The first people to step on to the surface of Mars won’t arrive aboard the chemical-fueled rockets that delivered Apollo 11 to the moon — they simply don’t provide enough thrust to get to the Red Planet before exposing their crews to months of dangerous space radiation. Instead, NASA is turning to long-ignored nuclear-thermal rocket technology to deliver the first Martian explorers into history.
How NASA Built a Better Rocket
Nuclear-thermal rockets are by no means a new technology. In fact, researchers began discussing the prospect of utilising nuclear power to propel rockets and aircraft back in 1942 after Enrico Firmi’s successful fission reactor tests. By 1944 teams at the University of Chicago’s Metallurgical Laboratory and Los Alamos National Lab developed an early nuclear-thermal design that used a fission reactor to super-heat hydrogen gas which would then escape through a small nozzle to generate thrust. Since nuclear fuel is about 107 times more energy dense than their chemical counterparts and similarly powerful rockets would weigh only about half as much, nuclear-thermal rockets (NTRs) can carry load to fuel ratios from 1:1 to as high as 7:1 especially when used as the upper stage. This design piqued the interest of the US Air Force, which conducted highly-classified testing at Oak Ridge National Labs between 1947 and 1949.
Developments in this technology went quiet for a few years before Los Alamos National Labs began development of a nuclear-powered ICBM in 1955 and expanded into nuclear-powered ramjet engines in 1956. But by 1957, the USAF had determined the technology unsuitable for military deployment, instead recommending the non-nuclear aspects of the R&D, dubbed Project Rover, be transferred to a newly-formed National Aeronautics and Space Administration.
Project Rover ran from 1955 until its cancellation in 1972, however, by 1961 the program had developed so quickly and performed so successfully that Marshall Space Flight centre began making noise about being allowed to use one on a RIFT (Reactor-In-Flight Test) by 1964, a key development milestone that would see the construction and launch of a final stage prototype. In response, the Space Nuclear Propulsion Office was formed in 1961 to perform oversight and planning operations as well as facilitate cooperation between NASA, which focused on flight systems and engine design, and the Atomic Energy Commission, which developed the reactor technology. The SNPO’s first director, H.B. “Harry” Finger, was having none of that RIFT nonsense, however, and delayed the launch, demanding a stringent set of performance metrics be met before the engines ever got off the ground.
NASA’s side, Project NERVA (Nuclear Engine for Rocket Vehicle Applications), was specifically tasked with creating a space-worthy, mission-deployable, and — most importantly — real nuclear-thermal star ship engine. The more than 20 distinct rocket designs produced in this 17 year span constituted numerous design phases: the Kiwi, an aptly-named early design never intended for flight developed between 1955 and 1964; the larger, intermediate Phoebus design developed from 1964 to 1969; and the Pewee, which ran from 1970 to 1971 until they were replaced by the nuclear furnace design. The NRX (Nuclear Rocket Experimental) series saw concurrent develop with the Phoebus and Pewee platforms from 1964 to 1968.
A pair of nuclear reactors for each model were built at Los Alamos’ Pajarito Site — one for Los Alamos Lab’s zero-power critical experiments, a state wherein the reactor is undergoing a sustained fission reaction at temperatures low enough to generate insignificant thermal effects and one for full-power testing at the more remote Nevada Test Site (helpful when test engines exploded and sprayed nuclear material all over the place). Los Alamos’s super-secret Sigma complex handled the production of plutonium-238, a non-fissable cousin of the Plutonium-239 used in the nuclear bombs dropped on Japan.
The earliest iterations of the KIWI model test fired for the first time in mid-1959. Comprised of a stack of uncoated uranium oxide plates doused in liquid hydrogen, it was an engine by only the loosest of definitions but did produce an impressive 70 MW of electricity and generated 2683-degree Kelvin exhaust. The second iteration, the KIWI B, swapped the plates of uranium for tiny balls made of uranium dioxide, suspended in a graphite matrix, and coated with niobium carbide. Liquid hydrogen flowed through these bundles to generate exhaust. In addition to electricity and thrust, the early KIWI designs exhibited a couple of design flaws that were never fully resolved by the program’s end. For one, they rattled and vibrated — a lot. Enough to crack the fuel bundles, rendering them useless. It also became so hot that the super-heated hydrogen steam eroded the walls of the reactor.
