Remaining Worm Gear Problems

[adrianVmariano]

Since worm gear progress is stalled I thought I should try to capture outstanding issues before I forget what they are.

1. Mesh seems to be problematic, especially at high starts. Maybe because helical angles don't match?
2. Crowning appears to be broken, resulting in bigger teeth instead of smaller, and altering worm gear thickness and maybe also tooth angle
3. worm() and enveloping() worm should be merged and both should take matching args (angle and length), so that it's easy to switch out one worm for another
4. worm gear teeth are invalid in some cases, failing cgal (maybe due to self intersections at teeth valleys). See example below.
5. backlash is greater than zero even when no backlash is specified. This may be not a big deal, since it's small, but it's not how other gear code behaves, so raises questions.

On the matter of crowning: this is a method for shaving a bit off the faces of the teeth so that they are (more?) convex than before. It should make teeth smaller. And shaving something off teeth cannot possibly make the gear thicker!

This code is failing CGAL:

starts=1;
ps=0;
pa=30;
dist_ba=0;
gear_ba=0;
t = worm_gear_thickness(mod=3/2,teeth=60,worm_diam=44);
echo(t=t);
//bottom_half()
projection(cut=true)
//bottom_half()
{
  color("lightgreen")
 xrot(90) 
 zrot(-90)
 enveloping_worm(     mate_teeth=60,$fn=64,pressure_angle=pa, 
//                       worm(l=35,
       d=14, 
       mod=3/2,//circ_pitch=3*PI/2,
       starts=starts);
right(worm_dist(d=14,mod=3/2,teeth=60, starts=starts,profile_shift=ps,backlash=dist_ba,pressure_angle=pa))
  zrot(360/30*.25)
  intersection(){
    worm_gear(
       mod=3/2,pressure_angle=pa,
       teeth=60,crowning=0,
       worm_diam=14,profile_shift=ps,
       worm_starts=starts,backlash=gear_ba);
    //cuboid([200,200,-7+worm_gear_thickness(mod=3/2,teeth=60,worm_diam=44,crowning=0)]);
}
}

Backlash examples, cylindrical (top), enveloping (bottom).