≡ Menu

Laser Technologies for Starflight

How do we go about using photons to accelerate a spacecraft to a substantial percentage of the speed of light? Ten percent of c ought to do — it gets us to Alpha Centauri with a mission time of 43 years and would allow us to at least send flyby probes into that system, with the promise of larger decelerating probes as new technologies became available. After yesterday’s post on Young Bae’s photonic thrusters, I went back and read his 2012 paper “Prospective of Photon Propulsion for Interstellar Flight” (reference at the end). 10 percent of lightspeed is indeed the minimum he chooses. It works out to 30,000 kilometers per second.

Back in the 1950s, the German scientist Eugen Sänger talked about a ‘photon rocket’ as a way of reaching the stars, but his design called for annihilating electrons and positrons and using the gamma rays thus produced for thrust. The idea stumbled over the problem of controlling this kind of exhaust stream, for gamma rays penetrate anything we use to contain them. But the insight that photons themselves could be used to impart momentum to a sail was already in the air when Sänger was pursuing these notions, and it would not be long before beamed laser propulsion came to the fore as the subject of numerous papers on interstellar flight.

Beaming Energy to a Sail


Robert Forward’s insights are repeated by Bae in his 2012 paper, which notes that separating the photon generators and the spacecraft pays numerous dividends. Propulsion systems that do not use rockets allow us to keep the heavy parts of the vehicle, especially the energy source, here in our own Solar System, while sending the payload out on its mission. As became clear yesterday, however, beamed laser propulsion also carries its share of problems, including the fact that we need to beam the laser photons over large distances. If you could forgive the need for Kardashev Type II engineering (able to use the entire energy output of its star), you could live with Forward’s huge Fresnel lenses, needed to keep the laser beam collimated, for laser beaming is based on reliably known physics. But we’re not exactly at Type II yet.

Image: Robert Forward, whose work on beamed laser propulsion produced designs that pushed engineering to extreme limits. Credit: UAH Library Robert L. Forward Collection.

Even so, assume a laser beaming capability using abundant solar power from a station near the Sun. A beam is sent out into the Solar System, to be focused by a huge lens presumably built by some kind of nanotech means out of local materials. Forward wanted to coax an interstellar mission out of these ingredients, and various permutations followed, culminating in 1984 in a concept for an interstellar round trip with a human crew. The destination would be Epsilon Eridani, 10.5 light years from Earth, to be reached using a lightsail with a diameter of 1000 kilometers, with a Fresnel transmission lens of the same diameter somewhere past the orbit of Saturn.

The numbers are staggering in every direction. With a total vehicle mass of 80,000 tons, including 3000 tons for the crew and its equipment, the lightsail would reach an acceleration of 0.5 lightspeed in less than two years. It would be accelerated by a 75,000 TW laser system. More on this mind-bending energy problem and its consequences in just a moment.

But first, how do we decelerate when our spacecraft reaches the target system? Without another laser installation already in place, Forward came up with a ‘staged’ sail concept, in which the lightsail is divided into nested circular segments. Approaching Epsilon Eridani (about 0.4 light years out), the rendezvous part of the sail with attached payload including crew detaches from the main body of the sail, allowing laser light to be reflected from what is now a 1000 km ‘ring sail’ back to the rendezvous sail. It’s this reflected light that decelerates the smaller sail and brings it into the Epsilon Eridani system. The sail will be divided again after the crew has finished its explorations. Now laser light beamed from Earth strikes the remaining 320 kilometer ring sail and accelerates the small inner sail and its crew module for the return trip. A final burst of laser light decelerates the spacecraft back into our Solar System.


Image: A four stage roundtrip interstellar flight using Robert Forward’s ‘staged’ sail to decelerate. Credit: Young Bae.

All of this is purely conceptual, but it has led to numerous papers on the methods and materials of beamed laser propulsion, even as we’ve widened the scope of the research to include microwave beaming. Not long ago, Joe Ritter (University of Hawaii at Maui) explained to the 100 Year Starship Symposium his own work on ultra-large, highly adaptive space mirrors, which could have interesting sail implications. Perhaps some of these studies, along with advances in laser capabilities, will make beamed sails commonplace in the Solar System, a period of experimentation that could eventually lead to the first true interstellar missions.

