metal or ceramic foam

>>>The diameter of the mass driver is sized to make its emplacement a
realistic, near-term possibility. And it's important to remember that
there's a continuous flow of materials (well, maybe for 2 weeks out of every
four). It's a pretty narrow pipeline, granted, but you might be surprised
how many tons you could get through it over the course of five years or so.>>>

The amount of power that is required in order to operate a mass driver (albeit
for a fraction of a second), is fairly large.

v = u + at
t = v/a

If v = 2370m/s (lunar escape-v)
a = 20,000
t = ?
u = 0

therefore, t= 2370/20,000 = 0.1185sec

spec energy = (v**2)/2 = 2.808450 Mj/Kg

spec power of mass driver = spec energy/ acceleration time =
2.808Mj/0.1185sec

spec power= 23.7 Mw/kg

This amount of power could be provided by a modest sized nuclear reactor,
delivered to the moon in several packages. Alternatively, a mass driver used to
accelerate 30 gram payloads, would consume 710 Kw of electric power. This is
puts it within the power range of the NASA SP-100 space nuclear reactor. Fuel
cell type power storage systems are not practical on the moon. They are both
heavy and relatively complex and would have to be imported from Earth at
enormous cost. Mass driver systems are therefor practical near-term mechanisms,
capable of launching large numbers of very small payloads onto lunar escape
trajectories. Unless enormous amounts of power or extremely long mass drivers
are available, they are not likely to be used to export single, large
manufactured payloads from the moon. If the same mass driver were used to
accelerate a 1 tonne payload, some 23.7Gw of electric power would be required.
This is equivalent to the output of 20 large, PWR nuclear reactors.

Tony