Version 4.28 online.

[ArtF]

Hi Guys:

 Version 4.28 fixes a few bugs in meshing.

It also has a notable change to the epicycloi dics. Ken recently
asked why my diameters were larger than speced from a cutter vendor,
and helpfully sent me the specs listed.

  Investiga ting this showed me that while our Epi's where mathmatic ally
correct, they didnt match horologic al wheels. A lot of this seemed to have
to do with clearance s. So as a test this version changed to a more horologic al
methodolo gy with larger clearance s and is much closer to "standard". I can redact the changes
if it becomes an issue, but these new epi's do match diagramat ically to examples provided
in the standard.

     While GM's diameters will still not match, this is now due
to the fact that while one cutter can cut any involute wheel size for a module, an
epicycloi dic cutter actually needs to change for every tooth count. Since that
cant really be a solution, they use the same cutter for ranges of teeth and specify
the addendum for that range, since GM uses true epi's, its diameters will not
match except at certain tooth counts so the addedndum is  calculate d amount based
on the tooth width as corrected by the standard, and the addendum size as calculate d by
the epi curve generator . Larger wheels do not come to the point, but Im working on a solution
 to that, the tip reduction is due to some of my routines, and is done on curve
sections unused to increase clearance .
  At worst you should find these epi's suffer from no binding force.

Thx
Art
 

[John S]

Hi Guys:

 
Due to the fact that while one cutter can cut any involute wheel size for a module, an
epicycloi dic cutter actually needs to change for every tooth count.
 
Thx
Art
 

??  Involute cutters work in band of 8 to cover the full range, a hob will cut the whole range or am I not understan ding what you are saying


John S.

[ArtF]

John:

  Just me being me..  Your right of course, both require bands for each cutter,
I was refering to hobs. A hob is basically a rack so the sioze of the gear doesnt
matter, but the epicycloi dics unless Im wrong cannot be hobbed except as a band similar to the restricti on of a cutter.
   While a proper involute is formed by the hob ( since its just simulatin g a rack), I dont think a epi curve can be done as a smulated rack, all the hobb for a epi could do is simulate a epi curve for a certain tooth count, so I think a epi hob
would have to be done for a range of tooth count. I have no proof of that, just seemed intuitive as I went through the epi math.

Art

[John S]

A few years ago I saw a hobbing machine on a Chinese web site for epi gears.
It was very very expensive which led me to believe it wasn't just a question of timing the hob to the blank as in involute's.
Perhaps other actions came into play depending on tooth count ?

John S.

[ArtF]

John:

  yeah, thats what I was suspectin g. A rack will always produce a mathmatic ally perfect involute on any size gear as a function of the math.. but an EPI curve relies on the diameter of both the wheel and pinion, so the curve isnt a function of only the gear your cutting, but of the gear its going to mesh to as well, so the hobber you saw probably had selection s to change the shape of the hob based on the requested wheel AND pinion tooth counts. Sounds like a mechanica l nightmare to me.. but Im sure it used some sort of banding function to range the request to the closest epi curve to that set. ( or I suspect it anyway. )

Art
 

[John S]

Aroused my curiosity enough to go looking.
No pictures but found this descripti on.

Descripti on of Y3280 Cycloidal Gear Milling machine

is capable of moving in an eccentric circular path as well as vertiacll y along the work axis.

 
    it is used for producing different pin teeth cycloidal gears with a convex half-circular dise type Milling Cutter. it is a necessity for plants that produce cycloidal gear reducers in small batches ot high productio n volume.
      the machine is a variant of model y3180h gear hobbing machine. in order to generate the required cycloidal teeth profile, the cutter head, carrying the convex half-circular dise type milling cutter, is capable of moving in an eccentric circular path as well as vertiacll y along the work axis, while the work table carrying the workpiece indexes continuou sly. because of this arrangeme nt, the machine is endowed with both accurate indexing of hobbing and high metal removal rate of milling.

[Mooselake]

Hear's a fuzzy picture:

http://www.unitechmachinetools.com/products/machining-machinery/gear-processing-machine/y3280-general-purpose-cycloidal-gear-milling-machine.aspx

Kirk

[ArtF]

lol, kinda what I thought. The math doesnt work like an involute, so "special" handling is
required. .

Art

[Ken Condal]

Hi Art,

Thanks for tackling the epy issue we've been discussin g. I haven't downloade d the latest version yet but was wondering if an epy project built in the previous release will reflect the new calculati ons when loaded into the new release.

Thanks,
Ken

[ArtF]

Hi Ken:

   No, I dont think they will. However, since the center distance hasnt changed,
the mechansm will still work as designed, but the gears wont have the tolerance s the new ones have in terms of clearance .

 Art