how to make [almost] anything mas.863.12   work about
3d scanning + printing: week_3  

physical -to- pixel -to- plastic

The focus of this week was an introduction to both 3d scanning and 3d printing technologies. In the spirit of keeping some common thread through these projects I chose to 3d scan a complex physical model that was developed earlier in the term. The object is a sphere composed of a fiber material and binder with both a hollow interior as well as various porosities.



After an hour of calibrating and training on the handheld 3d scanner in the arch shop I concluded that the dot application method as well as the low quality mesh generation was not going to be well suited for this object and rather than compromising the object itself I pursued other scanning methods. I chose to test the limits of the recent development from Autodesk 123d Catch. It is part of a larger 123D launch that include some software decedents from the cba. I downloaded the app to my phone but you can also upload pictures via their online site as well. The interface is simple and has instructions to get you started. Use your phone or camera to take 20-40 images of the objects from as many angles as possible without moving the object itself. Pretty simple and after 4 varied techniques and background conditions I was able to capture a decent representation and model. The model can be downloaded from the 123d site as an .obj and opened in many platforms. While moving more fully to osx I decided to try the meshlab freeware for some initial mesh cleaning.

Above is the .obj file saved out of meshlab and brought into rhino 5 beta for osx. I prefer to use rhino for the final .stl file prep because of its precision measuring and familiarity. This mesh was then cleaned and trimmed a bit more in rhino to begin to move away from a pure scan of the object into a new object to be sized to the 2x2 bounding box.

With the new object scaled and modified I offset an interior version of the mesh to best utilize my 2x2 volume and then added a shell or thickness to the mesh for the dimension printer in the arch shop. I imagined these shells being thin so that they would have some affect on light quality based upon varying material thickness but the initial 1/32” thickness was not enough to be read by the dimension. During a second edit I increased the thickness to 1/8” just to err on the safe side in order to have a print back in time.

The single print with support material removed

Both the exterior shell and the interior shell side by side showing surface details.

The final parts: In future printing of shell-like abs prints I would like to push a bit more the possibility of varying material thickness between perhaps 1/16” in light letting areas and degrees up to 1/8” in more structural regions.
  -- jared laucks -- © 2006-2012 --