RepRap: Blog

Another reason for obsessing on round parts built with a quadratic engine is that good circles are hard to make. If we had a radial engine, then the challenge would be in making a square box.
 

How did you get the neat reddish cast to the chalice on that stemmed glass?
 

Oh! (cleans reading glasses) Now I see it. It's got wine in. Sorry. :-p
 

So, then, the ultimate test of circles and overhangs would be, what, a hollow torus?
 

Now we only need a heating liquid dispenser.

And a computer voice interface.

So we can say:
Computer, tea , old gray, hot
 

No no no... remember, RepRap has been looking into peristaltic pumps. Now, once we get to RepRapping some low pressure valves, I believe a full CNC wet bar is not far off.
 

Hmmmm.... did I say quadratic? I meant cartesian.
 

steve: A torus requires too much of an overhang for the RepRap to do without support material. The wine glass is pushing it, and 45 degrees seems to be the limit.

Vik :v)
 

vik:

would you mind trying a torus just to see what happens? i'm sure you could find a model online, and that it would be interesting to see exactly how it fails.

after all, failure can sometimes be more informative than success.
 

I reckon a torus is printable, as long a the walls are thick enough. Would be interesting to know how thick they need to be. Thinking about printing something with embedded pipes, for example the darwin corners ...
 

I fear Vik's right - the torus problems would be at the very beginning and the very end, when the overhang is effectively flat. Remember the significance of the RepRap teardrop logo...
 

Yeah, I figure the best you could hope for would be a pseudo-torus with a 45-degree ridge.
 

Adrian :
Agreed for the lower part, you´ve got to give it a bottom with a rectangular cross-section. But I still think you can close the top side of a torus without too much distortion.
The technique required remains to be developed. Lowering the nozzle closer to the surface should make the last layers more flat, allowing for more oerhang. That´s going to be tricky for the software, though.
 

Normally I'd have to agree that a torus is not possible, but I see some exceptions. We've had some demonstrations of several centimeters of parallel airborne dispensing of HDPE (by mistake and on purpose... Forrest did some looking at this IIRC) but these always had each segment land itself on a solid on one end if not both. This would lend itself to some 3D printing (Reprap FDM is technically 2.5D printing) of self-supportive arches, but these would have to be laid down underneath the main torus structure and then the torus could be concentrically built.

Unfortunately, I'd see a thick wall torus being difficult without the arches because, at some point, you have to go "airborne" to lay down the ring layers. Unless you modify the design with an internal "step" and start going across the inner and outer wall gap, thereby squaring the upper corners and probably squaring the bottom of the torus, you will not be able to generate an equivalent shape.
 

This is really cool, I'm very happy to see this project evolving so quickly.

I'm interested in the interface though: what file format does it print? Could I extend it to use my format of choice? Is the interface itself scriptable (with something like LUA or Python) so that I can write a humane one for it?
 

The file format is the industry-standard STL, as generated by ArtOfIllusion, Blender et al.

You have to make sure that the shape is closed and walls do not pass through each other - ArtOfIllusion has a "Solid Editor" tool that does this nicely.

Vik :v)
 

Stunning Vik.