Saturday, October 6, 2012

Printed Fins, Futures Base

Here's a project tangent: 3D printing fins.

When I first got this buggy plastic extruder a year and a half ago, I printed a couple tiny fins before the original motor began to fail. Just recently I pulled the contraption back out and fixed it up enough to print out some replacement/upgrade parts to make it much more reliable (the white gears in the picture). So now I'm back in the 3D printing game, with this extruder temporarily bolted onto the CNC router.

The quality with this setup is just good enough to start making mechanical parts for surfboard-related schemes I'm working on - but also good enough to play around with printing small fins.

I've got about 4.25" of vertical travel with this setup. So a fin with a half-inch deep Futures box base can only be about 3.75" tall. That's what you see here:



Anybody see the big problem with this fin? No - not the little crack at the trailing edge. Not the slightly textured surface. There's something wrong with the base. The base is almost perfect. But ...



it's backwards. That's right, I designed the base with the notches switched around. But it does fit nicely in the box, and only took a little touch up sanding work. So, yeah, fail, but success anyway.

I've only got this one early board with Futures boxes, and the box mount job itself is a bit funny. I mounted too shallow and sanded too much, so the boxes aren't as deep as they should be. That's why the base sticks up a bit here. The grub screw pins it down though, and the fit is tight, so it's secure.


Now, about 3D printing: the technology to print a surfboard has been here for a while. It's probably been done, though I haven't seen it. Maybe you have? Maybe secret prototypes are out there incubating, and we'll see the printed surfboard start-ups emerge soon.

What I'd like to see is some cheap, tall, RepRap-derived 3D printers built in garages, apartments, dorm rooms, able to print a 6' board overnight from maybe $50 worth of biodegradable - maybe even self-recyclable - plastic. It looks like that's possible today. Maybe a first step would be to build a printer with a 12"x4"x6" envelope, and print a board in sections.

Add that to my tangents list. The main focus now is on building a laser cutter, so I can develop more detailed cardboard core designs, and designing and printing parts for a servo-driven coelostat, for the SolarCNC project.



Tuesday, January 10, 2012

Nat 64 - Reckless Play With Natural Fibers


This is an ugly, junky experimental board. If this is your first encounter with my work, then great! Check this out. But also look at the other boards, which aren't quite this crude.

These are tests with natural fiber cloths.

On the bottom is a single layer of cotton muslin. It's easy to work with. The weave is just tight enough that you can pour a small amount of epoxy directly on the dry cloth and quickly spread it, without it dripping through. It takes a lot of little pour-spread, pour-spread movements, but it's much faster than the brush technique I'm using with fiberglass.


That's the finished lamination. Good results.

We know cotton doesn't have the strength or impact resistance of fiberglass. I easily dinged the rail throwing the board in the van the first time. Had it in the water about 20 minutes before noticing. So this board has already gone through a dry-out and repair cycle.



Got a little help here on the quick patch job.

Though weak, muslin might have a role as a first lam material, since the wet-out technique is faster, and the finer weave will reduce pinholes. A second lam with something stronger might give back enough impact resistance and stiffness to balance it out.


On the deck we've got a base of fiberglass, and then full coverage with a layer of silk, with burlap (jute) rail patches (from green coffee bags).

Silk traps bubbles underneath when you try to apply it as a second lam. Don't know if there's a way to work them out through the fine weave. Silk seems too limp and droopy to lay directly over the core for a first lam. And silk isn't good at conforming to the compound curves around rails - it wants to fold instead of warp, and won't stretch. Not pleased with the silk. Too fine and tight.

But burlap is great - raw and loose. It's a great way to build up a thick section. The thick fibers and loose weave can work as traction if you don't fill the weave all the way with hot coats. You can try to saturate it minimally, or just keep pouring resin into it, depending on what you're after.

It's brown. It's crude. It's functional. I'll be using it again.




