• US Air Force Stability and Control Data Compendium - Starting in the 1920's, the US Air Force collected data for predicting the aerodynamic behavior of aircraft. This link is to the 1978 text version of the compiled data, known as the USAF DATCOM, which is a 3134 page, 113 MB .pdf file. Since then, versions of the data and formulas have been compiled into software modules, which can be run by themselves, or linked into more comprehensive airplane design software.
  • US Standard Atmosphere 1976 - This book gives information and tables describing composition, pressure, temperature, and other properties of the Earth's atmosphere under "standard" conditions. The actual atmosphere varies from these standard conditions due to weather, the solar cycle, and secular changes such as increasing CO2 concentration. The reference data is still useful for doing calculations, and formulas based on this data have been incorporated into software programs and modules within programs.

Nuclear Rockets


Previous US Nuclear Rocket Program


Solid Core NTR Development


McDonnell Douglas has a long association with nuclear propulsion, beginning with nuclear aircraft propulsion studies in the mid-1950s.[1]. Additional work includes NERVA nuclear vehicle interplanetary flight studies, and solid core vehicle integration studies under contract to NASA in the 1960s. In 1971 studies were conducted on a Shuttle-launched Nuclear Shuttle System, which used the PEEWEE reactor.

In 1972 LANL undertook a study of a small nuclear engine using NERVA technology and based on the PEEWEE reactor design, that could be used in the newly designed NASA Space Shuttle.[2] This engine would have weighed 2,555 Kg, using composite UC-ZrC-C fuel, with other parameters including the following:

  • Total Operating Period hours 2 1
  • Operating Cycles 20 3
  • Specific Impulse seconds 860 875
  • Thrust kN 71.6 73.0
  • Hydrogen Flow kg/sec 8.5 8.5
  • Thermal Power MWt 354 367

Various other NERVA-derived propulsion concepts were considered, and LANL evaluated a range of mission applications. For instance, it was concluded that a nuclear-propelled orbital transfer vehicle could significantly reduce propellant launch requirements, based on mission models that envisioned from 145 to 872 Orbital Transfer Vehicle flights. The authors concluded that developments in non-nuclear nozzle and turbo-pump technology in the Space Shuttle Main Engine program have enhanced the viability of NERVA derived engines.

LANL mission analyses have identified significant reductions in Initial Mass in Low Earth Orbit (IMLEO) through the use of NERVA technology:[3]

Propulsion System IMLEO (tons)
Chemical (all propulsive) 2,100
Chemical + Aerobrake 715
NTR (all propulsive) 760
NTR + Earth Only Aerobrake 540
NTR + Aerobrake 420

The costs of developing a flight-ready nuclear rocket was estimated to be $4-5 billion, which included the cost of rebuilding the 1960s NERVA capability, which was estimated to include:

Component Cost Notes
Engine Design & Construction $1,218 M 80% of NERVA
Technology $ 377 M 50% of NERVA
Test Facilities Capital $ 460 M
Operations $ 210 M
TOTAL $2,265 M in 1985 $

This analysis also considered issues associated with testing, noting that:

"The major obstacle to testing at NTS will be the reduced levels of radioactive debris which are allowed to transport into the public domain. The levels are more stringent than those present during the NERVA program. The current exposure limits of 150 m Rem to civilian personnel may restrict the tests of the NTR to low power levels and mass flows in the reactor... A simple solution to this problem may be to utilize one of the Pacific Ocean Islands owned by the United States -- namely Johnston Island... (an) ecological desert of ocean surrounds the area due to the stagnation of the return of the Japanese current..."

ANL Cermet Program

From 1961 through 1967 Argonne National Laboratory conducted a development program for nuclear rocket fuel elements independently of the Rover/NERVA program.[4]Although no engines were fabricated or tested, extensive work was conducted on testing ceramic fuel configurations. Thermal shock tests suggested that these cermet fuels could have substantial tolerance to the effects of nuclear excursions. This program was terminated in 1967, prior to its completion.


Two major problems were identified during the Rover/NERVA program:

Core disassembly due to vibration, accompanied by cracking of the fuel matrix and loss of material into the propellant flow;

Loss of fuel matrix uranium and carbon due to coating erosion and cracking, and through diffusion through the coating.

The first problem was resolved by changed designs which reduced vibration and matrix cracking.

However, the second problem, of fuel element corrosion, proved less tractable:[4]

"Corrosion was most pronounced in the mid-range region, about a third of the distance from the cold end of the fuel element. Fuel operating temperatures were lower here than the fabrication temperatures, hence thermal stresses were higher than at the hot end. Also, the neutron flux was highest in this region..."

