Orbiting fusion reactors (was: Mirrors versus diffuse reflection)

[snikt]
> Creating a fusion reactor that is 20 miles long should be even easier
> conceptially. The main difficulty is the economic resources that need to
be
> mustered to build it. If you talk to a fusion scientist, I'm pretty sure
he
> can come up with a working design. Changes are that the fusion powered
Island
> three could even beam surplus fusion power to Earth if in orbit. Large
> structures are easier to build in orbit than on the ground as they require
> less support material.
>

In particular, in orbit you could implement fusion reactor designs that are
basically unviable on Earth. For example, one could build what I call a
"mass drive-powered fusion reactor". Instead of fighting to tame a plasma
cloud trapped inside a magnetic field (which "is like sustaining a candle
inside a tornado", quoting a Japanese fusion scientist whose name I can't
remember), you can simply build two very, very long (tens of kilometers,
perhaps) mass drivers, each one pointing to the output of the other, and use
them to accelerate tiny droplets of hydrogen (perhaps metallic hydrogen, or
liquid hydrogen wrapped in an iron sphere, in order to maximize magnetic
interaction) towards the center of the structure. There, the two droplets
would collide against each other at hundreds or thousands of kilometers per
second, and the heat and pressure created during the collision would fuse
the hydrogen, converting the droplets in a micro-nuke. Of course the tiny
explosions would occur inside a hollow sphere of tungsten or whatever, and
the sphere would become red-hot, and the heat could be used to produce steam
and move turbines, just like in a conventional atomic power plant.

>
> Tom Kalbfus
[snikt]

Lucio Coelho