Gregory Benford and Larry Niven solved the problems with Shkavdov thrusters for a propulsion system for moving stars

Gregory Benford and Larry Niven have created as a ‘modified’ Shkadov Thruster for a propulsion scheme capable of moving stars.

Information from Centauri Dreams

Gregory Benford describes his modified Shkadov thruster.

Shipstars engines are Smart Objects–statically unstable but dynamically stable, as we are when we walk. We fall forward on one leg, then catch ourselves with the other. That takes a lot of fast signal processing and coordination. (We’re the only large animal without a tail that’s mastered this. Two legs are dangerous without a big brain or a stabilizing tail.) There’ve been several Big Dumb Objects in sf, but as far as I know, no smart ones. Our Big Smart Object is larger than Ringworld and is going somewhere, using an entire star as its engine.

Our Bowl is a shell more than a hundred million miles across, held to a star by gravity and some electrodynamic forces. The star produces a long jet of hot gas, which is magnetically confined so well it spears through a hole at the crown of the cup-shaped shell. This jet propels the entire system forward – literally, a star turned into the engine of a “ship” that is the shell, the Bowl. On the shell’s inner face, a sprawling civilization dwells. The novel’s structure doesn’t resemble Larry’s Ringworld much because the big problem is dealing with the natives.

The virtue of any Big Object, whether Dumb or Smart, is energy and space. The collected solar energy is immense, and the living space lies beyond comprehension except in numerical terms. While we were planning this, my friend Freeman Dyson remarked, “I like to use a figure of demerit for habitats, namely the ratio R of total mass to the supply of available energy. The bigger R is, the poorer the habitat. If we calculate R for the Earth, using total incident sunlight as the available energy, the result is about 12 000 tons per Watt. If we calculate R for a cometary object with optical concentrators, travelling anywhere in the galaxy where a 0 magnitude star is visible, the result is 100 tons per Watt. A cometary object, almost anywhere in the galaxy, is 120 times better than planet Earth as a home for life. The basic problem with planets is that they have too little area and too much mass. Life needs area, not only to collect incident energy but also to dispose of waste heat. In the long run, life will spread to the places where mass can be used most efficiently, far away from planets, to comet clouds or to dust clouds not too far from a friendly star. If the friendly star happens to be our Sun, we have a chance to detect any wandering life-form that may have settled here.”

This insight helped me [Gregory Benford] think through the Bowl, which has an R of about 10-10!

Image: Artwork by Don Davis, as are all the images in this article.


Shdakov thrusters aren’t stable. They are not statites, Bob Forward’s invention, because they’re not in orbit. Push them, as the actual photon thrust will do, and they’ll fall outward, doomed. So how to build something that harvests a star’s energy to move it and can be stabilized?

I worried this subject, and thought back to the work my brother Jim and I had done on speeding up sails by desorption of a “paint” we could put onto a sail surface, to be blown off by a beam of microwave power striking it. This worked in experiments we did at JPL under a NASA grant, with high efficiency. Basically, throwing mass overboard is better than reflecting sunlight, because photons have very little momentum. The ratio of a photon’s momentum to that of a particle moving at speed V is

(V/c)(2Ep )/EM

where Ep is the photon energy and EM the kinetic energy of the mass M. So if those two energies are the same, the photon has a small fraction of the mass’s momentum, V/c.

Big human built objects, whether pyramids, cathedrals, or skyscrapers, can always be criticized as criminal wastes of a civilization’s resources, particularly when they seem tacky or tasteless. But not if they extend living spaces and semi-natural habitat. This idea goes back to Olaf Stapledon’s Star Maker: “Not only was every solar system now surrounded by a gauze of light traps, which focused the escaping solar energy for intelligent use, so that the whole galaxy was dimmed, but many stars that were not suited to be suns were disintegrated, and rifled of their prodigious stores of sub-atomic energy.”

Creating and steering a giant standing solar flare

The key idea is that a big fraction of the Bowl is mirrored, directing reflected sunlight onto a small spot on the star, the foot of the jet line. From this spot the enhanced sunlight excites a standing “flare” that makes a jet. This jet drives the star forward, pulling the Bowl with it through gravitation.

The jet passes through a Knothole at the “bottom” of the Bowl, out into space, as exhaust. Magnetic fields, entrained on the star surface, wrap around the outgoing jet plasma and confine it, so it does not flare out and paint the interior face of the Bowl — where a whole living ecology thrives, immensely larger than Earth’s area. So it’s a huge moving object, the largest we could envision, since we wanted to write a novel about something beyond Niven’s Ringworld.

For plausible stellar parameters, the jet can drive the system roughly a light year in a few centuries. Slow but inexorable, with steering a delicate problem, the Bowl glides through the interstellar reaches. The star acts as a shield, stopping random iceteroids that may lie in the Bowl’s path. There is friction from the interstellar plasma and dust density acting against the huge solar magnetosphere of the star, essentially a sphere 100 Astronomical Units in radius.

