Consider a hard-SF (slower than light) starship. Boosting it up to speed takes enormous energy.

There are good reasons to want to not take all the fuel with you, but to collect energy on the run. Current thinking is that the Bussard ramjet, as popularized by the fiction of Larry Niven, would not gain energy in excess of its drag. Too bad. ☹

But the ideas of Robert L Forward are sound: use a laser to beam power from collecting points close to the sun.

You can beam power to a relativistic craft under way. But what about material goods? I have contemplated the idea of launching the craft and then sending pellets of matter after it. These would supply momentum when they are caught, and they can be released again or bounced back to avoid gaining needless mass, or retained if they are valuable useful types of atoms that you need.

More generally, how might you resupply a starship? And why would you send matter later rather than launching it initially? Since energy is m•v2, it’s more costly to make it catch up as opposed to including the item in the initial launch.

Note: two Answers have focused on Aldrin cycler, but that has zero applicability here. This craft in question will be under thrust or coasting at relativistic speeds and is interstellar; that is, between stars.

  • 4
    $\begingroup$ If your mission is planned far enough in advance, and your supplies are non-perishable (fuel, spare parts, freeze-dried/concentrated food), you could launch them in advance, at a lower speed, and let the ship catch up with them. Then again, capturing these slow-moving projectiles would slow down the ship. $\endgroup$ – Peregrine Rook Jun 20 '16 at 1:10
  • $\begingroup$ Current thinking is that the Bussard ramjet, would not gain energy in excess of its drag. - totally false - I have to explain why? And are you really stick to that idea, or you ok with any explanation how real non ftl starship fly? And add some inside about starship tech level which is comfortable for you. $\endgroup$ – MolbOrg Jun 20 '16 at 2:23
  • $\begingroup$ Got a citation for totally false? The wikipedia page I linked to agrees with my summary. So yes, explain why. I remember reading some years ago that calculations and numerical number crunching pretty much nailed the coffin of the ramjet; the only reason to suppose it might be possible after all is with some new twist. But, otoh, it could be used as a good brake! $\endgroup$ – JDługosz Jun 20 '16 at 3:32
  • 4
    $\begingroup$ Shooting stuff at spaceships using relativistic speeds is something you do to enemy ships, not your own. If you get the delta-V slightly wrong, you'll only be supplying your ship with new openings. $\endgroup$ – Cyrus Jun 20 '16 at 9:36

Launching matter to a spaceship is generally costlier than sending it in the ship in the first place. You have to lift not only your payload, but fuel and mahinery as well, so the less launches you do, the more efficient you tend to be.

You only do multiple launches if:

  • Your payload is too heavy for a single launch, im which case you'll want to do some orbital assembling of parts;
  • Your mission involves life support for an indefinite amount of time and you can't recycle food, water and gases.

The ISS is the best IRL example of a multilaunch mission.

The best fictional example I can think of is the Earth to Mars mission ship from the book and movie The Martian: the mission had enough food for its planned length, but due to an emergency they had to double the span of time in space. So they sent food for the ship to catch during a flyby around Earth.

Sending materials to a ship that is going away from you, except in an emergency, is a no-no. I think it is generally better to assemble everything you need in orbit, then send it away.

If your mission is doing the rounds between planets, then these planets may send it ressuplies for it to catch along the way, during flybies. This is explored in Joe Hadelman's "bound" series (Marsbound, Starbound, Earthbound).

Now, if the trip is too long, then the destination planet (or point in an asteroid belt, comet cloud, dyson sphere etc) may launch such packets so that they double as a way to reduce approaching speed.

As for ways to catch ressuplies:

  • During flybies, the packet is sent in an escape trajectory that matches the trajectory of the ship as closely as possible. The goal is to have the relative speeds of the ship and the packet close to zero on their closest approach. The closest approach should also have them a few dozen meters apart, for safety. You don't want to bang them against each other if you miscalculate. To catch te packet, have it approach the ship with small thrusts. You can then use a mechanical arm or claw to catch it, or have ot dock to you, or a combination of both. Watch some ressuply missions to the ISS videos for inspiration. If these methods fail or are unavailable, send a tethered astronaut to catch it.

  • Incoming packets from destination worlds need to decelerate greatly prior to interception, otherwise they become unintended ballisic missiles. The goal is to have their relative speed close to zero, but they should still impart some momentum to the ship. The safe relative velocity depends on the technology involved - it can be as fast as you can do without damaging neither the packet or the ship. To catch, have the packet position itself for interception, then unfurl some really big net.

