a 9/16"-16 leadscrew, and it does appear that most of the wear is on about 1/3 of the width of the teeth.

I may make the gear from a softer material, perhaps aluminum, brass, or even Delrin, so perhaps I can get by using a piece of 4140 alloy steel round stock, which I have, rather than getting tool steel (silver steel across the pond).

Here is my thread dial disassembled, before cleaning:

http://enginuitysystems.com/pix/tools/Lathe_Thread_Dial_1715.jpg

and a better shot of the worm gear, after cleaning:

http://enginuitysystems.com/pix/tools/Lathe_Thread_Dial_1720.jpg

dp
02-06-2015, 03:38 AM
If you make the gear from Delrin it will last forever and won't wear out your lead screw should you leave it engaged 100% of the time. Should you accept the challenge of making a thread dial there are examples in the archives here and at PM that are not only functional but beautiful examples of metal working craftsmanship.

Here's another setup for hobbing using the free wheel method with a mill.

http://pounceatron.dreamhosters.com/shop/hob/index.html

kitno455
02-06-2015, 08:21 AM
Why do you want to replace the gear? It is normal for the threading dial to cover 4 inches of leadscrew per rotation, which would be 64 teeth on a 16 pitch screw...

allan

Carm
02-06-2015, 10:39 AM
I too am curious why you would replace the thread dial gear; the pic looks usable, and Kitno has unassailable logic for your leadscrew.
BTW, thread dial gears can have a variety of tooth configurations, it's translation only. Could be 32 pins on the circumference. I personally wouldn't leave any gear engaged all the time- a Delrin gear can embed swarf as shown by your pic.

"Also, I think it may be possible to make regular spur gears by making a hob with concentric grooves and lands with the proper tooth profile, and then spinning it in the chuck as the work is moved toward the middle of the hob. "

Can't see how you would get any drive to the work that way. Relying on carriage steps (assuming lathe) would seem tedious and/or fraught with error, but mebbe I'm misunderstanding. It seems a longer "hob" using the hypothetical advancing carriage from tailstock method would be prone to deflection, which may or may not affect the work or one's nerves.

"One more thing, using this method, the same gear tooth profile on the hob may work for any diameter gear. The hob would have a profile that matches a rack, and as the gear diameter (and number of teeth) is reduced, the tooth profile will automatically change to the sort of bulb shape with undercut."

For a HSM of the hobbyist sort that could have some merit, but having oddball gearforms weighed against expediency/cost negates it for myself.

I don't think your multi-tooth cutter would work on a shaper as a single point tool is far easier and therefor faster to sharpen. But I'm not a shaper hand.

dp
02-06-2015, 11:17 AM
The worm gear appears to be too wide for the need and is as you suggest not precisely the correct one for your lathe. There's no need to but you can slim down the gear for aesthetic reasons. There is also no requirement the worm gear have radiused teeth for the application since there is no power being transferred. It is enough that the teeth have the correct helix angle. That is common practice on Asian lathes.

The number of teeth on the worm gear can be any multiple of the lead screw TPI. The choice will affect the number of meaningful scribe lines on the existing dial which means you will have scribe lines that have no alignment with the half-nuts. 16/64 will align the lead screw and half-nuts at quarter-turn marks on the dial. Note too the thread dial doesn't work at all for cutting metric threads on an imperial lathe.

lakeside53
02-06-2015, 11:41 AM
You can just use a notched disk for the "gear" of a threading dial... or if you want a slight helix, do as Emco does on their ($$$) Super 11 - three 3mm thick "discs" for the various metric pitches.

http://i238.photobucket.com/albums/ff150/lakeside53/misc%20linked%20uploads/emcoS11-threadingdial_zps4c9b3d5a.jpg (http://s238.photobucket.com/user/lakeside53/media/misc%20linked%20uploads/emcoS11-threadingdial_zps4c9b3d5a.jpg.html)

PStechPaul
02-06-2015, 05:04 PM
Good information. I might try making a thin gear like that from Delrin to see how it works. I might even be able to heat the lead screw with a hot air gun to a temperature above the melting point of the plastic, and form the teeth by rolling the blank along the screw. I might want to cut some rough notches at the correct spacing first, to make sure it lines up along the entire face circumference.

The reason for making a 32 tooth gear is to provide twice as much space between the engagement points so it may be easier to engage the half-nut at the correct point. This might be a problem only for the range of large pitch (low TPI) threads (8-14 TPI), like the 3/4-8 I cut for the 4 jaw chuck. For those, the change gears turn the leadscrew at (up to) twice the speed of the spindle, so there will always be half-thread (for 8 TPI) or other wrong positions that are difficult to discern with 8 points between each of the 8 marks on the dial. But it looks like it may be difficult to engage a half-size gear with the present housing, unless I remove some of the casting for clearance.

