Aiming for the Stars: The Forgotten Legacy of the Westinghouse Astronuclear Laboratory

Before his death, renowned science fiction writer, inventor, and futurist Arthur C. Clarke (1917–2008) confidently declared the space age had not yet begun, and would only commence when reliable nuclear-powered space vehicles become available to drastically reduce the cost of moving humans and heavy payloads from the surface of the earth to the farthest reaches of the solar system. It is a little appreciated fact that Pittsburgh’s Westinghouse Electric Company played a central role in bringing that vision much closer to reality through its participation in the Nuclear Energy for Rocket Vehicle Applications (NERVA) program between 1959 and 1973. With recently renewed interest in the human exploration of Mars and destinations in the outer solar system, attention is once again focusing on the remarkable accomplishments that Westinghouse made in the development of the largely untapped potential of the nuclear thermal rocket.

As early as 1949, the Los Alamos National Laboratory, Los Alamos, New Mexico, conducted research to develop a solid core nuclear thermal rocket engine to power intercontinental ballistic missiles. The idea of a nuclear-powered rocket had already captured the imagination of many serious science fiction writers, evidenced by Robert A. Heinlein’s 1948 novel Space Cadet that featured a sleek nuclear-powered rocket ship that inspired the 1950 CBS television series Tom Corbett, Space Cadet, starring Frankie Thomas (1921–2006). With encouragement from science advisor Willy Ley, in 1951 Joseph Lawrence Greene, writing under the pseudonym Carey Rockwell at the publishing house of Grosset and Dunlap, launched Tom Corbett, Space Cadet, a juvenile novel series that fired the imagination of an entire generation of America’s youth with images of a streamlined manned single-stage-to-deep space atomic-powered rocket called the Polaris.

Similar to the nuclear rocket engine eventually developed under the NERVA program, the Polaris employed turbo-pumps to supply propellant to a uranium-fueled reactor core. Virtually all of the single-stage rockets of the golden age of science fiction were described at the time as using some form of atomic energy for propulsion. In a classic example of scientific theory inspiring art and, in turn, inspiring practical engineering concepts, by 1957 Los Alamos Laboratory had acquired a test facility at Jackass Flats, Nevada, to test the first KIWI series of nuclear rocket engines as part of Project Rover. Because these were ground tests rather than actual flight tests, the early engines were named after the flightless Kiwi bird endemic to New Zealand. The trials were conducted with the engines mounted upside down on their test stands with the rocket plume firing upward into the atmosphere.

In 1959, the Westinghouse Electric Company of Pittsburgh and its Bettis Atomic Power Laboratory in nearby West Mifflin, also in Allegheny County, were busy building nuclear reactors for the U.S. Navy and had also designed the nation’s first commercial nuclear power plant at Shippingport, Beaver County, that went online in December 1957. In anticipation of landing more lucrative government contracts, John Wistar Simpson, Frank Cotter, and Sidney Krasik convinced Westinghouse CEO Mark W. Cresap Jr. in 1959 to approve the creation of the Westinghouse Astronuclear Laboratory (WANL) to investigate the feasibility of building nuclear rocket engines.

Authorized in May 1959, WANL officially became a Westinghouse division on July 26, 1959, and consisted of just six employees with Simpson at the helm. Krasik, a Cornell University physicist, served as technical director and Cotter worked as Simpson’s executive assistant and marketing director. Born in 1914, Simpson graduated from the United States Naval Academy, Annapolis, Maryland, joined Westinghouse in 1937, and earned an MS from the University of Pittsburgh in 1941. Working in the switchgear division of Westinghouse’s East Pittsburgh plant, Simpson helped develop electric switchboards that could survive the extreme impacts experienced by naval vessels under bombardment in the Pacific Theater during World War II. In 1946, he took a leave of absence from Westinghouse to work at Oak Ridge National Laboratory in Oak Ridge, Tennessee, to familiarize himself with atomic power. Upon his return three years later, he became an assistant manager in the engineering department of Westinghouse’s Bettis Atomic Power Laboratory. He subsequently managed the construction of the Shippingport Atomic Power Station in 1954 and the following year was promoted to general manager of the Bettis Laboratory. He was elected a Westinghouse vice president in 1958. By 1959 Simpson and his team had become enthusiastic about taking on the new challenge of building nuclear-powered rockets to explore the solar system.

