While looking for stuff on drawing bevel, helical and worm gears, I came across a new gear generator, posted by Rainer Hessmer three days ago. I tried it out, and thought that the following info might be helpful to others.
In my search, I came across a site with beautiful graphics that gives techniques using Rhino for creating simple spur wheels, bevel wheels, worm pinions, and helical gears. This is the “Guerrilla guide to CNC machining, mold making, and resin casting” © 2013 by Michal Zalewski,
http://lcamtuf.coredump.cx/gcnc/ch6/
Zalewski gives good ideas on drawing some complex gear forms in Rhino, that I found useful. But extra to this, he also has a graphical technique for deriving the tooth form based on the rolling motion of the teeth when they mesh - a different approach from the usual computation of an involute curve.
So here’s the point. Rainer Hessmer has just developed an online "Involute Spur Gear Builder"
http://hessmer.org/gears/InvoluteSpurGearBuilder.html
It outputs the drawing to dxf, which you can have open directly in Rhino from the browser. It uses an iterative computational method based on Zalewski’s approach. This means it will properly handle undercuts in pinions of low tooth count.
I thought I would compare this with GearGen, a script by Thomas Anagnostou that for many years has been old standby for Rhino
http://www.rayflectar.com/Rhino/RhinoScripts-Gallery.htm
(Incidentally, there is now a tutorial for Anagostou’s plug-in, written by a third party, at
http://makearchitecture.wordpress.com/people-2/varvara-toulkeridou/involute-gear-plug-in-demo/)
I compared the filesize of a gearwheel drawn by Hessmer and that by Anagnostou. The wheel was N=72 teeth, PCD=36" pitch circle dia, and 20deg pressure angle. Anagnostou’s drawing was made from 288 arcs and nurbs curves. Hessmer’s was made from 1800 straight line segments. Anagnostou’s version saved at 139kB, while Hessmer’s version saved at 90kB.
One issue was that this 72 tooth wheel only worked at ‘draft’ resolution in Hessmer’s online gear maker. It froze when I tried ‘normal’ resolution. Perhaps he will fix it soon. It worked ok for wheels with less teeth though.
When the two outputs were superimposed, the tooth forms were almost the same but not quite. Hessemer’s dedendum was slightly deeper, and there was more rounding at the base. However when I compared the two drawings of a pinion of 7 teeth, the difference was very pronounced. Hessmer’s drawing had much deeper undercuts at the tooth base. This shows the value of using Zalewski’s method rather than the standard involute calculations as used by Anagnostou.
Hessmer’s interface currently asks not for a pitch circle diameter like GearGen does, but asks for circular pitch (i.e. tooth spacing). This = pi * PCD / N, and needs a calculator, so is less convenient. But the fact it produces more accurate tooth forms, and that it might create a smaller file, means it could be a better option for some purposes.
Cheers.