Hybrid Missions with the Photonic Laser Thruster

Young Bae believes that beamed laser propulsion when combined with his own photonic laser thruster (PLT) can reduce the power requirements of Forward-style beaming by a factor of ten to one hundred, while using a short wavelength laser can reduce it yet again. Thus an infrastructure that would have become available no sooner than 2500 (Bae bases this on world power production and its yearly growth, recently analyzed by Tau Zero’s Marc Millis) becomes an actuality by the year 2100. His proposal is to perform initial flights of exploration but to follow them up with the construction of the ‘photonic railway’ we looked at yesterday. From the paper:

…the Photonic Railway… aims [at] enabling routine interstellar commutes via Spacetrains. The Photonic Railway, if successful, would radically depart from the conventional spaceship concepts, in which a single spacecraft carries both an engine and a large quantity of fuel. Rather, the Photonic Railway would have permanent reusable space structures that propel Spacetrains, which would consist of mainly crew habitats, and navigation and crew safety equipment. The technological foundation of the Photonic Railway lies on a strategic combination of BLP by Forward and PLT, which is named here PLT-BLP. It is predicted that the development of PLT-BLP can be further expedited by incorporating the anticipated development in x-ray lasers and advanced material science and technologies, and the interstellar PLT-BLP is projected to be within reach in a century.


Image: Young Bae’s ‘photonic railway.’ It consists of four PLTs: two for acceleration and two for deceleration. The spacetrain will have small thrusters for attitude control and most of the onboard spacecraft resources will be dedicated to crew comfort and safety. Credit: Young Bae.

The photonic railway is, then, infrastructure, a permanent transportation system that would link nearby stars. Rather than a single laser beam from Earth, the beam is generated along the way. The structural parts for the lensing and laser systems must be delivered from Earth, but Bae envisions assembly and activation by self-directing robotic systems at the destination world. His paper illustrates a photonic railway using four photonic laser thrusters, two for acceleration and two for deceleration. The vehicle, or ‘spacetrain,’ would have thrusters for attitude control, but most of its resources would be dedicated to crew safety. The paper continues:

One important factor is that the Photonic Railway PLT needs to operate [a] much shorter distance than the distance between the earth and the exoplanet. Typically, depending on the Spacetrain acceleration condition (the optimal case would be 1 g acceleration for maximum crew comfort), the system operation distance would be at least a factor of 3.2 shorter than the flight distance. Because of this, the Photonic Railway optical system can be at least a factor of 10 smaller in size than the PLT-BLP optical system.

The paper goes on to sketch out this permanent transport structure, one based on the idea of photon propulsion that would act much as the transcontinental railway systems did at creating investment interest and economic opportunities. Building the photonic railway would involve a four part process beginning with the use of photonic thrusters in satellite maneuvering and leading to a lunar photonic railway, an interplanetary version, and ultimately the interstellar objective. Bae believes it is critical that each step provides return investment, and cites factors such as NEO mitigation, lunar mining and space solar power as early drivers toward this end.

Ultimately, Bae takes beamed laser concepts that have been extensively studied and combines them with photonic laser thruster technology, adding expected developments in x-ray laser and advanced material science to reduce the power and engineering requirements demanded by the missions Forward looked at. The goal is a photonic railway that is permanent and efficient. The citation is Bae, “Prospective of Photon Propulsion for Interstellar Flight,” Physics Procedia (2012), pp. 253-279. The classic Robert Forward paper is “Roundtrip Interstellar Travel Using Laser-Pushed Lightsails,” Journal of Spacecraft and Rockets 21 (1984), pp. 187–195.


Comments on this entry are closed.

  • Alex Tolley October 22, 2013, 15:30

    I think that the role for light in interstellar travel is not for transporting mass (at least for most of the transport). It makes a lot of sense to transport the initial probe or industrial kit to the target, but after that, it should be for transmitting information.

    Consider the probe as a factory, perhaps an advanced 3D printer. Initially small, it acquires local resources to build a transmitter and all the ancillary technology it needs. Information is then returned to earth.

    All the latest designs and programming (even AI) can be constantly beamed to the local probe to manufacture robots to study the star system and its planets, and beam the information home. If during this period, we develop AGI, our intelligent proxies can go too.