The danger of a catastrophic failure known as a containment breach was very real during testing. These failures — caused by the orbiter impacting the ground, fission runaway, or design flaws — in either the atmosphere or orbit could rain down radiation over a huge swath of land. So in 1965, researchers purposely exploded a KIWI reactor in the middle of Jackass Flats, part of the Nevada Test Site. The resulting explosion dumped enough fallout to kill everything within 180m and poison everything within 600m. The amount of fallout depends on the format of fuel the engine runs on (discussed below) with solid fuel rods and spheres entombed in carbon matrices spreading far less radiation than their gaseous or liquid counterparts.
After five years of developing the KIWI, NASA moved on to a much bigger engine, the Phoebus series. The rocket’s initial test run in 1959 produced 1064MW of power and 2000C exhaust over its 10-minute run. Those stats jumped to 1500MW over 30 minutes in 1967 and earned the title of “the most powerful nuclear reactor ever built” when the Phoebus-2A Project Rover engine dumped a staggering 4000MW electrical load in a mere 12 minutes. That’s 4GW of power, equal to the total production capacity of Chernobyl — enough to power three million homes — generated in less than a quarter of an hour.
On the other end of the power scale were the stout 500MW Pewee models based on the original KIWI design. They were created to test a new zirconium carbide coating to replace the original niobium carbide. They were also used as the basis for modern 11,000kg/f nuclear-thermal rocket (NTR) designs known as NERVA-Derivative Rockets (NDRs). The Pewee 2’s core design further reduced corrosion caused by the fuel by a factor of three. During this time NASA also tested a distinct rocket design cooled by water, known as the NF-1 (Nuclear Furnace).
Another offshoot of the original KIWI designs was the NERVA NRX (Nuclear Rocket, Experimental), which began testing in 1964 and evolved into the NERVA NRX/XE — a nuclear rocket tantalyzingly close to flight readiness. The SNPO tested the XE engine a total of 28 times in 1968, firing it downward into a low pressure chamber to crudely mimic the effects of space’s vacuum. Every test, the engine generated over 1100MW of energy as well as 34,000kg/f (334kN) of thrust — the baseline output the SNPO had demanded Marshall meet before authorising a RIFT launch as well as the amount the agency needed to effectively get astronauts to Mars. During testing, the engine ran for over two hours in total, 28 minutes of which at full power, and typically only stopped when they burned through all 17kg of fuel.
This success, in combination with Los Alamos resolving three niggling materials issues, spawned a whole menagerie of potential uses for the new nuclear-powered rockets. Some wanted them to replace the J-2 boosters used of the second and fourth stages of Saturn I and IV. Others wanted them utilised as “space tugs”, towing objects from LEO to the upper orbital tracts, the moon and further. Unfortunately, none of these ideas ever got off the drawing board because the entire project was cancelled at the end on 1972, effective in the second quarter of 1973.
America had already put a man on the moon at by that point, the Apollo era was quickly transitioning into the Shuttle era, public opinion was beginning to sway against nuclear technology, and US Congress, quite frankly, had lost its nerve when faced with what the effects of financing a manned mission to Mars would have on the national budget. And without a mission to Mars, there really was no more reason for NTR development. So despite having met all but two of the required flight metrics — restart 60 times and run for a total of 10 hours — the Rover/NERVA project was shelved.
How Nuclear Thermal Rockets Work
All of the Rover/NERVA rockets ran on Plutonium-238, a non-fissible isotope with a half-life of 88 years. With such a short half-life and the relative difficulty of separating the specific isotopes from the clumps of naturally occurring plutonium, Pu-238 is typically synthesised using the same method originally employed by Berkely Lab researchers Glenn T. Seaborg and Edwin McMillan in 1940 — bombard a sample of Uranium-238 with deuterons.