Last but not least: remember that in space you don't move in straight lines. You are always orbiting something. You are orbiting a star when going from one planet to another, you orbit the galaxy when going from one star system to another. Aiming your ressuplies involves a lot of curves, and these curves change whenever you change your speed. If you are going to send multiple packets to a ship that is going from one planet to another, each packet is going to have a different trajectory. Food for thought.

P.s.: if you are both a space nerd and a gaming masochist like me, I suggest you play Kerbal Space Program. You'll learn a lot about fuel costs, orbital maneuvers and how to go from one planet to another, in a rather practical way. Helps a lot in understanding how these things really work.

  • $\begingroup$ I'm referring to an interstellar relativistic craft, not a transfer orbit "doing rounds". In the case of a permanent transfer craft, you boost only the people and cargo for that trip and leave the walls and fittings on the orbit. The case of The Martian turned into this, as well as a slingshot. None of that applies to my situation. $\endgroup$ – JDługosz Jun 20 '16 at 3:44
  • 1
    $\begingroup$ Btw, I learned about how orbital mechanics works by writing my own games, over 30 years ago. My Lunar Lander was more like the real DSKY than a video arcade game! $\endgroup$ – JDługosz Jun 20 '16 at 3:47
  • 1
    $\begingroup$ Carching packages sent from the destination end is a useful twist... that could make the difference! $\endgroup$ – JDługosz Jun 20 '16 at 3:48
  • $\begingroup$ @JDługosz you've just got a fan! $\endgroup$ – Renan Jun 20 '16 at 10:58
  • $\begingroup$ Thanks. (I think.) $\endgroup$ – JDługosz Jun 20 '16 at 12:56

Launch pellets (of fuel or propellant) or pallets (of other ship's stores) using an accelerator of some sort (e.g. coilgun, railgun, starwisp, light sail). Make sure the fuel pellets are the proper isotope mix ratio for you fusion power plant and carry enough "other stuff" for propellant. You should also charge the pellets so that your electrico-magnetic scoops can "catch" them.

A pre-seeded trajectory

A preseeded trajectory is one in which the fuel and/or propellant for a Bussard Ramjet like craft is launched prior to spacecraft launch. The velocity and launch time for each pellet is carefully calculated so that the spacecraft which follows it encounters it under ideal conditions.

The problem with a Bussard Ramjet is that the reaction cross-section of a $4 H^1 \rightarrow He^4$ reaction is very small. This means it takes a long time for the reaction to occur, which means that if the reaction occurs at all, it likely occurs after the fuel left the engine of your spacecraft. Using a fuel with a larger cross-section can greatly improve the gain possible from the reaction which might make the Bussard Ramjet possible.

Your pellets should contain the proper isotope mixture of fusion fuel (e.g. tritium + deuterium). Bear in mind that you need not only consider Hydrogen isotopes. $H^1$ fuses with $B^{11}$ for a primarily aneutronic reaction ( $H^1 + B^{11} \rightarrow 3He^4$ ).

In this case your pellets would just provide the $B^{11}$ portion of the fuel. You would scoop the $H^1$ component for this reaction from the interstellar medium. You could also get all of your reaction mass requirements from the interstellar medium's $H^1$. Such a system would provide a "clean burning" (aneutronic), "fast burning" (higher reaction cross-section) fusion reaction with higher thrust (extra mass provided by the interstellar medium) than a pure fusion reaction.

This technique eliminates the worst problems of a Bussard Ramjet and enables an interstellar craft to make the trip without accelerating and decelerating huge masses of fuel and propellant.

You must accelerate the pellets and time their launch very carefully and the spacecraft using this trajectory must not deviate from it by very much or it'll lose all benefit.

Although most of the benefits possible to achieve by such a system are realize if just the fuel and propellant are supplied this way, it should also be possible to seed the trajectory with other supplies. Some mechanism from grappling those pallets of other supplies must be included and the ship using the seeded trajectory must know it is coming or normal operations of the ship would destroy the goods.

This method has a few big benefit from the perspective of the crew:

  • All resupply shots are fired before committing to the launch of the spacecraft (your resupply is laid down before launch).
  • The prelaunched fuel would provide some visibility into the interstellar medium – thus alerting the trailing spacecraft of unseen hazards (e.g. brown dwarfs).

