NERVA

• 1. Introduction

  ›

  NERVA's most advanced engine was certified for a human mission to Mars at the end of 1968. NERVA, the Nuclear Engine for Rocket Vehicle Application, was a joint project of the Atomic Energy Commission and NASA. Barely four years later, the program was gone. The engines never flew. Its official objective, as the program's own documents stated, was to establish a technology base for nuclear rocket engine systems for space mission application. The question of how a program certified for Mars came to be cancelled without a single flight has no tidy answer. It involves Cold War politics, tight budgets, and a test program that stretched from Eisenhower's presidency to Nixon's.

  52573437

• 2. Nuclear Ideas Take Root

  ›

  Stan Ulam and C. J. Everett wrote a 1946 paper on the use of atomic bombs as a means of rocket propulsion. Both had worked at the Manhattan Project's Los Alamos Laboratory, where the first atomic bombs were designed. Ulam's colleagues Frederick Reines and Frederic de Hoffmann had speculated along the same lines. Their bomb-propulsion concept would eventually feed into Project Orion, a separate program. The same question was also being pursued independently in the United Kingdom.

  Val Cleaver was the chief engineer of De Havilland's rocket division. He and Leslie Shepherd, a nuclear physicist at the University of Cambridge, independently arrived at the problem of nuclear rocket propulsion and became collaborators. Their papers appeared in the Journal of the British Interplanetary Society in 1948 and 1949, outlining a rocket with a solid-core graphite heat exchanger. They concluded that nuclear thermal rockets were essential for deep space exploration but not yet technically feasible.

  In 1953, Robert W. Bussard, a physicist at the Oak Ridge National Laboratory, produced a detailed study on nuclear energy for rocket propulsion. He had drawn on Cleaver and Shepherd's work, the research of the Chinese physicist Hsue-Shen Tsien, and a February 1952 report from Consolidated Vultee. Only 29 copies were printed, and it was classified as Restricted Data. Published in Oak Ridge's Journal of Reactor Science and Technology in December 1953, still classified, it reached key figures including Darol Froman at Los Alamos and Herbert York at Livermore. Both established committees to investigate nuclear rocket propulsion, and Froman brought Bussard to Los Alamos for one week each month.

  John von Neumann formed an ad hoc committee on nuclear propulsion of missiles. Mark Mills of Livermore chaired it. Its members included Norris Bradbury from Los Alamos, Edward Teller, Abe Silverstein of the aeronautics agency NACA, and Allen F. Donovan of Ramo-Wooldridge. In March 1955, the committee recommended development proceed toward a nuclear rocket upper stage for ballistic missiles. The Armed Forces Special Weapons Project followed in March 1956, recommending $100 million over three years for feasibility studies. All work was consolidated at Los Alamos under the codename Project Rover. Senator Clinton P. Anderson of New Mexico secured funding in January 1957. On the 2nd of October 1957, the AEC proposed cutting Project Rover's budget. Two days later, the Soviet Union launched Sputnik 1.

  12303435

• 3. Building the Program

  ›

  Eisenhower created the National Aeronautics and Space Administration in 1958 to oversee civilian space development. NASA absorbed NACA as part of its formation. Abe Silverstein, who had served on the Mills committee, now headed NASA's Lewis Flight Propulsion Laboratory in Ohio. He was particularly eager to bring Project Rover under NASA's direction. The non-nuclear components of the program were officially transferred to NASA on the 1st of October 1958, the day NASA itself became operational.

  On the 29th of August 1960, NASA created the Space Nuclear Propulsion Office, known as SNPO, to manage the joint program. Harold Finger was appointed as its head. Milton Klein from the Atomic Energy Commission served as his deputy. SNPO headquarters was co-located with AEC headquarters in Germantown, Maryland. A formal management agreement was signed by NASA Deputy Administrator Robert Seamans and AEC General Manager Alvin Luedecke on the 1st of February 1961.

  Finger called for bids to develop a nuclear engine based on Los Alamos's Kiwi reactor. Eight companies submitted proposals: Aerojet, Douglas, Glenn L. Martin, Lockheed, North American, Rocketdyne, Thiokol, and Westinghouse. A joint NASA-AEC evaluation board rated North American's bid as the best overall. Westinghouse had the strongest reactor proposal, however, and Aerojet the strongest engine proposal when assessed separately. Aerojet promised NASA administrator James E. Webb it would put its best people on the project. Webb then told the selection board that North American was deeply committed to Apollo and the board might consider combining other bids. On the 8th of June, Webb announced that Aerojet and Westinghouse had been selected. By 1963, Westinghouse alone had 1,100 staff working on NERVA.