WANL was first headquartered in a shopping mall in the Pittsburgh suburb of Whitehall. By 1960 its staff and the leaders of Aerojet General had pooled resources to compete for the lucrative NERVA program contract from NASA’s Space Nuclear Propulsion Office (SNPO). Aerojet and Westinghouse won the contract to develop six nuclear reactors, twenty-eight rocket engines, and six Rocket In Flight Test (RIFT) flights the following year. With a substantial contract in hand, WANL increased its staff to 150 and relocated to the former site of the Old Overholt Distillery. By 1963, Westinghouse and its collaborators employed eleven hundred individuals on the project, based near the small town of Large, thirteen miles south of Pittsburgh in Allegheny County. Large was named for a former distillery founded during the early nineteenth century by Joseph Large. Together, Aerojet and Westinghouse developed the NRX-A series of rocket test engines based on an 1120 megawatt Westinghouse reactor. Assembled at Large, the reactors were loaded on rail cars for delivery to the nuclear test facility at Jackass Flats for field testing.

The initial objective of the NERVA program was to build a rocket engine that could deliver at least eight hundred seconds of specific impulse, fifty-five thousand pounds of thrust, at least ten minutes of continuous operation at full thrust, and the ability to start-up on its own with no external energy source. Seventy pounds per second of liquid hydrogen pumped from the propellant tank into the reactor nozzle would provide regenerative cooling for the rocket nozzle. The cylindrical graphite core of the nuclear reactor was surrounded by twelve beryllium plates mounted on control drums to reflect neutrons. The drums, also containing boral plates on opposite sides to absorb neutrons, were rotated to control the chain reaction in the core. The core consisted of clusters of hexagonal graphite fuel elements, the majority of which consisted of six fueled element sectors and one unfueled sector. The fuel, pyrographite-coated beads of uranium dicarbide, was coated with niobium carbide to prevent corrosion caused by exposure to hydrogen passing through the core. Each fuel rod cluster was supported by an Inconel tie rod that passed through the empty center section of each fuel rod cluster, and a lateral support and seal was used to prevent any of the hydrogen from bypassing the reactor core. Inconel is a high-temperature alloy, one version of which was being used at the time as the skin on the famous X-15 rocket plane.

The solid core nuclear thermal rocket used highly enriched uranium embedded in a graphite matrix. As the highly fissionable uranium 235 atoms absorb a neutron they split to form lighter elements, more neutrons, and a large amount of thermal energy. The nuclear rocket uses the thermal energy generated by a nuclear chain reaction to heat hydrogen, forced through narrow channels in the reactor core. The hydrogen propellant is delivered under pressure to the reactor core using turbo-pumps. The nuclear chain reaction in the reactor core causes the hydrogen to become superheated and expelled through the rocket nozzle at extremely high velocity as an explosively expanding reaction mass resulting in a high specific impulse of 825 seconds. In a chemical rocket, where a fuel (such as liquid hydrogen) and an oxidizer (such as liquid oxygen) are brought together and burned in a combustion chamber, the maximum specific impulse achievable is only about 450 seconds. Specific impulse is a measure of efficiency of a rocket and is defined by Konstantin Tsiolkovsky’s rocket equation as the pounds of thrust produced for the pounds of fuel consumed per second and is expressed in seconds.

With a high specific impulse, the ability to conduct multiple shutdowns and restarts, and a highly favorable energy to weight ratio, the nuclear rocket was the kind of vehicle that the early rocket pioneers Robert Goddard, Herman Oberth, Wernher von Braun, and Tsiolkovsky had long envisioned. As early as 1903, Tsiolkovsky, a Russian mathematics teacher, had hoped that it might be possible to somehow extract atomic energy from radium in order to power a rocket, but it was not until 1938 that Otto Hahn in Germany first succeeded in causing uranium to fission. Hahn’s former colleague Lise Meitner, living in exile in Sweden, realized the significance of what he had done—and the door to the atomic age flung open!

The power density of traditional chemical rockets is puny compared to the extraordinarily high power density of a nuclear rocket engine. Chemical rockets consist of numerous throwaway stages and require an enormous volume of their mass devoted to carrying both a propellant and an oxidizer. A nuclear rocket can be built as a single-stage vehicle, and requires no oxidizer because it heats a propellant that serves as the reaction mass, and is also able to undergo numerous shutdowns and restarts, making lengthy missions to the ends of the solar system both possible and economical. While the inefficiencies inherent in chemical rockets result in nominal costs of $3,500 to $5,000 per pound to deliver payload to low earth orbit, the more favorable propellant to payload mass ratio of the nuclear rocket promises costs in the range of just $350 to $500 per pound.

After radiation safety concerns were raised by SNPO at NASA over launching nuclear-powered rockets directly from the earth’s surface, von Braun at the Marshall Space Flight Center in Huntsville, Alabama, developed a proposal to boost a nuclear-propelled second-stage NERVA rocket to the edge of space using his Saturn V first-stage before firing the nuclear rocket engine after it was well above the densest part of the atmosphere. There is some debate as to whether this precaution is necessary for a well-designed nuclear rocket, but the prevailing cautiousness regarding anything nuclear renders it unlikely that direct ascent from the earth’s surface will be found acceptable anytime soon. The early NERVA rocket engine tests were, in fact, open atmospheric tests.