A post-seeded trajectory

A similar mechanism by which pellets of fuel and propellant and/or pallets of other stores are shot at the receding spacecraft. This technique has the added benefit of providing positive momentum transfer to the spacecraft (thus accelerating it), however, each shot must be carefully timed and calibrated to keep from hitting the spacecraft like a hypervelocity weapon instead of a gently captured resupply pallet.

Once again, the biggest benefit comes from resupplying the fuel and propellant. Other ship's stores would be nice to add but not nearly as critical to overall mission mass.

If something happens to your civilization after spacecraft launch but before all the necessary supplies are launched, then your crew will die. There's simply nothing they can do about it.

Important Note

It's important to note that in either system, the group launching the supplies must be very careful about the supply's trajectory.

  • Shots not in line with the ship's path provide no benefit.
  • Shots fired are the wrong velocity either fail to reach the craft or hit it so fast that they would likely destroy it.

Also the spacecraft can only stray from the designated trajectory by a little or it'll fail to get the vital supplies it requires.

  • $\begingroup$ I like the idea of seeding fuel for the Ramjet. Falling behind would still be a death sentence, but it should be easier to recover by burning the previously collected fuel reserves to catch up. $\endgroup$ – Cyrus Jun 20 '16 at 22:09

The answer (as far as I can see) is there is no safe way to send matter after your ship to resupply it. Any slight deviation from the planned course would mean the death of the crew, either from a collision or from starvation.

To illustrate this, suppose the ship is underway for a year when the beam powering it loses 20% of its energy due to an unexpected dust cloud between the ship and the origin system, and that happens to be more than the 10% beam loss accounted for in the design.

Due to this, the ship can't accelerate as planned and falls behind on its itinerary, meaning they're now outside the focus of the beam and lose more power. The crew immediately sends messages with this new information as well as their updated trajectory back to the origin planet, but it will take weeks for that signal to reach the beaming station and weeks more for the adjusted beam to reach the ship again.

If the ship were self-sufficient, this would be a minor setback. Past the halfway point, it could delay the deceleration phase a bit and vector the thrust slightly to still arrive at the destination.

However, if the ship needs those resupply pellets, it's dead.

Those pellets were launched months ago, most likely within a few weeks after the ship itself launched. They had to be, since otherwise the pellets would have to be launched at such speeds that they need huge rockets and fuel supplies just to decelerate down to the ship's speed. The origin can do nothing to help the ship at this point.

The ship may be able to signal the incoming pellets with updated course commands and for the next incoming pellet that may work, but the others will most likely not carry enough fuel to adjust course once the ship falls behind, since they have a huge correction to make. One by one, they overshoot the ship and that's it, game over.

The only resupply that might work is a full-sized drone ship sent out ahead of the main ship, carrying not only supplies but a huge amount of fuel for itself. This ship would aim for a meeting point roughly halfway (or when the supplies are needed) and would get course adjustments directly from the main ship all the time, allowing it to rendezvous in a much more controlled way. After transferring the supplies and any fuel left over, it would be abandoned. Without the need to have it arrive at a particular destination (and particular velocity), the drone ship can carry much more than the main ship with the same propulsion system and parameters.

For the cost of a complete second ship (and beaming station), this setup has very few advantages over just attaching the extra storage space to the main ship and detaching it halfway to save on fuel needed for deceleration. I could see it being used if two separate power beams for two ships is somehow more feasible that one double beam or if the unmanned ships could use a different propulsion system that is deadly to humans.

  • $\begingroup$ I'm not entirely sure the later supply packets (pellets, or whatever you want to call them) would necessarily be unable to adjust course to match. Yes, they would need to be going faster to catch up with the ship, but they are also farther out, and the greater the distance, the less delta-v needed for the same change of position at some later time, because the propulsive maneuver has more time to take effect. Without having done the math, it seems plausible that these two effects largely balance out in the end. $\endgroup$ – a CVn Jun 20 '16 at 20:34
  • $\begingroup$ I'm not sure either, but considering that the ship (after suffering from a period lower than planned acceleration) would be falling more behind its own itinerary with every passing day, I don't think the extra distance will compensate enough. $\endgroup$ – Cyrus Jun 20 '16 at 21:49

You have two problems with interstellar travel. 1. Getting up to speed (and slowing down) is very energy intensive and 2. it is really hard to carry everything you need because mass slows you down but mid-flight resupply is difficult because the distances are vast and the speeds are very high, making an intercept either very energy expensive or potentially explosive.

Assuming your ship is moving at a good chunk of light speed I don't think you can easily arrange for physical intercept. As mentioned, shooting relativistic packets of mass at a ship is something you do to enemy ships, not your own.

