GearHeads Forum

Hi All:

  just a progression update. Im halfway through the new escapements module, when done Ill add
a video and do a release. Again I find too much variation in standards, so Ive adopted the numbers
from the British Horological Society. Initial release will not include Grasshoppers as I want more time to
investigate a few better ways of doing them, so I will offer 3 escapements in gradually increasing
difficulty for the maker. The easiest will be the recoil escapement, intermediate will be the new graham
calculations, and finally a chrono escapment for those that want to play with kinetic sculptures.

Hope to have something out within a couple weeks..

Art

yeah, ball end dropped slightly id depth with current code woudl likely work..I dont think special coding for it woudl make it much better..

Art

ArtF wrote: Kirk:

Well, on the bright side you can now do sprockets on 4th axis, but due to the way the arcs in a sprocket work the rooting is fairly nasty, BUT with a very small depth per pass , it will actually do a pretty decent job at a sprocket, just some dressing with a file
I suspect.. In most 4th axis work under the new system, you can get alias artifacts on all but involutes, that will be typically approximately the size of the depth per pass settings and can be dressed away Id think....This is why I never allowed sprockets
and such in the past, but then as someone pointed out to me, better to have it half or 3/4 ass'ed than not at all..


Art

I specialize in fractionally assed stuff anyway   Would a ball end mill option be worth considering, or does it make the math too complicated?  You'd still need to file, but extra time in small stepovers might not make it too bad.

Kirk

Edit:  sorry, I'd mentally switched the axis and was having the ball end bit plunge in the Z direction.  Sorry, too many (legally prescribed) percocets.

It's slitting saw, Plick

John T:

  Thanks for the report, it really is as important for me to hear the sucesses as it is to hear ( and fix) the failures. It helps me to guage how systemic a failure is , though Ill admit I tend to regard no report as a success report. ( Which is probably not a good thing to do., )

  Art

John:

  And while your here...

  I may have trouble with terminology... While Im OK with calling the main 4th routine a "Tangental Bit" routine, Im not so sure "saw" is proper or even "Fly Cutting" for that matter..its just the terminology Im used to using. Please correct me if the terms in your opinion would be better as something else. Terms like "Involute Saw" to indicate a side based fly cutter may be misleading...or not.. As always you experts can chime in anytime. Remember that Im a simple coder, not a real machinist, though Ive sat at the big boys table for long enough to have picked up a feel for the lingo and proceedures, I often find I have to make up a word or phrase, and while things like "Charge Pump" or "Kernal Speed" take off and become part of the lingo, term like "Tangental shaving" may just sound silly to people who cut steel all their lives, so whenever you feel Ive strayed to terms simply too weird for you to adapt to, yell at me to use a better term like "???".
  Ive become far too complacent about making up terms to suit a condition, in Mach3 that was OK as the world
tended to adopt my silly phrasology, but if anyone has terms they feel more adequately describe something we
have as an option, please chime in, there will be no hard feelings. Truly it wasnt that long ago that I had you to thank for explaining to me what the heck a hobb was and one "Hobbed" a gear. In that regard many would be justified in blaming you for GM ever being built.. lol .. Now that I think of it..I can imagine some poor slob sitting around wondering if people thought him silly for calling a multiheaded gear cutter a "hob". On the other hand maybe his name was Julious Hobb or something..... Maybe Ill just call the process of cutting tangentals sideways an Art....but then isnt all gear making an "art" lol

Art

John:

  You can. Were it me, Id use a ball end and instead of zeroing it, Id zero the bit and set to Z to .2 or some number that makes the ball go just a touch lower then spec. While it would cut that much lower, the small offset should
lessen the aliasing artifacts in the root and make it easier to dress.

    I imagine there are solutions we can find to all this, but in terms of speed efficiency, using a ball end, with
a small dpeth per pass on the rooting should make most things very close to the end desired result of matching
the core shape as close as we can. Obviously though, this is all about time. Any program that tries
to do a perfect curved sprocket root when doing side to side cuts will in the end have the same artifacts
GT will put out, the only solution would be to do a great many contours acrosss the blank which would take
near forever. Dressing would be more efficicent.

  I also question whether we need to worry about having unidirectional cutting, while it may help finish a bit
on the edges of a tooth, the time involved to rise the bit, cross over, then lower for the next cut likely is a time
loss greater than the increased finish is worth. I guess thats something worth discussing as well.

Art

ArtF wrote: Kirk:

  Well, on the bright side you can now do sprockets on 4th axis, but due to the way the arcs in a sprocket work the rooting is fairly nasty, BUT with a very small depth per pass , it will actually do a pretty decent job at a sprocket, just some dressing with a file
I suspect..


Art

Can you use a ball end mill to ease the root problem ?

Hi Art
The simulations I have tried all work as advertised - Thanks

John

Craig:

  I may.... Ill look into the SDK to see what it woudl take..

Art