After a period of indifference, in June 1939 a demonstration of the Power Jets WU was made before a delegation of the Air Ministry, notably Dr Pye, Director of Scientific Research. The demonstration was so successful that the Ministry quickly arranged to buy the engine to give Power Jets working capital, lending it back to them for testing. At the same time, a contract was placed for a "flight engine", the W.1. Unlike the Whittle WU, that began bench testing in 1937, the W.1 was a symmetrical engine designed to facilitate, after development, installation in an aircraft. The W.1 used a double-sided centrifugal compressor of Hiduminium RR.59 alloy, reverse-flow 'Lubbock' combustion chambers and a water-cooled axial-flow turbine section using 72 blades with 'fir-tree' root fixings; the turbine was later modified to use air-cooling. The turbine blades were of Firth-VickersRex 78, a stainless steel developed under Dr. W. H Hatfield. Design rating was at 16,500 rpm, increased to 17,750 rpm above 4,000 feet. With the W.1 aircraft manoeuvring would subsequently be limited by a compressor casing stress to 2g. Maximum jetpipe temperature was 597°C. As development of the new design dragged on, it was decided to build a test unit "early engine" using any components that were deemed unairworthy along with test items. This was assembled to become the one-off W.1X. This officially unairworthy unit powered the Gloster E.28/39 on a short 'hop' during taxiing trials in April 1941, with flight trials taking place a month later with a definitive W.1 engine, the W.1A. After a visit to England in 1941, General Henry H. Arnold arranged for the W.1X to be flown to the U.S in October 1941, along with drawings for the more powerful W.2B engine, together with a team from Power Jets. The former became the prototype of first the General Electric I-A and then the General Electric I-16, and by April 1943 the latter had been developed to produce 1,650 pounds thrust. In 1941, experiments with boosting the W.1's thrust by introducing a liquid coolant were initiated, the first fluid tried being liquid ammonia which proved too effective, resulting in the engine overspeeding and pushing the thrust and rpm indicators off the scales, before later trials changed to using water, and water-methanol. A system to trial the technique in the E.28/39 was devised but never fitted.
