I had an interesting idea for increasing the Specific Impulse and Thrust of a nuclear rocket engine; it's probably not practical - best case scenario 20% increase in exhaust velocity, with proportional increase in thrust, at the cost of 1.5x to 2x weight increase, and likely much lower - but I still thought the concept was interesting enough to see what people here think. Even if it wouldn't work, I think I'd learn a great deal from knowing why. From my reading, the performance of a nuclear thermal rocket is strongly tied to it's operating temperature. It seems that making a nuclear reaction produce more power is trivial, it's getting rid of the heat you make that causes all of the problems. My idea was that instead of passing the reaction mass through the reactor once, you pass it through twice. Rockets already do this to an extent, cooling the rocket with a small amount of cryogenic fluids and passing the resulting hot gas to power the turbopump. In short, instead of just cooling the engine with some of the gas, we pass all of the fluid through the reactor, up to a turbine, and back down to the reactor to be reheated and ejected. Instead of the turbine just powering a pump though, it now powers a large generator as well. We can than use this electrical power to heat the reaction mass further after it passes through the reactor the second time. The only part that seems iffy to me is electrically heating the gas after the second pass. The other stuff would likely have practical issues, but seems theoretically sound to me. As far as the amount of power one could potentially add vs a normal nuclear thermal rocket, I think it's capped at 50%, and adding in other inefficiencies would lower that further. I arrived that that by saying that all of the power that the reactor produces needs to be cooled by the incoming fuel, be extracted by the turbine or show up in the exhaust after the second pass, and creating a relation with Carnot's theorem to determine the temperature after the turbine and before the second pass. Not totally sure my equation is valid though. Any thoughts? Is this theoretically sound, or am I missing something?