Greetings all;

In order to reduce code duplication in this drive, I am about to embarque
on a make 3 modules out of two edit, by splitting the inner facing
spline code into two modules, shareing a third outer module because that
is already identical, or is supposed to be.

And because in a harmonic drive, the difference in the number of inner
teeth, or splines puts more stress on the splines due to the difference
in tooth heights, I've chosen an arbitrary lower limit on the effective
gear ratio of 30/1. Commercial steel versions are generally more like
100/1, but that puts a much more extreme precision requirement in the
printers rendering of the splines. That probably puts a 50/1 ratio as
the maximum practical limit for a device this small size. At 30/1, I can
get nice clean rendition of the splines w/o changeing to a much smaller
nozzle than the default .4mm.

My mental question then, since I am generating the spline as a cylinder
with an $fm=3 to make the triangular spline, then rotating the inner
facing pattern 180 degrees, then merging the bases at 6 degree
increments for the 60 tooth, and at 62, an angular increment of
5.8064516129.

But this puts the all the change in the effective spline height in the
puter radius of the valley where the splines bases merge, and ideally it
seems like for a least stress engagement, I should adjust the translate
statement in the 62 tooth inner face, by some of that difference. It
does however, leave the tip of the spline at a constant radii. Maybe
thats best because that will allow full engagement of the spline tips.
IDK.

The difficulty in doing this is complicated by the need to keep cura out
of retraction hell by keeping the size of the merged area of the spline
large enough to allow an infill connection from spline to spline. So
there is a very firmly enforced by cura limit to how small that merged
area can be, and I have to assume any slicer does this. This condition
does warn you because it will at least double the estimated render time
and will probably clog up the printhead from thousands of retractions
long before the render is finished.

BTDT quite a few times and that fix is a full diss of the hot end and
burning the blockage out in the cookstove burners gas flame. Not my fav
pastime for obvious reasons.

But I have zero idea what the math to do that would look like, so here I
am, hat in hand. Can anyone advise me? Assume I have just enough algebra
to be dangerous.

Thanks all;

Cheers, Gene Heskett

Greetings all;

In order to reduce code duplication in this drive, I am about to embarque
on a make 3 modules out of two edit, by splitting the inner facing
spline code into two modules, shareing a third outer module because that
is already identical, or is supposed to be.

And because in a harmonic drive, the difference in the number of inner
teeth, or splines puts more stress on the splines due to the difference
in tooth heights, I've chosen an arbitrary lower limit on the effective
gear ratio of 30/1. Commercial steel versions are generally more like
100/1, but that puts a much more extreme precision requirement in the
printers rendering of the splines. That probably puts a 50/1 ratio as
the maximum practical limit for a device this small size. At 30/1, I can
get nice clean rendition of the splines w/o changeing to a much smaller
nozzle than the default .4mm.

My mental question then, since I am generating the spline as a cylinder
with an $fm=3 to make the triangular spline, then rotating the inner
facing pattern 180 degrees, then merging the bases at 6 degree
increments for the 60 tooth, and at 62, an angular increment of
5.8064516129.

But this puts the all the change in the effective spline height in the
puter radius of the valley where the splines bases merge, and ideally it
seems like for a least stress engagement, I should adjust the translate
statement in the 62 tooth inner face, by some of that difference. It
does however, leave the tip of the spline at a constant radii. Maybe
thats best because that will allow full engagement of the spline tips.
IDK.

The difficulty in doing this is complicated by the need to keep cura out
of retraction hell by keeping the size of the merged area of the spline
large enough to allow an infill connection from spline to spline. So
there is a very firmly enforced by cura limit to how small that merged
area can be, and I have to assume any slicer does this. This condition
does warn you because it will at least double the estimated render time
and will probably clog up the printhead from thousands of retractions
long before the render is finished.

BTDT quite a few times and that fix is a full diss of the hot end and
burning the blockage out in the cookstove burners gas flame. Not my fav
pastime for obvious reasons.

But I have zero idea what the math to do that would look like, so here I
am, hat in hand. Can anyone advise me? Assume I have just enough algebra
to be dangerous.

