integration of Jupyter with qt-applications
https://blog.jupyter.org/plug-your-appl ... 5e48918801

I will try to explain the procedure in the next days. For now there is one more file to try. Not a perfect gear, but at least it should be working from a theoretic point of view. code for animation: from PySide import QtCore def update(*args): App.ActiveDocument.dd.time += 0.02 App.ActiveDocument.re...

the 2. try to create the "involute" part of the tooth succeeded. I have documented the workflow here if anyone is interested: https://github.com/looooo/freecad.gears/blob/master/examples/worm_cutting_tool/cutting_tool_worm_assembly.ipynb the file I used and the code: import numpy as np imp...

my attempt didn't work out as the equation to solve for the contact-line is not correct. The derivative of the transformation is simple an axis of current rotation (function of t). Now trying another attempt by projecting the pitch point (line) onto the moving surface of the rack. Therefore I need t...

for helical gears it's easy to see that the rack is always straight (involute) for all cross sections normal to the gear-axis. But this is not true for a worm gear. Also there is no way to create a surface where the tooth can be constructed in 2d (like it is possible with bevel-gears, there the toot...

and another reference: https://bibliotekanauki.pl/articles/2173604.pdf

some more thoughts for the method I use: 1. the undercut surface should be easy to compute. it is created by the tip edge of the cutting tool (a part of a helix projected via the kinematics) 2. the "involute" part of the tooth is more difficult. One idea I had was this: one property of the...

I think the kinematics would be OK as it is the same as hobbing a gear. When you hob a spur gear, the hob is angled to its helix angle so its teeth are orthogonal to the gear's teeth. If you use a 'hob' with no helix angle (a circular rack) there is no need to angle the cutting tool. In addition, a...

Revolve that profile around the centreline of the circular rack I'm unsure, but I believe this is merely an approximation and won't result in a kinematically precise outcome. Maybe it's time to think about a gear-cutting-simulation again like it was already tested some years ago: https://forum.free...

Edit: for anyone interested, note that the animation in looo's post shows a multi-start worm. Another layer of complexity...

Multi start worms are already possible with the gear-workbench: parameter teeth