Encyclopedia Astronautica
Nerva Alpha


DoE nuclear/lh2 rocket engine. 71.7 kN. Study 1972. The final Nerva Alpha flight engine reference configuration as documented at the end of its development. Isp=860s.

The final Nerva Alpha flight engine reference configuration as documented at the end of its development was a smaller engine that could be launched together with its stage and a payload in a single space shuttle launch.

The engine had a mass of 2550 kg and was designed to operate at a specific impulse of 875 s for one hour at a thrust of 73 kN or at a specific impulse of 860 s for two hours at a thrust of 71.7 kN (the Alpha engine). The growth potential of the engine to a 975-s specific-impulse version incorporating carbide fuel elements was studied (the Gamma engine), as was the potential of the reactor to double as a heat source for a long-term electrical power supply.

The principal function of the Alpha nuclear engine was to heat the hydrogen monopropellant to an average temperature of ~ 2700 K so as to achieve a maximum net specific impulse of ~ 875 s. The required heat energy was generated by fissions of uranlum-235 in a small reactor core. The reactor fuel was contained in hexagonal elements (figure 2), measuring roughly 1.9 cm across flats; these elements were made up of a composite mixture of graphite and a solid solution of UC-ZrC, and contained 19 flow passages each. Hydrogen was heated by passing through these holes, which were coated with a layer of ZrC to inhibit hydrogen corrosion. The reference design contained 564 of those elements in a core that had a diameter of O.655 m and was O.89 m long. The total core uranium loading was 60 kg (92.5% enriched uranium). Heating the 8.5 kg/s of hydrogen flow to produce the nozzle-plenum condition at maximum specific impulse required a total thermal power of ~ 367 MW and resulted in 72,975 N (16,406 lb) of thrust.

Nuclear criticality in this nuclear reactor was achieved by the following three means:

  • One out of every three positions in the hexagonal core array was occupied by an element that combined the functions of neutron moderator and core support (figure 3). These elements (figure 4) were of the same size as the fuel elements, but contained a zirconium-hydride moderator, a low-density ZrC insulator, and a metal tie tube which supported the entire core against the pressure-drop load of the flowing hydrogen. These 241 elements were cooled regeneratively by passing 47.5% of the total hydrogen flow through the tie tubes; the flow then turned to cool the moderator and was subsequently used to drive the main turbine before it was returned to the core inlet.
  • The core peripheral space, between the irregular outline defined by the hexagonal fuel pattern and the circular outline desired for the reflector, was filled primarily with beryllium slats, which were protected from the high core temperatures by low-density ZrC insulators. These slats were regeneratively cooled by 7.5% of the total hydrogen flow, in parallel with the tie tubes. These slats were, neutronically speaking, part of the reflector.
  • The core was surrounded by a 0.l42-m-thick annulus of beryllium metal, which was very effective as a neutron moderator and reflector.
Nuclear criticality control was achieved by 12 circular solid drums of beryllium, which were covered over a 120-deg sector of their periphery with a boron-containing "poison" plate. Gang rotation of the drums changed the position of the boron plates, thus controlling the number of neutrons in the reflector which could return to the core without capture. The total control span was roughly 9.0% of reactivity. These drums and their drive mechanisms were the only moving parts housed within the reactor pressure vessel.

The reactor nozzle was a conventional, regeneratively cooled, U-tube design. The nozzle and the reflector were cooled in series by 45% of the hydrogen coolant flow. This flow entered the nozzle at a torus located at the 25:1 area-ratio point, passed through thin-walled inconel tubes toward the core, and discharged into the reflector aft-end plenum. After passing through the reflector, this flow was mixed with the turbine discharge flow, and the combined flow then cooled the shield, the core-support plate, and the core in series. The flow path is shown in Figure 5.

The aft end of the nozzle, from the 25:1 area-ratio point to a 100:1 area ratio, was a lightweight uncooled graphite-fiber structure. This nozzle skirt could be unlatched and turned back on hinges to be positioned alongside the cooled nozzle portion so as to reduce overall engine length for stowage inside the cargo bay of the space shuttle vehicle.

Hydrogen pressure was supplied by a turbopump with a nominal discharge pressure of 6O3 N/em (875 psia) corresponding to a nozzle chamber pressure of 310 N/cm2.

