potential to be isothermal and is thus quite attractive. Coupled to a superconducting electric motor, it could have a very high efficiency with low hydrocarbon emissions. The water vapor emissions can be reduced to zero, though at some expense in weight.

System weight and thermal efficiency are the fundamental system issues. The three curves in Figure D-1 show cycle efficiency as a function of the ratio of the heat rejection temperature (T2) to the heat absorption temperature (T1). The Carnot cycle is clearly the most efficient, with the Otto cycle second and the Brayton cycle the least efficient. The comparison is somewhat misleading, however, since efficiencies for the Carnot and Otto cycles are per pulse and not the average over the cycle. Carnot and Otto cycle efficiencies are somewhat lower in practice. An additional curve in Figure D-1 shows the efficiency of the Otto cycle if power is produced during only half the cycle.

The isobaric cycles can be continuous or intermittent since the pressure falls during combustion processes. Each process has its limitations. For example, the efficiency of the Brayton cycle is limited by material limits. That is, the compressor exit temperature is limited by high temperature material properties. Nevertheless, the continuous or open Brayton cycle provides a high power density and a relatively simple structure.

Relaxing the constraint on adiabatic compression and expansion (e.g., by using regeneration) almost doubles the number of options. For example, it may be possible for heat

FIGURE D-1 Thermal efficiency of the Otto, Brayton, and Carnot cycles.

removed during compression to be added during expansion in the turbine. With the exceptions described above, the current state of the art for heat exchanger technology is based on surface heat removal, not volume heat removal. The thermal boundary layer is thin even in an axial flow compressor, meaning the heat exchanger is likely to be heavy and create additional pressure losses. Hence until a volumetric heat exchange process is invented, the regenerative part of a modified Brayton cycle seems impractical.

Also, a cooling system upstream of the compressor could inject mist into the air stream. This would have two advantages: The stream would be cooled by evaporation and the total pressure would be increased. This is Ascher Shapiro’s aerothermo compressor. It might even be possible to inject the mist ahead of each compressor rotor and stator blade. In principle the cooling could be adjusted such that the temperature remains constant during compression. In practice, it may be possible to increase the effective compressor efficiency while reducing compressor outlet temperature by 50 to 150 °F. Another option would be to allow combustion in a suitably designed turbine stage, so that heat could be added in a more nearly isothermal fashion. This modified Brayton cycle seems attractive enough to warrant further research.

In summary, there seem to be a large number of alternative propulsion schemes. However, at present few alternatives seem to be practical; only a very few, including modified Brayton cycle engines, seem to warrant more than passing attention.


Zwicky, F. 1959. Future Prospects of Jet Propulsion. Volume XII, Section L, Jet Propulsion Engines: High Speed Aerodynamics and Propulsion. Princeton, N.J.: Princeton University Press.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement