When I code this: |include <BOSL2/std.scad> sinewave = [for(i=[-10:10:360 + 10]) [i/40,4*sin(i)]]; path_sweep2d(square(1), sinewave); | I end up with this: The left part of the sine is pretty, but the right side has a kink.  Any idea why this is so?  Any idea how to fix the kink?

I ran into this just today. Play with the attach parameter in the sweep. On Fri, Feb 11, 2022 at 9:17 PM jon <jon@jonbondy.com> wrote: > When I code this: > > include <BOSL2/std.scad> > sinewave = [for(i=[-10:10:360 + 10]) [i/40,4*sin(i)]]; > path_sweep2d(square(1), sinewave); > > I end up with this: > > The left part of the sine is pretty, but the right side has a kink. Any > idea why this is so? Any idea how to fix the kink? > _______________________________________________ > OpenSCAD mailing list > To unsubscribe send an email to discuss-leave@lists.openscad.org >

include <BOSL2/std.scad>
sinewave = [for(i=[-10:10:720 + 10]) [i/40,4*sin(i)]];
td = zrot(135, translate([0,-7.7], teardrop2d(r=2,ang=15)));
path_sweep2d(td, sinewave);

On 2/11/2022 7:16 PM, jon wrote: > > When I code this: > > |include <BOSL2/std.scad> sinewave = [for(i=[-10:10:360 + 10]) > [i/40,4*sin(i)]]; path_sweep2d(square(1), sinewave); | I think Adrian is on East Coast time and I don't know where Revar is, so I'll take a stab at this. I think your problem is that square() has one corner of the square at the origin, and it is that corner that is swept along the path.  That asymmetry means that the resulting form is asymmetrical.  The square will always be on one side of the path, and so the result will be different at the tops of the sine waves than at the bottoms. Try square(1, center=true). Note that the example given in the documentation uses circle(), which is centered around the origin. Perhaps this will be more obvious if I use a teardrop pointed at the origin, instead of a square. include <BOSL2/std.scad> sinewave = [for(i=[-10:10:720 + 10]) [i/40,4*sin(i)]]; td = zrot(135, translate([0,-7.7], teardrop2d(r=2,ang=15))); path_sweep2d(td, sinewave); (And with respect to the magic number -7.7, I suspect that there's a way to get BOSL2 to put the tip at the origin using anchors, but so far I have failed.)

Yes, this is exactly the answer to the problem. To position the teardrop you do teardrop2d(r=2,ang=15,anchor=BACK) or alternatively, teardrop2(r=2,ang=15,anchor=[0,1]). The reason TOP didn't work is that it points in the Z direction. I am wondering if it would make sense to have some special case handling that converts TOP and BOTTOM into BACK and FORWARD for 2d objects. Here's a visualization of what's happening in the original example where the red line shows your sweep path and you can see how the constructed object follow the path. include <BOSL2/std.scad> sinewave = [for(i=[-10:10:360 + 10]) [i/40,4*sin(i)]]; %path_sweep2d(square(1), sinewave); color("red")stroke(path3d(sinewave),width=.1); On Fri, Feb 11, 2022 at 11:30 PM Jordan Brown <openscad@jordan.maileater.net> wrote: > > On 2/11/2022 7:16 PM, jon wrote: > > When I code this: > > include <BOSL2/std.scad> > sinewave = [for(i=[-10:10:360 + 10]) [i/40,4*sin(i)]]; > path_sweep2d(square(1), sinewave); > > > I think Adrian is on East Coast time and I don't know where Revar is, so I'll take a stab at this. > > I think your problem is that square() has one corner of the square at the origin, and it is that corner that is swept along the path. That asymmetry means that the resulting form is asymmetrical. The square will always be on one side of the path, and so the result will be different at the tops of the sine waves than at the bottoms. > > Try square(1, center=true). > > Note that the example given in the documentation uses circle(), which is centered around the origin. > > Perhaps this will be more obvious if I use a teardrop pointed at the origin, instead of a square. > > include <BOSL2/std.scad> > sinewave = [for(i=[-10:10:720 + 10]) [i/40,4*sin(i)]]; > td = zrot(135, translate([0,-7.7], teardrop2d(r=2,ang=15))); > path_sweep2d(td, sinewave); > > (And with respect to the magic number -7.7, I suspect that there's a way to get BOSL2 to put the tip at the origin using anchors, but so far I have failed.) > > _______________________________________________ > OpenSCAD mailing list > To unsubscribe send an email to discuss-leave@lists.openscad.org