Take the 2-minute tour ×
Physics Stack Exchange is a question and answer site for active researchers, academics and students of physics. It's 100% free, no registration required.

Freeman Dyson proposed that the power needs of an advanced civilisation would eventually require the entire energy output of a star to be collected, so that the star would end up surrounded by a dense network of satellites extracting power from the radiation. In science fiction this idea has often mutated into a solid shell that completely surrounds the star. My question is, if it were possible to build such a shall, what effect would it have on the dynamics of the star inside?

I suspect the effect would be catastrophic, but I'd like more details. Here's my reasoning: Wikipedia implies that the temperature gradient between the core and outer layers of a star plays an important role in its stability. The shell around the star would reflect or re-emit a lot of the star's radiation. (I'm assuming the shell isn't composed of perfect solar collectors that would simply absorb all the radiation.) This reflected radiation would reduce the loss of heat from the outer layers of the star. This would reduce or even eliminate the temperature gradient, which I guess would cause the star to expand. My question is whether this is correct, and if so whether it would be enough to disrupt the process of fusion in the star's core. Or would there be some other, less obvious effect on the star's dynamics?

A closely related question is, would stars be stable in a static or contracting universe? In this case all of space would be filled with the radiation emitted from other stars, and I'd be interested to know what effect this would have on stellar dynamics.

share|improve this question
An important point here is that "stability" in this case simply means convecting versus non-convecting for almost all stars. Of course stellar physics people spend a lot of time on the exceptions because that's where all the fun is. –  dmckee Feb 5 '12 at 20:16

3 Answers 3

up vote 4 down vote accepted

Main sequence stars of roughly one solar mass are--in a lot of ways--really, really simple objects.

They have non-convecting cores and non-convecting envelopes and may have a convecting layer in the middle.

About the only things that a uniform change in the external temperature could do is tweak the location and existence of the convecting band and marginally increase the overall temperature.

That might have a measurable effect on the star's lifecycle (how fast it burns, how much of the outer hydrogen makes in down to the core to fuse before the switch to helium burning), but to first order I would expect the changes to be mediated by $T_{ext}/T_{star}$, and so to be small. (BTW--I'm suspect, but can't prove, that the $T_{star}$ you want to use depends on what you're looking at (i.e. the middle layers for convection band tweaks, but the core temperature for fusion rate tweaks).

share|improve this answer
Thanks for the answer. I've accepted it, but I'm still interested in what happens if the constant outside temperature is really, really high. In the case of a static universe (or a perfectly thermally insulating Dyson sphere) the radiated energy would have nowhere to go, and consequently the vacuum temperature would reach the stars' surface temperature. At that point the stars can no longer keep their outsides cool by radiating, so it could increase still further, up to the ~$10^7$ K core temperature and beyond... –  Nathaniel Feb 6 '12 at 15:08
The other thing I guess I'm kind of curious about now is, if you wanted to destroy a star (i.e. stop the fusion process) while expending a minimal amount of energy, how would you do it? I guess that's a bit frivolous to make into a proper question though. –  Nathaniel Feb 6 '12 at 15:11
@Nathaniel: What happens as it gets hotter is much like what happens in the core of a bigger star...it burns faster (and by faster we mean the rate grows exponentially with the temperature), switches to helium burning much sooner goes red giant with enough force to lose a lot of mass and so on. That said, destroying a star is really hard without some kind of science-fictional technology. The Charles Stross book Iron Sunrise posits one of way doing it that is pretty rigorous aside from the not-related-to-any-physics-we-know triggering mechanism. –  dmckee Feb 6 '12 at 18:55

Interesting question - I don't know much about stellar dynamics but my guess is it will have very little effect.

My reasoning: right now the Sun is receiving thermal radiation from a heat bath which is dominated by the cosmic microwave background, at a temperature of about 2K. If the Dyson sphere were to be built, presumably it would be designed to operate at a temperature comfortable for human life / technology, around 300K.

So the Sun's background will rise by several hundred degrees - but this should be negligible when compared to it's own enormous temperature, so there should be little or no changes.

share|improve this answer
The shell temperature will be dependant on its radius. The energy radiated by the shell will be equal to the energy radiated by the star, so the increased surface area of the shell versus the star will control the temperature required for equilibrium. –  Colin K Feb 5 '12 at 16:43
@ColinK In the case of a thermodynamically perfect system, the equilibrium temperature would be controlled by the outer radius of the shell. The other limit is controlled by the inner radius. A "real" Dyson sphere probably falls somewhere between and presumably is close to the perfect that worst case. –  dmckee Feb 5 '12 at 17:01
@dmckee: Hmm, I was thinking of it as having no thickness, but you're right. –  Colin K Feb 5 '12 at 18:09
The radius of the sphere would determine the temperature of the outside of the sphere, assuming the star's output remains constant. The temperature inside the sphere can be hotter if the sphere is not made of a perfect thermal conductor. I'm most interested in the case where it's a perfect insulator, i.e. it reflects all of the star's radiation back at it. Naturally this wouldn't result in a habitable environment on the inside - this is just a thought experiment to see whether the "cold radiation bath" is important for the stability of a star. –  Nathaniel Feb 5 '12 at 18:48

Building it as a solid shell is a really terrible idea: it would just plummet into the star due to slight imbalances in gravity. - this is wrong, see comments.

In terms of stellar dynamics, I don't think it would have any major impact on the star. There would still be a temperature gradient as you can't really heat up a vacuum. There maybe some radiation pressure due to the reflected photons but this will be pretty insignificant. The star might eventually expand but I think this will be well beyond its main sequence phase anyway.

share|improve this answer
Er...a solid spherical shell has neutral stability in a $1/r^2$ field (i.e. both gravitation and the radiation), so your first sentence is not really correct. It would require proportionally less stabilization than a ringworld. –  dmckee Feb 5 '12 at 19:43
If you think about it, the shell's center of mass is directly in the middle. If that is even slightly misaligned with the CoM of the star, one side of the sphere will be attracted to the star more than the other and it will move towards that direction. As the side gets closer, it is attracted even more, causing it to crash into the star. It is explained on wikipedia better than I can: en.wikipedia.org/wiki/Dyson_sphere#Dyson_shell –  Durand Feb 5 '12 at 21:03
Durand: No, the wikipedia says that uncorrected drifts are a problem, and it is right because the sphere is not stable, but it is also not un-stable. There is no net gravitation force between the star and the sun for any relative position with the sun inside. Yes, one side is closer, but there is more stuff on the away side and it is just enough to balance out. Really. This is a well known problem. Follow the link in the wiki article so Shell theorem. By contrast a Niven ring is unstable in it's own plane (but stable out of plane). –  dmckee Feb 5 '12 at 21:07
Mm, ok. I see what you're saying. I always end up confusing myself when I think about problems like this! –  Durand Feb 5 '12 at 21:10

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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