Plutonium-238 is a valuable commodity for deep space exploration where insufficient amounts of sunlight render solar panels useless. NASA’s radioisotope thermoelectric generators (RTG) that most of power these missions instead run on a nugget of Pu-238. While plutonium is a poor conductor of electricity, its emission of alpha particles as part of its decay process generates a terrific amount of heat to run the RTGs. The famous Voyager probes, the Cassini spacecraft, the Curiosity Rover, and the New Horizon’s probe all rely on nuclear power for their continued operation.
Solid Core
The simplest core design uses a solid fuel (like the plates and pellets that powered the KIWI and Phoebus) to superheat the hydrogen working fluid. The amount of heat, and thereby thrust, that this design can produce ranges from -250C to over 2700C and is really only limited by the melting point of the reactor components around it. Working with liquid hydrogen propellants, a solid core can produce a specific impulse of 850 to 1000 seconds — double that of the Space Shuttles main engine.
Liquid Core
If, instead of entombing the nuclear fuel into graphite matrices, one were to mix the fuel pellets directly into the working fuel itself, the resulting liquid-core engine would be able to generate temperatures beyond the melting point of the nuclear fuel — theoretically at least. No one’s even been able to build one yet. Trapping the radioactive fuel in the engine while allowing the working fluid to exit is proving quite difficult, however rotating designs similar to terrestrial “pebble bed” reactors that use centripetal force to separate the two have shown a great deal of promise.
Gas Core
A Gas Core reactor is even more difficult than a liquid, requiring a spinning pocket of uranium gas surrounded by hydrogen vapour. Since the fuel wouldn’t ever come in contact with the heat-sucking core chamber walls, it should become intensely hot (on the magnitued of several tens of thousands of degrees K) and produce 30kN to 50 kN over 3000 to 5000 seconds.
NASA Goes Back to the Future
After a multi-decade hiatus, both NASA and the Russian Federal Space Agency (which developed many of its own NTRs during the Cold War but never physically tested their designs) announced in April 2012 that they would be revival of nuclear-engine powered rocket technology and coordinating a new $US600 million joint engine project along with potential involvement from France, Britain, Germany, China and Japan.
Marshall Space Flight centre is also forging ahead on its own Nuclear Cryogenic Propulsion Stage as part of the upcoming Space Launch System. This upper stage would be super-chilled by its supply of liquid-hydrogen fuel and be unable to initiate a fission reaction until safely out of the atmosphere. However, since above-ground nuclear testing has been universally banned since the last time NASA tinkered with NTRs, researchers are instead using Marshall’s Nuclear Thermal Rocket Element Environmental Simulator (NTREES). This model can accurately simulate the interactions between various components of an NTR engine, allowing rocket scientists to tweak design and engineering aspects without the risk of spreading nuclear fallout.
“The information we gain using this test facility will permit engineers to design rugged, efficient fuel elements and nuclear propulsion systems,” NASA researcher and Manager of the NTREES facility, Bill Emrich, said. “It’s our hope that it will enable us to develop a reliable, cost-effective nuclear rocket engine in the not-too-distant future.”
Outside of developing revolutionary engine technologies, NASA is also facing something of a fuel shortage. See, America hasn’t produced plutonium-238 since the 1980s and we’ve been raiding our existing stocks pretty regularly for powering RTGs. Some estimates figure we’ll deplete the entire supply by the end of this decade.
Which is why NASA isn’t taking any chances and has announced that the DoE will once again manufacture Pu-238 starting in 2017. “We have turned the spade in starting the project for renewed plutonium production,” said Wade Carroll, deputy director of space and defence power systems at the DoE, said during last March’s Nuclear and Emerging Technologies for Space (NETS) conference. “It’ll take probably five or six years before the next new plutonium is available.”
The DoE plans to produce a total of 1.3kg to 1.8kg of the isotope annually, enough to satisfy our robotic planetary science missions. All we need now is an interplanetary spaceship. No problemo.
[Wikipedia 1, 2, 3, Popular Science, Space Travel, Science Daily, Los Alamos National Lab, EPA, Mother News Network, Idaho National Lab, Today I Found Out]