"No fuel element geometry or fuel material ever totally solved the NERVA fuel element degradation problem. Mass loss of both uranium and carbon continued to limit service life by causing significant perturbation to core neutronics during the tests. Crack development in the fuel element coating was never completely eliminated.... Non-nuclear testing of coated fuel elements revealed an Arrhenius relationship between diffusion and temperature. For every 205 K increase in temperature (in the range 2400 to 2700 K), the mass loss increased by a factor of ten... resulting in loss of 20% of total uranium in approximately 5 hours of testing at 2870 K."

A number of other program management lessons emerged from this period. One analysis notes that:[4]

"One overriding lesson from the NERVA program is that fuel and core development should not be tied simply to a series of engine tests which require expensive nuclear operation. Definitive techniques for fuel evaluation in loops or in non-nuclear heated devices should be developed early and used throughout the program..."

Solid Core NTR Tests


The initial series of nuclear rocket tests was conducted by Los Alamos National Laboratory, under the KIWI program, which eventually cost $177 million (in then year dollars).

Graphite was chosen as the internal structural material for these nuclear reactors for several reasons. It has excellent strength at high temperatures and its strength actually increases at higher temperatures. In addition, in contrast to metals which are strong neutron absorbers, graphite acts as a moderator, reducing the amount of enriched uranium required in the core. The great disadvantage of graphite, however, is that it quickly erodes in the presence of hot hydrogen. While this erosion cannot be eliminated, techniques are available to reduce erosion to acceptable levels over the operating lifetime of the reactor, which is measured in minutes to hours.

Westinghouse was the prime contractor for the reactor component of the NERVA program, while Aerojet was the contractor for engine elements such as pumps and nozzles. The NERVA program resulted in an investment of $662 million (then-year dollars) for development and testing of flight engine prototypes.[4]

The first nuclear rocket test, conducted in July 1959, used uncoated UO2 plates as fuel elements. This test reached a maximum temperature of 2683 K and a power level of 70 MWt. Vibrations during operations produced significant structural damage in the reactor core.

The first nuclear reactor tested, KIWI-A, successfully demonstrated the principle of nuclear rockets, but it used unclad fuel plates that were not representative of later tests.

KIWI A test
This test, conducted in July 1959, incorporated significant modifications in the core design used in KIWI A. The fuel consisted of short cylindrical Uranium Oxide elements in graphite modules, with four axial channels coated with Niobium Carbide using chemical vapor deposition. The reactor ran for 6 minutes at power levels as high as 85 MWt.
KIWI-A prime
tested in 1960, replaced the fuel plates with NbC plated graphite modules with 4 micron diameter UO2 particles embedded in the graphite matrix. However, some structural damage occurred in this improved design during its 6 minute test.
The subsequent KIWI-A3 reactor used a higher temperature Chemical Vapor Deposition process, resulting in a thicker NbC coat with improved adherence. Some core damage occurred during the 5 minute test in October 1959, which reached power levels of 100 MWt, with some fuel elements showing blistering and corrosion. Generally this reactor test was considered successful.
Using the same UO2 fuel as KIWI A3, the fuel element design was changed to a 7 channel configuration 66 cm long, with niobium carbide coatings. The December 1961 test, which was intended to reach 1100 MWt, only reached 300 MWt, and was terminated after 30 seconds due to a fire caused by a hydrogen leak in the reactor exhaust nozzle.
This September 1962 test, which was essentially a repeat of the KIWI B1A test, achieved a power level of 900 MWT, but was terminated within a few seconds when several fuel elements were ejected from the reactor exhaust nozzle.
Design configuration not tested.
Design configuration not tested.
This reactor incorporated substantial redesigns based on the failure of the KIWI B1B configuration. The fuel elements were fully extruded hexagonal graphite blocks 1.32 meters long and 19 millimetres in diameter, with 19 cooling channels, each 2.3 millimetres in diameter. However, the test run in November 1962 was terminated when bright flashes in the exhaust stream indicated vibration induced damage to the core was leading to its disintegration.
Design configuration not tested.
Design configuration not tested.
Although modifications in the design of this reactor eliminated the vibration which had marred previous tests, the May 1964 test was terminated after about 60 seconds at full power due to the rupture of a nozzle cooling tube.
This reactor was the first to use coated Uranium Carbide (UC2) fuel, in place of the Uranium Oxide (UO2) fuel previously used. To avoid oxidation of the Carbide fuel, it the uranium fuel particles were coated with a 25 micron layer of pyrolytic graphite which exuded water vapor. This pyrolytic carbon layer subsequently was also used to enhance fission product retention, though this was not initially its purpose. This reactor was operated for 12 minutes, including 8 at full power. The duration of the test was limited by the available liquid hydrogen storage capability.
This KIWI-B type reactor was deliberately destroyed on January 1965 by subjecting it to a fast excursion. This test was intended to confirm theoretical models of transient behaviour.
Phoebus 1A
This first test of a new class of reactors on June 1965 included over 10 minutes of operations at 1090 MWt, with an exhaust temperature of 2370 K.
Phoebus 1B
This February 1967 test built on the previous Phoebus 1A test, reaching power levels of 1500 MWt for 30 minutes, with an additional 15 minutes at lower power levels.
Phoebus 2A
The most powerful nuclear reactor of any type ever constructed, with a design power level of 5,000 MWt. Operations in June 1968 were limited to 4,000 MWt dur to premature overheating of aluminium segments of pressure vessel clamps. At total of 12.5 minutes of operations at temperatures of up to 2310 K included intermediate power level operations and reactor restart.
This small reactor was intended to be a reactor testbed, incorporating Zirconium Carbide coatings on some fuel elements, instead of the Niobium Carbide used on Phoebus. With a peak operating power of 503 MWt at 2550 K, it achieved new levels of core power density (2340 MWt/m3 average and 5200 MWt/m3 peak), demonstrating a specific impulse of 845 seconds.
Nuclear Furnace 1
The final phase of NERVA fuel development was the Nuclear Furnace (NF-1) reactor, a heterogeneous water-moderated beryllium-reflected reactor for high-temperature nuclear testing of fuel elements and other components. 47 of the 49 fuel elements used Uranium Carbide and Zirconium Carbide Carbon composite fuel, while the remaining 2 fuel elements used Uranium-Zirconium Carbide. Tests in 1972 of composite fuel elements with various carbide contents, thermal expansion coefficients, and thermal stress resistance demonstrated that minimizing the mismatch in thermal expansion coefficient between the fuel and coating would reduce coating cracking and carbon erosion. With a peak operating power of 44 MWt at 2500 K, it achieved new levels of core power density (4500 to 5000 MWt/m3).

