Stabilizing Design for Laser Pushed Light Sails

$100 million has been put into the Breakthrough Starshot project to push a sail up to 20% the speed of light and send an unmanned probe to Alpha Centauri in about 20 years. There is progress on a stabilizing design for a sail to ride the laser beam.

An optical beam rider making use of a light sail comprising two opposing diffraction gratings is experimentally demonstrated for the first time. We verify that the illuminated space-variant grating structure provides an optical restoring force, exhibiting stable oscillations when the bigrating is displaced from equilibrium. We further demonstrate parametric cooling by illuminating the sail with synchronized light pulses. This experiment enhances the technical feasibility of a laser-driven light sail based on diffractive radiation pressure.

Physical Review Letters – Experimental Verification of a Bigrating Beam Rider

29 thoughts on “Stabilizing Design for Laser Pushed Light Sails”

  1. Sailbeam concept right there; probably the highest thrust to power ratio we can get esp if you factor in using some sort of a deuterium micro sail pellets…

    Lasers are hugely wasteful if easy to model for gram or under sized missions. Particle beam or sailbeam would be the natural choice for any larger mission, esp on a non insane energy budget.

    “The primary reason for switching from lasers and lightsails to particle beams and Magsails is roughly six orders of magnitude reduction in the power required during initial acceleration, and a like reduction in spacecraft cost and complexity. Speci.c bene.ts found with this approach are: (1) improved electrical e/ciency of particle beam generators relative to lasers (50% vs. 25%), (2) two to three orders of magnitude increased force on the sail for the same beam power, and (3) elimination of a separate deceleration system since the acceleration Magsail can serve dual purpose.”

    Interstellar propulsion opportunities using near-term technologies – Dana G. Andrews

  2. Why ?.

    Any species as incredibly advanced as being capable to traverse interstellar distances would have mastered fusion, antimatter, etc. … they will laugh at the thought of stealing Earth’s measly resources.

    In fact they would have already discovered us centuries ago. Do you think that we are the only ones in the Universe with SETI-like programs ?.

  3. It would be a 20-year journey

    20 years * 20% speed of light = 20 * 20 / 100 = 4 light-years (distance Earth/Alpha Centauri)

  4. That self-reproducing tech is what I expect space to give us. Not because you need to be in space to do it, but because you need it to be in space in a big way.

    The push to automate on Earth is a thousand times stronger in space, because even with SpaceX, it’s going to be expensive to ship something to Mars. And living in space requires a much higher ratio of infrastructure to people than life on Earth needs.

    I’m a tooling engineer, designing production systems is my job. And it is my opinion that we are quite a bit closer to cracking this nut than the average person realizes. Not the compact version, but the big clunking replicator, that’s just a bunch of automated factories building each other.

    We just need a sustained effort focused on doing it. And expanding into space will more or less automatically give us that.

  5. To be fair, the tech for synthesizing novel enzymes and DNA is getting (t)here (and maybe some of the other things too). We can already synthesize polypeptides and somewhat large DNA constructs (thousands of bp? tens of thousands? don’t know the latest numbers).

    The scale still needs to improve, but there’s been rapid progress in that in recent years. What we’re really missing is the ability to design new enzymes and predict their function. That takes computing power, among other things.

    But anyway, that’s less important in the big scheme of things. What would really open up the way to K1 and beyond is compact, self contained, self-replicating manufacturing tech. Most likely and preferably Atomically Precise Manufacturing, but a macroscopic version could work too. That may be developed this century. That then allows the sort of huge-scale construction that K1+ requires.

    Access to space merely provides the materials (and energy).

  6. “But that requires long term dedication,” … not to mention the 4.1 year delay from α-Cen in detecting a need for a slight deceleration beam azimuth change ‘cuz the probe needed to move a bit, to avoid the target civilization’s plethora of nearly undetectable (from here) bits of asteroid parts. 

    ⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
    ⋅-=≡ GoatGuy ✓ ≡=-⋅

  7. Thanks.
    I think I would have gone with “science fantasy”, because wow.
    But I am talking more about cosmic ‘rays’ than instellar hydrogen as far as damaging the sail material. Sorry for confusing the issue. Not so much a wind- and not really interstellar (exclusively at least).

  8. Yah.  
    I guess so. 

    The idea that within 100 years, we’d be “well on our way to a K–2” civilization (my paraphrasing) depends almost entirely on whether the progress-of-technology curve continues along as it has for the LAST 100 years. 

    In 1920, we had airplanes.  
    And capable trains.
    And radio, X-rays. 
    Nuclear power was being sussed out.
    Medicine hadn’t yet discovered DNA, or the genetic code. 
    But some antibiotics discovered.

    We had no-frills cars
    We had skyscrapers with elevators.

    What’d we get in 100 years, to now?

