Is it time to give interstellar travel a shot?

[Camacha]

A small fission reactor will suffice to keep it powered, I think.

Any numbers? I would be interested to see how a Earth bound reactor would differ from one for insterstellar use

You can use the engine bell of a Daedalus-type engine as radio dish, or you can (and tbh, should) use lasers for communication, if just for the bandwidth.

I am not sure why you would resort to optical communication when radio waves are equally fast and a little more robust. Bandwidth might be an argument, but I do not think that has been necessary on the shorter range (within the solar system). I guess sticking to proven technology where possible might be prudent.

No, I don't think so. I think it establishes that we can with comparative ease build things that will overcome the stresses they are expected to be subject to, and often exeed them.

Yes, on a totally different timescale. We are used to building things for 2, 10 or even 20 years. Stuff just does not get built for two or five centuries and certainly not for that timespan in space, one of the most hostile environments. When anything survives that long it is down to luck. It is a different ball game alltogether and one we know little or nothing about.

We've been over this a couple of times now, so I will leave it at that - save new arguments or information.

[SargeRho]

Radio waves become too dispersed somewhere around 2 lightyears or so out from anything but very large radio telescopes like Arecibo. Lasers don't suffer from that nearly as much because they don't disperse as quickly...I think, anyway.

[Stargate525]

We are used to building things for 2, 10 or even 20 years. Stuff just does not get built for two or five centuries and certainly not for that timespan in space, one of the most hostile environments. When anything survives that long it is down to luck. It is a different ball game alltogether and one we know little or nothing about.

We've been over this a couple of times now, so I will leave it at that - save new arguments or information.

Yes, but is that because we CAN'T, or because we don't see a need to?

[Camacha]

Radio waves become too dispersed somewhere around 2 lightyears or so out from anything but very large radio telescopes like Arecibo. Lasers don't suffer from that nearly as much because they don't disperse as quickly...I think, anyway.

That sounds like a fair argument I am curious whether other problems might not also arise, such as minute bits of dust accumulating and blocking the signal, or that acquiring a target might become nigh on impossible.

What about focusing the radio beam tightly?

Yes, but is that because we CAN'T, or because we don't see a need to?

At the moment that is the same thing. There is no need so we don't, but we can't because we won't.

[SargeRho]

A parabolic dish does basically just that, focusing the radio waves into a beam. I don't know if you can actually turn radio waves into a laser, or if their long wavelengths would interfere with that. Space is pretty cold, and exept for planets and stars, infrared light isn't too common. An infrared laser will probably do the job nicely. Having a 360° detector for the laser pulses will be enough to aquire a target with the laser. That far out you won't need to aim very carefully anyway. The beam will be wider than Earth.

[rkman]

Interstellar travel is not just a matter of deciding that it is time to do it. First we need to improve technology.

Like i said in that other thread: it is like setting out to discover America, before the invention of sails.

The difference being that back then America was behind the horizon, so we did not even know it was there. But now the horizon is billions of light years away, so we can see many places that we can't get to.

[Camacha]

A parabolic dish does basically just that, focusing the radio waves into a beam.

That was what I was referring to. Are you sure the two lightyear radio fadeout is about direct communication and not the radio signals leaking into space from Earth?

Having a 360° detector for the laser pulses will be enough to aquire a target with the laser. That far out you won't need to aim very carefully anyway. The beam will be wider than Earth.

Would that not negate the advantages compared to radio signals? Also, knowing where to point your dish is actually a bit relevant when finding a target with current technology and procedures.

[SargeRho]

It wouldn't negate the advantages. While the beam fans out, it's still a lot easier to pick up and laser transmissions have a much higher bandwidth. HD images from the edge of the solar system anyone?

I don't know if it's about leaking radio signals, or transmissions from direct comms as well. A data laser is also lighter and smaller than a large radio dish and transmitter I think.

[AngelLestat]

Interstellar travel is not just a matter of deciding that it is time to do it. First we need to improve technology.

Like i said in that other thread: it is like setting out to discover America, before the invention of sails.

But that is the key, we already have the "Sail" again

[Camacha]

But that is the key, we already have the "Sail" again

Well, we don't quite have it yet I see you give some estimations. Do you have, by chance, any calculations that go along with them?

