Gregg, I've been playing with doing almost exactly what you are talking about. I've even cut some proof-of-concept involute gears using only a gcode program that generates the gear.
Here's some images of the test part (aluminum): http://i.imgur.com/0AYZbpi.jpg and http://i.imgur.com/LeW10nX.jpg Here's the "finished" part: http://i.imgur.com/tCYRMjJ.jpg The gear closest to the end is 18 teeth and the one farthest from the end is 17 teeth. The cutter was just a 1/8" diameter end mill. The aluminum is ~2" round stock that I cast from dirty scrap aluminum so there is some obvious porosity. The gcode program is attached to this message. I'm kind of hesitant to post my code because it's likely difficult to follow (even I have a hell of a time debugging it, and I wrote it. Unlike using a form cutter to generate both sides of a tooth space at the same time, my program uses a straight sided end mill and forms one side of the tooth space at a time. With a form cutter you move the cutter only in a horizontal plane (or vertical in the case of a fly cutter) while rotating the gear under it. My program moves the cutter at an angle from the horizontal that is equivalent to the pressure angle. While my program generates involute gears, they are not even close to standard DP of modules, but gears made with the same parameters but with different number of teeth will mesh correctly. With my code you specify what I call the root width (the width of the flat on top of the equivalent rack teeth, which is also assumed to be the diameter of the cutter), the tooth height (the height of the gear rack teeth), the pressure angle (anything positive and non-zero should work), number of teeth, and the "quality" (number of passes for each side of the tooth, bigger numbers take longer, but make more accurate forms, especially on forms with undercut). Things I would like do add (once I get one of those elusive round tuits) are the ability to make actual standard DP and Module sizes, as well helical and knuckle gears. -Chris P.S. My 4th axis may or may not rotate the right (standard) direction, so you may have to change a few signs to make the gcode work. On Thu, Jul 4, 2013 at 8:12 AM, Matthew Herd <[email protected]> wrote: > To Andy's point, it will work, it's just a matter of lots and lots of > passes. For instance, see > http://neme-s.org/Shaper%20Books/Michael_Moore/shaper%20gear%20cut.pdffor how > it can be done on a metal shaper with a rack form tool. There is > no need for an undercut. The same basic method could be used for a > ball-nose endmill but you'd need to make even more passes because you > wouldn't have the correct taper to the sides of the tool or the correct > nose radius. There's really no need to reverse if you just complete a full > rotation to bring you to wherever you need to start for the next tooth, so > that eliminates the backlash problem in your 4th axis. However, I suspect > you'd also have a problem with wear on the endmill because of the > relatively high RPM and numerous tiny (and slow) passes at a low chip-load > required to generate acceptable surface finish. You should be able to do > helixes without a problem by adding in the spiral motion, but this would > take some thinking. > > You could also use a form ground grinding wheel as a slitting saw, as in > Dave's method, to finish the gear. It's essentially a single tooth hob > with an infinite number of teeth. You could do this post-heat treat if you > wanted automotive quality gearing. One form wheel could handle any gear > tooth count in the same module/pitch diameter and you can do helixes if you > can set your head over at an angle. You can even dress the gear with the > CNC machine to ensure that it is of known diameter and accurate form. > Obviously I've given this a great deal of thought ... maybe a project I'll > attempt someday. > > Matt > > ------------------------------------------------------------------------------ > This SF.net email is sponsored by Windows: > > Build for Windows Store. > > http://p.sf.net/sfu/windows-dev2dev > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users >
gearcutB.ngc
Description: Binary data
------------------------------------------------------------------------------ This SF.net email is sponsored by Windows: Build for Windows Store. http://p.sf.net/sfu/windows-dev2dev
_______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
