In my story I have a space ship, constructed in orbit around earth by an advanced artificial intelligence, large enough to be a generation ship. This ship was never designed to land on a planet. The construction of the ship (or ships, but we're focusing on just one for the story) uses up the majority of the earth's resources, and make a large part of the population rather unhappy.

Rebelling against the AI, the population not aligned to it, manage to take the ship and travel in search of a new planet. Finding this planet, they are forced to land the spaceship (not figured out how or why).

They are able to land the space ship, but not without it sustaining some damage. I intend that, generations later (100-500 years, I'm undecided) the ship forms a large part of a city (repurposing some of the onboard structures and systems) for the colonists.

However I want a point of tension in the story to be that this new city-ship was never designed to land or support its own weight and is slowly collapsing under its own weight.

At what size would a space ship be, using modern materials, that could still orbit a planet (without burning up, or being pulled apart from tidal forces), but when it lands it can only just support its own weight. I want that part to then mean any extra weight in the city-ship or damage tips it over the limit.

I'm happy to assume some hollow structure, but not entirely hollow (I.e. no Dyson spheres or O'Neill cylinders.).

I'm happy to expend some handwavium in how the ship landed (parachutes? artificial gravity?) or transformed into a 'city' etc, as this is not hard science, but still science based, so long as that doesn't interfere with the risk of it collapsing eventually..

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – L.Dutch
    Commented Jan 29, 2021 at 3:44
  • 1
    $\begingroup$ for closers - it is a legit question about physical limits for a structure in a gravity field. to op - your gauge are skyscrapers, so less than a few km size. it significant enough to allocate a significant portion of current metal production for a year, but not even close to making a significant dip in the total of resources available. Also as a tip, the moon is a good if not better resource source for space construction, considering interstellar aspect hydrogen for water and such u can take it from Jupiter. so u can read and ask at the place to tidy your plot in general, it needs it. $\endgroup$
    – MolbOrg
    Commented Jan 30, 2021 at 20:53

2 Answers 2


The upper boundary on the size of the ship would be constrained by two things acceleration forces and rotational forces

The first is unlikely to be an issue unless you envision the ship was able to obtain multiple G accelerations, which in most setting is not needed simply because much gentler rates of acceleration still allow you to reach your desired final V in a time frame that is irrelevant given the decades/centuries it would take for a 'conventional' ship to reach another star system. (A few extra weeks/months spent on acceleration/deceleration is irrelevant to a generation ship.)

That leaves rotational forces (for 'gravity'). And these are an issue because there are physical limits on how big the diameter of a rotating body beyond before the strain imposed causes structural failure. Using the strongest materials we are currently aware of (e.g. some form of carbon nano-fiber) the maximum diameter of a rotational habitat called a Bishop Ring is estimated to be about 1000K with a width of 300 to 400K. Although you would probably build one with a smaller diameter to allow for some margin or error in the quality of the building material.

So your ship could be quite large. In fact more than large enough to form the basis of a city. The issue is landing something that big. A controlled landing barring handwavium technology would be next to impossible. Which leaves a crash landing sometime after your colonists have descended to the surface with all their stores and equipment and set up shop leaving the ship more or less vacated.

Perhaps it simply breaks up into several enormous pieces that impact around the surface at different locations to be explored and occupied regardless of the weird angles the structures may be lying at compared to their normal up/down/top bottom configuration in space.

  • $\begingroup$ Would you call a Bishop Ring a hollow structure? It's not the sort of space ship I envisioned, but it's sometimes a good thing to have your expectations challenged! $\endgroup$ Commented Jan 30, 2021 at 9:43
  • $\begingroup$ The Bishop Ring merely serves as a theoretical template for the maximum possible size of your ship. You asked how large such a ship could be and I gave an example of the largest structure it might be possible to build in space. The key metric is how much rotational force can be placed to a rotating object before it breaks apart, not how hollow or not it is. In engineering terms at the moment the answer is about 1000 K. You could in theory have a ship smaller than that with thousands of decks. The key issue is how much stress is placed on the hull by rotation not the 'density' of the ring. $\endgroup$
    – Mon
    Commented Jan 30, 2021 at 10:23
  • $\begingroup$ I mention hollow as I specifically called out hollow structures in my question such as Dyson spheres and O'Neil Cylinders. I hadn't heard of Bishops Rings until now $\endgroup$ Commented Jan 30, 2021 at 15:22
  • 1
    $\begingroup$ In any event you not try to use something the size and shape of a BR as a generation ship. The cross section is wrong for a start. It's basically a giant doughnut and you want a more conventional cylindrical shape to minimize the profile of the vessel along its line of flight. This is because you want to reduce the risk of kinetic impacts with dust & gas molecules etc during the voyage as much as possible. A narrow frontal profile also makes it easier to avoid any threats detected in advance and/or use active defenses to stop them. The BR represents the upper limit in size that's all. $\endgroup$
    – Mon
    Commented Jan 31, 2021 at 22:57

There is no practical lower bound to this...

A sphere made of aluminium foil the size of a basketball will be a practical spaceship (mouse pilot, or AI). Pressurized at 0.2atm it will keep its shape quite nicely, and be quite maneuverable if it had external thrusters. Put it on the planet and it'll be crushed.

Your lower bound for this happening is the smallest spaceship possible based on crew or fuel or mission requirements - ie. no lower bound.

Or upper...

Look at the ISS, it's made by bolting bits together. While those individual pieces are strong, a 1g force is highly likely to shear the connections. Why waste weight making the connections stronger than they need? This scales up to the ludicrously large too. If my city is buildings connected to a base (eg Atlantis from Stargate), why waste weight on bolts able to withstand 1g when they're always going to be in zero g? Or windloads when they'll never be in an atmosphere?

When building floors, you choose beams based on the expected load of the floor, this includes gravity in the maths. If your ship is expected to be in microgravity, why waste weight on building the floors to a 1g building code?

Basically, everything designed to work only in space will fail on a planet.

With really large ships you won't be able to lay them flat on the surface. A ship 2000km in diameter can't lay flat on either the surface or ocean - it will have to bend and distort due to earth curvature. This would damage the ship badly unless it was specifically designed for this landing.

Your upper bound is the upper bound of spacecraft that can fit on the planet - an earth sized spacecraft can't land on earth.

  • $\begingroup$ Your answer highlights some assumptions I missed out on specifying in my question. Once I've been able to think them through more, would you be ok with me adding them to the question? Things like wanting this to be able to be a city (so being a basketball sized city won't work) etc $\endgroup$ Commented Jan 28, 2021 at 14:18
  • $\begingroup$ @Pureferret Of course :-) $\endgroup$
    – Ash
    Commented Jan 28, 2021 at 14:49
  • $\begingroup$ I really appreciate that, I've updated my question now with extra details and some small tweaks. Let me know is that oversteps and should be a new question. $\endgroup$ Commented Jan 29, 2021 at 1:40

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .