Part I : 1945 - 1950

3. Hydrogen-Oxygen for a Navy Satellite



Aerojet and Martin Design Studies


[42] Aerojet's contract that began July 1946 called for furnishing detailed design information to the North American Aviation and Glenn L. Martin design study groups on a hydrogen-oxygen rocket engine suitable for their vehicles. The thrust of the rocket engine was specified as 1.33 meganewtons (300 000 lb), the exhaust velocity 4165 meters per second, and the mass not more than 1814 kilograms. Aerojet chose a combustion pressure of 34 atmospheres and a hydrogen-to-oxygen molar mixture ratio of 3 to 1. The combustion chamber and nozzle were to be made of porous stainless steel for transpiration cooling. Young's flared tube design concept (fig. 8) was to be used. A greater unknown than the thrust chamber was the turbopump design, and Aerojet concentrated its initial effort there. By mid-October, pump characteristic curves had been determined and a pump speed of 10 000 revolutions per minute selected. Although larger than any previously designed for a rocket engine, the pump would be about the size of the turbines in turbojet engines of the period and not beyond current technology.


The Aerojet design study was completed and reported by the end of March 1947-in time for use in the Martin study but too late for the North American analysis. The...

three rocket nozzle designs

[43] Fig. 8, Aerojet's experimental flared-tube engine (top) had less than a Tenth the combustion volume of a conventional plenum chamber engine (middle) of the same size throat, nozzle, and thrust (4.5 kN or 3000 lb). Below: Aerojet's application of the flared tube concept to the design of a large engine (1.3 MN or 300 000 lb thrust) where the nozzle dwarfs the combustion chamber. Note difference in scales.


....thrust chamber resembled a huge ice-cream cone some 7 meters long; the combustion chamber at the small end was dwarfed by the large conical nozzle (fig. 8, bottom). The inner wall, porous stainless steel, was cooled by hydrogen flowing through it into the combustion chamber. The mass of the chamber, turbopump, and assorted valves and lines added up to 1762 kilograms, comfortably within the specifications.25


[44] The Glenn L. Martin Company had the same general guidelines as North American Aviation for designing a single-stage rocket to orbit a satellite, but they too found that it could not be done within these guidelines.* In striving to do so, Martin's structural designers developed a remarkably ingenious and lightweight structure using pressure-stabilized, thin-wall tanks. With initial vehicle mass only 5 percent greater than specified in the guidelines, they managed to increase the payload by 50 percent over that specified.26


A comparison of the North American and Martin designs is given by table 1. Martin increased the wall thickness of Aerojet's thrust chamber and used a heavier engine than Aerojet furnished. In addition to the thin-wall, pressure-stabilized tanks, the Martin design made the large thrust chamber an integral part of the aft liquid-hydrogen tank, and added four small auxiliary rockets around the nozzle exit for stability and control. The small rockets eliminated the need for external aerodynamic stabilizer fins and movable fins in the hot exhaust stream for thrust-vector control. The idea of surrounding the thrust chamber with the tankage was remarkably similar to Tsiolkovskiy's hydrogen-oxygen spaceship of 1903 (fig. 9).


Using the same basic design, Martin analyzed a family of vehicles with initial mass from 13 600 to 72 600 kilograms with payloads varying from 136 to 780 kilograms. With these the Bureau of Aeronautics had a range of vehicle sizes for possible development.


  TABLE 1. -Comparison of' Single-Stage-to-Orbit Rocket Designs.


North American





Initial Mass (Navy)

45 360

45 360

Payload (Navy)



Engine (Aerojet)

2 268

1 762


Initial Mass

59 000

47 468


52 510

42 484

Final Mass

6 490

4 984

Mass ratio (initial-to-final)







2 268

2 044


3 768

1 791

Instruments for control




* Martin used the same JPL satellite study as North American but chose an initial to flinal mass ratio of 9.52, rather than the 9.09 used by North American.