The Aircraft Reactor Experiment

Suraya Omar
March 24, 2012

Submitted as coursework for PH241, Stanford University, Winter 2012

Introduction

The Aircraft Reactor Experiment was a concept designed at Oak Ridge National Laboratories (ORNL) in the 1940s to use molten salt reactors to propel a plane. Nuclear reactors using molten sodium as the coolant operate at high temperatures, close to approximately 1500 °F, and can be constructed small enough to incorporate into the engine of a plane. Originally, the idea was to incorporate the nuclear component between the conventional compressor and turbine sections to heat up and expand the air getting forced through the engine. This addition would theoretically allow for cruising distances almost twice as far as standard engines of the day would permit. While a molten salt reactor was never actually used on a plane - the project was eventually shut down due to corrosion caused by high temperatures and levels of zirconium tetrafluoride in the coolant - an ARE was built at ORNL and achieved criticality in 1954. [1] As described by E.S. Bettis et al., the specific objective of the experiment was to "create a high-temperature, low-pressure circulating fuel reactor using materials suitable for a high-power reactor." [2]

The ARE Reactor

The molten salt reactor finally constructed was operated at 1-3 MWt. Liquid uranium tetrafluoride (UF4) fuel, using uranium enriched to 93.4% 235U, was circulated through the core through 66 tubes surrounded by bryllium oxide (BeO) moderators. The sodium fluoride (NaF) coolant was pumped through the spaces between the moderators and the fuel pipes. Molten salt is advantageous as a coolant because of its low vapor pressure, which allows the reactor to operate at high temperatures without pressurization - the lack of a pressurizer was another reason molten salt reactors were viable for aircraft propulsion. In order to extract maximum heat, the fuel and coolant were pumped through separate fin-tubed heat exchangers. The reactor shell and tubing was made of Heliarc welded Inconel, a nickel alloy typically resistant to corrosion by molten fluorides. [3]

Design Considerations

Fuel

Designers of the ARE found early on that liquid fuel was necessary because of its negative temperature coefficient of reactivity. Solid fuel pins cause the reactor to be unacceptably unstable due to the increased reactivity of xenon at high temperatures. Hence, molten salt reactors with solid fuel could not safely be translated to high-power operation. [3] Liquid UF4, on the other hand, expands when it is heated. Since the thermal volume expansion of the fuel (3 × 10-4/°C) is bigger than that of the reactor walls, sharp power increases will cause the fuel to expand outside of the reacting zone within the core. If a fraction of the fuel mass Δm/m is expelled, the reactivity coefficient k can be expected to decrease by Δk/k. For the ARE, the ratio (Δk/k)/(Δm/m) is 0.24. While expansion alone would render a negative temperature coefficient of 10-4 (Δk/k)/°C, the ARE had a negative temperature coefficient of 1.75 × 10-4, indicating that loss of fuel from the critical lattice contributes to the mitigation of reactivity. [4]

One initial problem with liquid fuel was that stagnant fuel has a very large temperature gradient because of its poor thermal conductivity: fuel at the center of a 2-mm-wide (inside diameter) pipe could be many hundreds of degrees (F) hotter than the wall of the tube. Consequently, the ARE was redesigned so than the fuel itself circulated through the core and through heat exchangers. It was found that 66 parallel passes through the core was optimal for keeping the Reynolds number above the transition temperature in order to ensure turbulent flow. Additionally, the pipes were double-walled, with helium in the outer chamber for better heat distribution. [3]

Control Rods

Reactor control was achieved with shim rods of boron carbide. Because the configuration of moderator blocks makes the fuel undrainable, this "poison" must be neutron-absorbing enough to make the reactor subcritical at room temperature. The rods were held above their slots in the moderator blocks by electromagnets so that in case of power failure, the rods would fall down into the core, causing an automatic shutdown, or scram. If the control rod fell, compressed gas formed a pneumatic shock absorber at the bottom of the channel. The rods were constructed in short segments so distortion of the channel would not cause the shims to stick. [3]

Coolant

Fluorides were used as coolant salts because fluorine has a small absorption cross-section; at thermal neutron energies it is close to that of beryllium. Fluorides also give the highest attainable average log energy loss per collision - 0.10 - while still allowing a reasonable critical mass. [4]

© Suraya Omar. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.

References

[1] J. C. Warf, All Things Nuclear (Figueroa Press, 2005).

[2] E. Bettis et al., "The Aircraft Reactor Experiment-Operation," Nucl. Sci. Eng. 2, 841 (1957).

[3] E. S. Bettis et al., "The Aircraft Reactor Experiment-Design and Construction," Nucl. Sci. Eng. 2, 804 (1957).

[4] W. K. Ergen et al., "The Aircraft Reactor Experiment-Physics," Nucl. Sci. Eng. 2, 826 (1957).