The Nuclear Furnace test facility included provisions for remote controlled replacement of core elements, as well as a reactor effluent scrubber system to remove radioactive contaminants from the propellant exhaust.

Nuclear Furnace 2
This reactor was built but not tested, due to the cancellation of all work in this area in 1972. The goals of this experiment included testing of a novel coated particle fuel, using a graphite fuel matrix with a coefficient of thermal expansion closely matched to that of the coating, to reduce thermal stress and cracking.
This September 1964 engine test included 5 minutes of operation at half to full power levels. Duration of the test was limited by available hydrogen storage capacity. At full power of 1100 MWt the engine demonstrated a specific impulse of 760 seconds.
This April 1965 test operated for 8 minutes, including 3.5 minutes at full power. The first test was terminated by a spurious trip from the turbine overspeed circuit. The reactor was restarted on May 1965, and operated at full power for 13 minutes, and subsequently restarted for low to medium power operations for an additional 45 minutes. In total, 66 minutes of operations were accumulated, including 16.5 minutes at full power.
This engine operated for a total of 110 minutes, including 28 minutes at full power of 1100-1200 MWt, on five different days in February 1966.
This 1100 MWt engine was operated in June 1966 for 30 minutes at full power, with the test duration limited by available hydrogen storage capacity.
This 1100 MWt engine was operated in December 1967 for 60 minutes at full power, exceeding the NERVA design goal.
This 1100 MWt engine was a prototype engine, the first to operated in a downward firing position. It accumulated a total of 28 start cycles in March 1968 for a total of 115 minutes of operations. Test stand coolant water storage capacity limited each full power test to about 10 minutes.
  1. Haloulakos, V.E.; et al. Nuclear Propulsion: Past, Present, and Future, Fifth Symposium on Space Nuclear Power Systems, Albuquerque 11-14 January 1988. pp. 329–332. {{cite book}}: Explicit use of et al. in: |author= (help)
  2. Bohl, R.J., and Boudreau, J.E. (January 1987). Direct Nuclear Propulsion: A White Paper. Los Alamos National Laboratory.{{cite book}}: CS1 maint: multiple names: authors list (link)
  3. Howe, Steven (10–14 June 1985). Assessment of the Advantages and Feasibility of a Nuclear Rocket for a Manned Mars Mission, Manned Mars Mission Workshop. Huntsville, AL: Marshall Space Flight Center. Preprint LA-UR-85-2442.{{cite book}}: CS1 maint: date format (link)
  4. a b c d Horman, F.J.; et al. (4–6 September 1991). Particle Fuels Technology for Nuclear Thermal Propulsion, AIAA/NASA/OAI Conference on Advanced SEI Technologies. Cleveland, Ohio. Paper AIAA 91-3457. {{cite book}}: Explicit use of et al. in: |author= (help)CS1 maint: date format (link)