[SargeRho]

Solar Sails have actually been tested in space. And I remember reading about another solar sail probe being launched in the next few years.

[DonLorenzo]

The issue at the core of the 'yes we can build long term stuff' and 'we can't' argument is that indeed, it is conceivable that current tech could last hundreds of years. What's uncertain is if we can make it so that that is actually likely. The lightbulb example is a great one. It clearly demonstrates that a lightbulb can last over a century, however it's only one out of countless lightbulbs that actually have. If you absolutely need to light a space for a century or longer, lightbulbs do not become the preferred option just because they CAN succeed. It has to be likely.

The problem is ascertaining that likelihood. The timescales are just too long. Another nice example is the shelf life of military grade canned food (made in the 60s for bomb shelters), we don't know because it's not gone bad yet. It might be forever, it might be 60 years.

This is actually a very similar problem to the 'well new ships will overtake it anyway'-scenario. No one in their right mind would advocate a trial-run of a century before committing to a build, with what then would be outdated (but proven to be reliable) tech. You can mitigate a bit with multiple redundancy, but intuition tells me that as timespans increase the adding of redundancy will start to have diminishing returns (as they all age together).

All that said, considering that our day-to-day stuff routinely lasts a decade, our space stuff regularly outperforms mission durations and 'oops, component X failed'-failures are relatively rare I do think that with careful engineering and a project firmly grounded in a longevity mentality we would be capable of constructing something with decent odds of making it through a century.

As an unrelated side note; space, while harsh is a lot more predictable than our terrestrial environmend, which should help in engineering longevity.

[Camacha]

All that said, considering that our day-to-day stuff routinely lasts a decade, our space stuff regularly outperforms mission durations and 'oops, component X failed'-failures are relatively rare I do think that with careful engineering and a project firmly grounded in a longevity mentality we would be capable of constructing something with decent odds of making it through a century.

I think a century might be feasible, although reliable operation will be hard to achieve. The problem is that we probably will need a lot more time and, as you indicated correctly, it is most likely a game of diminishing returns. I can not reasonably imagine a craft that will have a likely operational lifespan of 200 to 300 or even 500 years. With current tech, that is.

Solar Sails have actually been tested in space. And I remember reading about another solar sail probe being launched in the next few years.

Tested in space, much like Gustave Weisskopf came up with powered flight. What we are proposing here is akin to a modern airliner. Solar sails are still very experimental and we barely ever used it as main or significant propulsion on a mission that was not dedicated to testing it.

[Rakaydos]

Radio frequency lasers exist- they're called Masers. (microwaves are a frequency of radio that happens to generate a harmonic in water)

[Red Iron Crown]

Radio frequency lasers exist- they're called Masers. (microwaves are a frequency of radio that happens to generate a harmonic in water)

Microwave is not radio, though they're adjacent parts of the EM spectrum. Radio waves penetrate Earth's atmosphere, microwaves don't.

[Seret]

100ish years trip time to alpha centauri. It's not going to be too big of a problem building electronics that last that long actually.

As a reliability engineer I'd say that the electronics would be the weak point when trying to design for that kind of life. The problem is that performance and efficiency pull electronics toward greater miniaturisation, but that works against long-term reliability.

I'm not saying we couldn't build a system conservatively enough to last that long, but it wouldn't be as easy as you seem to be suggesting.

intuition tells me that as timespans increase the adding of redundancy will start to have diminishing returns (as they all age together).

Your intuition is correct. In very general terms components follow a trend called the Bathtub Curve. At the wear-out phase of their life things fail a lot, and they do in real life all fail together in batches. The fact that redundant components might not be working hard throughout is no guarantee that they'll last longer either. Some types of failure mode are related to how hard something works, some are just about how old it is.

Edited March 12, 2014 by Seret

[Camacha]

As a reliability engineer I'd say that the electronics would be the weak point when trying to design for that kind of life. [...] I'm not saying we couldn't build a system conservatively enough to last that long, but it wouldn't be as easy as you seem to be suggesting.