I can always just keep the half-nut engaged for the entire screw-cutting operation as is necessary for metric threads. Probably it would be a more useful project to make a back gear for lower RPM (120 is my present lowest), or use a variable speed drive. Easiest way might be to add another set of pulleys to the belt drive:

http://enginuitysystems.com/pix/tools/Lathe_Change_Gears_Grizzly.jpg

Rosco-P
02-06-2015, 05:39 PM
I can always just keep the half-nut engaged for the entire screw-cutting operation as is necessary for metric threads. Probably it would be a more useful project to make a back gear for lower RPM (120 is my present lowest), or use a variable speed drive. Easiest way might be to add another set of pulleys to the belt drive:


Would it make sense to do those projects on a machine with a lower speed, low end? How many more reasons are needed to continue to take adult classes in machine shop? Get access to: professional instruction; machines too big for your shop; too expensive to own; used too infrequently; large power requirements; expensive tooling; etc.

PStechPaul
02-06-2015, 06:23 PM
I do have access to such machines, at least to the end of this semester, but it is a 30 mile drive, and only available for a few hours Friday mornings. I think it is more important to modify my own machines, if possible, to do what I would like, whenever. Then again, my shop is unheated, so I can't do much now that it's literally freezing. Don't need as much coolant, however. :rolleyes:

dp
02-06-2015, 06:39 PM
Good information. I might try making a thin gear like that from Delrin to see how it works. I might even be able to heat the lead screw with a hot air gun to a temperature above the melting point of the plastic, and form the teeth by rolling the blank along the screw. I might want to cut some rough notches at the correct spacing first, to make sure it lines up along the entire face circumference.

You should definitely gash the blank so that it isn't dependent upon BSL to self-form the correct pitch. When the hob first makes contact the length of the contact point is very small and prone to skidding. My first gear blank was ungashed and it came out one tooth shy of 24.


The reason for making a 32 tooth gear is to provide twice as much space between the engagement points so it may be easier to engage the half-nut at the correct point. This might be a problem only for the range of large pitch (low TPI) threads (8-14 TPI), like the 3/4-8 I cut for the 4 jaw chuck. For those, the change gears turn the leadscrew at (up to) twice the speed of the spindle, so there will always be half-thread (for 8 TPI) or other wrong positions that are difficult to discern with 8 points between each of the 8 marks on the dial. But it looks like it may be difficult to engage a half-size gear with the present housing, unless I remove some of the casting for clearance.

Some of the functionality of the dial is to help you create multi-start threads which very few of us need. A 16/32 ratio will probably work for you for single-start threading for the rest of your life. You will still have the 64t gear if you need more. Tag it, though, so you don't forget what it is for.


I can always just keep the half-nut engaged for the entire screw-cutting operation as is necessary for metric threads. Probably it would be a more useful project to make a back gear for lower RPM (120 is my present lowest), or use a variable speed drive. Easiest way might be to add another set of pulleys to the belt drive:

I got around the speed problem when a friend gave me a 2-HP VFD. I bought a 1-HP, 3-phase motor and replaced the fixed-speed motor on my lathe. That is a really nice upgrade.

gbritnell
02-06-2015, 07:11 PM
Hi Paul,
Just a little clarification. A worm and wheel aren't the same as a helical gear set although the worm has it's teeth formed as a helix around the axis the wheel has basically a straight sided tooth form cut at an angle to the axis.
Whereas a helical gear set has the teeth on both gears cut as a helix around the axis. Helical gears don't all have to be 45 degrees to each other either. By changing the helical angles the pitch diameters can be altered to give variable sizes.
The gentleman that came up with the fixture and process on MEM for cutting helical gears is named Chuck Fellows. When I was building my Ford 300 six engine I followed his instructions, made the fixture and cutter and created my own set of helical gears to drive the distributor from the camshaft.
gbritnell
http://youtu.be/blaZ5tz0_6E?list=UUPvNzXJm9KOlaQwjAmYW9Xw

PStechPaul
02-06-2015, 09:18 PM
That is pretty sweet! Looks like you can cut each tooth on just one pass, which saves a lot of time.

It may be possible, although probably ill-advised, to make a helical gear using a single-point tool on a lathe and set the advance so that it moves one diameter per revolution of the spindle for a 45 degree helix angle. This would involve a major modification to the change gears and/or QCGB for anything other than tiny gears. On my lathe the fastest setting is for 8 TPI which moves 1/8" per revolution of the spindle. A 45 degree helix angle is formed when the pitch = pi*d so this would work for a gear with diameter of 0.039" (which has a circumference = pitch). For a more normal small helical gear with diameter of 0.5" the carriage would need to turn 0.5*pi or 1.571" per revolution, and with a lead screw of 16 TPI it would have to turn at 1.571 *16 or 25.13 times the spindle speed. :eek:

I think I have talked (or typed) myself out of making such an attempt. ;)

Thanks for the good video.