I just saw a Twilight Zone episode where people were in suspended animation for 100 years. This sort of thing is common in science fiction (STNG: The Neutral Zone The Neutral Zone (episode) | Memory Alpha | Fandom for example). As an engineer, I’m always skeptical about these sort of stories, because I know who hard it is to get a device to run flawlessly for a year, let alone centuries.
So, if you had to build a machine that would work without human intervention for a thousand years, would you make it mechanical (like the Long Now clock), or electronic (like the cryogenic storage ship in STNG)?
Electric… fewest moving parts as possible. Many redundancies as possible faults to protect against surges and what not. Many power sources and backup battery.
I don’t see a battery lasting for 1000 years. Heck, we’ve all seen batteries rendered useless after not being used for the winter. I’d make it mechanical, and I’d make it big, with loose tolerances, so that part wear would be less of an issue. With an electrical system, oxidation of the conductors will be a huge concern (although I suppose we could put it in a vacuum. Aren’t there still mechanical traps in the Pyramids that are still “sprung”, so to speak?
Electric are mechanical. You just can’t see the moving parts.
:smack:
what was i thinking! i still say electric is the way to go.
Moving parts are bad. Anything you can do to reduce the number of moving parts is a good idea.
For electric power, a radiothermal generator would be the best bet. You could choose an isotope with a half-life in the hundreds or thousands of years, though it might not have the greatest power density.
It’s an impossible question to answer. Too many variables.
A lot of people think that, in general, mechanical systems are less reliable than electrical systems. While it may appear this way, there’s more to it than meets the eye.
While it is true mechanical systems tend to have a linear wear pattern, they also tend to be very robust, meaning that they continue to function despite lots of wear and tear. Many electronic systems, by contrast, will tend to fail if just one component fails. Furthermore, electronic failures tend to be abrupt and “show-stopping,” whereas mechanical failures can often be prognosticated.
Having said all of that, electrical components are almost always more reliable than their mechanical counterparts. I know of at least one exception: a small reed relay - which is an electromechanical device - is actually more reliable than a Hall Effect switch, which is a purely electrical device. I always thought that was interesting.
Then your best bet is a very large thermos bottle.
You need to shop the item not the store. For this project you should choose parts for reliability not type or cost. (I’d pick a bank of relays and leaf switches over an 8255 PIA or a 6235 RIOT any day.)
If all humans died in one day Hoover Dam would still be working for several afterward lighting up the Vegas strip. Until muscles clogged up the cooling line inlets at least. Then it would automatically shut down. Not bad for a complicated electro-mechical system of that size.
Couldn’t you reduce the steroid dosage?
Not that this means anything, but my greater fear is that my car start suffering electronic problems rather than mechanical problems. It’s an engine; what can go wrong with it? It’s a transmission (a robust one); what’ll happen? Instead it’ll be a fuel injector, mass airflow sensor, gear shifting solenoid, door keypad, something like that which stops the car dead first, versus anything strictly mechanical. Sure, all the ball joints and tie rod ends that I’ve replaced are mechanical, but (1) I drive on Michigan roads, and (2) they don’t stop the car dead.
How many thousand year old mechanical systems have we made to date? Not even counting maintenance periods, I don’t recall any off hand.
A look at the return shelf at the local electronics store will set you straight on the overall reliability of electronic systems.
The fact is that millennial operation lifetimes haven’t been a priority in human activity, since the Pyramids, and all they had to do was just sit there. We don’t even have many actively human maintained and operated systems that old.
Tris
Even if there are, then their success isn’t down to being mechanical, but because they haven’t had to move.
Fair enough, but electrical systems don’t have to move, per se, and I still think it’d be a real challenge to design one that could last for 100 years, let alone 1000.
Some data:
Here’s some mechanical systems that have lasted on the order of a thousand years. I imagine they’ve recieved maintenance, though: The Water Wheel Album: Page Five
Think the bearings could do with a little grease. Wouldn’t want to sleep next to the thing! http://www.metacafe.com/watch/579105/ancient_water_wheel/
The oldest working clock in the world was striking the hour from 1386 till around 1884. That’s doing pretty well as far as I’m concerned. Salisbury Cathedral clock - Wikipedia
If you’re talking about systems including integrated circuits, I agree. The current densities in those little conductive tracks are phenomenal! Then there’s thermal cycling, and even cosmic ray damage to worry about, among other things.
On the other hand, there’s food for thought here - a sealed, batery-powered electric bell that’s been ringing for 160 years! Oxford Electric Bell - Wikipedia
If you’re talking about good old Victorian-era overengineering, with thick copper wire insulated with wax and bitumen, then a really long lifetime might be feasible. But such a system isn’t going to be fully electrical - there will be relays, contacts opening and closing that may arc and/or accumulate dirt, springs flexing. Sealed reed switches and mercury relays will do rather better.
Really, I think there’s insufficient data to answer the question, simply because we’re talking about an untried design philosophy. Thermodynamics is going to bite you whatever you do; - corrosion, erosion, fatigue, slow reactions, recrystallisation, UV degradation… Any device that doesn’t repair itelf WILL stop working eventually. And yet we know of all sorts of materials that degrade extremely slowly - noble metals, fired ceramics, glasses… Amber is an example of a natural polymer that has an impressive lifetime if it isn’t exposed to UV!
I’d like to think that a Swiss watch made from hard platinum-iridium alloy, with jewel movements and powered by air-pressure fluctuations like an aneroid barometer, would run for a century without a problem. Maybe even a millenium. But who knows?
I had a similar thread (strictly mechanical, no electric vs. mechanical debate) a few weeks ago. Best answer was the noria wheels, but I’m sure those received regular maintenance.
Cool article about the Oxford Electric Bell, matt. Odd nobody kept any notes on what the dry pile was made from.
What if the whole thing is submersed in a non conductive liquid?
What do you want the machine to do?