Thanks all;

Cheers, Gene Heskett

Hi Gene,

I guess these are the sort of shapes you are after

//gene
module point(){
translate([20,0,0])cylinder(50,d=5,$fn=3);
}

for(j=[1:60]) rotate ([0,0,6*j])point();

module in_point(off){
translate([20+off,0,0])rotate([0,0,180])cylinder(50,d=5,$fn=3);
}

#for (j=[1:62]) rotate([0,0,(360/62)*j])in_point(4);

If it is 60 deg points, then you can calculate the height of the tips,
increase size if bigger diameter/etc. (the internet, will give the
answer for dimensions of triangles in circles). So, chose a big enough
triangle/cylinder (or do 2d and linear extrude). To make the 'rim'
thinner, intersect with a cylinder for the external ring, difference a
cylinder for internal. The 'off' value will be filament/printer
specific.

Cura's 'retraction hell' - no idea what you are talking about.

There
are settings to reduce retractions, so that the filament is not ground
by the feed wheels, and there are plenty of tutorials on line (in the
search engine of your choice, just enter "cura retraction settings"
and read a few of them to get the idea). and for a shape like this
there should be few retractions. As I keep saying, you need to set the
printer up properly for your filament, and then you need to make the
correct adjustments, temperature, cooling, speed, etc,. within cura
(or any other slicer).

I've attached a jpeg of the cura retraction settings area, and if you
install the settings guide extension, you'll get a clue about their
function. I expect combing mode would help.

hth

On Sunday 25 July 2021 21:24:17 Ray West wrote:

Hi Gene,

I guess these are the sort of shapes you are after

//gene
module point(){
translate([20,0,0])cylinder(50,d=5,$fn=3);
}

for(j=[1:60]) rotate ([0,0,6*j])point();

module in_point(off){
translate([20+off,0,0])rotate([0,0,180])cylinder(50,d=5,$fn=3);
}

#for (j=[1:62]) rotate([0,0,(360/62)*j])in_point(4);

If it is 60 deg points, then you can calculate the height of the tips,
increase size if bigger diameter/etc. (the internet, will give the
answer for dimensions of triangles in circles). So, chose a big enough
triangle/cylinder (or do 2d and linear extrude). To make the 'rim'
thinner, intersect with a cylinder for the external ring, difference a
cylinder for internal. The 'off' value will be filament/printer
specific.

I also have "keys" to lock rotation without depending on fit grip to do
it.

Cura's 'retraction hell' - no idea what you are talking about.

Thats where it does a retraction to move from point to point as it
infills the point, could be 2 or more times a second and 6 or more per
each of 60 tooth*.2mm layers and its 12mm tall. Double or more, the
render time.  Wears out the filament too.

I'll do that code mod later today when I've got both eyes open.

There
are settings to reduce retractions, so that the filament is not ground
by the feed wheels, and there are plenty of tutorials on line (in the
search engine of your choice, just enter "cura retraction settings"
and read a few of them to get the idea). and for a shape like this
there should be few retractions. As I keep saying, you need to set the
printer up properly for your filament, and then you need to make the
correct adjustments, temperature, cooling, speed, etc,. within cura
(or any other slicer).

Still working on that but mking good progress.  cura's default feed flow
is now down to 82% The prusa has direct drive with no geardown, so can
retract at double or more the default speeds, breaking the string and
reducing stringing which is a major problem with petg. But with 3 of the
9 parts completed, it has certainly proved the worth of the e3d all
metal hot end, rated to 300C so I think I'll retrofit it to at least the
CR10S Pro V2, along with a bmg extruder I already have and maybe on the
BIQU BX, as it turns out that while the famous H2 head the BX wears is
nice, its also a pla only head, petg temps destroy the capricorn tubing
in it in 2 jobs.

I've attached a jpeg of the cura retraction settings area, and if you
install the settings guide extension, you'll get a clue about their
function. I expect combing mode would help.

Combing is off, as it seems to enhance the buildup of plastic on the
bottom of the nozzle when squirting petg, eventually catching and
destroying the job.

I think thats on by default in the AppImages I'm running. Nice little
quite verbose popups.

hth

Definitely helps Ray, thank you.

Take care now.

Cheers, Gene Heskett

"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.

• Louis D. Brandeis
  Genes Web page http://geneslinuxbox.net:6309/gene

OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org

I've had good results with PETG by using plastic repellent paint on
the nozzle and its surroundings.