    Whatever our desires for human star flight, it seems to me that machines (intelligent or otherwise) will be the dominant form in the galaxy because their replication at each target star can be very rapid, and if they do develop human level intelligence, individuals can transport themselves quickly and cheaply between stars with mind uploading.

    If or when humans do travel to the stars, we can hope that robots have built the infrastructure we need to live, so that colonization doesn’t have to look like small, “primitive” communities trying to start afresh.

  • CoolStuff October 22, 2013, 16:52

    Cool stuff. But, if the probe is doing 0.1c when it gets to Alpha Centauri that means it will spend less then 20 hours in the inner solar system (orbit of Jupiter). That is going to be a hard sell with how much something like that will cost. Unless the system was built for something else first.

    I wonder if the probe could pass close to both Proxima and Alpha Centauri?

  • Astronist October 22, 2013, 20:39

    I really don’t understand this concept at all. At first I thought Bae was talking about repeatedly bouncing photons between a mirror close to the laser and a mirror on the spacecraft, but this clearly won’t work over interstellar distances unless the mirrors are improbably perfect, so that can’t be it.

    Stephen A.
    Oxford, UK

  • David Cummings October 22, 2013, 21:41

    Photonic Railway. Plastic Phantastic Lovely Idea.

  • Eniac October 22, 2013, 22:34

    Alex Tolley:

    Consider the probe as a factory, perhaps an advanced 3D printer. Initially small, it acquires local resources to build a transmitter and all the ancillary technology it needs. Information is then returned to earth.

    I fully agree that that is the way to go. Slight problem, though: Most estimates of the minimum size of this “industrial kit”, or technology seed, are anything but small. Try to start thinking it through: You need to gather rock, pulverize it, separate it, smelt it, refine several different materials from it, form them into parts, etc. etc. The sheer number of tools needed for all that will make for a hefty collection of equipment, and many of those tools will not miniaturize well. Furnaces, for example. 3d printing will help, but only with a small subset of processes. Maybe nanotechnology using radically different processes can ultimately make it small, but that could be a long way off, if ever. Anything we can imagine building now would be in the hundreds of tons, if not much more.

    Still, this is something we can get to work on right now, and we definitely should. There are non-star-travel uses, as well….

  • Dispatcher October 22, 2013, 22:56

    I understand the concept of using a light sail to accelerate a craft. However, I’m confused as to how such a sail, split for deceleration, can decelerate the craft at the destination, since the laser light is first hitting the reflector. How is it that the light hitting the reflector AND then the deceleration sail does not net a neutral effect on the craft speed? Photonic pressure is being applied to both the reflector and the final sail at the same time (in opposing directions).

    Is this the result of the final sail being made out of a non reflective material? Thanks in advance.

  • CharlesJQuarra October 22, 2013, 23:47

    @AlexTolley: That sounds good in theory, but there are drawbacks once you look at the issue more closely

    If you have a chance, look at Dr. Messerschmitt talk in SC2013 or just read his paper about interstellar communication http://arxiv.org/abs/1305.4684

    To summarize the highlights; communication bandwidth goes inversely proportional with power. If you want to transmit, say, an compressed file with your brain connectome (so you can be uploaded at remote destination), you are going to either consume a lot of power or wait a lot of time. The exact figures I don’t know, so I’m not arguing that it wouldn’t be competitive (a rough calculation would have to be made), but that the gains are not as straightforward as it might seem. Given that matter can hold a huge amount of information in a gram of mass, it seems that above a certain amount of information/bandwidth, it is preferable to send mass

  • Brett Bellmore October 23, 2013, 6:38

    Strikes me as akin to demanding that Lewis and Clarke build a transcontinental railroad to do their explorations. This is a technology for rapid transit between two built up star systems, not for exploration.

    I’ve got my doubts there will EVER be a need for rapid transit between built up star systems, but it’s certain we’ll be in exploration mode for a long time before this occurs.

  • Paul Gilster October 23, 2013, 9:30

    Dispatcher writes:

    How is it that the light hitting the reflector AND then the deceleration sail does not net a neutral effect on the craft speed? Photonic pressure is being applied to both the reflector and the final sail at the same time (in opposing directions).

