I decided to take some time today and experiment with hairbrush designs. The problem is that my hair doesn't do well in this humidity, and I've been breaking the bristles on my brush. The campus printers were huge, fast and accurate so I'm used to just making a new one when I lose too many bristles, but my printer takes a day and a half. And the ones I make here break too often.
The main problem is the 0.02 error rate. It doesn't sound like much but it means that in a cubic millimeter of solid matrix there are 20 or more dots missing or out of place. Blue dots are 100 micron cubes studded with polymer fingers. In a warm ion solution under gentle UV illumination the fingers straighten out and the solid form dissolves into a slurry. The print head deposits them into a matrix for the desired form, and then cooling and flushing cause them to mesh together. The dots align with their neighbors, the fingers close on each other, and the result is a rigid solid.
The problem is that errors propagate through the cubic crystalline structure. A few dots out of place push their neighbors out of skew as well and make fault planes. You can see them as reflective surfaces inside the otherwise semi-translucent plastic with the familiar blueish tint. Since the solid is rigid, the flexible bristles of my favorite hairbrush design are made by putting a spring system of interlocking 3D zizzags at the bottom of each spine. Using the micro-PSI scanner I can see that the tiny springs are riddled with fracture planes. No wonder they snap off when they encounter even a mild knot.
I'm trying a helical screw design for the bristle spring. That was part of the shock system in the mounting brackets and seems to have done really well in the field so far. I examined one and it looked like the fracture planes were such that they never cut across the entire width of the screw. That may have to do with the orientation relative to the matrix, or it might be due to the relative largeness of the screw compared to the old spring design.
I used to dabble in mechanical hacking like this back in high school, but I gave it up once everyone who did it seriously was so much better than I was. I showed some other kid in the dorm a micro-solenoid I built and he rolled his eyes at me. It's been a fun distraction, but if it still fails (I'll know in about 28 hours) I'll just look up a competent design once the Roam comes online.
Comments