discuss@lists.openscad.org

OpenSCAD general discussion Mailing-list

View all threads

Re: Small Handle

L
larry
Fri, May 24, 2024 9:08 PM

On Thu, 2024-05-23 at 22:17 +0200, Marcus Poller wrote:

You should do as much in 2d as possible. That is easier to handle
mentally, easy to describe in pictures.
For instance:

    include <BOSL2/std.scad>
    projection()
    hull() {
      sphere(6);
      right(24) sphere(4);
    }

should be

    include <BOSL2/std.scad>
    hull() {
      circle(6)
      right(24) circle(4)
    }

I sometimes think in 2D, but just as often in 3D.

That is why we need the slices (slices labeled in attachment).

If you want to do a path sweep (or a hullSeries), you need to
describe slice1 and slice2. I attached a few samples what the might
look like on page 2. You need to estimate their shape and get
measurements. The shapes I propose are:

 (a) cylindrical
 (b) spherical
 (c) two spherical (requires 3 measurements: axis, r_greater,
r_lower)

There is another component for path sweep - that is centerline
(bottom line and top line will do as well, whatever is easier for you
to measure). Is your bottom line straight?
If not, you will need to provide an aproximation. Pushing your handle
onto a piece of squared paper, drawing the outer line with a pencil,
and taking coordinates from the paper will help.

Having a cut-through 2d shape (slice1 and slice2) and a path
(centerline), you can recreate the handle using path sweep.

OK... I have that figured out. I have printed a sheet of graph paper,
and printed a large outline of the shape. I will be cutting the large
shape out of the paper, and placing it on the graph paper, then using a
pencil to mark points along the slice1 and slice2 lines, each time
marking point in pairs.

I will then find the mid-point of each pair, and use those coordinates
to generate a path. I already have a shape to sweep along the path.

For the knob you need to measure two cylinders. I call them circle1
(larger diameter) and its height1 and circle2 (smaller diabeter) ind
its height2. Those measurements will provide the rounding of the
knob.

I already have the knob. BOSL2 generates it easily, including the
roundion on top and bottom.

The thing hardest to catch is slice3. I will need a sketch from you
of its shape, labeled slice3. Is slice3 constant along the
circumference of circle3?
You will also need to provide an approximation of circle3's radius.
The piece of plastic sticking to circle3s circumference will be
another path sweep.

If I understand you correctly, you are advocating doing two path
sweeps, one along slice1, and the other along slice2. If I have that
wrong, please let me know.

There is a lot of detail in this handle, it will take time, patience,
measurements and iterations to get it right. Are you still shure you
want to put that much effort into it?

I'm sure. Path Sweep is something I've been avoiding, as I have yet to
fully understand it

On Thu, 2024-05-23 at 22:17 +0200, Marcus Poller wrote: > You should do as much in 2d as possible. That is easier to handle > mentally, easy to describe in pictures. > For instance: > >     include <BOSL2/std.scad> >     projection() >     hull() { >       sphere(6); >       right(24) sphere(4); >     } > > should be > >     include <BOSL2/std.scad> >     hull() { >       circle(6) >       right(24) circle(4) >     } I sometimes think in 2D, but just as often in 3D. > That is why we need the slices (slices labeled in attachment). > > If you want to do a path sweep (or a hullSeries), you need to > describe slice1 and slice2. I attached a few samples what the might > look like on page 2. You need to estimate their shape and get > measurements. The shapes I propose are: > >  (a) cylindrical >  (b) spherical >  (c) two spherical (requires 3 measurements: axis, r_greater, > r_lower) > > There is another component for path sweep - that is centerline > (bottom line and top line will do as well, whatever is easier for you > to measure). Is your bottom line straight? > If not, you will need to provide an aproximation. Pushing your handle > onto a piece of squared paper, drawing the outer line with a pencil, > and taking coordinates from the paper will help. > > Having a cut-through 2d shape (slice1 and slice2) and a path > (centerline), you can recreate the handle using path sweep. OK... I have that figured out. I have printed a sheet of graph paper, and printed a large outline of the shape. I will be cutting the large shape out of the paper, and placing it on the graph paper, then using a pencil to mark points along the slice1 and slice2 lines, each time marking point in pairs. I will then find the mid-point of each pair, and use those coordinates to generate a path. I already have a shape to sweep along the path. > For the knob you need to measure two cylinders. I call them circle1 > (larger diameter) and its height1 and circle2 (smaller diabeter) ind > its height2. Those measurements will provide the rounding of the > knob. I already have the knob. BOSL2 generates it easily, including the roundion on top and bottom. > The thing hardest to catch is slice3. I will need a sketch from you > of its shape, labeled slice3. Is slice3 constant along the > circumference of circle3? > You will also need to provide an approximation of circle3's radius. > The piece of plastic sticking to circle3s circumference will be > another path sweep. If I understand you correctly, you are advocating doing two path sweeps, one along slice1, and the other along slice2. If I have that wrong, please let me know. > There is a lot of detail in this handle, it will take time, patience, > measurements and iterations to get it right. Are you still shure you > want to put that much effort into it? I'm sure. Path Sweep is something I've been avoiding, as I have yet to fully understand it