The overall length of the engine was 3.165 m with the nozzle skirt folded back, and the mass was about 255O kg . Engine mass statement was as follows:

  • Reactor core and hardware: 868.2 kg
  • Reflector and hardware: 568.8 kg
  • Shield: 239.5 kg
  • Pressure vessel: 150.1 kg
  • Turbopump: 40.8 kg
  • Nozzle and skirt assembly: 224.5 kg
  • Propellant lines: 15.4 kg
  • Thrust structure and gimbal: 27.7 kg
  • Valves and actuators: 206.8 kg
  • Instrumentation and electronics: 158.8 kg
  • Subtotal: 2500.6 kg
  • Contingency: 50 kg
  • TOTAL: 2550 kg
The engine was mounted as closely as possible to the hydrogen tank. Thrust-vector control was achieved by glmballing the entire engine. Chamber pressure (and thus thrust) was controlled by adjusting the turbine bypass valve, whereas chamber temperature (and thus specific impulse) was controlled by adjusting the reflector control-drum rotational position.

The Alpha Engine design was based on operation at either of two full-power conditions. In Condition A, the engine could be run for a total of 2 hours cumulative duration, with 20 start/restart cycles, a specific impulse of 860 seconds, and a thrust of 71.724 kN. In Condition B, the engine could be run for a total of 1 hour cumulative duration, with 3 start/restart cycles, a specific impulse of 875 seconds, and a thrust of 72.975 kN. Condition B was based on a maximum calculated fuel material temperature of 2880 K and a corresponding average fuel-element exit-gas temperature of 273 K. The fuel-element weight loss by diffusion of carbon through the coatings was assumed to be, based on calculations, 12.3 g/element/hour under these conditions. To provide for a two-hour duration, the fuel-element exit-gas temperature would be reduced by 60 K. This would result in a lower fuel-element mass loss rate (~ 9.2 g/element/hour) and would provide a wider margin between the operating temperature and the limit. Calculations near the end of the project indicated that the actual specific-impulse predictions for these state points exceeded the target values by about 6 seconds.

Due to thermal-stress limits of the fuel elements and of other carbide reactor core components, the startup and shutdown transients were limited to roughly 83 k/s. Final engine mapping and transient calculations were made with the engine computer model to estimate startup and shutdown transients for minimum propellant use within material temperature constraints. Indications were that the specific impulse degradation during startup and shutdown might be decreased by a factor of two over early rough estimates that were based on linear thrust and linear temperature ramps. The run profiles are given in Figure 6. Approximate values of the flow-rate and thrust integrals for these intervals were:

  • Engine conditioning (3O s): 15.0 kg hydrogen, and an impulse in N-sec equivalent to 0.81 x full thrust power
  • Startup (26.4 s): 95.5 kg hydrogen, and an impulse in N-sec equivalent to 8.75 x full thrust power
  • First 30 s of shutdown: 15.0 kg hydrogen, and an impulse in N-sec equivalent to 9.75 x full thrust power
The shutdown of a reactor had to be followed by a cooldown period to remove the residual heat generated in the reactor core as a result of the radioactive decay of fission products formed during the full-power run. This heat was generated in three types of decay: (l) fission chains triggered by delayed-neutron precursor decay, (2) beta decay, and (3) gamma decay. The fission heat dropped off quickly and was virtually zero within about ten minutes, but the beta and gamma decay tailed off more slowly and required some reactor cooling for a period of a day or longer, depending on run duration. The heat release rate was predictable and depended strongly on run duration. During cooldown the specific impulse had to be decreased gradually as the thrust decreased, to maintain material temperatures below operating limits.

Total cooldown impulse and total cooldown coolant mass required as a function of run duration are given in Figs. 7 and 8, respectively.

The usefulness of the cooldown impulse to the mission depended greatly on where the thrust occurred and this was dependent on the specific mission being analyzed. A practice which was developed for estimating payloads was to define a cooldown effectiveness factor: the reduction of required full-power impulse attributable to the cooldown divided by the cooldown impulse. Typical values varied from 0.6 to nearly unity. If a fraction of the cooldown impulse was effectively utilized then an average specific impulse for the startup, full-power, shutdown, and cooldown sequences could be determined.

A parallel effort to the nuclear-engine study at LANL was a design study of the associated vehicle, or stage, performed for NASA's Marshall Space Flight Center by the McDonnell Douglas Astronautics Co. (MDAC) under contract NAS8-27951. This nuclear Stage Definition study was a comprehensive, continuing effort that was redirected from similar studies performed for the NERVA engine.