But you CAN have a circling packet ship, perpetually orbiting between the two systems at a high speed, that the cruise ships are timed to come close to with a matching velocity to allow for easy transfer. But of course that means the resupply ship was ALREADY loaded with stuff and was boosted to a high speed, if you could do that then why not just do it with the cruise ship (maybe it was boosted using the Orion drive that isn't compatible with human occupants)? The packet ship is robotic and can grow its own biomass to impart to passing cruise ships, or be resupplied itself when it is close to a star system when energy requirements are not as limiting.

So in order to maintain this perpetually circling packet ship system you would need pre-positioned laser boosters along the interstellar route. They could be boosted in the solar system and now maintain station along the route the interstellar ships use. They may even be able to maintain position by shooting each other (if there are enough of them so they are within range of each other). So your cruise ship is laser boosted in the solar system, then as it passes each laser booster it gets another nudge. At the halfway point it starts getting laser nudges to help it slow down. This system can also keep the packet ships moving and allow for some ability to speed up/slow down to widen the intercept window between packet ship and cruise ship.

This will remove a LOT of onboard fuel/reaction mass requirements for both ships, allowing the ship to carry more supplies for the crew or cargo.

A really wacky solution is a series of high gravity nodes strung along your path so the ship gets the benefit of gravity slingshots along the way. Obviously this would need a crazy high level of technology to construct but once in place it could be used by relatively low tech cheap ships.

The general idea is to spend a lot of time/energy creating an interstellar "hi-way" that can then be used very cheaply. The packet ships are there for mid-transit resupply of critical parts, food, maybe even fuel if they are capable of producing collecting it. All of this allows the cruise ships to get up to speed as fast as possible by needing less mass for fuel and potentially food stuffs to make the trip. Close to a solar system energy is plentiful, so the packet ships can be resupplied there, but in the cold inbetween spaces they are lifeboats of food/spare parts/fuel that has already been "brought up to speed" so transfer to a cruise ship is as effortless as possible.

  • $\begingroup$ Midway lasers are interesting. But how are they powered? $\endgroup$ – JDługosz Jun 21 '16 at 19:34
  • $\begingroup$ Fusion reaction? They could be automated and sent out at a much slower velocity, then maintain station by lasers at the other end (so you build it back to front. Obviously this would imply the route between systems is quite old as it would take years to set this up. You'd need thousands of them to cover interstellar distances but still be in range of each other, but it could provide a nice steady(ish) thrust the entire way. If you don't have anti-gravity this will eliminate the need for spinning rings on the ship. $\endgroup$ – Jason K Jun 21 '16 at 20:43
  • $\begingroup$ Sending the waypoint takes more fuel than just taking the fuel with you directly. If it was practical to carry fuel (to run the laser), the ship could just take it. Btw, taking fuel for fusion is only barely possible, according to the tyranical rocket equation. $\endgroup$ – JDługosz Jun 21 '16 at 20:51
  • $\begingroup$ The normal thought for lasers is to have them generated as close to the sun as you can manage, to be solar powered. They don't need to be along the route as the beam will go a long way. $\endgroup$ – JDługosz Jun 21 '16 at 20:54
  • $\begingroup$ @JDługosz But you have to consider the SPEED. A big fat fuel rich laser waystation can be slowly accelerated in the home system and sent on its way at a (comparatively) low velocity towards the target star. Decades later it gets close enough to be slowed and "parked" by a laser in the target system. Then the waystation does the same to the next one, repeat, all the way back to the home system. The cruise ship is trying to be FAST, which requires way more power, but it can't carry all that fuel without being too massive which requires more thrust, then more fuel, then more thrust, etc. $\endgroup$ – Jason K Jun 21 '16 at 20:56

My answer is simple. To resupply a starship, make it part of a fleet of starships. Most of this fleet will be tanker ships and freighters carrying the supplies your, I presume, manned starship(s) needs. Since they are all flying together they will share the same velocity, and are at relative rest with respect to each other. The empty freighters and tankers can be abandoned once their supplies are gone. Always remember to give them a little boost to the side to avoid them crashing into anything inhabited in the target system.

Another way could involve your starship carrying one mouth of a wormhole. The other mouth is back in your home system. The supplies will need to be boosted up to the same velocity as your starship before they can be sent through the home system's wormhole mouth. This suggestion may break your sublight requirement, but wormholes are speculatively plausible. This is not too dissimilar to the teleportation back and forth from a relativistic spacecraft in Poul Anderson's The Enemy Stars.