  On the 12th of April 1961, the Soviet Union launched Yuri Gagarin into orbit on Vostok 1. A few days after that, Kennedy launched the Bay of Pigs Invasion of Cuba, which ended in failure. On the 25th of May, Kennedy addressed a joint session of Congress. He committed to landing a man on the Moon before the decade was out. He also told Congress that 'an additional 23 million dollars, together with 7 million dollars already available, will accelerate development of the Rover nuclear rocket.' Nuclear propulsion, he said, offered promise of exploration 'perhaps beyond the moon, perhaps to the very end of the Solar System itself.'

  What those engineers were being asked to build demanded new materials, new sensors, and approaches to radiation shielding that had never been attempted before.

  537344652

• 4. The Physics of the Engine

  ›

  Hydrogen was chosen as NERVA's propellant for a specific reason: its low molecular mass maximizes the thrust produced for a given heat input. In a chemical rocket, the fuel provides both the energy to heat the working mass and the working mass itself. A nuclear reactor separates these two functions. The reactor provides the heat; the propellant can then be chosen purely for performance. Nuclear thermal engines generally aim for at least twice the specific impulse of a chemical engine.

  Plutonium-239 was considered first as a reactor fuel but rejected: it forms compounds easily and cannot reach the high temperatures that uranium can. Uranium-233 was also evaluated; it is slightly lighter than uranium-235, releases more neutrons per fission event on average, and has a higher probability of fission. Its radioactive properties make it harder to handle, and it was not readily available. Uranium-235 was therefore the choice.

  Tungsten emerged as the leading metal candidate for the reactor's structure, but it was expensive, hard to fabricate, and had undesirable neutronic properties. Graphite offered a compelling alternative: it is cheap, gets stronger as temperatures rise to 3,300 K, and sublimes rather than melts at around 3,900 K. Graphite was chosen. The reactor was controlled by a ring of drums around the core, each coated with beryllium on one side and boron on the other. Rotating the drums adjusted power output: beryllium facing inward reflected neutrons back to sustain the reaction, boron facing inward absorbed them to suppress it.

  The Aerojet Nuclear Systems Company developed NERVA's radiation shielding material, known as BATH, an acronym for its components: boron carbide, aluminum, and titanium hydride. Titanium hydride moderated neutrons; boron carbide absorbed them. The three were mixed in powdered form and extruded into shape by a commercial machine. BATH proved strong, with a tensile strength up to 28,000 psi, resistant to high temperatures, and effective as a radiation shield. The Los Alamos design team worked through concept names including Uncle Tom, Uncle Tung, Bloodhound, and Shish. By 1955, they had settled on a 1,500 MW design called Old Black Joe. In 1956, that design was scaled up to 2,700 MW as the intended upper stage of an ICBM.

  All of these design decisions would eventually be tested at Jackass Flats in the Nevada Test Site. Even the railroad connecting the test stands had been built to handle nuclear cargo.

  555133563

• 5. The Nevada Test Ground

  ›

  Reactors for Project Rover were built at Los Alamos Technical Area 18, also known as the Pajarito Site. Each was first tested at very low power before being shipped to Jackass Flats in the Nevada Test Site. Construction at Jackass Flats began in the middle of 1957. All materials and supplies had to be brought in from Las Vegas. Test Cell A was built around a farm of hydrogen gas bottles, with a concrete wall three feet thick to protect the electronic instrumentation from radiation. The control room was located two miles away.

  The reactor maintenance and disassembly building, designated R-MAD, was 250 feet long, 140 feet wide, and 63 feet high. It was large enough for engines to be moved in and out on a railroad car. That railroad connected R-MAD to the test stands and had two locomotives. The remotely controlled electric L-1 was the primary vehicle; the diesel-electric L-2 served as a backup and had radiation shielding around the cab. Construction workers were initially housed in Mercury, Nevada. Thirty mobile homes were later brought to the site and arranged into a settlement named Boyerville, after the site supervisor Keith Boyer. Construction was complete by the fall of 1958. NASA planned to expand the area eventually into a community of 2,700 people with 800 dwellings.