Great, an expert I was planning to ask a couple of friends who are electronics engineers to comment on this matter, but this might be even better. Could you give us some insight into the difficulties to be expected? And for what kind of timespan are specialist longlife Earthbound systems built, for what application? I can imagine that systems that are hard or impossible to maintain (sealed or dangerous environments et cetera) are designed for longevity.

[AngelLestat]

Well, we don't quite have it yet I see you give some estimations. Do you have, by chance, any calculations that go along with them?

I follow very closely all graphene advances since the time their was discovered, Is one of the branches of sceince which is receiving the biggest investments.

We already have products that use graphene, now their are improving the quality and size of the samples made to study. It would be like plastics, in 50 years we would see it in everywhere.

About the quarter wave mirror, it was already made in laboratory. You can search it by "perfect mirror". Another recent discovery said that graphene can absorb 100% of light with only 1 layer if some properties are added.

Graphene can resist until 5000k of temperature, even without reflection, it can be very close to the sun and not sublime.

Something so light can not fall faster towards sun. For that reason it would need to be sent by something heavy, and then detach in the periapsys at 0,08 Au.

The main problem is all the other things.

I guess there is not real issue with time degradation in electronics, it all depends on the design.

But that is the key. It has to be a new design of electronics and everythings.

You need to imprint the electronics in the sail (the side opposite to the mirror), if you add layers over layers of electronics, you get the risk of being damage by interstellar particles or not resist the high G forces in the acceleration phase. Then we have all issues with comunications and AI software (to be able to manuver by their own, also to try restart comunications in case we lost it).

If we were in times of war, and this it will be a weapon. I guess it can be ready in 10 years. Like this is not the case. My estimation goes to 50 years.

But it will be very usefull launch this kind of probe, also to measure how much particles there is in the interstellar medium. Any flaw, will teach us how to not mistake in the next tries.

If we dont have that information, we are just delaying even more these kind of missions.

Edited March 12, 2014 by AngelLestat

[Nuke]

That just confirms my point. One of our most reliable probes is a mere 36 years old. With current or even realistic future technology we need to span a lot more years than that. It is nice that we have a 100 year old lightbulb, but building a highly complex spacecraft that can reliably do the same is another matter completely. Yet we need it to be more reliable than that.

Unfortunately this works both ways. The larger your transistors, the less advanced/quick they will need to be and the more power they will consume. Not great in an interstellar void without many external energy sources. Sending a probe out on a journey for hundreds of years means going very big when it comes to transistors and die size. And let us not forget that transistors have only been around since 1947. Any evidence on 60+ years operation is at best anecdotal and beyond that we simply do not know. What about the other couple of hundred years? We know literally nothing about the wear, degradation and other problems during that time.

We are not even talking about the (un)reliability of storage and the problem of standards changing back on earth over such a period.

Do you have any numbers for that estimate?

its not far fetched that an interstellar ship might have its own small scale fab on board. especially if it was a colony ship, you would want as much manufacturing capability you could muster to bootstrap industry at your destination. you could produce new chips and storage devices while in transit.

that said you could go all butlerian jihad and just not use computers. people are plenty capable of running numbers in their head. you can use analog control systems, mechanical computers, etc. we might develop better optronics too, which should work well in radiation rich environments.

[Camacha]

its not far fetched that an interstellar ship might have its own small scale fab on board. especially if it was a colony ship, you would want as much manufacturing capability you could muster to bootstrap industry at your destination. you could produce new chips and storage devices while in transit.

True, and that is one way of doing it I do see working - taking production capacity and workers along. I think generation ships are at the moment the most viable long range option anyway. The whole discussion on reliability was aimed at unmanned probes that cannot be serviced after launch.

[Seret]

Great, an expert I was planning to ask a couple of friends who are electronics engineers to comment on this matter, but this might be even better. Could you give us some insight into the difficulties to be expected? And for what kind of timespan are specialist longlife Earthbound systems built, for what application? I can imagine that systems that are hard or impossible to maintain (sealed or dangerous environments et cetera) are designed for longevity.

Uh-oh, you know what they say about experts...

