<P> The most obvious technology, the centrifuge, failed, but electromagnetic separation, gaseous diffusion, and thermal diffusion technologies were all successful and contributed to the project . In February 1943, Groves came up with the idea of using the output of some plants as the input for others . </P> <P> The centrifuge process was regarded as the only promising separation method in April 1942 . Jesse Beams had developed such a process at the University of Virginia during the 1930s, but had encountered technical difficulties . The process required high rotational speeds, but at certain speeds harmonic vibrations developed that threatened to tear the machinery apart . It was therefore necessary to accelerate quickly through these speeds . In 1941 he began working with uranium hexafluoride, the only known gaseous compound of uranium, and was able to separate uranium - 235 . At Columbia, Urey had Karl Cohen investigate the process, and he produced a body of mathematical theory making it possible to design a centrifugal separation unit, which Westinghouse undertook to construct . </P> <P> Scaling this up to a production plant presented a formidable technical challenge . Urey and Cohen estimated that producing a kilogram (2.2 lb) of uranium - 235 per day would require up to 50,000 centrifuges with 1 - meter (3 ft 3 in) rotors, or 10,000 centrifuges with 4 - meter (13 ft) rotors, assuming that 4 - meter rotors could be built . The prospect of keeping so many rotors operating continuously at high speed appeared daunting, and when Beams ran his experimental apparatus, he obtained only 60% of the predicted yield, indicating that more centrifuges would be required . Beams, Urey and Cohen then began work on a series of improvements which promised to increase the efficiency of the process . However, frequent failures of motors, shafts and bearings at high speeds delayed work on the pilot plant . In November 1942 the centrifuge process was abandoned by the Military Policy Committee following a recommendation by Conant, Nichols and August C. Klein of Stone & Webster . </P> <P> Electromagnetic isotope separation was developed by Lawrence at the University of California Radiation Laboratory . This method employed devices known as calutrons, a hybrid of the standard laboratory mass spectrometer and cyclotron . The name was derived from the words California, university and cyclotron . In the electromagnetic process, a magnetic field deflected charged particles according to mass . The process was neither scientifically elegant nor industrially efficient . Compared with a gaseous diffusion plant or a nuclear reactor, an electromagnetic separation plant would consume more scarce materials, require more manpower to operate, and cost more to build . Nonetheless, the process was approved because it was based on proven technology and therefore represented less risk . Moreover, it could be built in stages, and rapidly reach industrial capacity . </P>

Who are the scientists involved in the development of atom