Faster Cars and Smellier Air

Now that Discovery is happily back safe and sound, it’s hard not to be dissapointed about another ginormous chunk of foam detaching from its external fuel tank after everyone worked so hard to prevent it.

There were some successes here. Notably, the repair process worked well. The centerpiece of this success was Canadarm, the newly-extended remote manipulator arm contributed by our friends to the north.

I agree with astronaut Scott J. Horowitz and others that the design of the shuttle is simply too complicated. If you amortize the total cost of the program, each launch burns about a billion dollars, and human resources account for much of it.

There is really no effective way to test the shuttle. Other manned rocket systems are launched dozens of times without people in them before they are considered safe. Only the pieces of the shuttle can be tested. As Richard Feynman pointed out in the Challenger investigation chapter of his excellent book, when you assemble 100 components that each are 99% reliable, you should expect the resulting machine to be less than 36.6% (0.99 ^ 100) reliable. The only ways around this problem are to use fewer components, or to rigorously test the finished assembly.

When considered against the initial goal of routine access to space, one has to conclude that the space shuttle program has not been a success. In the near future, a return to more conventional rocket technology is appropriate. However, we can’t drop the ball developing the promising scramjet technology. If we ever have spaceships like the kind they have in Firefly, they’re going to use scramjets. We should take all the money we save by scrapping the next generation space shuttle and invest heavily in non-military scramjet research. And, of course, continue the highly successful unmanned exploration programs to Mars and the outer solar system.

My brother and sister-in-law were nice enough to indulge my train fixation with a LEGO “My Own” locomotive for my birthday.

This is a pretty good LEGO set.

It was fun to put together. It features two distinct, non-trivial configurations.

It includes stickers.

And best if all, it’s operable. There are swivelling wheel trucks with spinning wheels. It’s a little disappointing that no reciprocating pistons drive those wheels. But unlike LEGO’s rigid frame 1903 Wright Flyer, this simplification doesn’t seriously compromise the historical accuracy of the model (like the particle beam cannon that I constructed out of the extra pieces does, to give another example.)

One complaint: the box the track came in clearly depicts my happy, LEGO-sized train-conducting counterpart. However! Opening the box did not reveal him. Danish toy-making friends: please include a conductor in future track boxes, or failing that a disclaimer starburst.

All aboard!