First of all, a couple of caveats: I'm not an electronic engineer, although I'm more gripped up on electronics/MEMS than most of my mechanical brethren. So you might want to consult your electron-herding friends anyway. Second, ensuring reliability over the kind of timescales we're discussing here is not something I've ever had to do professionally. From a bit of a scrounge around the libraries I suspect that's because nobody else has either. Papers discussing extreme lifespans don't seem to be jumping out at me.

What I can tell you is that commercially available electronics (even those intended for aerospace or military use that we deem "high reliability") are not up to scratch. Part of the problem is you've got priorities pulling in opposite directions: you're going to need a spacecraft with a high level of autonomy, which necessitates substantial processing power, but the way you achieve increased processing power and efficiency in integrated circuits is by shrinking them. As structures get smaller and smaller you get all sorts of problems, including "Weird Quantum ****". For example, sputtering a thin film onto a substrate invariable puts atoms into the wrong places in the wrong energy states. That causes intrinsic stresses in the film, couple that with current flow causing thermal stresses and you get unavoidable mechanical processes within the film. Eventually it will fail. Just like with corrosion, we're talking about devices that only work if their atomic components are in a state other than what nature would prefer them to be in. Eventually mother nature will reorganise your work back into a chemical or electrical state that she's happy with, and your gizmo stops working.

The way we get around this in practice is to not make things any more complex than they need to be. Despite 22nm chips being commercially available a lot of what you'll find in even brand new industrial kit is built at the 130nm node. It's well-understood and reliable. It still wouldn't last 100 years in a hostile environment like space, though.

Professionally I'd be really interested in how the QA and design bods would tackle the problem of testing kit for a 100 year service lifespan with no chance of repair. AFAIK no one has ever done it with electronics. We have really good predictive theories for mechanical reliability that work over those kind of lifespans which we use on things like bridges, and even aircraft (some DC-3s are pretty old, and the USAF is planning on flying their B-52s until they're nearly 100) so I'm confident we could design a spacecraft that would handle that life structurally. But designing electronics for that kind of life isn't something that anybody seems to have ever done, so it would be a whole new field. We would have to wind the scientists up and get them to do some fairly fundamental research to map out the way forward.

[78stonewobble]

This thread took an interesting turn.

The only thing I know of being engineered for the really long haul is this stuff:

http://longnow.org/

A clock designed to work for 10.000 years and a record of human languages on some sort of microfilm disc.

Edited March 13, 2014 by 78stonewobble

[Nuke]

rest assured the computers we use in space long term will have more in common with a computer from the 80s than it would with your gaming rig.

i know there are things like the elevator hoists and the water pumps at the empire state building are all original equipment, installed in the 30s, and still work so good that the though of replacing them with something modern is outrageous. then there are various steel bridges and buildings that have been around a long time, even ancient structures like the pyramids and roman aqueducts teach us something about long term engineering. concrete is known to last a long time if you dont reinforce it with steel. steel structures also last a long time. we know if we build a structure more robust than it needs to be, it will last a long time. and we know stone structures are pretty much immortal. now granted we dont think of these things as space faring tech, but they are important.

you can also get around the problems of radiation if your ship is really big, by burying critical system deep within the bowels of the ship. when i say big i mean asteroid/ice dwarf scale, using much of the object's mass as propellant. charon would make a good candidate for such a ship. with it you could take a city sized population and enough industry to fix anything along the way utilizing the natural resources of the object. the other side of it is that its going to be a very long trip, so the idea has merit even if its an emergency evacuate the solar system type mission.

[Seret]

A clock designed to work for 10.000 years

Note that they've gone mechanical for this. Like I said, we've got really good predictive models for things like fatigue and corrosion.

and a record of human languages on some sort of microfilm disc.

Interesting they've gone for something readable at 1000x magnification. That's actually pretty big.

[78stonewobble]

Note that they've gone mechanical for this. Like I said, we've got really good predictive models for things like fatigue and corrosion.

Interesting they've gone for something readable at 1000x magnification. That's actually pretty big.

Well obviously I won't be around to see if it worked.

Yeah it is... I guess the assumption is that it will be easier to "reinvent" a microscope than it will be to "reinvent" a computer around whatever dataformat they could have picked.

Just look at nasa's problems finding an actual machine to read tapes from the 60's.