JF make 293
JF Nav
JF Nav
Creation Date: 2003-10-07
Greets, I wish I could tell you about all the stuff I'm doing, but I'm kinda torn between talking about good stuff and doing good stuff. You see, Robat taught me lessons, but I'm repeating those mistakes already even though I tell people that I will not. You see, instead of building an ornithopter, like I had originally set out to do, I spent time defending the theory that a flapping robot using muscle wires could fly. Ornithopter builders, electronics experts, programmers, and every person off the street said that Robat was impossible. None gave a concrete reason. Most of their reasoning was simply: muscle wires cannot provide enough strength. Back before I understood levers, I did the numbers: the wing was ~32g, the muscle wire could lift 1kg. Simple, huh? Well, the muscle wire would only contract ~0.1% which means that a long piece of wire would contract only a few mm. So to get the wing to move the required 0.2m, I would need to use a lever. I created it and it worked fine. But the lever would require ~20kg to lift it. No problem, just hook up 20 muscle wires in parallel. There wasn't really a problem with electricity with 20 long wires. So, the problem is solved. The next problem was stability. Using Robat as a glider showed the inherit instability of my design. It would nose dive for two reasons: The center of mass was in front of the center of lift and the wings had unstable drag during forward non-flapping movement. The center of mass could be brought back easily by changing the weight configuration. The wings having unstable drag could be solved by changing the shape of the wings. Those problems were solved. The thing that sunk the design of Robat was the flapping downwards instead of using normal aeroplane physics. You see, planes use a propellor to gain speed and the wings to generate lift. Birds use their wings to gain speed and to generate lift. Robat, on the other hand, was designed to use the wings to generate lift, but not gain speed. You see, I forgot to design speed into Robat. Forgive my stupidity, I was young and I learned my lesson. The idea was to have a hummingbird type system where it could hover just because it was flapping enough to cancel gravity (very hard). You see, it was fairly heavy. Not compared to planes which have very heavy engines. In total it was ~100g. Flapping wings at 3 Hz will not sustain that. However, it is my opinion that Robat could be redesigned to support forward motion and gain lift from wing shape instead of flapping. The flapping would advance forward motion and the forward motion would promote lift. This is a much more traditional model for ornithopters. Also, a much lighter, smaller, energy-efficient model could be made to achieve flapping flight with muscle wire. For example, the wings could be made shorter to decrease the energy needed to propulse them. The wing materials could be more light. The body could be made of thinner plastic. The electronics could be made very simple. A very efficient energy/weight battery could be used. One day I will open up XFig and design it, then I'll buy $100 of materials and build it. The day after, I will fly it and it will fly away. But here I am doing what caused the previous failure: telling you how it will work instead of making it work.

So now you all know about Robat. That's the reason that I haven't actively advertised Hack Mars or Javantea's Fate all over the place. I will later when there is something for consumption. Of course, someone could say that there is something for consumption here at JF. My friend says I should advertise JF all over the place. $20 will buy 200 1/6 page color flyers which I can post all over telephone poles here in Seattle. I can post on Craig's List (even though I feel it would be a sell-out) to get a megahit thing. So the MoJF of today is nothing less than a design for such a thing. It won't be the thing because it's just vector graphics. I'll make it 3d and then it'll be cool. I won't tell you what the picture means. It's supposed to attract some attention. If you want to print this out and advertise JF, I have the PDF version. It's nice because the pdf is so small; it's all vector graphics after all. Talking about vector graphics, GIMP 1.3.21 came out and I downloaded it. I installed it and loaded up an SVG. It worked. I was sidetracked, so I couldn't use it. I tried to load an SVG as a set of paths and it crashed. That's how beta software goes. But I have hope that the GIMP will become a premiere graphics editing system. Usually, I like to seperate raster and vector graphics, but if they can do a good implementation of vector graphics, I will gladly use it.

JF Nav
Home Characters Making Of Technical Mail News Links |< First < Prev Next > Latest >|  bandwidth version Goto Scene