  Test Cell C was added as the program grew, equipped with two giant cryogenic storage dewars. Each held 500,000 gallons, supplied by the Chicago Bridge and Iron Company. A test stand designated ETS-1 allowed an engine to fire downward into a reduced-pressure chamber, partly simulating the vacuum of space. It was constructed from pure aluminum, which does not become radioactive when irradiated by neutrons. The structural work alone required 120,000 pounds of steel, 8,700 pounds of welding wire, and 6.5 miles of welds. During a test, 234 tubes would carry up to three million gallons of water. To cut cabling costs, Aerojet moved the ETS-1 control room to a bunker 800 feet away.

  A minor accident caused a violent hydrogen release at Jackass Flats on the 7th of November 1961, and the complex did not become fully operational until 1964.

  2750544438

• 6. Running the Reactors

  ›

  The first phase of Project Rover's engine tests was called Kiwi, named after the New Zealand bird that cannot fly. The Kiwi engines were not intended to fly either: their purpose was to verify the design and test the materials. In a series of tests between July 1959 and October 1960, three Kiwi A reactors were built and tested. They proved that hydrogen could be heated in a nuclear reactor to the temperatures required for space propulsion and that the reactor could be controlled. Kiwi A was regarded as a successful proof of concept.

  Kiwi B1A fired on the 7th of December 1961. The next test, Kiwi B1B on the 1st of September 1962, caused extreme structural damage. Fuel module components were ejected as the reactor was brought up to full power. A full-power test, Kiwi B4A on the 30th of November 1962, along with cold flow tests, identified the cause. The problem was vibrations produced when hydrogen heated up during its passage through the core, not when the reactor was running at full power. Unlike a chemical engine, which would likely have exploded after such damage, the nuclear rocket remained stable and controllable even in these conditions.

  President Kennedy visited Los Alamos on the 7th of December 1962 for a briefing on Project Rover. It was the first time a sitting president had visited a nuclear weapons laboratory. The following day, the party flew to Jackass Flats, making Kennedy the only president to visit a nuclear test site. The program had received $187 million in 1962, and AEC and NASA were asking for another $360 million in 1963. Harold Finger declared there would be no quick fix for the vibration problem. He assigned three SNPO staff to Los Alamos to enforce new management practices and assembled vibration specialists from NASA centers, Los Alamos, Aerojet, and Westinghouse. The in-flight test program, known as RIFT, was cancelled in December 1963.

  Kiwi B4D, tested on the 13th of May 1964, was started and briefly run at full power with no vibration problems. Kiwi B4E, on the 28th of August, ran for twelve minutes, eight of them at full power. On the 10th of September, Kiwi B4E was restarted and run at full power for two and a half minutes. This was the first demonstration that a nuclear rocket engine could be shut down and then restarted. Clustering tests in September placed two Kiwi B4 reactors at distances of 16 feet, 9 feet, and then 6 feet apart. Neutron exchange between the reactors was negligible, measured as 3, 12, and 24 cents of reactivity respectively. Nuclear engines could be clustered just as chemical ones often are.

  SNPO selected the Kiwi-B4 design, with 75,000 pounds of force and a specific impulse of 825 seconds, as the baseline for the NERVA NRX. Unlike the Kiwi engines, the NRX required a complete flight-type configuration: actuators, gimbals, a hydrogen-cooled nozzle, and radiation shielding. New temperature sensors had to be developed; the available ones were accurate only to 1,710 degrees Celsius, far below the reactor's operating range. The new sensors were accurate to 2,376 degrees Celsius, even in a high-radiation environment. The first test, NERVA A2, took place on the 24th of September 1964. Aerojet and Westinghouse increased power in stages, through 2 MW, 570 MW, and 940 MW, before reaching full power at 1,096 MW. The engine had to be shut down after 40 seconds only because the hydrogen supply was exhausted. The confirmed specific impulse was 811 isp. Solid-propellant rockets achieve a maximum of around 300 isp; liquid-chemical rockets seldom reach more than 450 isp. Aerojet and Westinghouse took out a full-page advertisement in the Wall Street Journal, captioned: 'On to Mars!'

  NERVA A3, on the 23rd of April 1965, was intended to run and restart at full power. A coolant line clogged during a planned shutdown, and power briefly rose to 1,165 MW before the line cleared. Tie rods designed to operate below 378 degrees Celsius may have reached 1,100 degrees. The engine ran on undamaged for thirteen minutes at 1,072 MW. Testing of the NRX/EST, the Engine System Test, began on the 3rd of February 1966. The engine was started eleven times and ran for nearly two hours in total, including 28 minutes at full power. That exceeded the operating time of all previous Kiwi reactors by nearly a factor of two.