Figure 10 indicates the characteristics of a typical nuclear stage as described in the MDAC study. It was a reusable stage consisting of a command and control module (CCM) docked to the forward end of an insulated and meteoroid-protected tank. Thrust-vector control was achieved by gimballlng the engine. The stage was designed for hydrogen-propellant resupply in orbit, and was sized to fit within the payload bay of the Space Shuttle Orbiter leaving a 2.43-m space allowance for payload. A typical design for the complete stages studied had 17,783 kg mass, including 12,8l4 usable propellant. This stage could be used either alone or in conjunction with one or more propellant modules to increase the available propellant if needed for a particular mission. The maximum propellant module size that could be placed into orbit by the Space Shuttle Orbiter was 18.29 m, although smaller units could also be considered. Such a propellant module would have a mass of 23,181 kg, including 21,265 kg of usable propellant.

Radiation naturally represented a unique operational characteristic of the Nerva stage. In the Prerun Environment, prior to the first burn phase, the reactor contained no fission products and was innocuous as a radiation source.

The payload radiation environment during run was another matter. By power-plant standards, the engine reactor had a very high power density and was only lightly shielded. Thus it represented a large radiation source to its environment. Because the payload was forward of a large tank of hydrogen, the payload was very effectively shadow-shielded by the tank, its remaining hydrogen and all the engine components being forward of the core. Because the engine was operated in space, there was no scattering of the side-leakage radiation by the environment into the payload region. The 240-kg shield provided in the design was included primarily to reduce engine component heating rates by roughly a factor of ten. It also reduced propellant tank heating by roughly a factor of two and payload doses by a factor of nine. The dose rate at the payload increased by a factor of 50 as the hydrogen propellant was drained from the propulsion module tank. The total dose was obtained by integrating the rapidly changing dose rate over the entire profile, with 99% of the total being accumulated over the last half of the full-power run and its cooldown. Estimated payload doses were: maximum dose rate: 10^7 n/cm^2/sec for neutron radiation (energy > 1 MeV), 24 rad/sec gamma radiation; total dose: 10^6 to 10^9 neutron radiation, and 13,000 rads of gamma radiation.

In the post-run radiation environment the radiation field around the reactor decayed rapidly during cooldown, and after about 0.5 h became basically a gamma field which diminished with time and distance from the reactor. Both in the tank and forward of it the field was depressed due to the shadowing effects of the shield and other structure

Prior to launch, each engine would have been operated only briefly and at a negligible power level to determine the exact critical control-drum position for that particular core. After this operation, poison wires would be installed in the reactor core to provide an absolute safety precaution against a neutronic startup which would result from an inadvertent outward rotation of the control drums or from water flooding the reactor core. The poison wires would remain in the reactor through the prelaunch, launch, and subsequent man-tended operations in space. They would be removed remotely prior to the first startup. Once removed they could not be replaced.

The engine could be used in a low power mode in which a low flow of hydrogen was heated to 700-1000 ,K. The turbopump was not used so that the hydrogen was either pressure-fed or fed from a small electrical pump in the cooldown line. Possible uses of this mode were: tank prepressurization; orbit trimming; preburn maneuvers; midcourse corrections; and propellant settling.

It was certain that at least the following two types of operation would be required prior to the 30-s bootstrap startup to full power:

  • Neutronic startup, accomplished by a programmed positioning of the control drums to bring the reactor to critical, bring both power and core temperature up to the prestart level, and execute the transition to temperature control. This operation would require a maximum of roughly ten minutes.
  • Thermal conditioning of the turbopump, by trickle flow through the turbopump for a period of a few minutes. This or other trickle flow could also be used to perform propellant settling in the tank.
After a lengthy full-power run, there would be intervals during which the engine could not be restarted, due to the buildup of the radioactive isotope Xenon-135, which was a neutronic poison. These exclusion intervals would start three to eight hours after a full-power run and would last from a few hours to a day or slightly more. The duration would increase as the run duration increased. These intervals were not precisely known. These deadbands apparently would not interfere with a normal one-module geosynchronous mission. They might affect the selection of intermediate orbits for perigee propulsion on escape missions.

The design objective of the Alpha Engine was to derive the greatest benefit from the technology developed over the 17 years of the nuclear propulsion program and thus to minimize development risks. The resulting engine combined the best and most straightforward application of the successful developments of LASL Rover and NERVA programs. In general, the design did not require any major technology advancements to achieve the desired goals.