    Forward’s idea here is that the outer ring-sail reflects laser light back to the now smaller payload sail after the two have separated. What he says in the 1984 paper is this: “The central payload section of the sail is detached from the larger stage and turned around so that its reflecting surface faces the reflecting surface of the ring-shaped portion… The light reflected from the ring sail is focused onto the smaller sail, now some distance behind.”

    In other words, he’s assuming the inner sail is turned so that it can take advantage of the light reflecting off the inner ring sail.

  • Alex Tolley October 23, 2013, 9:58

    @Eniac – I agree that the probe would need to be quite large, certainly not a Voyager class probe with a MakerBot on board. Having said that, 100 tons is vastly less than would be needed for a crewed sail ship.

    My thinking is that we should be aiming to reduce the size of the probe as far as possible, to minimize energy to propel it. Ultimately, what I would like to see (and this now has me thinking along Dyson’s lines) is technology that mimics life. The probe becomes a seed, that can grow and develop. A hybrid approach where a probe seeds the local environment to produce the necessary materials through biosystems, then the bots start printing the parts to self assemble from these “living” materials and so on, bootstrapping each stage. This technology may be very far out, or not. I suspect the key is adapting technology to avoid trying to bulk refine the necessary resources. For example, I read that NASA is thinking about using 3D printers to use lunar dust to build habitats, rather than transporting bulk materials to the lunar surface.

  • Alex Tolley October 23, 2013, 10:09

    I agree with the bandwidth issue. It goes with the “don’t underestimate the bandwidth of a van loaded with music tapes speeding down the highway”.

    If we could encode information at the atomic level, we could transport minds in grains of diamond and propel them cheaply with the beamed sails. Slower, but perhaps more feasible than beaming the information directly. Again, I am assuming that a robotic mind will be easier to encode than a human one, although we could still argue about copies and originals, that bedevil discussions of human mind uploading. We could encode even the most advanced computer systems we have today in very little space with atomic level encoding, should that be possible. Perhaps, rather than using a 3D object, we should encode a 2D layer of the sail material instead, for easier I/O.

  • Eniac October 23, 2013, 22:47


    A hybrid approach where a probe seeds the local environment to produce the necessary materials through biosystems, then the bots start printing the parts to self assemble from these “living” materials and so on, bootstrapping each stage.

    Living organisms could be used to seed only places that are suitable for them. It is unlikely that we can modify our Earthly lifeforms to thrive on an alien planet, much less in the vacuum of space. Even if we could, would it ever result in much more than a bunch of weeds? A communication system? I don’t think so.

    It seems more likely that we will learn to build regular machines which can thrive on rocks and light. Much like plants, but able to operate in vacuum and fully programmable to make anything we want them to. For example, communication devices, habitats, and spacecraft. Their biomass would be primarily steel, glass, aluminum, and silicon, all of which can be won from space rocks, by well understood methods of smelting and refining. The energy from this would come from the sun, via photovoltaics and/or polished aluminum mirrors for process heat. Adding carbon based organisms (or even materials) to such a setup will be more trouble than it is worth, in my opinion.

    I am with you on the need to minimize the mass of such an industrial seed. There is much opportunity for progress, along three dimensions:

    1) miniaturize the components (toy-sized)
    2) reduce the number of different components (Lego-like)
    3) use highly generic processes (one process to produce many different components, like 3d printing, robotic manipulators, machining, and casting)

  • Alex Tolley October 24, 2013, 16:25

    If we try to extract resources by fairly brute force means, then I agree, you don’t need life as part of the process. On earth, we had utilized life to be one of our main mechanisms to extract and produce resources, until the industrial revolution replaced this approach largely with non-biological processes.

    While I don’t expect organisms to produce iron ingots, I do expect that they can concentrate metals that can be further processed, as metal powders for 3D printing via sintering. Alternatively, have them produce materials than can replace the iron in bulk applications.

    Life in vacuum. While Dyson has speculated about this, I tend to agree that life, even synthetic life, will need aqueous conditions. But this isn’t hard to do. For example, an icy comet could have solar heated hot water pumped under the surface, forming a bubble of water water that organisms could use to process the cometary material. No exposure to vacuum as the bubble will create a self sealing ice container. This should even work on worlds like Mars, by tapping the subsurface frozen aquifers/glaciers.

