Nuclear Rocketry, Anyone?

One of the deal-breakers with fission-based rocketry was the need to cool the reactor with the propellant fluid, which, first, made the whole scheme subject to the fundamental Carnot limitations of heat engines and, second, put huge stresses on the reactor itself, since it has to withstand the vibration generated by the working fluid flowing through it.

This approach is an attempt to build a "dusty plasma" reactor, where the fission products themselves can either become the reaction mass for the rocket, or can be decelerated in an electric field to directly generate electricity for some kind of ion propulsion. Using the fission products as reaction mass could yield a specific impulse of about 1,000,000 seconds (that's about a factor of 100 better than the best ion engines, BTW).

On the downside, this rocket works by emitting nuclear waste at 3-4% of the speed of light. This qualifies it as the second-dirtiest rocket ever envisioned. (Don't forget Project Orion!)