Supply of Liquid Hydrogen
 From the first tests in 1945 through the second series of rocket experiments in 1947, Aerojet had to use gaseous hydrogen because liquid hydrogen was not available. Starting in early 1946, Aerojet enlarged its facilities to handle gaseous hydrogen and oxygen. Gaseous hydrogen under a pressure of 136 atmospheres was available directly from a trailer of high- pressure tubes with a capacity of 800 cubic meters (at atmospheric pressure) and from a stationary bank of high pressure tubes of about the same capacity. Gaseous oxygen at pressures up to 163 atmospheres was supplied from two trailers with a capacity of 560 cubic meters. The total quantity of gases from these sources allowed only a few minutes of operation-a situation conducive to continued frustrations, as the following incident illustrates. One day the test crew was ready to run the rocket and waiting impatiently for a commercial firm to deliver some needed gas. When it came, the crew quickly connected the trailer to the pipes leading to the test cell and ran the test. Meanwhile, the truck driver had gone to the office to get the delivery ticket validated. On his return he was told the trailer was empty and could be taken back. Used to leaving such trailers for a considerable time at other places, the  driver simply would not believe the crew until it was explained rather forcibly to him. He departed with the trailer, shaking his head.33
By early 1947, the Aerojet group was planning ahead to the next phase of hydrogen-oxygen experimentation and acutely felt the handicap of not having a supply of liquid hydrogen. Envying their former associate, Marvin Stary at Ohio State University, with his assured supply of liquid hydrogen from the Johnston liquefier, they decided to attack the problem directly. They discussed liquid hydrogen with several possible users on the West Coast and the idea blossomed into a proposed cooperative venture among several government agencies, universities, and industrial firms. Confident that they could get liquid hydrogen-and having gone to as high a thrust as was reasonable with gaseous hydrogen-the Aerojet engineers proposed to use liquid hydrogen in their third series of experiments starting in July 1947. They went even further and proposed to build a flyable rocket engine, complete with its own controls and turbine-driven pumps. They also recommended that the government build a medium-scale hydrogen liquefier on the West Coast.
Aerojet got its new contract in July 1947, but immediately faced a problem: the cooperative venture to get liquid hydrogen failed to materialize. Aerojet decided to try to interest private industry in supplying liquid hydrogen, and if that failed, to get authority and funding from the Navy to build a liquefier. The first step was to get an estimate of the amount of liquid hydrogen needed . The Jet Propulsion Laboratory agreed to participate and estimated a need for 600-900 kilograms a year. Aerojet added their needs and settled on a 3600-kilogram total requirement for two years. Three possible commercial sources were then queried. The Shell Development Corporation could not supply liquid hydrogen, but had a surplus of high-purity gaseous hydrogen for sale. The National Cylinder Gas Company believed that the sale of liquid hydrogen was neither economical nor safe and recommended liquefaction at the point of consumption. The Linde Air Products Company submitted an oral bid for $62 per kilogram at their plant in Los Angeles, but later lowered the price to $55 per kilogram for the first 1800 kilograms and $44 thereafter.
While soliciting industry, Aerojet began investigating the possibility of building a liquefier modeled after Johnston's and estimated that it would cost $100 000, including the cost of the liquefier, materials, and labor for producing 3630 kilograms of liquid hydrogen. This was half the revised Linde estimate and had the added advantage of being under Aerojet control and located near the rocket test stand. Aerojet officials became enthusiastic over the prospect and set about convincing the Navy. By late September they received oral approval, which was formalized on 16 December 1947. Aerojet engaged Johnston as a design consultant; he was also to supply parts of the lique'lier. Herman L. Coplen was the principal Aerojet engineer for design, construction, and operation.
Aerojet expected to have the liquefier in operation by late spring or early summer. As so often happens, the optimistic schedule fell victim to late equipment deliveries. However, the liquefier produced its first liquid hydrogen-12 liters-on 3 September 1948. The initial operation turned up the usual number of bugs; the second operation on 21-23 September produced 120 liters. Of this, 75 liters were shipped to the Jet Propulsion Laboratory for rocket tests there.
 Aerojet was pleasantly surprised to find that the actual capacity of the liquefier was 30 liters per hour instead of the design value of 25. The increased capacity came from a larger hydrogen compressor; the Johnston-built heat exchangers were oversized. This led Aerojet to propose, in early 1949, the doubling of the liquefaction capacity by installing additional hydrogen compressors.
At first, the liquefier was operated intermittently. Beginning on 8 November, a two-shift operation was begun to meet the needs of the rocket test engineers, and from 27 December three shifts were employed. By the end of 1948, 7500 liters (535 kg) of liquid hydrogen had been produced, over 90 percent of it in November and December. Only about 30 percent of the hydrogen liquefied was used in test operations; the bulk was lost during storage and test delays.
In the first three months of operation, the liquefier was shut down twice, but the troubles were quickly fixed; the time lost was four days. Overall, the liquefier was highly successful and made possible the testing of pumps and thrust chambers.
By the end of March, Coplen had added two more compressors and the liquefaction rate rose to 80 liters (5.67 kilograms) per hour. But early March had brought catastrophic news to the liquid hydrogen producers. On 2 March 1949, the Bureau of Aeronautics directed Aerojet to change fuels from liquid hydrogen to anhydrous hydrazine, which is a liquid at room temperature and pressure.* The directive allowed Aerojet to continue liquid hydrogen testing until the end of June, but the irony was that the switch came just as the producers of liquid hydrogen were finally prepared to meet rocket test needs.
In its operations through June 1949, the Aerojet liquefier produced 47 000 liters (3357 kilograms) of liquid hydrogen at an estimated cost of $29.72 per kilogram. The cost of commercial gaseous hydrogen and liquid nitrogen were major expenses.
Sometime after the contract ended in mid-1949, Aerojet received a government directive to dismantle and prepare the liquefier for shipment. Very few at Aerojet knew, but the liquefier was destined for reassembly on a remote Pacific isle for use in the first test of a thermonuclear device, the predecessor of the hydrogen bomb.
* The author has been unable to pin down the reason for this sudden change, but it is not surprising. Hydrazine is storable and considerably easier to handle than liquid hydrogen, its performance is high, and interest in it during the 1940s and 1950s was high. For example, Canright, in his analysis of relative importance of exhaust velocity and density, preferred hydrazine to hydrogen even though hydrogen gave higher performance (pp. 47-48) .