Yeah, this is a single fin fish. Haven't heard any advocates for this kind of thing, probably for good reason. But you've got to try it to know.


As long as we're doing things wrong, let's really do it wrong. A 7.5 inch epoxy-sawdust holey baby hatchet. (The first board I've made with some kind of accommodation for a leash - seven, in fact.)


Instead of fiberglass strands (fin rope) along the base, tried some jute twine. Too crude, twisted, stiff and lumpy. At this point I'm done playing and just want to get this in the water. The divot at the front of the base got filled and faired with wax. Let's get it wet.



Still nice with light behind. See where the two missing pieces go?

Fish 64 Build


Recent Fish 64 build.
3 a.m. and I want to get these photos done, so no suburban paradise yard shots this time.


Don't think that left fin is really that crooked - but maybe it is. No - just checked again - both fins are pretty much straight down and straight ahead - and meatier than anything you'll find at a surfboard supermarket.


Cosmetic flaw, in unflattering light, and then the same section backlit. There's a point when the epoxy is fluid enough to saturate the cloth, but not enough to fully saturate each fiber of the cloth. There's plenty of warning beforehand - you notice that it's getting thicker and taking longer to sink in. Don't push it - just dump it and start a new batch. Otherwise you'll get one of these embarrassing white stripes. Seems to be structurally sound - it's just not the look were going for. It's on the deck. Conceal with wax.


First vent install I've done. Looks okay on the deck. To hold the vent plug flush with the deck as I glassed it in, I supported it on the bottom with a little paper and tape tube, which will peek out the bottom like this forever.

Hard sharp rails on the tail, for the first time. Learned how to do that from Make Magazine #19. More old plywood fins - NACA0008 foils carved with the router. Figure it's 50 year old redwood alternating with some other pale pine.


Backlit.

Monday, January 9, 2012

Rib Squish for Simmler



Maybe you noticed there's a mini Simmons design on the surfcorr album (Simmler), but no kit available. No? Didn't notice? Didn't even know about surfcorr? Go check it out, then (and read the liner notes). Simmler is the current R&D subject. Cutting now.

An hour into cutting, noticed the notches were coming out in two different sizes. Usually they are all the same. Checked the collection of parts cut so far and they all had the same problem. What now?

Ahh, but that's not a problem - that's rib squish. Prepared these files months ago, and forgot that Simmler has a core pattern variation that requires different notch sizes.

Rib squish: Compare the quarter isogrid pattern above to most of the other board core top views, and you'll see that the diagonal pieces have less slope to them. The pattern is kind of squished in the direction of the stringers. Normally, all parts intersect at the same angle. But with rib squish, you get two different intersection angles, so some notches have to grow and others shrink.

Why rib squish for Simmler? I forget why. Think it was just time to try it. It looks like I also spaced out the stringers wider on Simmler, so it has two fewer stringers than other short boards. Squishing the ribs keeps the hexagons from getting too big in this case, and makes up for some of the lengthwise strength lost by having fewer stringers.

If you look at the art for Glass Fish on surfcorr, you'll notice the center board core has negative rib squish - or rib stretch. The point there was to get the core pattern to fit better in the tail section.

Retractable Pen



Retractable pen mechanism for the router, for numbering parts. It uses skate bearings on a central hollow tube, the pen fixed in one end of the tube, and a plastic zip tie fixed to the other end, connected to a wheel on a servo. In the down position, the pen is free to ride over irregularities on the surface, since there's some springiness in the zip tie that pushes the pen tube down.

It's got 2 inches of vertical travel. When the pen is down and writing, the router bit is about 0.5 inches above the material surface. When the pen is up, and the router is cutting, the pen is about an inch above the surface.

Auto numbering with this mechanism was supposed to save time, but it ends up taking longer than doing it by hand. It's worth it for the neat numbers, and the ability to make any other kind of mark on the surface and cut during the same run. This makes router production a bit more like laser production.