Westinghouse Astrofuel’s fabrication plant at Cheswick, Allegheny County, supplied nuclear fuel for the NERVA project. Fuel element corrosion was tested by heating the fuel elements by their own resistance, first at the Large site, and later at a new facility at Waltz Mill, Westmoreland County. In order to ensure fuel corrosion resistance and the stability of dimensional tolerances to several thousandths of an inch, the materials in the core elements were extruded into a bar possessing a hexagonal cross section having nineteen longitudinal holes. The extrusion was then polymerized, baked at a low temperature, and graphitized at a higher temperature of about 2200 degrees Centigrade. The resulting unfinished fuel element was subjected to a high-temperature chemical vapor process to coat the surfaces of the longitudinal channels with a gas mixture of niobium pentachloride, hydrogen, and methane. This mixture reacted with the graphite to form a niobium carbide coating intended to prevent corrosion of the core when it was exposed to the hydrogen propellant. The great challenge was to achieve a good match between the thermal expansion coefficients of the graphite and the niobium carbide to prevent cracking.

On September 24, 1964, the NRX-A2 established proof of concept by providing six minutes of power. By April 23, 1965, Aerojet and Westinghouse tested the NRX-A3 nuclear rocket engine at full power for sixteen minutes and demonstrated a three-minute restart. Pulse cooling was also introduced at this time in which bursts of LH2 were used to cool the reactor core. This was followed by a test of the NRX/Engine System Test (EST) engine equipped with Aerojet’s new nozzle and turbo-pump mounted next to the engine in place of the earlier Rocketdyne pump that had been housed separately behind a concrete wall. This permitted full operational testing of all of the equipment in a high radiation environment typical of an actual spaceflight. In 1966, Aerojet and Westinghouse commenced an additional series of tests to demonstrate ten startups on the NRX-A4/EST and full power operation of the NRX-A5 engine for two periods totaling thirty minutes of operation. On December 13, 1967, the NRX-A6 reached sixty minutes of operation at full power. According to data compiled by Aerojet and Westinghouse, on June 11, 1969, the XE engine was started twenty times for a total of three hours and forty-eight minutes, eleven of which were at full power. By 1970, the proposed NERVA I concept vehicle that evolved out of this work was projected to be capable of delivering 1500 MW of power and 75,000 pounds of thrust. It also had a projected lifetime runtime of ten hours and could be started and stopped 60 times while delivering 825 seconds of specific impulse for each hour of continuous operation. Especially encouraging was the fact that it was projected to have a total weight of less than fifteen thousand pounds.

Capable of starting up on its own in space and reaching full power in less than one minute, the design operating temperature of the reactor was 2071 degrees Centigrade and its reliability was projected to be at least 0.997. The .003 projected failure rate covered all forms of operational deficiencies, not just a catastrophe such as a crash or explosion. In one test conducted at Jackass Flats on January 12, 1965, a KIWI-TNT nuclear rocket engine reactor was intentionally exploded to more accurately assess the consequences and cleanup implications of a truly catastrophic launch pad accident. Off-site radiation from the test was judged to be statistically insignificant, adding just 15 percent to an individual’s average annual exposure at a distance of 15 miles from ground zero, and technicians were able to thoroughly clean up the site at ground zero within a matter of weeks.

Aerojet and Westinghouse prepared to begin construction of five reactors and five NERVA I rocket test engines for actual flight testing from the Kennedy Space Center on Merritt Island in Florida beginning in 1973, the year the federal government terminated the NERVA program. Total government expenditure by that time on the combined Rover/ NERVA program from 1955 to 1973 had reached more than $1.45 billion (equivalent to roughly $4.5 billion today). As a result of the cancellation of this program, a NASA plan to use a NERVA-type vehicle to place humans on Mars by 1981 was quietly shelved.

Based on the rapid improvements made to the design of the NRX engines in little more than a dozen years, it has been argued that with subsequent improvements in materials science, coupled with a better understanding of physics, the solid core nuclear thermal rocket would have been improved to the point where it could have delivered at least 1000 seconds of specific impulse, 3000 MW of power, and been capable of perhaps 180 recycles. Such a rocket would have been capable of continuously cycling back and forth to Mars about fifteen times with each transit taking as little as 45 to 180 days depending upon the transfer orbit configuration chosen, instead of the six to nine months required for a chemical powered rocket to make the same trip. The faster transit would actually lower astronauts’ exposure to radiation from cosmic rays, the van Allen radiation belts, and solar flares; it would also make it possible to launch heavier vehicles with larger crews and better shielding against cosmic radiation.

