Experiments with hydrogen and oxygen in a rocket began at Ohio State University on 2 April 1947 and ended 29 May 1950. Similar tests were also underway at Aerojet General Corporation in California from 1945 to 1949 and at the Jet Propulsion Laboratory of the California Institute of Technology from 1948 to 1949, which will be described in the next chapter.
At Ohio State, the first twelve tests were made with liquid hydrogen and gaseous oxygen, because the installation of a liquid-oxygen tank at the test cell had been delayed. On 13 June 1947, Stary and his staff made the first rocket engine test in the United States using liquid hydrogen and liquid oxygen. The engine produced 471 newtons (106 lb of thrust) at a chamber pressure of 21.1 atmospheres with an oxygen-to-hydrogen mass ratio of 4.2. Exhaust velocity was 2405 meters per second, or 82 percent of the theoretical performance for that ratio (according to the theoretical performance given in the Lemmon report). Following this, an additional 118 runs were made with the same engine, and beginning in September 1948, 38 runs were made with an engine five times larger. One of the most significant accomplishments was a series of 37 runs at the smaller thrust using an engine regeneratively cooled with liquid hydrogen, starting of 26 August 1949. These will be discussed later.
An early problem for all rocket experimenters was satisfactory instrumentation to measure thrust, mass flow rates of fuel and oxidizer, and combustion pressure. From  these the exhaust velocity at a given mixture ratio can be obtained and compared with theoretical calculations.* Of these measurements, the mass flow rate of liquid hydrogen was of most concern. It was determined by measuring the pressure differential across a sharp-edged orifice-a time-honored method of measuring flow rates. The accuracy depends upon the density of the fluid being measured; for liquid hydrogen, the large density changes with temperature are lessened at the high pressures used in rocket experiments. Measurements of hydrogen properties by Johnston, David White, and others were going on in parallel with the rocket work. Using Ohio State's temperature-density data, Johnston and Doyle reported that if the temperature of liquid hydrogen in a tank increased from its normal boiling point to the critical point at 25 atmospheres, the flow measurement would be approximately 10 percent too high. For this reason and because improvements in measurements were made as the tests progressed, strict comparisons of the various runs were not made, but qualitative comparisons were made to show trends.9
The major design element affecting high performance is the propellant injector. Stary came to Ohio State from Aerojet General Corporation where impinging jet injectors had been successfully used with other propellants. In this type of injector, streams of propellant are directed so as to impinge on each other to break up the liquid stream into fine droplets and mix the fuel and oxidizer (fig. 5). This was the prevailing design philosophy of the period, and it is not surprising that major emphasis was placed on this type of injector for hydrogen-oxygen at Ohio State. In fact, 18 out of 20 injectors at Ohio State used impinging jets for at least one, and usually both, propellants. The two exceptions, a concentric tube and a "showerhead," were not tested. Ironically, these were later found by the Aerojet team and other investigators to be best for liquid hydrogen.
* Rocket experiments used specific impulse (thrust divided by total propellant flow rate) for determining performance, which is equivalent to exhaust velocity used in this text (appendix B)