Why dont we have end mill gear cutters ?

And Duffer's Mk1 patent Medieval version:

Seriously though, Neil pointed out in another thread that I'd over-relieved the teeth on this one. Quite right.

Nonetheless I successfully used the cutter to make eight aluminium and two brass gears, each 1/4" across. The cutter is still sharp but a careful look at the picture shows that one of the lower teeth has lost it's tip. I expect more damage later this afternoon when I try it on a steel blank.

I'm impressed by this method. The cutter is a lot easier to make than an equivalent form tooth for fly-cutting.

Dave

Agreed. I see how this method works for a gear with a large number of teeth, even the Brown & Sharpe cutters assume that anything greater than 135 is the same as a rack. I can also see that the interference from the cutting teeth modify the form either side of the primary horizontal tooth and would be sufficient to give a good form for teeth down to, say, 60. Does anybody have a formula for calculating the adjustment required either side of the horizontal to give a good approximation to the involute down to e.g. 16 teeth? How does a Sunderland shaper do that?

Cheers,

Rod

Rod,

I think the 'Base Circle' link that I posted might help your understanding.

MichaelG.

First 5 minutes of this video are quite entertaining and not all outside the realms of a home workshop with a 4 axis CNC.

https://www.youtube.com/watch?v=tqaQ-sD2rBs

Rod,

The principle works right down to the smallest of gears. If you think about it, a tiny gear (say 8 teeth) will still roll perfectly along a rack, although its teeth will be undercut to the extent of resembling tulip flowers. Rarely seen in metal but look at small plastic pinions end-on.

You reproduce this rolling motion by moving the rack form cutter along a set number of steps. Say you take four cuts, the cutter should move 1/4 of a tooth spacing and the blank should be turned round 1/4 of a tooth before travelling round the blank a second time.

In practice, a gear of > about 30-40 teeth will a have several cuts taken by a single pass of the cutter and indexing the blank means every tooth gets all of these cuts. As long as one cutter tooth is at centre height to give a full-depth to the spaces the teeth will have a decent form.

As the blank gets smaller the single pass will take fewer cuts, but even a 12-tooth gear gets three cuts per tooth.

If you rotate the blank by half a tooth and move the cutter half a tooth and go around again smaller gears will get a total of six cuts and larger ones twelve or more.

Every one of these gears was made with a single pass, they are 0.7 mod and they run together freely at their 'designed' PCDs despite some of them being on over/undersized blanks. I lie - one of them needed a light fettle as it had a small step near the bottom of the tooth flanks.

I think my cutter may be slightly over-size (i.e. the end flat may be slightly too wide so the cutter feeds in a bit too far), but the gears are more than adequate for my needs.

If I was making gears of over 1mod or 20DP I would probably take two passes for gears where the cutter only makes three cuts per pass.

Neil

OK. Thanks, I think I'might getting there

I need to draw it up

Rod

Wrong again. My iffy cutter hacked out a mild steel gear in two passes without breaking and it's still sharp.

However, although the gears it makes mesh successfully the profile I'm getting isn't particularly involute.

I think this is because my cutter teeth go to a single point when they should be flat topped: consequently they don't cut as John Stevenson's rolling diagram shows. Therefore. if anyone else has a go making one of these cutters it should look like Neil's version rather than mine. On the other hand, it's interesting that even my deviant cutter form still produces gears that mesh workably with commercially made ones.

I have a book that claims the Sunderland method can generate more accurate involutes than any other system. This is because accuracy can be progressively improved by making repeated passes round the gear after rotating the blank slightly. (i.e. along the lines of Neil's "If you rotate the blank by half a tooth and move the cutter half a tooth and go around again smaller gears will get a total of six cuts and larger ones twelve or more. "

The book doesn't mention any disadvantages but I guess hob gear cutting machines are faster and therefore cheaper for mass production. For amateur use I don't understand why the Sunderland system isn't much more popular.

Dave

Edited By SillyOldDuffer on 23/10/2016 17:08:28

.

What machine would you run it on Dave ? Very specialised to cut say 3 or 4 teeth then index round to accurately line up to cut the next 3 or 4 teeth.

Whereas on a mill, be it vertical or Horizontal it's not hard to organise the rotation of the hob to the blank, either by gearing or electronic means.

My old Victoria U2, bought off the scrap man for £100 has spent the last 10 or 12 years doing nothing but gear hobbing with the odd few specialised splines thrown in for good measure

I worded that badly John! I didn't mean that amateurs should be building and using actual Sunderland machines.

What I was trying to suggest is that it's rather easy to make a rack-form cutter like Neil's that can be spun by an ordinary mill to machine a reasonable gear from a blank held by an ordinary dividing table. It's a simple manual method that works if you're not up to organising the relative movement of a hob and blank. Also, compared with making a tooth form fly-cutter it's a doddle to make a rack-form cutter with a lathe.

So a useful technique for the junior end of the hobby, not a wonder process that renders everything else obsolete!

Dave

OK, I misunderstood your post.

Yes I do agree and if you had CNC it would be even easier to program the 10 or so steps, like a Sunderland in keeping with the blank rotation.

Ah,. The flat top is important, if it is accurately sized it sets the depth of the cutter. If it is too narrow you can feed the cutter in a bit extra,. if its too big (like mine is) the gear spaces will be slightly too big.

In the past, I have made D bit gear form cutters when we had no gear cutter at the time. It was quite easy to make as it is only 1 side of the profile to be produced. It was for a 1 off job and worked well enough at the time.

Neil

Neil, you beat me to it, I was also going to suggest that a D bit would do the job.

Ian S C

Are hypoid gears not cut with a an endmill cutter?

https://www.youtube.com/watch?v=A-xVdUukC-I

It seems not

Reference: Theory of Machines, L. Toft & A.T.J. Kersey, Pitman, London, 6th edition, 1949.

That's a very optimistic tooth profile on the cutter!

Neil

.

In what sense, Neil ?

Aside from it being economically un-resharpenable [i.e. a throwaway]; I don't see any real problem.

... cutting suitable materials, of course.

MichaelG.

I don't think so? Although the pitch surface of a hypoid gear is a hyperbolic surface rather than a cone, the size of the tooth still varies across the face of the gear, exactly the same as a straight tooth, or spiral, bevel gear. So I don't see how they could be cut with a shaped endmill.

Of course thry can be machined using a ballnose cutter on a 4 or 5-axis CNC mill, but that's not the same thing.

Andrew

Since this post was originally started I saw on another forum that  "end mill gear cutters" are available

Edited By JasonB on 13/03/2017 10:09:44