Although graphite fuel elements had been used during most of the program, composite fuel had been the predominant development effort at LASL after 1967. Composite fuel elements had been tested successfully in one reactor test (Nuclear Furnace-l) and in many hundreds of electrically heated runs over a wide range of conditions bracketing the design-point conditions. These highly reliable fuel elements were superior to the graphite elements originally developed. Neither the basic cross-sectional size nor their shape were changed for the Alpha engine design, and all the manufacturing and test equipment was set up and operating at LASL. The shorter elements of the Alpha Engine (compared to Nuclear Furnace elements) were expected to present only routine development problems.

The core design, particularly the crucial use of zirconium-hydride center elements for neutron moderation, was successfully tested in a very similar size and configuration in the Pewee-1 reactor. The tie-tube design regenerative core support cooling principle was tested successfully in the Phoebus IB and Phoebus 2A reactors.

Low-density ZrC insulators were used in the core instead of the traditional pyrolytic-graphite insulators. The change was made because it was believed that ZrC was more reliable and inherently capable of longer life. The development of low-density insulators was progressing well, and such insulators would be incorporated throughout Nuclear Furnace-2 then being fabricated.

The beryllium reflector and rotary drum control principles were tested in all but three of the 19 reactors in the nuclear rocket program. The turbopump design appeared to be straightforward although two areas might require development. These were the pumping of slush hydrogen and the further development of radiation-resistant bearing retainers in the event that the conventional Armalon retainers proved to be marginal. Based on the limited experience with slush hydrogen, no large problems were anticipated. Existing radiation-damage data for Armalon indicated that these retainers would be adequate for the intended application, but confirmation was required.

A standard U-tube regeneratively cooled nozzle using conventional materials appeared to be adequate, although confirmation of low-cycle thermal-fatigue resistance was required. Development of the uncooled skirt was believed to be straightforward, but would be confirmed by scale-model testing.

Development of the actuators was based on existing technology but would require radiation-hardening, particularly in the area of lubrication of gears. A sound technology base also existed for all other components including structures, valves, and instrumentation. Existing test facilities at the Nuclear Rocket Development Station on the Nevada Test Site could be used for ground-testing. The main new addition required was a "scrubber" effluent cleanup system, which has been tested successfully in l/7th size during the Nuclear Furnace-1 test.

The proposed Alpha Engine development schedule included a reactor test in 1976, seven engine ground tests, and a prototype flight test in 1979. An extensive component test program would provide suitable qualified components as required in the engine program. This schedule would permit delivery of the first mission-qualified flight engine in 1982.

Engine: 2,550 kg (5,620 lb). Area Ratio: 100. Propellant Formulation: Nuclear/Slush Hydrogen. Restarts: 20.

Status: Study 1972.
Height: 4.46 m (14.64 ft).
Diameter: 1.24 m (4.06 ft).
Thrust: 71.70 kN (16,119 lbf).
Specific impulse: 860 s.
Burn time: 1,200 s.
First Launch: Designed 1972.

More... - Chronology...


Associated Countries
See also
Associated Manufacturers and Agencies
  • AEC American agency overseeing development of rocket engines and rockets. Atomic Energy Commission, USA. Responsible for development of nuclear weapons and nuclear power applications. Became part of the newly-created Department of Energy in 1971. More...

Associated Propellants
  • Nuclear/LH2 Nuclear thermal engines use the heat of a nuclear reactor to heat a propellant. Although early Russian designs used ammonia or alcohol as propellant, the ideal working fluid for space applications is the liquid form of the lightest element, hydrogen. Nuclear engines would have twice the performance of conventional chemical rocket engines. Although successfully ground-tested in both Russia and America, they have never been flown due primarily to environmental and safety concerns. Liquid hydrogen was identified by all the leading rocket visionaries as the theoretically ideal rocket fuel. It had big drawbacks, however - it was highly cryogenic, and it had a very low density, making for large tanks. The United States mastered hydrogen technology for the highly classified Lockheed CL-400 Suntan reconnaissance aircraft in the mid-1950's. The technology was transferred to the Centaur rocket stage program, and by the mid-1960's the United States was flying the Centaur and Saturn upper stages using the fuel. It was adopted for the core of the space shuttle, and Centaur stages still fly today. More...

Associated Stages
  • Nerva Alpha Nuclear/LH2 propellant rocket stage. Loaded/empty mass 17,783/4,969 kg. Thrust 71.70 kN. Vacuum specific impulse 860 seconds. Nuclear stage designed to fit into the space shuttle payload bay. Additional propellant modules could be added in orbit. Such propellant modules would have a mass of 23,181 kg, including 21,265 kg of usable propellant. Given goahead in 1972, it would have been flight tested by 1982. More...

Home - Browse - Contact
© / Conditions for Use