On Fri, 19 Dec 2014 12:24:41 -0500, "Jim Wilkins"
wrote:

"Ed Huntress" wrote in message
.. .

One question about your project: Are you planning to use involute
cutters, like the shape used in gear milling, or are you trying to
replicate a Fellows gear shaper and generating the teeth?

It it's the former, it sounds like a big challenge, but doable. If
it's the latter, abandon hope all ye who enter here.

--
Ed Huntress

Does CNC make feeding a rack-tooth-shaped cutter across the
synchronously rotating blank practical?
-jsw

A good thought, but the short answer is "no."

The problem with an incremental-stroke approach to generating gears is
that it requires multiple storkes at very short intervals, as the
blank rotates.

A hobber effectively does this by having multiple cutters displaced
around a cylinder, and by rotating both the tool and the work to have
a sequence of cutters entering the work at close intervals.

A Fellows gear shaper does it with a lot of tool strokes at brief
intervals, as the gear blank rotates slowly.

There was a machine built by Gleason in the late '70s, called the
G-Track, that used rack-type cutters as you describe. But there were
dozens of them chained together. The chain moved continuously as the
blank rotated, acting like a shaper with multiple "racks." Or you
could think of the tool as something like a push broach, with one
cutter quickly following the previous one.

The relationships are mechanically simple, so CNC only gives you one
thing: quick-change versatility. You can set up for a different gear
very quickly with CNC. But it doesn't simplify the geometry, except to
replace one or both gear trains with some kind of servos. That's now
most modern gear hobbers work. I haven't seen a gear shaper for 30
years, but I assume they've applied the same technology to them.
CNC is a big setup-time saver, but that's about it.

--
Ed Huntress