Test pieces for exploring both laser cuts and etches as channels for fluid transport and preferential diffusion.  Results soon.

While folding the hexagonal parabolic surface whose crease paths I had etched on a laser cutter, I began to notice and think about the etch marks in a little more detail.  The etch is a thinned path in the paper left by the laser combusting only a fraction of the paper’s thickness rather than burning completely through.  In a cross section of the paper, this would look a bit like a trench dug into the surface of the sheet.  A channel.  At these small scales, perhaps these etched channels could work as capillaries (at least capillaries on a different scale than those composing the porous microstructure of paper).  By wicking fluid into a piece of paper into which a pattern was etch, maybe the fluid would preferentially flow or accumulate in these channels.  And if you control the layout of the channels, maybe you could make some interesting things happen.  Lab-on-a-chip applications (lab-on-a-post-it-note?) come to mind, but first, an experiment to test the wicking/fluid flow in paper etched with a pattern.  As can be seen in the images, not all of these samples are pure etches— some are closer to cuts through the paper, as the light-transmission indicates.

Ever since I learned that laser cutters could be used at low powers to etch materials, I wanted to try using etch paths as crease paths in folded paper structures.  To test this, I picked a design that woud normally be very difficult to fold by hand:  a hexagonal parabolic surface.  After some trial and error, I settled on 4 Watts at full speed on NYC Resistor’s 35 Watt Epilog laser cutter for 65lb paper.  Since the etch is only on one side of the paper, folding is easier in one direction than the other:  mountain folds with the crest on the etch side are easier than valley folds with the valley on the etched side.  With this in mind, an even better way to lay out crease paths would separate mountain and valley folds and laser etch them onto the two sides of the paper.  This would require careful alignment, but is totally doable.  For this simple test, I just etched one side, which was totally sufficient for me to fold the etched test piece into a parabolic surface, which I was quite happy about.  

While I finished this project in May, I never finished sharing it.  I am working on updating the reprap wiki page and my main website with documentation of the project.  More soon.  

In the meantime, here are some images of the finished Bottlebot and a bottle-opener that I attempted to print in my recycled HDPE filament.  HDPE has a high coefficient of thermal expansion/contraction, so it warps quite a lot, which made printing complete objects very difficult, but I was happy with the proof-of-concept.  The system is way too Rube Goldberg-esque for practical use, but I think it was a good exploration of the issues surrounding personal production of printer feedstock.  

More folding.  Also experimenting with wicking liquid into the paper.  Need to explore how folding/compression/cutting affect the progression of the fluid through the paper and the patterns it leaves behind.

First results from my experiment to attempt to control fluid flow/diffusion with laser-etched channels in a porous media (paper).  The two samples that delivered the highest amount of directed flow (as evidenced by the dark residue along etch paths) are those that also had the deepest etches, made with 10% and 14% laser power (3.5 Watts and 4.9 Watts, respectively) on an Epilog Helix.  A control of scored paper perfromed poorly, as did the lower power (4% or 1.4 Watts) etched sample.  I will have to try this with some simpler geometries to build a more definitive case for which parameters yield the best direct flow.

A couple weeks ago, a paper that I wrote came up in a conversation about pattern recognition with a student studying emergence theory at NYU.  In the paper, I investigated Hausdorff dimension analysis of tree branches to derive their fractal dimension (fractals have a non-integer dimension) as a method of species differentiation and identification.  As the bar graph indicates, my efforts were mostly successful, though the lack of a guarantee of dimensional uniqueness calls into question the ultimate utility of the approach.  

My original paper can be found here.

Laser cut coaster.  I used Nervous System’s Radiolaria tool to generate geometries for laser-cutting.  They really have some wonderful tools on their website.  I can’t wait to make some larger geometries for light fixtures.

I had some time with the forge this past Fall and experimented more with a stencil-resist corrosion technique I’ve been working on.  I used tape to lay out a simple design on a disc of steel and then apply a clay slurry to the exposed metal.  I then removed the tape and slowly dried the clay.  While the clay was drying, I cleared the tuyere and build up a good bed of coke.  I then placed the workpiece in the forge over the now large heating area and ran the blower at full power to maximize oxidation.  Every few minutes, I flipped the piece to develop good, oxidized surfaces on both sides.  After about 20 minutes or so, I removed the piece and quenched it immediately to shock off most of the scale.  I then reheated to cherry red and let cool before buffing off the remaining scale with a wire brush wheel.  (Though it isn’t pictured, I then hammered this piece into a bowl using a recessed oak die and large, ball-faced striker.)  

Starting to work a bit with paper.  It’s an interesting material when approached as something structural.  I’d like to try some of this with steal shim stock some time.  Maybe with a little help from a hydraulic or laser cutter.