    Once the organisms make the desired products you need, these are then fed to the printers, analogously to cotton or silk thread being fed to looms. From there you can form structures that can be used in more industrial strength processes. The key is to pick production processes that fit with this model. Obviously smelting steel in a furnace won’t work, but perhaps focusing sunlight on a NiFe asteroid might get you molten metals that can be tapped in some way. Hand waving here :)

    What I want to emphasize is that it might be possible to bootstrap large scale engineering starting from a very small scale, using some sort of biology and a small fab.

  • Joseph Paul October 25, 2013, 12:47

    Are there any ideas on going faster than .1c or is that pretty much it for lightsails/magsails etc? I fi ti s not what is the most that could be squeezed out of that technology?

  • Paul Gilster October 25, 2013, 15:46

    Joseph, some of Forward’s concepts ratcheted things up a bit. I believe Starwisp, which was microwave-beamed, was thought capable of 20 percent of lightspeed, while his Epsilon Eridani expedition with return capability assumed close to 50 percent of lightspeed — it also assumed that huge Fresnel lens in the outer system. You get into major issues, of course, with protecting the craft from matter in the interstellar medium as you push this high.

  • Andrew Palfreyman October 25, 2013, 18:43

    I am still stuck on the direct sunlight vs. laser issue. Jim Benford replied to me thus:
    “Coherent radiation can be focused to a spot size that varies inversely with the size of the radiating antenna or optic. Incoherent radiation can be focused only to the angular size of the radiating source as seen at the focusing lens.
    For example, the sun can be focused by an optic to the apparent diameter of the sun in our sky. But no further. Therefore a large lens close to the sun will produce a poorly focused beam because the sun will take up a large segment of the sky. That’s why there’s only so far you can go with focusing the sun for solar energy.”

    I think a parallel beam is what’s needed. Furthermore, beaming should be done with mirrors and not lenses. Why then go to the expense of building high power lasers when there’s 10^26 W of raw sun power available? With that kind of a resource, even gross inefficiencies are tolerable.

  • Brett Bellmore October 25, 2013, 21:32

    The principle is the same, whether you use mirrors or lenses; Your light is coming from a distribution of angles, it leaves with a distribution of angles. No getting away from that if you don’t have coherency.

    In this case, the gross inefficiencies are enough to make even 10e26 watts of power inadequate over interstellar distances.

  • Andrew Palfreyman October 27, 2013, 0:51

    Um, no. A parabolic mirror does what you claim is impossible. And I also don’t think you’ve done the maths (just a suspicion) to support your statement about inefficiencies.

  • Mark Wakely October 27, 2013, 13:15

    There’s one perfectly awful way for an alien, interstellar self-replicating probe to operate without depending strictly on an on-board foundry. If it sought out worlds with metals already processed from raw materials and scooped them up in a kind of interstellar “smash and grab,” it wouldn’t have to resort to pulverizing rocks and all that entails. It could just recycle the metals it steals from other civilizations for its own use. As the probe approached a planet like Earth, it could zero in on those objects that contained the refined metals the probe needs for replication and grab them. Easiest would be the things not anchored to the ground, such as boats and airplanes. Even better would be our satellites and “space junk,” since they’re already in orbit. And our own distant probes such as the Voyagers, Pioneers, etc., could not only provide raw materials a self replicating alien probe might need, but help point the way to where plenty more refined metals can be found, back to Earth.

    Anyway, just a thought.

  • Eniac October 28, 2013, 0:12

    Andrew Palfreyman:

    I think a parallel beam is what’s needed.

    This is exactly correct. And the sun will not provide that, no mater what mirrors or lenses you use. A parabolic mirror will make a parallel beam ONLY from a point source. The sun is not a point source. Lasers create FAR better parallel beams than any that could be made with sunlight and mirrors. Many orders of magnitude. It’s the proverbial “laser focus”.

  • Ron S October 28, 2013, 10:16

    Parabolic mirrors cannot be perfect. Keep that in mind.

  • Andrew Palfreyman November 4, 2013, 3:36

    @Eniac: That helped a lot. So it comes down to a practical matter of mirror accuracy. The next decision is sun->photovoltaics->laser or sun->solar-pumped laser. Surely the latter is best? Can the former be justified on a cost basis, or indeed on any other basis?