Part I : 1945 - 1950

2. Air Force Research on Hydrogen



Rocket-Engine Cooling


[21] In rocket experiments, the measurement of the heat flowing from the combustion gases to the engine walls and the use of this information to devise satisfactory cooling of the engine are second only to obtaining maximum performance. Without cooling, a flight-weight rocket engine would be heated to its melting point in a second or two. Major factors affecting this heat transfer are gas temperature, density, and velocity; all three of these are much higher in rocket engines than in other internal combustion engines. These factors, plus gas composition, are functions of the propellants, engine design, and operating conditions. The particular fuel and oxidizer, the proportions used, combustion pressure, and combustion efficiency determine gas composition, temperature, and density. Injector design, propellant proportions, mass flow, and combustion chamber design affect gas velocity. The rocket engineer seeks a design giving both high performance and a cooling method for steady-state operation. He is aided by combustion characteristics, for peak performance usually occurs at a fuel-rich mixture where the heat transfer is lower than at a leaner mixture.


drawing of injector designs

[22] Fig.5. Experiment rocket engines using hydrogen-oxygen, Ohio University, 1947-1950.


[23] Heat transfer measurements at Ohio State used two techniques common in rocket experiments. In the "heat-sink" method, the combustion chamber and nozzle are made from a high-conductivity material, usually copper, in which a thermocouple to measure temperature is buried in the thick, uncooled wall. During rocket operation, the high thermal conductivity of the copper keeps the inside wall from melting as the heat rapidly flows into the interior of the mass. This allows a rocket to operate for a few seconds, and sometimes as long as 30 seconds. After the run, the temperature of the copper mass comes to equilibrium and by measuring this temperature, the total amount of heat absorbed can be calculated from the known mass and specific heat of the copper. In the second method, a water jacket surrounds comparatively thin engine walls and a high-velocity water flow keeps the walls cool. The average heat transfer can be obtained by measuring the water flow and its temperature rise. Using these methods, Ohio State measured average heat transfer rates of about 1.6 joules per second per square meter (1 Btu/sec-sq in) for the combustion chamber and about twice that for the nozzle. These values were on the same order as found in high-performance rocket engines using other propellants, but are several times higher than heat transfer rates in other types of internal combustion engines and are, for example, from 20 to 200 times higher than in steam plants.


In mid-1948 a mechanical engineer from Aerojet, Irwin J. Weisenberg, joined the Ohio State rocket staff under Stary and specialized in heat transfer and cooling experiments. The first attempt to use hydrogen as a coolant was to employ a porous combustion chamber wall and force hydrogen through the wall into the combustion chamber.10 This type of cooling, called transpiration or "sweat" cooling, was popular at the time and work with it was under way at several other rocket laboratories.


In the first part of 1949, another engineer at the Ohio State rocket laboratory, Clair M. Beighley, made a theoretical analysis in which a temperature ratio involving combustion gas temperature, wall temperature, and coolant temperature was related to dimensionless flow parameters. A porous combustion chamber was tested later and the experimental data agreed with the theoretical predictions. Porous wall chambers with uniform permeability were difficult to make, however, and the Ohio State rocket engineers turned to regenerative cooling when an analysis showed it to be feasible. In this method, hydrogen is circulated in coolant passages surrounding the engine prior to injection and burning.


In the midst of preparations to try it experimentally (in June 1949) Stary returned to Aerojet and still another Aerojet engineer, Dr. Willard P. Berggren, arrived at Ohio State as the new chief engineer for rocket experiments.11


The experimental thrust chamber for regenerative cooling was designed to produce 445 newtons at a chamber pressure of 20.4 atmospheres (fig. 6). Liquid hydrogen in the coolant jacket would be well above this value and hence far above its critical pressure of 12.8 atmospheres so that no boiling could occur in the coolant passages. The first successful regenerative cooling run was on 26 August 1949, when the thrust chamber operated for 60 seconds at an oxygen-to-hydrogen mass ratio of 4.1 and produced an exhaust velocity of 3190 meters per second-about 93 percent of theoretical performance.


In all, 33 successful runs were made, over half of which operated for 60 or more seconds; one operated for 159 seconds. The runs covered a range of mixture ratios and...


technical drawing of thrust regenerative motor

[24] Fig. 6. Rocket thrust chamber of 445 newtons designed to use liquid hydrogen-oxygen and be regeneratively cooled by the liquid hydrogen, Ohio State University, 1949. Scale and dimensions are inches. (Courtesy of 1. J. Weisenberg.)


....the maximum exhaust velocity for the series was 3270 meters per second.* In general, performance with the regeneratively-cooled engine was considerably higher than that obtained with the water-cooled chambers. The experimenters attributed this not only to the elimination of heat losses, but also to a lower-density hydrogen entering the combustion chamber, which produced improved mixing and higher combustion efficiency. Figure 7 shows the regeneratively-cooled rocket operating in December 1949 during the series of tests. The frost on the chamber indicates that it was well cooled.12


photo of rocket test in laboratory

[25] Fig. 7. Liquid hydrogen-oxygen rocket engine regeneratively cooled by the hydrogen. Ohio State University, December 1949. Note the frost on the outside of the rocket chamber and the shock diamonds in the exhaust. (Courtesy of I.J. Weisenberg.)


* The highest performance run lasted 90 seconds at a fuel-rich mixture (0/ F,4.7), 21 atm, and a relatively low overall heat transfer rate of 2.1 J/s . m2. In contrast, the longest run (159 sec.) was at the stoichiometric mixture (0/ F,8), 19.6 atm, much lower exhaust velocity (2800 m/s), but almost triple the overall heat transfer rate (5.2 J/s . m2). The comparison illustrates that peak performance does not come at the same operating conditions as maximum heat transfer. It also shows that hydrogen cooling handled the higher heat load.