Space Station Solar Equipment Showing Damage

bhmit1 writes "The latest space walk has turned up some bad news for the problematic solar panels: metal shavings. From the article: "The rotary joint, 10 feet in diameter, has experienced intermittent vibrations and power spikes for nearly two months. Space station managers were hoping a thermal cover or bolt might be hanging up the mechanism. That would have been relatively easy to fix, so they were disheartened when Daniel Tani radioed down that metal shavings were everywhere. 'It's quite clear that it's metal-to-metal grating or something, and it's widespread,' Tani said.""

Comments

Re:Towing in space

[Migraineman]

Towing in space has been done before. Grumman sent North American Rockwell an invoice for towing their crippled spacecraft home. The rate per mile seems pretty reasonable too.

All joking aside, this is going to be a bear to fix. The best scenario would be that the drive gear was munching an insulation blanket. The debris would be friendly to space suits, and should only be labor intensive to clean out. If the gears are grinding on each other, the debris will be sharp and hard. That would be "bad" and I'd expect NASA to seriously consider returning the entire assembly to earth for repair. Expensive, but much less likely to kill someone.

I'm of the opinion that the drive system on this beast is probably over-engineered. It should resemble a Ford F-150 differential - loose tolerances, and designed to run for many millions of rotations without much maintenance. There's absolutely no need for the solar array to have precision pointing capability. I really do hope that the problem isn't due to over-engineering, but I wouldn't place a bet.

Re:Will a replacement fix it?

[mha]

So how would you propose these things are done? I mean, things that no one's ever done before, and which you can't really simulate?

An anecdote from my days of working for a huge German company (240000 employees) at Oracle (first job after university): I was part of the 32-64bit porting team. The question came up, are customers going to need additional or larger hard drives for the 64bit version of Oracle?

The answer from the Germans: Well, you've got the source code. Examine all structures in the code that end up on disk and count the bytes. (we know how many Bytes an "int" takes up on 32 vs. 64bit, etc.)

The answer from the Americans: Well, you've got the source code there. Just compile it and see what happens!

You know, while the German approach (I *am* German) sounds a lot more "scientific" and exact I would say the American way was not just better, but the only one that actually WORKS outside a simulated computer environment with a limited number of known-in-advance factors.

So again, how would YOU go about discovering the unknown? *I* would do just what NASA does, and what humans have done for millenia: Try, fail and try again, never approaching any ideal solution but something that works for now, until the next unforeseen thing happens.

Of course, in the western world everything that even LOOKS like risk has to be eliminated: from hot coffee to horses with tourists on them going any faster than a slow walk (I'll NEVER go on any tourist expedition on a horse in the US again, in Germany my friends who've never been on a horse before were forced to "survive" gallop several times in a 2 hour tour - and did so with relative ease).

Space travel, the trial-and-error way

[Moraelin]

Space travel, the trial-and-error way, eh? You mean, Wan Hu style? ;)

"Early in the sixteenth century, Wan decided to take advantage of China's advanced rocket and fireworks technology to launch himself into outer space. He supposedly had a chair built with forty-seven rockets attached. On the day of lift-off, Wan, splendidly attired, climbed into his rocket chair and forty seven servants lit the fuses and then hastily ran for cover. There was a huge explosion. When the smoke cleared, Wan and the chair were gone, and was said never to have been seen again."

Dunno about you, but I'd rather have it designed by the kind of people who'd rather sit down and calculate, instead of just doing the first dumb thing that comes to mind and see if it works.

Yes, _sometimes_, for a very narrow class of problems (like counting the bytes) the simplest way is to just do it and measure the number. But when you actually have to design something more complex, that quickly becomes a horrible idea. Even for something as simple as a watch, the probability to get it right by just throwing some parts together repeatedly and seeing if the result works, is close to nil. At some point you have to sit down and calculate the size of those cogs.

Plus, there is a lot of other stuff around you that happened only because someone sat down and calculated stuff, instead of good ol' dumb trial-and-error.

There's no way to invent a laser by trial and error, for example. The probability that just accidentally you'd have the right kind of material, and the right kind of coating at the ends, and the right light wavelength to excite it, and it's cut at the exact right length, is completely negligible. Humans have been cutting and setting rubies for millenia, and there are exactly _zero_ that started just emitting a laser beam by trial and error. It took someone calculating what happens there, before you could even know that a laser is possible, and how to make one.

In fact, pretty much the last major invention (that I know of) that was perfected by trial and error, was the light bulb. And, at least according to Tesla, that was a monumentally wasteful undertaking. To quote Tesla, "His method was inefficient in the extreme, for an immense ground had to be covered to get anything at all unless blind chance intervened and, at first, I was almost a sorry witness of his doings, knowing that just a little theory and calculation would have saved him 90 percent of the labor. But he had a veritable contempt for book learning and mathematical knowledge, trusting himself entirely to his inventor's instinct and practical American sense."

But even that stopped working for anything more complex than a light bulb. There's no way you could design a CPU nowadays by trial and error, for example. Even with specialized tools and massively simulating everything ahead, just one glitch sunk 3DFX. (Their Voodoo 5 was supposed to compete with the Geforce 1, but due to having to fix the malfunctioning chip design produced by their tools, it ended up competing with the Geforce 2 instead.) Now picture doing that layout by dumb trial and error instead. I wouldn't even try.

Heck, even in the job of counting bytes, sometimes the "American approach" (*) you describe would give the awfully wrong results until you've fully ported it. E.g., if the code was written with hard-coded constants for the saved data (which probably wouldn't be the case in Oracle's code, but I've seen it happen in other places), then compiling and running it would give the wrong results anyway. E.g., if someone saved an int by writing exactly 4 bytes to disk, it would still be 4 bytes for 64 bit code... and the very incorrect answer.

(*) As a side note, I hate thinking of those as "American approach" and "German approach", as it's really just the approach of whatever person gave that answer. I know there are plenty of Americans too who will stop and calculate, because otherwise the