After the NERVA program ended, the Westinghouse Astronuclear Laboratory in Pittsburgh continued to work on several other projects, including the development of a nuclear-powered artificial heart. Amidst a changing political climate concerned with finding “green” energy sources, the laboratory became the Westinghouse Advanced Energy Systems Division (AESD) in 1976. Engineers at AESD experimented with a heliostat and worked on the Solar Total Energy Project in Shenandoah, Georgia, that used five acres of solar collectors to power a knitting factory. AESD also worked on a prototype for a magnetohydrodynamic system which reuses exhaust gases to increase the electrical output of a coal-powered plant by 30 percent. Following Westinghouse’s shuttering of AESD, several former employees formed Pittsburgh Materials Technology Inc. in 1993 at the former Westinghouse Astronuclear Laboratory. Pittsburgh Materials Technology specializes in producing high temperature specialty metal alloys for government and industrial customers.

During the 1970s, Westinghouse Electric Corporation sold its home appliance division and oil refineries, and in 1988 closed its East Pittsburgh manufacturing plant. In 1995, the company purchased CBS and the following year acquired Infinity Broadcasting. Renaming itself CBS Corporation in 1997, it sold off the nuclear energy business to British Nuclear Fuels Ltd. which, in turn, sold it to Toshiba in 2006. Under the wing of Toshiba, the nuclear energy business continues to operate under the name Westinghouse Electric Company and, because of rapid expansion in overseas demand for nuclear power plants, moved its corporate headquarters in 2009 to a new larger campus in Cranberry Township, Butler County.

In 1963, when Cresap died, Simpson was responsible for eighteen major Westinghouse divisions. Six years later he became president of Westinghouse Power Systems. He earned the Westinghouse Order of Merit and was elected to the National Academy of Engineering in 1966. In 1971, he won the prestigious Edison Medal. A member of the board of governors of the National Electric Manufacturers Association (NEMA) and chairman of NEMA’s Power Equipment Division, he was also a fellow of the American Nuclear Society where he served on the board of directors, on the executive committee, and as chairman of the finance committee. In 1995, the American Nuclear Society published his book Nuclear Power from Underseas to Outer Space, in which he recounted his experiences at Westinghouse. The book includes a detailed description of the company’s astronuclear program. Simpson died at the age of ninety-two on January 4, 2007, at Hilton Head, South Carolina.

The Westinghouse Astronuclear Laboratory was a product of an era of bold optimism in the promise of science and technology to solve problems and to bring to fruition a vision long shared by rocket pioneers Sergei Korolev, Stanislaw Ulam, Freeman Dyson, Tsiolkovsky, Goddard, Oberth, von Braun, and many others to eventually spread mankind across the vast solar system. Much of the science fiction of the era, such as the Tom Corbett television and juvenile novel series, was grounded in hard science as it was understood at the time. Overtaken by the social and political upheavals that accompanied the growing disillusionment with the Vietnam War and social dissension at home, the NERVA program nonetheless achieved remarkable successes that were ultimately cut short by shifting political events and a narrowing of national horizons. Despite a long hiatus, those successes are now inspiring a new generation of aerospace engineers to once again think boldly and embrace the difficult challenges articulated by President John F. Kennedy, a strong early supporter of the NERVA Program, at Rice University, Houston, Texas, in 1962: “We choose to go to the moon in this decade, and do the other things, not because they are easy, but because they are hard.”

The collaboration of Westinghouse Electric and Aerojet General in tackling the difficult work of developing a viable solid core nuclear thermal rocket engine is a down payment on the eventual human exploration and settlement of the solar system. The full utilization of such nuclear technology will make possible the fulfillment of the dream first enunciated by Tsiolkovsky who more than a century ago proclaimed, “The earth is the cradle of mankind, but a man cannot live in the cradle forever.” Nurtured by the dreamers in the cradle of western Pennsylvania’s Three Rivers Valley for a brief but shining period of fourteen years, the dream of one day boldly setting off into the new frontier moved a little closer to reality.

For Further Reading

Bruno, Claudio. Nuclear Space Power and Propulsion Systems, AIAA Progress in Astronautics and Aeronautics Series No. 225, 2008.

Dewar, James A. To the End of the Solar System: The Story of the Nuclear Rocket. Lexington: University Press of Kentucky, 2004.

Simpson, John W. Nuclear Power from Underseas to Outer Space. La Grange Park, Ill.: American Nuclear Society, 1995.

Willis L. Shirk Jr., Lancaster, is an archivist for the Pennsylvania State Archives, administered by the Pennsylvania Historical and Museum Commission. He holds a B.A. in history from Millersville University and an M.A. in American studies from the Capitol College of the Pennsylvania State University. An avid astronomer, the author is a member of the American Institute of Aeronautics and Astronautics, British Interplanetary Society, Mars Society, Meteoritical Society, and the National Space Society.