In article .com,
Allen Thomson wrote:
Supposing you wanted a reactive propulsion system -- say a
nuclear rocket, electrical thruster, mass driver or other
throw-mass-out-the-back widget that had continuous thrust =
1 kN, ISP = 10,000 sec (= 100 km/sec Ve) and a total run
time = 1 megasecond or greater...

Mmm... Even at 100% efficiency, this requires 50MW of continuous power,
which will not be easily had. (Jet power is 0.5*thrust*exhaustvelocity.)

That kind of exhaust velocity is almost certainly impractical with a mass
driver or anything similar.

Most electric thrusters are out too. Ion thrusters can get up into that
range easily enough, although they are more usually optimized for lower
Isp; you will need a whole bunch of them to get 1kN. The power source and
thruster hardware will be quite heavy.

There is no way that solid-core nuclear can do 100km/s. Even gas-core
probably tops out around 50km/s. 100km/s or more should be feasible with
systems that don't try to separate fission fuel and propellant -- NSWR or
imploded-pellet fission, for example -- but operating costs will be high
and the exhaust generally rather dirty.

Fusion or antimatter rockets can deliver the performance, but fusion is
beyond what we can build now, and antimatter is beyond what we can fuel
now (the hardware technology is rather easier than fusion [!!], but
large-scale antimatter production would require massive infrastructure).

Does current technology or anything that can be reasonably
foreseen in the next 20 years support anything like that? If
so, what might it be?

Current technology can stack up an array of ion thrusters, but powering
them is problematic. That is much too big for current-technology solar;
the problem is not the solar cells but the structural dynamics of enormous
lightweight solar arrays. 100MW space reactors are not off-the-shelf
items either. Solving either in the next 20 years is conceivable, but not
a small project, and probably will not happen without specific need.

100km/s dirty-fission or antimatter systems are probably buildable without
major new technology, but again, it would be a big project and is unlikely
to happen within 20 years unless someone makes a major effort to do it.
--
"Think outside the box -- the box isn't our friend." | Henry Spencer
-- George Herbert |