Buzz Aldrin once proposed a relay system of vessels between earth and mars that, once in orbit and up to speed, would require virtually no additional fuel/force to keep there. The idea was that you sent a ship up, attached to this already moving vessel, and hitched a ride to mars or back to earth. You could build a similar relay system along a few, well known or populated routes, to create fuel depots for these vessels. Then you could use smaller, faster vessels, to deliver the goods locally to ships (as a cargo load of munitions is going to be smaller, and easier to get up to speed, than an entire warship).

  • $\begingroup$ Thar applies to orbits, not to an interstellar relativistic craft. @renan covered this innhis answer earlier. $\endgroup$ – JDługosz Jun 20 '16 at 21:46
  • $\begingroup$ @JDługosz orbits are everywhere even intergalactic travel will have orbits )) $\endgroup$ – MolbOrg Jun 21 '16 at 14:45
  • $\begingroup$ The individual stars embedded in a disk don't have simple repeating orbits, but act on each other locally more like a gas as well as being confined to the disk. They bounce up and down like carosel horses, as well as make a Galactic Year of a quarter billion years. A Cycler would not work the same way, but would end up wandering like any other star in the same open cluster. $\endgroup$ – JDługosz Jun 21 '16 at 19:33
  • $\begingroup$ So add in solar panels and some sort of energy based (but slow) engine to do course corrections. If you are picking up and dropping off cargo you will be losing speed anyway and will need to "throttle up" at some point. It is still cheaper than conventional shuttling. $\endgroup$ – Marshall Tigerus Jun 22 '16 at 13:26

Conceptually the answer is to send pellets after the starship. The ship uses some sort of momentum exchange system to capture the pellets and their momentum, i.e. the ship accelerates as it slows the pellets down.

This can be done in many different ways, including having the pellet slam into a pusher plate (something like the ORION nuclear pulse drive) or using a magnetic or electrostatic field decelerate the pellet. The main issue here is that early in the flight, where the pellets are moving much faster than the starship, the energy exchange will likely vapourise the pellet and the starship will only see a cloud of plasma in the capture zone. This isn't a show stopper, since early on in the flight, the ship can consume on board supplies.

As the ship continues to accelerate, the relative velocity between the ship and the incoming pellets becomes much lower, so it becomes feasible to capture the pellets from the momentum exchange system, and take the materials on board for consumption. Since the propulsion is via momentum exchange, then it really doesn't matter what the pellets are actually made of with respect to propulsion. In theory, you could send canned peas and other supplies via these means, although there are probably technical issues in ensuring the pellet launch system can handle and accelerate the pellets, ensuring the pellet isn't being vaporized inside the launcher and isn't going to be affected by the interstellar environment causing it to miss the starship.

There is a great deal of literature on this idea, although AFAIK few people have considered this as a means of resupply rather than just propulsion. Starting points:





You could set up a net work of space stations between plaints, the space station could served as a island port do when ocean sailing on our planet. The space stations would be places to restack on trade and refuel.

  • 1
    $\begingroup$ you have to stop to take rest, in space there is no friction you have to work to stop. $\endgroup$ – MolbOrg Jun 20 '16 at 2:37
  • $\begingroup$ @MolbOrg That not hard just have two sets of engines one to again speed and the other pointed in the opposite direction used to slow down. $\endgroup$ – Bryan McClure Jun 20 '16 at 3:01
  • 1
    $\begingroup$ The "rest" would be pointless, as you use the same energy yo stop at the waypoint as you would if you continued directly. And you don’t need two engines: just flip the ship around. $\endgroup$ – JDługosz Jun 20 '16 at 3:35
  • 1
    $\begingroup$ So the point is that they may need to stop for supplies like water and food, even at the expense of energy/reaction-mass? $\endgroup$ – JDługosz Jun 20 '16 at 3:52
  • 1
    $\begingroup$ @BryanMcClure On the contrary, it's very hard. Remember how Apollo 13 had to go around the Moon to get back to Earth, even though the oxygen tank explosion occured well before they got to the Moon? Same basic problem: they didn't have enough fuel to turn around. (Or well, technically they did - the crew even had an appropriate abort pad onboard - but doing so would have required them to give up the lunar module, which they needed as a lifeboat. So they didn't.) See Is this a correct understanding of Tsiolkovsky's rocket equation? on Space Exploration. $\endgroup$ – a CVn Jun 20 '16 at 20:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.