    TV, electronics, hand-held computing, nuclear power, Internet, world-wide communication, medical tech. The world’s middle class can FLY anywhere, affordably. Semiconductors revolutionized lighting, computing, control systems, power conversion. Robotics -or China- makes our stuff.

    Yet, meat still grows on cows, chickens, pigs, fish, sheep and similar critters. 
    And, vegetables in dirt. 

    We can’t synthesize, a novel enzyme, or the DNA in a cell to make it. 
    Nor do we have flying cars. 
    Nor solid-state refrigerators. 
    Nor any real (BIG) longevity-enhancing tech.

    … so 

    I’d say ‘100 years of low-hanging fruit’.
    And the next 100 years of juicy fruits are MUCH higher up.
    So… same rate?
    Because of space?

    I don’t think so.

    ⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
    ⋅-=≡ GoatGuy ✓ ≡=-⋅

  9. Yeah, that’s always been the hard part.

    Forward solved it by using the larger part of the sail as a retro-reflector to allow the center portion to decelerate using the beam from the Solar system. But that requires long term dedication, it would be more satisfying to use a self-contained system.

  10. Heliospheric stellar wind ion density braking (magnetic) is just about the only hope of slowing down somewhat after the journey. Its velocity (V) is perhaps 1000 km/s. 

    When one’s coming in at 20% of 300,000 km/s or 60,000 km/s … well 60,000 + 1000 OR – 1000 … is barely any difference from really fast, at all. 

    To substantially slow down, the probe would also have to displace a sizeable fraction of its own mass in ions. It isn’t likely that the stellar wind is actually thick enough, even with a near-whole-system pass thru. 

    Our mean density is 4 atoms/cm³ or 4×10⁶ per m³. Or, if they’re all hydrogen and helium, 7×10⁻²³ kg/m³. If the magnetic deflection sheath is 1,000 m in radius, with a 100 AU thru-wind path, well … that’s only ⅓ of ONE kilogram of stellar-wind intercepted.  Total.

    See what I mean?
    Nowhere near enough. 
    Just isn’t.  

    Worrying about the sail’s film integrity is kind of ‘false worry’ since the whole question is SO deeply into the Science Fiction realm.

    ⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
    ⋅-=≡ GoatGuy ✓ ≡=-⋅

  11. I think we could plausibly be a K2 civilization within 100 years. The amount of material necessary to build a statite array capturing all the Sun’s power is quite feasible to extract from the asteroid belt, and maybe some mining on Mercury.

    Assume that we build the array at 0.75 of Mercury’s orbit. 4.3 10^10 meters.

    Place a reflective bubble around the Sun, and the amount of light inside the bubble will simply keep rising until radiation to the exterior equals the Sun’s output. So using reasonable material choices we can count on the statites being supported by 10 time current solar luminosity while still tapping the full flux to power some sort of heat engine. That gives us a mass budget of 7.8g/m2, and a total mass of 1.8 10^20 kg.

    A fifth the mass of Ceres! Totally doable.

    Mind, we’d probably want a good idea what being enclosed in that bubble would do to the Sun’s dynamics. And I’m not really sure how long it would take the Sun’s surface to reach equilibrium with the light within the bubble. There’s a lot of mass there to warm up.

    But assuming we make the transition to a space faring species, and don’t screw up, we could be a K-2 civilization faster than most people realize.

  12. ⊕1 of course. SOMEONE reads my math! Yippee!!!

    The point of the 150 kg/m³ part was to adopt the “well, we can make quite perforated-on-the-whole metamaterials that’ll reflect exactly one narrow range of beam wavelength efficiently, and allow anything else to pass right thru unimpeded” idea.  

    I’m just not sure whether 0.1 µm thickness … kind of gets into a different stretch-physics set of questions. But no matter… stretch physics is OK for speculation.

    The 100+ petawatts (i.e. 100×10¹⁵ W) is definitely Kardashev I level power. The earth receives 55×10¹⁵ W of sunlight, more-or-less across its illuminated face. So… deploying ALL of that sunlight as laser light, times 2.  

    Yah, unobtainium. 

    But that doesn’t stop speculative physics people. What’s a few Kardashev’s between friends?

    Again, thanks
    ⋅-=≡ GoatGuy ✓ ≡=-⋅

  13. 113 PW is pretty close to Kardashev I level – all to accelerate a measly 100 ton payload to 20% of c.

    15e-6 kg/m^2 / 0.1e-6 m = 150 kg/m^3. Pretty low. About the same as balsa wood, so not unobtanium, but it needs strength too. Less than 1/10th of graphene density. Maybe some sort of perforated graphene? Or make maybe it thinner. Its specific strength should allow it.

    pi * (80000 km)^2 * 15e-6 kg/m^2 = ~300 tons, so the sail is heavier than the payload. Not too surprising.

