NASA engineers put crewed Mars transit at 335 days. The hard part is where to test the reactor.

Put the surface stay back in and the arithmetic tidies up in a way the authors never state: 335 days in transit plus 30 days on Mars is a 365-day mission, against 650 for the baseline ^[2]. Kurt Polzin, chief engineer of NASA's space nuclear propulsion project at Marshall, and Robert Schleicher of General Atomics make that case as a systems argument rather than a performance one, arguing that a shorter transit simplifies engineering problems as well as protecting crew health ^[3][4]. The long version carries life support with no resupply, more elapsed time for accidents, and a cumulative dose from microgravity and radiation ^[5].

The physics underneath is old and not in dispute. Hydrogen burned with oxygen tops out near 450 seconds of specific impulse ^[6]. Hydrogen forced through a fission core and heated to at least 2,700 kelvin comes out at 900 seconds or better ^[7], because uranium releases far more energy per kilogram than combustion does and the exhaust temperature is therefore not set by chemistry ^[8]. Roughly a factor of two ^[3] is the entire reason a fast trajectory is worth drawing. The idea was recognised as early as 1946 ^[12].

What is load-bearing in the piece is its account of the last failure. Rover and NERVA got their data from open-air ground tests ^[9]. By the early 1970s the hardware was mature enough that flight tests were being planned, and cancellation arrived in 1973 on political and budgetary grounds ^[10]. Nothing in that account describes an engineering wall. It describes a test capability retired at the point where it had the most to prove, alongside a flight record that still consists of one reactor, SNAP-10A, launched in 1965 ^[11].

Two cautions on the 335 days. It is stated as an objective rather than a result: the authors write that they want to make it possible ^[2]. And one of the two is chief engineer for nuclear technologies and materials at General Atomics ^[3], which is not disqualifying but is worth reading with. The text supplied here also stops after introducing nuclear electric propulsion and the ion thrusters flown on Dawn ^[13], before any account of how the synchronal bimodal design produces the number ^[14]. Until that arrives, the figure is a target with a plausible mechanism behind it and no published trajectory.

On this evidence the reactor is the tractable half of the problem. The other half is institutional, and institutions are what closed the work in 1973 ^[10].