On Saturday my mechanically-inclined associates hosted a soiree in commemoration of Halloween.

Jon (center) dressed as Michael J. Nelson and featured theater seating, Tom Servo (not pictured), and Croooooow T. Robot (left).

I assumed the countenance of John D. Zoidberg, M.D.

Photos courtesy Connie.

Lessons:

• Measuring out the proper shape to cut not only saves material but makes what you have a lot easier to manage.
• Compress if you can, I couldn't here.
• You can go for the heavily laminated look, or put very little resin on so you can see the features of the weave. In this case I chose the latter, you can see the detail pretty well in the full size image.
• The mold release was clouding the resin on previous tests. No more.

Next up... Refinish the windshield? Front fender? Fairings? I just don't know when to stop.


Oh, and here's Zeus.

What do you do when it's unhealthy to go outside and network television is an incessant stream of yellow journalism? Fun in the garage of course.

I picked up some carbon fiber weave and epoxy resin to see if compression could make up for my shameful lack of an autoclave. Additionally, I tried curing the resin at 500F.



The resin runoff cured in the oven within minutes and turned a deep amber; neither of these things is desirable. I removed the bake assembly from the oven and let it cure at room temperature, surprisingly it turned out to be visually identical to the full room temperature cure (seen below on left and right). Nonetheless, my procedure will hence be room temperature curing.



Straight out of the press, the composite surface is a bit hazy on account of the mold release. With a bit of wet sanding and a layer of clear coat it looks pretty good. I have a two layer carbon-carbon that is rather flexible. I also created a carbon-glass-glass-glass-carbon-carbon that is quite rigid.

A chain guard is currently drying in the garage, we'll see how it turns out.

Regarding the fires, I was out checking on a friend's house and decided to skirt the perimeter of my area to see if I'd need to worry. The one blaze I saw was pretty small, particularly in comparison to three years before about a mile west.



Apparently the most significant damage was done farther north in Rancho Santa Fe and east in Ramona.

So we took the atc2k out for the second first time. Work must be done on how it's affixed, but otherwise it worked well. The Fletcher's break was pretty brutal, enough so that Erik's recent acquisition lost most of its skin on a wave.




On the plus side, there was a board sale at the shore that day and he picked up a mighty fine replacement. E was undecided until he learned that it was a benefit for San Diego's youth, he immediately laid down some cash money.

In other news San Diego is burning again, I'm back over .500 in fantasy, and I hate the following equation: transmission + final drive + differential = transaxle.

ATC2K came. I strapped it to the Duc with sunlight fading. Obviously I could have done a better job of securing it but here's an idea of what it looks like... (ed. blogger-hosted videos, now gone)

This will prove highly entertaining for motoring events, surfing, and paintball. I'm rather impressed with the value. The camera feels sturdy, waterproof, and comes with quite a few mechanisms for securing it to your head, helmet, vehicle, etc. And there are lots of fun things you can do with one or more mini cams.

(May need to click to play)

I'm simulating a collective of independent agents. Each is controlled by a set of neural networks that determine the agent's action. The networks train themselves based on positive and negative action outcomes. Currently they eat, swarm, and reproduce. Above you can see them moving as a collective to consume food (green). Their population (plotted below) fluctuates pretty steadily and you can see the group expand and contract based on environmental stimuli. It's pretty processor-intensive, so I've had to set the environment to support only a few hundred organisms.



The algorithms are still pretty basic, development goals include:

• More interesting environment features.
• Physical traits beyond the decision networks and lifespan.
• A genetic algorithm for generating new agents.
• Faster pseudo-random and pseud0-gaussian number generators, with minimal flops.
• Heritage records to determine which founding organisms were most successful.
• Multiple antagonizing colonies.
• UI to allow realtime environment manipulation.Leave me your thoughts on algorithms and features.

Windy day, bright sun in the wrong spot. But if you have access to a nice lens, why not?
Next: Columbus Day off.