    (Coincidentally, I also got a fuel/payload mass ratio of ~3 when trying to optimize energy requirements for a fusion rocket. But I think my energy equation was wrong. Mixed reference frames.)

  14. Integrating (by formula!) and also by itsy-bits (Newton’s!), yielded exactly the same results. Two different techniques to cross-check results.

    And the findings are STUNNING in their consequence.

    I already said “the power needed” is gargantuan for real-world masses. Hundreds of kilograms to hundreds of tons.  

    But something quite interesting showed up, too. One is that the lightness of the film is crucial for “ultimate velocity”.  Not the payload so much! Because if the payload is a fraction of the total mass, well, thinner-lighter-stronger films make for larger sails. And larger sails can take more total power. And be accelerated at reasonable beam divergence coupling further. Further, bigger, more power, and longer. Cool!

    But to get to 20% of ‘c’ takes a breathtaking amount of power for a 100 metric ton payload (at 25% of total system mass).  

    0.1 µm film
    0.000015 kg/m² sp. mass
    113,000,000 GW beam power (!!!)
    1.0 G acceleration for 2 months
    Acceleration out to 37 AU (!!!)
    160 km sail diameter (!!!)

    Obviously stretching (terrible pun) limits of credibility. But… 20% of ‘c’. 

    ⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
    ⋅-=≡ GoatGuy ✓ ≡=-⋅

  15. Well… that’s kind of the point. 7.5×10⁻⁸ IS a small value! Yet, LASERs are not exactly like most-other electromagnetic wave beams. Or, maybe they are in a sense!

    Coherence is the key … all the waves are time-and-phase aligned in a laser beam, especially from those generated in either long crystal or extra-long path gas laser tubes. The real coherence can rather easily better 10⁻⁵ radian, at present.  

    However, when contemplating turning 100 to 200 gigawatts of electrical ‘excitation’ energy into a coherent laser beam, I’m kind of stunned by the level of ‘coherence’ that the generating plant itself would have to have, mechanically.  

    After all, tho’ it might sound big, 10⁻⁷ radian … at 400,000 km (earth-moon distance) is … let’s see … 400,000,000 m × 10⁻⁷ = 40 cm or about a 1⅓ feet in diameter. Sitting on Earth, pointing at Moon. 

    So yah… that is a pretty small divergence!

    As BRETT BELLMORE pointed out elsewhere, one could make hundreds of thousands of smaller ‘throw-away’ sails, and accelerate them to fairly high velocity ((0.1% of c) over the at-target impact velocity of the ship), to impact … and accelerate the thing, continuously.  

    I like it.

    ⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
    ⋅-=≡ GoatGuy ✓ ≡=-⋅

  16. What practical divergence for a laser have been achieved? 0.75*10-7 radians seems like an awfully small number…

  17. You couple the sail to the payload by accelerating a lot of small sails, and steering them to impact a pusher plate on the payload. IOW, the sails are largish particles in a mass beam propulsion system.

    I find being transmitted as data and reassembled at the destination philosophically unsatisfying, but suspect the universe will be colonized by people who don’t share my unease with it, given how much faster it’s likely to be.

  18. philosophically unsatisfying” — hmmm… I guess so. 

    Here’s the challenge, Brett. Payload-to-sail mass ratio. 

    For a signification-fraction-of-c … photonic thrusting is irreducibly energy-inefficient (except for the first 5% of the trip) and while inefficient, the sail becomes huge to comport REAL living thinking payloads … plus meals, oxygen, life support, science payload, out-of-photonic-reach thrusters, spaceship, engines, energy sources, recycling doo-dah, impact mitigation and repair, robotics, … the list goes on and on.  

    Let’s say that a 5 man, 20 full-AI robot, 100 year, hibernation-mission is conceived to be desired. 

    1,000 kg? 1 ton. (payload)

    Seems WAY TOO SMALL for shielding from radiation and micro-impacts to make it.  

    In the 10 LY radius around Sol, there are 8 known stars.  
    In 15 LY, nearly 53.  
    In 20 LY, there are around 100 to 125. 

    However, at even 10% of ‘c’, 10 LY is 100 years transit, 15 LY takes 150 and 20 LY is 200.  
    Some hibernation!!!

    I’d rather bet that the total ‘payload’ mass would be closer to 100 tons. With “all the stuff” mentioned. 

    If getting to 10% ‘c’ requires a sail that without payload could achieve 20% c (i.e. equal mass…), then a 100 ton sail at 0.0015 kg/m² is 67,000,000 m² or 9.2 km in diameter

    Thats one big sail. 

    And it begs … how to couple its thrust to the payload?

    At least with the 1e-7 rad divergence laser, it could be accelerated out to 2.7 AU!

    ⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
    ⋅-=≡ GoatGuy ✓ ≡=-⋅

  19. A couple of points/questions. Not an authority on the subject, but-
    1- How durable will this sail material need to be to survive a 20 year exposure to interstellar “wind”. It needs to retain enough integrity to decellerate, assuming that is the objective.
    2- If the target is Alpha Centauri- is that in the direction of heliopause? Will the probe be driving against the interstellar medium?
    0.2c is very, very fast in such a harsh environment.

  20. The project is funded by The Church of Jesus Christ of Latter-day Saints. I heard theres an embeded message along the lines of “do you have a moment to talk about jesus christ”

    cannot confirm if true.

  21. Why are people who don’t want it done so arrogant to think they could dictate what somebody else does?

    Kind of silly, anyway, given our broadcasts. Any alien intelligence advanced enough to interact with us is likely advanced enough to detect us even without an affirmative attempt to contact them.

  22. Agreed, it would be pointless if the only goal was imaging, as we could get as good of images from here, quicker, by very large aperture optics, or launching a mission to the Sun’s gravitational focal point.

    With sufficiently developed molecular nanotechnology, a sail like this could potentially deliver a seed for self-reproducing infrastructure, and so you could build up beamers for slowing down relativistic spacecraft, enabling a manned mission. Or just send a load of data, and duplicate the people you were sending. Which is philosophically unsatisfying as a mode of interstellar colonization, but otherwise works.

  23. What’s the point of this? To take better pictures of nearby stars? Wouldn’t it be faster, easier and cheaper just to build larger telescopes? Either on Earth or orbital space? Optical long baseline interferometry seems promising.

  24. The link is behind a paywall. Is it a 20-year journey, or set to begin in 20 years? In either case, I doubt most of us will be alive by the time it whizzes by Alpha Centauri and sends back a handful of blurry pictures – which will take a second 20 years, I guess. Can’t wait…no, really, I can’t wait.

  25. I guess my questions would be, Who gave anyone permission to try to contact other intelligent life with such a mission? Our greatest minds including Hawkings, Musk, Gates etc. all say its a bad idea to try contacting other intelligent life forms. Why are these people so arrogant they can just brush those great thinkers off, and do as they please.?

  26. And while noodling, at 1 µm wavelength, to achieve 10⁻⁷ radian divergence

    diverg (radians) = 1.22 λ (m) / ( η ⋅ diameter (m) )

    where λ is wavelength ( 10⁻⁶ m = 1 µm )
    where η is degree-of-goodness. Like 75% of perfect or 0.75.

    So, invert it algebraically

    diameter (m) = 1.22 λ / η ⋅ divergence
    diameter = 1.22 × 10⁻⁶ / (0.75 × 10⁻⁷)
    diameter = 16 meters

    Cool! That’s a nice number to know. Given that the receiver is 250 m diameter, one might try a larger synthetic aperture, in order to get a tighter (smaller) divergence, allowing “until it falls off 95%” illumination for a greater distance. In fact, it is linear. 

    If a 16 meter infrared laser of 41 GW output can make it out to 11,000,000 km, then a 32 meter IR laser would make it out to 22,000,000 km. And so on. Straight linear extrapolation. 

    Lovely stuff.

    ⋅-=≡ GoatGuy ✓ ≡=-⋅

  27. I guess my questions would be, “how much transmitter power”, and “how big a sail”.  

    Gossamer fabrics — we have tough polymers that can be metalized and patterned.  
    Given some pretty reasonable illumination/absorption/reflectance characteristics,

    reflectance = 85%;
    absorption = 0.1%;
    areal sp. mass = 1.5 g/m²

    laser wavelength = 1.0 Um; (infrared)
    laser divergence = 10⁻⁷ radian; (0.02 arcsec)
    max film heating = 350 °K

    With a blackbody emission (who knows), using 

    P-spmax = σ T⁴  … σ = Stefan-Boltzman constant 5.67×10⁻⁸
    P-spmax = 5.67×10⁻⁸ × 350⁴
    P-spmax = 850 W/m² absorbed

    P-illum = 850 ÷ 0.001
    P-illum = 850 kW/m²


    F = P-illum × 2 × reflectance / c
    F = 850×10³ W × 2 × 0.85 ÷ 299,792,458 m/s
    F = 0.004825 N/m² at full illumination.


    F = ma, and 
    a = F/m
    a = 0.004825 N ÷ 1.5×10⁻³ kg/m²
    a = 3.2 m/s²

    And max useful distance would be what, 5% of original thrust?

    D = sail-diam / ( divergence • √( give-up % ) )
    D = 250 m / ( 10⁻⁷ • √( 0.05 ))
    D = 11,200,000 km, or 7.5% of one AU.

    Well … that can’t be right. 

    Total illuminating power

    P-tot = π(½ 250 m)² × 850 kW
    P-tot = 41.7 GW

    Well… order three of them!

    ⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
    ⋅-=≡ GoatGuy ✓ ≡=-⋅

Comments are closed.