The advanced civilization in my story is trying to reach as many galaxies as possible in the next few billions years before the expansion of space has sent those galaxies permanently out of reach without some form of FTL travel which my story does not have.

I plan to have von Neumann probe-like self replicating machines powered by laser relays from Nicoll-Dyson beams, possibly a double beam from opposite sides keeping the star in its location whilst sending probes in opposite directions.

Once in the new galaxies the self replicating machines will gather materials and turn the majority of stars into Shkadov thrusters, the collective thrust of the stars and the gravitational hold of the galaxy will move the whole galaxy towards the civilizations home galaxy. This assumption that whole galaxies can be moved was taken from the following videos: https://www.gregschool.org/gregschoollessons/2018/6/12/shkadov-thrusters https://www.youtube.com/watch?v=_VetAm7fCS0

My question is what percentage of the stars need to become Shkadov thrusters to move the whole galaxy? Would there become a point where converting more stars is taking more time than it is worth for the additional speed? The civilization isn't worried about the time the journeys will take, as long as the galaxies eventually arrive to their area of space.


1 Answer 1


This wouldn't work.

The big issue is that stars make up only a few percent of the Milky Way's total mass - estimates vary by a bit, but I've heard numbers in the 3-5% range. For instance, McMillan 2011 claims about 4.98%. Dark matter makes up a substantial amount, as do interstellar clouds - i.e. non-rigid bodies.

The point is, even if you attached a thruster to every single star in the galaxy, it wouldn't produce any significant gravitational effects on the rest of the galaxy, and it certainly wouldn't be enough to move the Milky Way.

You could try and find a galaxy that 1) has an extremely low amount of dark matter, like NGC 1052-DF2 or NGC 1052-DF4, and 2) is old and gas-poor, meaning most of its star formation is over and much of its mass is locked in stars and stellar remnants, not interstellar clouds. I'd have to learn more about stellar populations of these ultra diffuse galaxies to say anything intelligent on the subject, though.

  • 3
    $\begingroup$ Well. If you attached thrusters to EVERY star, you'd get the whole galaxy's worth of stars. $\endgroup$
    – Andon
    Commented Apr 30, 2020 at 13:48
  • 2
    $\begingroup$ @Andon True, but in terms of potential resources, it would be a pretty small fraction of the total available mass. Stars contain, optimistically, 20-30% of a galaxy's baryonic mass (depending on star formation rate and age), so you're missing out on a good chunk of things. Plus, that gas mass would be much easier to harvest, I think, than gas locked in stars. All you have to do to get hydrogen from an interstellar cloud is scoop it up, whereas starlifting (for example) is more complicated and less energy-efficient. $\endgroup$
    – HDE 226868
    Commented Apr 30, 2020 at 13:54
  • 4
    $\begingroup$ Interesting question I’ve never thought of: If you take all of the stars out of a galaxy, is it still a galaxy? Can you have a galaxy of nothing but black holes and dust? What is the sound of one hand clapping in the interstellar medium?? $\endgroup$
    – Joe Bloggs
    Commented Apr 30, 2020 at 14:27
  • 2
    $\begingroup$ @RodolfoPenteado It was the OP's requirement that the entire galaxy be moved - and besides, as I mentioned above to Andon, the rest of the galaxy contains a substantial amount of resources, too. $\endgroup$
    – HDE 226868
    Commented Apr 30, 2020 at 15:11
  • 3
    $\begingroup$ @LawnmowerMan Space is big and stars are small. Most of the gas is nowhere near a star; if you start moving the Sun, a gas cloud two light-years away isn't going to pulled along with it. It's the old inverse-square law at work. $\endgroup$
    – HDE 226868
    Commented Apr 30, 2020 at 22:42

You must log in to answer this question.