  A bird landed on the nozzle of NRX A5 during cooldown on the 8th of June 1966. It was asphyxiated by the nitrogen or helium gas purging the system. Westinghouse engineers rigged a television camera and a vacuum hose to remove it while remaining safely behind a concrete wall. The engine was restarted on the 23rd of June and run at full power for another fourteen and a half minutes. NRX A6, started on the 15th of December 1966, ran at full power of 1,125 MW with a chamber temperature above 2,000 degrees Celsius. It took 75.3 hours to cool the reactor with liquid nitrogen. On examination, the beryllium reflector had cracked from thermal stress. This led to the abandonment of plans for a more powerful NERVA II engine.

  With the A6 results in hand, SNPO cancelled the planned follow-on tests and turned its attention to completing ETS-1. All previous NERVA tests had the engine firing upward; ETS-1 allowed it to fire downward into a reduced-pressure chamber, partly simulating space. The final test engine, the NERVA XE Prime, was 6.9 meters long, 2.59 meters in diameter, and weighed approximately 18,144 kilograms. Its design thrust was 246,663 Newtons. A series of experiments ran between the 4th of December 1968 and the 11th of September 1969. The reactor was started 24 times and ran at full power for a total of 1,680 seconds.

  NASA and SNPO concluded that the XE tests confirmed a nuclear rocket engine was suitable for space flight. The question of whether it would ever fly was now entirely a matter of politics.

  3852574439

• 7. The Budget Battlefield

  ›

  In 1966, NASA's plans for NERVA included a visit to Mars by 1978 and a permanent lunar base by 1981. Probes to Jupiter, Saturn, and the outer planets were also envisioned, along with nuclear tugs to ferry payloads between Earth orbits. Congress defunded NERVA II in the 1967 budget. Johnson needed Anderson's support for his Medicare legislation; on the 7th of February 1967, he provided the funds for NERVA II from his own contingency budget. Defunding NERVA II ultimately saved about $400 million, mostly in new test facilities. AEC and NASA accepted this because the NRX A6 results showed NERVA I could perform all the missions planned for NERVA II.

  NASA also considered using NERVA and Saturn V on a Grand Tour of the Solar System. A rare planetary alignment occurs only every 174 years. One such alignment fell between 1976 and 1980, allowing a spacecraft to visit Jupiter, Saturn, Uranus, and Neptune in sequence. With NERVA, such a spacecraft could weigh up to 52,000 pounds. At a projected specific impulse of 900 isp, NERVA could place a 170,000-pound space station the size of Skylab into orbit around the Moon. Richard Nixon replaced Johnson as president on the 20th of January 1969, and cost cutting became central to his administration. NASA received $88 million for NERVA in fiscal year 1970 and $85 million in fiscal year 1971.

  In December 1970, the Office of Management and Budget recommended cancelling NERVA and Skylab. Nixon was reluctant. He calculated that cancellation could cost up to 20,000 jobs, mostly in California, a state he felt he needed in the 1972 election. He kept both programs alive at reduced funding. When Nixon tried to kill NERVA in 1971, Anderson and Smith retaliated by killing his pet project. They cancelled the Boeing 2707 supersonic transport, a significant defeat for the president.

  In fiscal year 1972, Congress allocated $69 million for NERVA even though the budget request had been only $17.4 million. Nixon spent just $29 million of it. In 1972, a bipartisan coalition led by Margaret Chase Smith and Howard Cannon appropriated $100 million for a smaller NERVA engine. It was designed to fit inside the Space Shuttle's cargo bay, with an estimated cost of about $250 million over a decade. The coalition stipulated that no further reprogramming of NERVA funds for other NASA activities would be permitted. The Nixon administration decided to cancel the program regardless. On the 5th of January 1973, NASA announced that NERVA was terminated. Staff at Los Alamos and SNPO were stunned; the small engine project had been going well. Layoffs began immediately, and the SNPO was abolished in June. After seventeen years of development, Projects Nova and NERVA combined had spent about $1.4 billion.

  In 1983, the Strategic Defense Initiative identified missions that could only be performed by propulsion more powerful than any chemical engine. A nuclear propulsion project, designated SP-100, was created in February 1983. The questions it raised were ones that engineers at Jackass Flats had spent nearly two decades trying to answer.

  4714415234