So the jet can be managed to adjust acceleration, if needed. If the jet becomes unstable, the most plausible destructive mode is the kink – a snarling knot in the flow that moves outward. This could lash sideways and hammer the zones near the Knothole with virulent plasma, a dense solar wind. The first mode of defense, if the jet seems to be developing a kink, would be to turn the mirrors aside, not illuminating the jet foot. But that might not be enough to prevent a destructive kink. This has happened in the past, we decided, and lives in Bowl legend

Mechanical Engineering

We supposed the founders made its understory frame with something like scrith–a Ringworld term, greyish translucent material with strength on the order of the nuclear binding energy, stuff from the same level of physics as held Ringworld from flying apart. This stuff is the only outright physical miracle needed to make Ringworld or the Bowl work mechanically. Rendering Ringworld stable is a simple problem—just counteract small sidewise nudges. Making the Bowl work in dynamic terms is far harder; the big problem is the jet and its magnetic fields. This was Benford’s department, since he published many research papers in Astrophysical Journal in the like on jets from the accretion disks around black holes, some of which are far bigger than galaxies. But who manages the jet? And how, since it’s larger than worlds? This is how you get plot moves from the underlying physics.

One way to think of the strength needed to hold the Bowl together is by envisioning what would hold up a tower a hundred thousand kilometers high on Earth. The tallest building we now have is the 829.8 m (2,722 ft) tall Burj Khalifa in Dubai, United Arab Emirates. So for Ringworld or for the Bowl we’re imagining a scrith-like substance 100,000 times stronger than the best steel and carbon composites can do now. Even under static conditions, though, buildings have a tendency to buckle under varying stresses. Really bad weather can blow over very strong buildings. So this is mega-engineering by master engineers indeed. Neutron stars can cope with such stresses, we know, and smart aliens or even ordinary humans might do well too. So: let engineers at Caltech (where Larry was an undergraduate) or Georgia Tech (where Benford nearly went) or MIT (where Benford did a sabbatical) take a crack at it, then wait a century or two—who knows what they might invent? This is a premise and still better, a promise—the essence of modern science fiction.

Our own inner solar system contains enough usable material for a classic Dyson sphere. The planets and vast cold swarms of ice and rock, like our Kuiper Belt and Oort Clouds—all that, orbiting around another star, can plausibly give enough mass to build the Bowl. For alien minds, this could be a beckoning temptation. Put it together from freely orbiting sub-structures, stuck it into bigger masses, use molecular glues. Then stabilizes such sheet masses into plates that can get nudged inward. This lets the builders lock them together into a shell–for example, from spherical triangles. The work of generations, even for beings with very long lifespans. We humans have done such, as seen in Chartres cathedral, the Great Wall, and much else.

Alex Tolley proposed An Alternative Structure – The Solar Pulse Jet without the need for scrith

Here is Alex Tolley thought on avoiding the need for scrith.

The bowl structure becomes a Dyson swarm, each orbiting the star. Some of these objects are habitats, most are solar energy collectors and a few are particle beam generators. The job of the collectors is to accumulate the solar energy needed for the beam pulse. One can think of this as being like the LHC or the NIF laser. When the particle accelerators are in the correct position, the collectors send their energy to them to fire a pulse at the star, creating the mass ejection for thrust. Perhaps those particles could be muons.

Each pulse only happens when the beamers are aligned so that the solar jet is generated when the direction of thrust is correct. Because of the time delay, there should be a gap in the swarm behind the accelerators. All the objects in the swarm must track their distance from the star and use some energy to redirect and accelerate the solar wind to adjust their orbits to track the movement of the star.

Another advantage of the swarm is that the timing of the firing can be adjusted, allowing steering in the orbital plane. For a Dyson swarm, with objects orbiting in different planes, the steering can be in 3 dimensions.

Detecting such objects should be relatively easy. Their spectral signatures will be pushed into the IR, and if their direction of travel is away from us, there should be regular transit intensity peaks (swarm gaps) and even stronger peaks as the star flares. The pattern should be regular over a number of years. and would be expected be more frequent than a nova to maximize thrust.

Such an approach might be less efficient than the bowl, but I think it offers a plausible way to move a sun without invoking massive structures. It will require a technical civilization to allow travel between the habs and to ensure that the many objects in the swarm are kept running correctly.

Gregory Benford likes Alex’s swarm idea, and Eniac’s right that there may be better ways of inducing fusion on the star surface. Benford has worked on relativistic beam dynamics and fusion effects, theory & experiment, and there’s continuing research into how to do that. Not easy, no. Without giving away a plot point, the jet is very intricately run, but not by the Folk of the Bowl. Read SHIPSTAR to see how.

Must admit Benford envisioned the Bowl and chose it over the swarm idea, early in the 2000s, because the Bowl is a striking image of vast potential. A swarm is far more likely, yes. That’s on the of tradeoffs hard sf makes! Fiction is not a grant proposal

If you liked this article, please give it a quick review on ycombinator or StumbleUpon. Thanks


Don’t miss the latest future news

Subscribe and get a FREE Ebook