Aerojet's Second Series of Experiments, 1946-1947
 In addition to the rocket engine design study, Aerojet's contract that began in July 1946 called for experiments with a gaseous hydrogen-liquid oxygen thrust chamber. The thrust was 4.5 kilonewtons (1000 lb) and the minimum exhaust velocity was specified as 2940 meters per second. Moreover, the engine was to operate continuously  for three minutes. The chief experimenters were Robert Gordon and Herman L. Coplen, reporting to David Young. By the end of the twelve month period they had met the specified performance.
 The thrust chamber had a waterjacket and an inner liner of porous material through which the water seeped and evaporated on the inner surface for cooling. The shape was the flared tube design (fig. 8), having in this case a chamber diameter of 5 centimeters and overall length of 21. The gaseous hydrogen was injected through a series of holes to form a cone in the chamber, and the gaseous oxygen was injected radially inward to intercept the hydrogen cone. The combination of this injector and the flared tube design produced very high heat transfer rates-several times higher than normally experienced in rocket experiments. This led to a separate investigation of the characteristics of the flared tube by Gordon using a smaller engine independently cooled with water. Gordon found that high performance (95 percent of theoretical) could be obtained with the design, but the combustion pressure was not constant as in a conventional plenum chamber; it dropped rapidly throughout the length of the flared tube chamber.27 The average heat transfer rates were much higher than those of a plenum chamber.*
Instead of reconsidering their basic engine design, the Aerojet men focused most of their attention on cooling. They tried a dozen different porous materials. Porous nickel made by the Amplex Division of the Chrysler Corporation proved to be the best. An attempt was made to match the water flow through the porous liner with the large variation of heat transfer rate along the combustion chamber and nozzle, but this was only partially successful. The best they could do was to use almost twice as much coolant as they had originally calculated to be necessary. This was a matter of some concern, as the water entering the combustion chamber diluted the propellant and lowered performance, for its mass had to be considered in determining thrust per unit mass flow or its equivalent, exhaust velocity. (One percent increase in water flow decreased the exhaust velocity by 0.75 percent.) To make up for the drop in performance, the combustion pressure was increased, which increased gas expansion and exhaust velocity. On 26 June 1947, four days after expiration of the contract, the performance objective was achieved on the 46th run, which lasted over three minutes.28
With these experiments, Young, Gordon, and Coplen were still confident that their 1.3 meganewton (300 000 lb thrust) design study was sound, although they had yet to operate a rocket using liquid hydrogen and oxygen or to cool a hydrogen-oxygen rocket with hydrogen rather than water.
* The average rate was 13 J/s . m2 ; the section just before the nozzle, a peak of 29 was measured. Pressure was 20 atm at the injector end and the mixture was fuel rich (oxidizer to fuel mass ratio of 5). The average heat transfer rate was about 6 times greater than Ohio State's values when the latter used a plenum chamber at about the same operating conditions and performance (fn.. p. 24). Some of the difference can be attributed to the much greater gas velocities in the flared tube as well as the different types of propellant injection.