Gear formulas question.

[putputman]

Steve, I am trying to picture this in my mind. When you take the final cut at .090 depth, you remove all the metal that was left from the previous cut at .085 depth. I think you get your "involute profile" from the cutting tooth above & below the primary cutting tooth of the hobb. It will not be a true involute profile, but would be suitable for our type of use.

I think someone with good 3D drawing skills could show this. I may try to draw it up with my 2D system.

 

[Cedge]

Putput
Yeah.... I'm trying to visualize it too. I keep coming up with the last cut erasing things. As you say, it isn't going to be a "true" involute, but close enough for any foreseeable needs I'd have. Some things are just hard to get one's head wrapped around without seeing it happen.

Steve

 

[ksouers]

Try this site.

http://www.metallmodellbau.de/GEAR-CUTTING.php

It's in German, but he has drawings. That should unwarp things a bit.

 

[Cedge]

Kevin
Excellent site!! The one thing I immediately noticed is that he's taking full dept of cut. The second was that what I was seeing in my mind's eye is correct. The next tooth over on the hob only clearances the bottom of the gear tooth. Made sense, but nice to see it confirmed.

Definitely high on my to do list..... if the sunny South ever becomes so again. I'm now in full withdrawal from getting no shop time.

Steve

 

[putputman]

Kevin, Thanks for that post. I will bookmark it. I can't read it but the photos tell the story pretty good.

That is pretty much how I imagined it would look. I thought the cutting tooth would be on center instead of straddling center. When I cut gears with an involute cutter, I always put the center of the cutter on the center of the gear. I wonder if the profile that he illustrated would look just a little differant if he put his cutter tooth on center of the gear.

 

[Cedge]

As long as the cutter is dead on the center line of the gear blank I don't suppose it would really matter much. One trick Duclos shared was to leave a tiny tip on the end of the mandrel to make setting the cutter height easier.

I can see where a small tapered point could make this guy's way work quite nicely. When both sides of the hob tooth touch the little taper, you have to be dead on centerline.

Steve

 

[ksouers]

As long as the cutter is dead on the center line of the gear blank I don't suppose it would really matter much. One trick Duclos shared was to leave a tiny tip on the end of the mandrel to make setting the cutter height easier.

I can see where a small tapered point could make this guy's way work quite nicely. When both sides of the hob tooth touch the little taper, you have to be dead on centerline.

Steve

He has a picture doing just that. He has a point held in a three-jaw chuck setting the cutter height.

Take a look around his site. I can't figure out if the guy just does amazing work or is just such a good photographer that it looks great. Maybe a little of both. There are online German translators you can use but most have problems with the compound words so prolific in the German language.

 

[gmac]

Run the site thru Google Translate and it makes even more sense.... :big: I'd been to the site before and had to find a German speaking friend to help me out, not much though, he's not too technical. Google did well enough.

Cheers
garry

 

[precisionmetal]

After the part has made one full turn, the cutter could be shifted half a tooth in Z, and the gear could be indexed half a tooth... then run around one more time. Eventually if this is done enough, it begins to approach what is accomplished in a hobbing machine.

 

[shred]

After looking at HobbyNut's videos and that German site, I'm noodling on the idea there's an easier way to set the geometry for the 'straight-hob' of any given DP/PA combo (yes, I know it might not truly be a 'hob' per-se, but I'm calling it one for convenience)

I started thinking about it when grinding the 40-degree bit (an otherwise fairly mindless task on a T&C grinder)-- I think the need to precisely grind the tip width as Hobbynut does is to set the infeed depth and thus generate the size of the flats on the outer diameter of the hob teeth. So, why not grind a pointed tool and just infeed it into a slightly oversize blank until it's all nice sharp crests and valleys? Then measure it and go back over it with a turning tool and turn off the peaks to get the right size flats.

Since the 40 degree angle is known, the spacing of the V's is known, and the size of the desired flat is known, the actual OD change required should be easy to figure out. Time to wrestle with some geometry.

 

[bob ward]

After looking at HobbyNut's videos and that German site, I'm noodling on the idea there's an easier way to set the geometry for the 'straight-hob' of any given DP/PA combo (yes, I know it might not truly be a 'hob' per-se, but I'm calling it one for convenience)

I started thinking about it when grinding the 40-degree bit (an otherwise fairly mindless task on a T&C grinder)-- I think the need to precisely grind the tip width as Hobbynut does is to set the infeed depth and thus generate the size of the flats on the outer diameter of the hob teeth. So, why not grind a pointed tool and just infeed it into a slightly oversize blank until it's all nice sharp crests and valleys? Then measure it and go back over it with a turning tool and turn off the peaks to get the right size flats.

Since the 40 degree angle is known, the spacing of the V's is known, and the size of the desired flat is known, the actual OD change required should be easy to figure out. Time to wrestle with some geometry.

I was thinking along these lines as I watched the videos. Another thought is to grind a sharp tool, do the sums and plunge it in the right distance to form the correct crests as you go.

Gear cutting from scratch is one of the things I want to try, and I'm starting to think that making a straight hob (or whatever it is more correctly called) will be easier than producing a single point cutter. Which probably means I'm overlooking something.

It was also interesting to learn that the pressure angle of a gear is more than a just an abstract concept, it actually has a physical counterpart, ie the hob for a 20° PA gear is cut with a 40° point.

 

[ironman]

On one of the sherline forums on Yahoo Groups is a guy named hobbynut. He has many videos on YouTube about cutting gears both ways. Under Sherline Gear cutting. He also has a program he wrote in the files section (of the forum) called Gear Code It. I have been using that program for over a year. Just put in a couple of things and it does the rest, including making the G-code to run on cnc mill. ironman

 

[Richard1]

I think these ideas were first documented by J A Radford of New Zealand in the 1960s and published in the Model Engineer. I have a book of his called "modifications and improvements to your lathe" or something like that and he goes into quite a bit of detail about making the hobs and form relieving them. If you could find a copy of the book or the articles it might be helpful.
Richard

 

[shred]

I finished up a quick 'straight-hob' the other day and tested it. Watching it cut was tedious, but the resulting gear-shaped-object would actually run against a 'real' 24DP gear, though it didn't engage all that well; I attribute this to a rather sloppily made hob.

That got me thinking again...

Not only could you (with CNC) use a simple 40-degree double-bevel cutter (think a T-slot cutter with the teeth beveled off to a <> shape as seen from the side) to shape each gear tooth (well, tooth gap), about as perfectly as you wanted, but you could use it to emulate the multi-tooth straight hob with or without CNC--

Run it around the blank (well, rotate the blank, but you get the idea) at center height, then move it up the CP and run it around again, and move below center by the CP and run it around again. Same cuts as if you had a 3-tooth straight hob. More cutting passes, but less work making the cutter, and a 40 degree cutter could easily be ground from HSS toolbits and used flycutter-style (which might make for a longer-lasting tool than hardened drill rod), or turned directly on the lathe without needing the intermediate 40-degree lathe tool. One flycutter would do for any gear size of a given DP, but I suspect you'd have to have the proper flat on the tip, preventing use of one flycutter for many different DPs.

A 40-degree V- engraving cutter might also work (and those are available off the shelf) applied to the top of the blank rather than one side, but it would have to spin faster than my machine is keen to. It would seem such a thing could cut helical gears on a CNC.

 

[Maryak]

Shred,

I've been following this with much interest as I intend to try the straight tooth hob method for the gears on my hit and miss.

Maybe I've missed something but you seem to have gone full circle and perhaps defeated the purpose of a hob as opposed to a single point flycutter.

I accept that straight sides are much easier to grind than an approximation of the involute curve; but it seems to me you end up with a series of ridges on the sides of the teeth, similar to using the hob and have introduced more potential for error moving up an down the Z axis and bringing backlash into the equation.

Just my thoughts on it. :-\

Best Regards
Bob

 

[Jasonb]

The gear cutting machine in this set of photos shows how using a cutter like a hob will be cutting several teeth in one go to create the involute and saves having to buy a full set for each profile, you just need to index the blank for the required number of teeth

Jason

 

[John S]

Not quite Jason,
That's a Sunderland gear shaper and it rolls the gear as the cutter moves down to create the true involute.
It then retracts, steps back and does the next few teeth.

John S.

 

[Jasonb]

Ah, missed the bit in teh middle, thinking about it now if the wheel was only indexed you would just get maybe three facets to teh involute curve.

Jason

 

[bob ward]

I accept that straight sides are much easier to grind than an approximation of the involute curve; but it seems to me you end up with a series of ridges on the sides of the teeth, similar to using the hob and have introduced more potential for error moving up an down the Z axis and bringing backlash into the equation.

Thinking about it a bit more, I can see that a straight cut hob is not going to cut good gears on a manual mill. But, as I understand it, a spiral hob will do that?

Is making a DIY spiral hob as simple as combining straight hob techniques with single point threading?

 

[shred]

Thinking about it a bit more, I can see that a straight cut hob is not going to cut good gears on a manual mill. But, as I understand it, a spiral hob will do that?

Is making a DIY spiral hob as simple as combining straight hob techniques with single point threading?

Maybe... except the "thread" pitch required usually involves numbers that can't be generated with the typical lathe. I read a discussion on it over on the HSM forum a while back. Seemed like PI was involved. Perhaps a decent approximation can be created. A proper spiral hob ought to be able to make 'correct' gears of any size.

Straight-hob cut gears will be 'good enough' for many if not most ME purposes, similar but with different tradeoffs to the traditional commercial or fly-cutter gear-cutters; even with a perfect grind, it's only 'correct' for one tooth count.

With a CNC (or manual and lots and lots of patience), and one 40-degree cutter (or straight hob) you could make the approximation go as many rounds as you'd like, though I think it will always remain an approximation of some sort.