HUNTSVILLE, Ala. — NASA’s Marshall Space Flight Center here is expanding the scope of its nuclear technology work to encompass the Obama administration’s shift away from the canceled Constellation back-to-the-Moon effort toward a more open-ended technology development portfolio.

During the Constellation years, Marshall worked with the Department of Energy on nuclear-power technology that might one day power a lunar outpost. While Los Alamos and other national labs handled the radioactive material, NASA experts here used heating elements to simulate nuclear fuel and concentrated on the power systems that would generate electricity on the Moon.

That work continues, but it has expanded to encompass another technology goal under the new Obama policy: advanced in-space propulsion. In a nondescript high-bay building, the power-plant team has installed a nuclear-thermal rocket environmental simulator, which flows gaseous hydrogen over heating elements that mimic different nuclear-fuel configurations. The idea is to test the way different materials react with the hydrogen at high temperature and pressure.

“Before you move into any nuclear testing, you have a good feel for how those elements might behave,” says project engineer Dave Houts.

The setup includes a mass spectrometer and optical pyrometers that monitor temperatures and materials performance during runs. Testbed operators retreat to a control center away from the pressure chamber during tests, but only to protect their ears from the noise they generate.

Although NASA conducted nuclear-rocket tests in the 1960s, none will be run here. The prescreening of simulated nuclear elements could help humans use radioactive fuel to reach Mars and other distant destinations faster, reducing the time they spend in the dangerous space-radiation environment.

“Because you can use hydrogen as a propellant, which has a very low molecular weight, a nuclear thermal rocket allows you to get a very high specific impulse even at reasonable material temperatures,” Houts says. “So for what we consider early systems, a first-generation system, we should be able to achieve a 900-sec. specific impulse or better, so you’re roughly twice that of a chemical engine. We think we can go up from there using some of the advanced materials, advanced cycles and advanced geometries in the actual system.”