Week 2: Computer-controlled cutting and body horror

This week I: (1) did a kerf-finding group assignment

2. drew and cut a laptop sticker using the vinyl cutter

3. designed (using Fusion360) and cut (XTOOL P2) a parametric construction kit to make weird, messed-up hands

Laser cutting + vinyl cutting

First I’ll go over my laser cutting parametric construction kit, then I’ll show you the sticker I vinyl cut for my laptop.

Parametric construction kit: hands

My inspiration for this project was a kinetic sculpture I made a while ago modelled after my own left hand:

Sketchy sketchy + fusion fun times

First I did some sketches to get a sense of what pieces I would be making and generally what they would look like.

Then, I was ready to move into CAD software (I’m using Fusion 360)! I received wise advice from my friend Ben: decompose each shape into basic geometric objects first to make dimensioning easier, then add the curves and nice touches later. So, after measuring my left hand I drew out these sketches for the simplified forms of the palm, joint, and phalange (finger segment):

I implemented these into a sketch for the base palm (ignore the curve at the bottom, I added that later but forgot to take a screengrab at this point so I captured this after all was done).

In another sketch, I parametrized the slot depth and width and used these to add slots to the extruded palm. I accounted for the kerf by subtracting the approximate kerf value from my initial (desired) slot width.

Finally I had a nice geometric palm to start with:

Next, I modelled the joint by similarly beginning with the very simple sketch I had made (in this case, a circle). I drew this sketch off of one of the inner faces of the slot in the palm to avoid having to assemble it later. Then I added one slot and used the circular pattern tool to make 6 evenly distributed slots around the circle.

After the joint I made the (intermediate) phalange, which is just a rectangle with two slots. Here I made a construction line across the middle of the rectangle and mirrored my slot across.

The terminal phalanges (with only one slot, e.g. the “fingertip”) were trivial to make, just a rectangle with one slot. In keeping with anatomical accuracy I made them slightly shorter than the intermediate phalanges. I then copy-pasted my components like crazy and used the Assemble functionality to put the pieces together. For each interface between a joint and a phalange I had to add three face-to-face constraints (one for each of X, Y, Z) to ensure the slots were fully slotted together.

(I realize now that I modelled the hand with one fewer joint in each of the pointer - pinky fingers, oops.)

Spiral design: laser cutting & learning as I go

In the name of Rapid Prototyping I cut and assembled this basic structure on the XTOOL P2 using 87% power and 37 mm/s speed, just to sanity check.

Mostly, everything looked good! I had a small hiccup where since my cardboard was bowed, though I had set the focus height of the laser to my measured material thickness, it still began to pull the cardboard piece around, so I had to stop the cut and manually quick-measure the highest point of the material, which ended up being more than twice the thickness of the material. Also, my joints were way too loose. I had measured my cardboard to be aroudn 0.16 inches thick so I’d set my cardboardThickness parameter to 0.16 inches in fusion, but everything was falling apart. So, I ran a small test to see what material thickness value would give me the joint clearance I was looking for:

I decided to go with 0.14”, since even after I deformed the cardboard as I tried to slot the slots together the joints still held well.

Smoothing edges, bending the palm

Of course, hands are soft. It was important to me that I make the hand more organic-looking, by smoothing out all the sharp corners from my base sketch. I added filleting to opening of the slots as well, and used the same fillet tool to smooth any corners. I also added a curved line to the bottom of the palm base. When all that was done, my model looked like this:

Next, I wanted to implement flexure across the palm. I was noodling with some online flexure patterns when Alfonso happened to walk by. He drew me a tiny drawing of the most fundamental flexure pattern, a grid of offset slots, and explained to me some of its more important properties (that the slots are positioned so there is a gap between them which is centered relative to the slot above it, that the spacing is sufficient to allow for this gap). He was even nice enough to draw it to scale, so I used approximate measurements from this very drawing to create my feature:

He showed me how to make the “molecule” of this flexure (two offset slots), extrude them into holes, and then use the rectangular patterning tool to pattern the features instead of the sketch objects. After some debugging I was able to use the pattern to make the slightly-diagonal row of flexure units I sought:

Hands, hnds, haaaaands

I laid out my .dxf files for printing on my last remaining bit of cardboard (I only used one sheet for this whole process!). I cut the cardboard into two pieces to reduce the bowing, but I still did the aimed measure anyways to avoid repeating my previous mistake.

Finally, my weird idea came to life! Behold, hands:

By slotting the phalanges into different slots on the joints, I could make poses, such as a peace sign:

Behold, hnds:

Behold, haaaaaaaaaaaaaaaaaands:

(My face, for scale)

After I finished this I realized it would have been fun to introduce flexure into the other joints as well. It wouldn’t be too hard to make a small piece with flexure with slots on both sides, but I decided to stop here.

Vinyl cutter sticker-making

For this part of the assignment I wanted to take the opportunity to draw myself a laptop sticker. One motif that has been floating around my head recently is a “fear” – a fish-bear. I thought I would draw myself a fear for my laptop. Here are the sketches I made in Procreate:

Here is the final design I went with, after inking it to be fully opaque:

I was already in the CBA lab with Ruipeng, so we decided to layout our vinyl cut designs in the Cricut software together so that we could reduce vinyl waste. Here was our layout:

After cutting and weeding, I used gridded transfer tape to transfer my design onto my laptop.

Ta-da!

Group assignment: kerf-finding with Ray and Sun

Full disclosure: thinking / Googling how to do this beforehand would have saved us a lot of time and effort. But, we had so much fun deriving our experiments from first principles i’m not upset at all! :)

For this assignment we used the XTOOL P2 cutter. Here is the result of all of our experiments, which I describe below:

Power-speed experiment

First, we wanted to know which power / speed to use. We arbitrarily chose a fixed speed (12 mm/s) and made a grid where we varied power from 1% - 100% in 10% increments. In retrospect, we should have made a matrix to test different speeds as well, but we were too locked in on taking the Shortest Path To Kerf Finding. For the XTOOL laser cutter and cardboard from Dan, 40% power at 12 mm/s was about right.

Kerf-finding round 1

First, we made a series of slots from 4.1-4.9mm thick, with 0.1mm increments between them. We measured our material to be 4.19mm thick (after cutting), and the 4.1mm slot fit the best. From this, we estimated our kerf to be around (4.19 - 4.1) / 2 = 0.045mm.

Kerf offset fun times

Next, we wanted to use the “kerf offset” settings on the XTOOL GUI (under the Cut menu) instead of manually setting the width of each rectangle. XTOOL has some confusing language on their interface - “inward” vs “outward” - and we couldn’t tell which direction was which so instead of looking the answer up like reasonable people, we ran another experiment. We wanted to also get better sensitivity on our kerf measurement, as a two-birds-one-stone kind of deal. We found that “outward” (positive kerf offset) makes slots looser, and “inward” (negative kerf offset) makes slots tighter We found that -.03 kerf offset was about right for our slot, for a 4.15mm thick square.

Experiments in joint clearance

Finally, to really test our kerf hypothesis on slot fitting, we deduced that we’d need to make our tabs slightly bigger (to account for the kerf) and our slots slightly smaller. So we made 4 squares, +.02, +.03, +.04, +.05, and 4 slots vertically and horizontally, -.02 to -.05. We found that all of them worked due to the pliability of the cardboard, with the +/-.05 having the least joint clearance (tight) and +/-.02 having the most (loose).