CENTRIFUGES

• Centrifuges
• Electromagnetic Separation
• Gaseous Diffusion
• Thermal Diffusion
• Separation Hazards and Wastes

Many scientists initially considered the best hope for uranium isotope separation to be the high-speed centrifuge, a device based on the same principle as the cream separator. The centrifuge operates using the familiar principle of centrifugal force. This force separates two gases of unequal masses in a spinning cylinder or tube. The heavier uranium-238 isotope collects at the outer edges of the cylinder while the lighter uranium-235 collects near the axis of rotation at the center. The part enriched with uranium-235 can be collected from one end, while the part with the concentration of uranium-238 can be removed through the other end. A high degree of separation, scientists hoped, could be attained by running these concentrations through a series, or cascade, of many centrifuges.

By the late 1930s, Jesse W. Beams at the University of Virginia had developed a high-speed centrifuge with which he achieved significant separation of chlorine isotopes. Beams nonetheless had encountered some real difficulties. Spinning a short tube was easier than spinning a long tube, although long tubes were more efficient. Speed was limited by the strength of the rotating tube. Particularly troublesome were the vibrations encountered at certain critical velocities that threatened to shake machines to pieces. Scientists did not think these difficulties too serious, however, and of the various possible methods the centrifuge alone seemed to offer much hope of success. With funding provided by the Navy in 1940, Beams spun tubes of various sizes and tested methods of applying the principle. When uranium hexafluoride became available, he achieved concentrations of uranium-235 but at a yield not as high as theory had predicted. At the same time, Harold C. Urey at Columbia University headed an effort to develop a centrifuge suitable for industrial operations. By early 1941, the team had designed an experimental unit of encouragingly short length that the Westinghouse Electric and Manufacturing Company agreed to construct.

Although the British in their influential July 1941 MAUD report were skeptical of the centrifuge method because it required precision machinery of a type that had been attained only in a laboratory instrument, American scientists remained optimistic. Arthur Compton's National Academy of Sciences committee reported in November that the centrifuge appeared practical and was further advanced than the method preferred by the British, gaseous diffusion. At its first meeting in January 1942, Eger V. Murphree's Planning Board for overseeing engineering studies and supervising pilot plant construction recommended that Westinghouse be authorized to proceed with 24 centrifuges to be installed in a pilot plant that was to be operated by the Standard Oil Development Company and located at the Standard Oil Bayway Refinery at Linden, New Jersey. Transitioning to a full-scale industrial plant, however, appeared problematic. Urey estimated that a plant producing a kilogram per day of very pure uranium-235 would need 40,000 to 50,000 centrifuges. A smaller plant producing 100 grams per day was deemed more feasible as it would require only 6,000 centrifuges, and Westinghouse indicated that it could produce centrifuges at a rate of 1,000 per month beginning six months after the government placed the order.

By fall 1942, when decisions needed to be made on full-scale production plants, the outlook for the centrifuge appeared grim. With Beams's experimental centrifuge separating at only 60 percent of the amount predicted from theory, even the 100-gram plant might have to be increased to 25,000 centrifuges. In addition, Westinghouse encountered difficulties with its full-scale model centrifuge. Test operations revealed severe instabilities at critical vibration frequencies. Repeated efforts to develop a gas-tight, corrosion-proof seal were unsuccessful, and frequent failures occurred in motors, shafts, and bearings at the high speeds required.

On November 12, 1942, the Military Policy Committee, acting on General Leslie Groves's and James Conant's recommendations, approved proceeding directly to full-scale plants with the gaseous diffusion and electromagnetic methods. The committee approved continued but limited support for research and development of the centrifuge process. Two centrifuge models were installed and operated at the pilot plant, but on January 19, 1944, Groves informed Conant that further extension of the centrifuge project was not justified.

• Centrifuges
• Electromagnetic Separation
• Gaseous Diffusion
• Thermal Diffusion
• Separation Hazards and Wastes

   Next