Here is the scheme: A craft from earth slingshots around the moon back towards the earth, gaining some speed (and slowing the moons orbit slightly). a few days later it expends this energy skimming the earths atmosphere, scooping some atmosphere into its tanks. It heads back towards the moon with the same speed it had initially, but more mass. (interestingly, its efficiency increases rather than decreases each time it gathers more mass, because more mass lets it grab more momentum from the moon with each slingshot allowing it to skim more earth atmosphere) Can anyone comment on the validity of this idea? Can slingshotting like this work? I initially thought it would add exactly the moon's orbital velocity of about 1km/s but Im no longer sure its that simple. Can aerobreaking maneuvers be designed that create the correct reduction in velocity AND the correct trajectory back towards the moon, especially given that the mass will probably be different each time? If it holds up theoretically, what about practically? eg how strenuous would the aerobraking be? What would the atmospheric pressure in the scoop be?.. (we would need to choose a width of the scoop to answer that of course, but I dont think details of the earth's atmosphere are required. Instead we are asking how many newtons to lose how much velocity in a semicircular low earth orbit.. I think) If it holds up theoretically, has it already been investigated? any reference? Im not expecting someone to solve all my problems for me (that is well beyond the scope of this thread ;) ) but any pointers would be appreciated.