petebooth[iaac]

Now that Shane and I repaired the milling machine it gave the last 9 groups the opportunity to mill their 3rd assignment for Digital Tools and Fabrication from last semester.

Maria and I jumped at the chance, being the first of the nine to mill our ripple on the sheet. We had to change our setting slightly as the tool bit that we wanted to use was not in the lab, so after a few adjustments were made we were milling away fantastically.

You can find all of our cad work in our previous post here.

For our CNC milling assignment we based out design on mirrore, radiating curves. This would provide a great variation between surface finished from one end of the piece to the other. The detail below shows connection curves outside the material boundary, joining all of the line together in order to maximize milling efficiency. To add further detail to the surface, one set of lines was elevated 1mm above the finished level while the opposing set were left directly on the finished level.

engraving lines and detail.

rough cut.

We reduced the rough cut levels in order to remove excess material as close to the finished surface as possible. Even with these settings, flat sections are still visible.

Engraving cut.

The lines that we created in AutoCAD were used as engraving paths, in 2 continuous passes.

This work was not able to be finished due the CNC machine technical problems.

Based upon 3-dimensional “cellular automata”, m.any displays the concept and prototypical realization of a complete digital work flow from parametric design to production on CNC-machines. An irregular spatial structure generated by algorithms was controlled and manipulated through an especially developed software. In addition to the calculation of the complex geometry, the software also generated necessary parametrical construction elements for the realization.

java generation of form

A masters project for a group of postgraduate students of Computer Aided Design (CAAD) at ETH Zurich, Switzerland, m.any was realized during a three month period. Intertwined with the programming process, construction studies and fabrication systems were developed. Using construction data directly derived from the 3D-model, m.any variations could be explored and tested. Using a combination of JAVA and RhinoScript, the group were able to generate, test and prototype the design process ensuring that it met the assessment criteria. 3D printed models were created during initial visualizations, followed by laser-cut, component based models.

rhino3d generation and 3d printer tests.

The final and realized spatial structure consists out of 1500 individual parts all generated and produced in a seamless digital workflow. 111 main assembly frames were CNC milled, and precisely 1368 MDF connection pieces were laser-cut. In total 34 hours were spent in the workshop cutting all of the pieces, in addition to 16 hours of installation and construction time within the gallery.

rhino3d layout of cutting paths.

fabrication and assembly

The project represents a quite remarkable fusion of two different CNC techniques, highlighting the different capacities of each of the two machines.

The Rip Curl Canyon, by Benjamin Ball and Gaston Nogues, consists of around 4000 sheets of cardboard and is intended to transform the way people understand and interact with the spaces they inhabit. Throughout the exhibition, at Huston’s Rice Gallery 2006, visitors were encouraged to climb on and around the installation, in order to truly experience the change the object produces in the gallery’s space.

The 4000 sheets of cardboard were industrial die-cut into over 20,000 curved strips during a 12 day period. when joined together, these laminated strips act as a semi-monocoque with an intermediate plywood armature. The armature is constructed of 90×45mm timber and plywood which were individually CNC routered to fit the curvature from the 3D model.

A specific language was developed to allow the pieces to be slotted together, allowing the vast quantity of pieces to be installed within 4 days. This system allowed for very little structural constraint and gave on-site freedom to make improvised choices when installing the cardboard.

The digital information from CAD software and the cut-outlines

on site installation process, including die-cut template

installation completed in gallery

The 3d printed object sat within a 6×6x6cm cube. The bell consists of a cylindrical outer and a contoured inner surface, while the horn is held in place by the connection pipes. The bell and pipes could have been much tinner, possibly 2mm. This would have created a much greater elegance within the object. I guess that is why hindsight is a wonderful thing!

The fabrication process consists of sending a STL file to the printing software, which in turn sends the file to the 3D printer. While modeling inside rhino it is important that all the faces are pointing the same way, thankfully as my model comprised of completely closed surfaces.
The printing process is a long and slow on, printing approximately 0.1mm of powder per layer, followed by a layer of glue. A 60mm hich object consists of about 600 layer of powder. once printing and drying is complete, the model needs to be ‘excavated’ from the print bed. Using a fine brush, the loose powder was removed from print. Once completely clean of powder the model needs to be sealed externally with glue to strengthen the form.

Georgia and I have completed the fabrication and construction of our first digital tools and fabrication assignment, we call it ‘the tube’ (pretty unoriginal we know). We worked collaboratively on the design and fabrication of the object, passing many rhino files between us.

There was a bit of a problem with the rhino model though, it wouldn’t cut the intersection tubes away from the component planes. At 2am we rebuilt the rhino model, it still didn’t work. We ended up doing it the long way, creating lines through intersection and manually joining all the lines. Time consuming but the end result was very satisfying.

We had a minor technical hitch in the laser cutting as well, autocad didn’t like the text format that rhino exported. The machine cut the first sheet, but left off the text. Luckily I realized before we removed any of the pieces [see approx 1:45 into the attached movie]. We were able to convert the text in autocad and re-cut only the text. It all would have been bad if we had realized any later though.

The final product, complete with pretty lighting.

Click the image above to download the PDF of the manual.
Click HERE to download the DWG laser cut file.

Our class for digital tools and fabrication has completed our 3d printing assignment.
We were required to create a form that fits within a 6×6x6cm cube, with the only real restraints being the reference surface and the interface at the edges. The object was modeled in rhino 3d and exported to the z-corp 3d printer.

Now that all of the tubes have been competed, Shane and Thomas spent the better part of 3 hours connection all 27 of them together in the lobby of IaaC.  The images here were taken by IaaC, as they also managed to set up a decent light.  normally the space is a lot darker.
see the IaaC post about the tubes here

Georgina and I have completed the documentation and construction of our tube for Digital Fabrication.
You can download the DWG here.
You can download the PDF manual here.

Georgia and I have completed the fabrication and construction of our first digital tools and fabrication assignment, we call it ‘the tube’ (pretty unoriginal I know). We worked collaboratively on the design and fabrication of the object, passing many rhino files between us.

There was a bit of a problem with the rhino model though, it wouldn’t cut the intersection tubes away from the component planes. At 2am we rebuilt the rhino model, it still didn’t work. We ended up doing it the long way but creating lines through intersection and manually joining all the lines. Time consuming but the end result was very satisfying.

We had a minor technical hitch in the laser cutting, autocad didn’t like the text format that rhino exported. The machine cut the first sheet, but left off the text. Luckily I realized before we removed any of the pieces [see approx 1:45 into the attached movie]. We were able to convert the text in autocad and re-cut only the text. It all would have been bad if we had realized any later though.