# Are the mirrors on O'Neill cylinders really feasible?

The classic O'Neill design for a cylindrical space colony has a cylinder four miles in diameter and 20 miles long, with three mirrors reflecting sunlight into the colony. To illuminate the whole colony, each would have to be at a $$45^\circ$$ angle to the cylinder axis and have a length of $$20 \sqrt{2}$$ miles. The outermost edge would be experiencing about a 10g acceleration, assuming 1g at the cylinder wall and a 2 mile radius for the cylinder.

Can we engineer a rigid 2 mile x 28 mile mirror able to take those forces? Is such a thing even remotely credible?

This is perhaps the canonical artist's conception of an O'Neill cylinder pair. The mirrors don't look to be at a $$45^\circ$$ angle, but in reality they would have to be or else part of the interior would be in shadow.

• Is there a diagram or picture or better way to understand the orientation of the mirrors? perhaps these or these or these? – uhoh Nov 9 '18 at 6:18
• mi = mile? I sure hope nobody uses miles anymore by the time we're ready to build those. Why 20$\sqrt 2$? – gerrit Nov 9 '18 at 10:12
• I'd prefer to see this posed as "Is this any more or less feasible than any other part of the O'Neill cylinder?" but that's just me. – Roger Nov 9 '18 at 14:45
• @gerrit - 20$\sqrt{2}$ was because of the 45 degree angle. And I bet the US will be using miles for a long time to come; anyway, those are the dimensions generally given for his proposal. – Mark Foskey Nov 9 '18 at 16:47
• @MarkFoskey I understand where the $\sqrt{2}$ comes from, but not where the 20 comes from. It's not too important. – gerrit Nov 9 '18 at 17:31

Remotely credible is had to quantify. We have never made any single structure even close to that big on Earth withstanding 1G of acceleration, making a 2mi by 28mi structure in space withstanding those forces is well outside existing capabilities. It's not just materials, it would be lifting huge structures into space or manufacturing them there that would require massive leaps in ability.

However, that doesn't mean it is impossible or we can't learn to make it. In 1920 the NY Times published an article saying that Robert Goddard's assertions on rocketry were "A severe strain on credulity", 49 years later they issued a retraction when NASA put people on the moon. It is credible that these structures could be built at some point in the future.

• re the retraction: 1, 2, 3, 4 and i.stack.imgur.com/oglWi.png – uhoh Nov 9 '18 at 10:26
• Interesting story! I'd never heard about that. – uhoh Nov 9 '18 at 10:43
• The uninformed media slammed him, really damaged his reputation and he never got the backing he deserved. – GdD Nov 9 '18 at 10:56
• Whoever wrote that NY Times article didn't understand conservation of momentum, pure and simple. The validity of Goddard's assertions were established in the late 1600s and should have been clear to many physicists by 1920. – WaterMolecule Nov 9 '18 at 17:56
• Absolutely @WaterMolecule, there were a lot of misinformed people in the press. Scientific understanding is more widespread now. – GdD Nov 9 '18 at 20:08

The mirrors will presumably be held in place by cables of some kind. So if we stick with the exact design shown, a 20 mile long habitat and mirrors at 45 degrees (all of which are things that could be varied) the longest cables will need to be strong enough to support the equivalent of 100 miles of their own weight plus 10 x the weight of their share of the mirror. Kevlar fibre has a yield strength of about $$3.6\, GPa$$ and a density of about $$1400\, kg/m^3$$ allowing it to support about 200km of its own weight at 1g. So it's just about strong enough. You then have to figure out the thickness and spacing of the cables based on the density and strength of the mirror material.

Of course there is no need for the mirrors to rotate at the same speed as the habitat. It would make much more sense to spin them much more slowly -- just fast enough for them to hold their shape. Then the whole problem goes away.

• @uhoh The question (and the first para of my answer) is premised on them rotating with the cylinder and needing an acceleration of up to 10g at the tips to keep them in their path. In the second para I'm suggesting that the mirrors would indeed not rotate, or rotate much more slowly (a little tension might be helpful), but that might create a flicker problem. – Steve Linton Nov 10 '18 at 21:13