Neddermeyer discarded Serber and Tolman's initial concept of implosion as assembling a series of pieces in favor of one in which a hollow sphere was imploded by an explosive shell. He was assisted in this work by
Hugh Bradner,
Charles Critchfield, and John Streib.
L. T. E. Thompson was brought in as a consultant and discussed the problem with Neddermeyer in June 1943. Thompson was skeptical that an implosion could be made sufficiently symmetric. Oppenheimer arranged for Neddermeyer and
Edwin McMillan to visit the
National Defense Research Committee's Explosives Research Laboratory near the
laboratories of the
Bureau of Mines in
Bruceton, Pennsylvania (a
Pittsburgh suburb), where they spoke to
George Kistiakowsky and his team. But Neddermeyer's efforts in July and August at imploding tubes to produce cylinders tended to produce objects that resembled rocks. Neddermeyer was the only person who believed that implosion was practical, and only his enthusiasm kept the project alive. , beside the
Bockscar B-29 that dropped the original device – black liquid asphalt sealant was sprayed over the original bomb casing's seams, simulated on the mockup. Oppenheimer brought
John von Neumann to Los Alamos in September to take a fresh look at implosion. After reviewing Neddermeyer's studies, and discussing the matter with
Edward Teller, von Neumann suggested the use of high explosives in
shaped charges to implode a sphere, which he showed could not only result in a faster assembly of fissile material than was possible with the gun method, but greatly reduce the amount of material required because of the resulting higher density. The idea that, under such pressures, the plutonium metal would be compressed came from Teller, whose knowledge of how dense metals behaved under heavy pressure was influenced by his pre-war theoretical studies of the
Earth's core with
George Gamow. The prospect of more-efficient nuclear weapons impressed Oppenheimer, Teller, and
Hans Bethe, but they decided that an expert on explosives would be required. Kistiakowsky's name was immediately suggested, and Kistiakowsky was brought into the project as a consultant in October. The implosion project remained a backup until April 1944, when experiments by
Emilio G. Segrè and his P-5 Group at Los Alamos on the newly reactor-produced plutonium from the
X-10 Graphite Reactor at
Oak Ridge and the
B Reactor at the
Hanford Site showed that it contained impurities in the form of the
isotope plutonium-240. This has a far higher spontaneous fission rate and radioactivity than
plutonium-239. The
cyclotron-produced isotopes, on which the original measurements had been made, held much lower traces of plutonium-240. Its inclusion in reactor-bred plutonium appeared unavoidable. This meant that the spontaneous fission rate of the reactor plutonium was so high that pre-detonation was highly likely and that the bomb would blow itself apart during the initial formation of
critical mass, creating a "
fizzle." The distance required to accelerate the plutonium to speeds where pre--detonation would be less likely would need a gun barrel too long for any existing or planned bomber. The only way to use plutonium in a workable bomb was therefore implosion. The impracticability of a gun-type bomb using plutonium was agreed at a meeting in Los Alamos on 17 July 1944. All gun-type work in the Manhattan Project was re-directed towards the Little Boy,
enriched uranium gun design, and the Los Alamos Laboratory was reorganized with almost all of the research focused on the problems of implosion for the Fat Man bomb. The idea of using shaped charges as three-dimensional
explosive lenses came from
James L. Tuck and was developed by von Neumann. The success of the bomb relied on absolute precision in all of the plates moving inward at the same time. To overcome the difficulty of synchronizing multiple detonations,
Luis Alvarez and
Lawrence Johnston invented
exploding-bridgewire detonators to replace the less precise
primacord detonation system.
Robert Christy is credited with doing the calculations that showed how a solid subcritical sphere of plutonium could be compressed to a critical state, greatly simplifying the task, since earlier efforts had attempted the more-difficult compression of a hollow spherical shell. After Christy's report, the solid-plutonium core weapon was referred to as the "
Christy Gadget". The task of the
metallurgists was to determine how to cast plutonium into a sphere. The difficulties became apparent when attempts to measure the density of plutonium gave inconsistent results. At first contamination was believed to be the cause, but it was soon determined that there were multiple
allotropes of plutonium. The brittle α phase that exists at room temperature changes to the plastic β phase at higher temperatures. Attention then shifted to the even more malleable δ phase that normally exists in the range. It was found that this was stable at room temperature when alloyed with aluminum, but aluminum emits neutrons when bombarded with
alpha particles, which would exacerbate the pre-ignition problem. The metallurgists then hit upon a
plutonium–gallium alloy, which stabilized the δ phase and could be
hot pressed into the desired shape. They found it easier to cast hemispheres than spheres. The core consisted of two hemispheres with a ring with a triangular cross-section between them to keep them aligned and prevent jets forming. As plutonium was found to corrode readily, the sphere was coated with nickel. (Fat Man test unit) being raised from the pit into the bomb bay of a B-29 for bombing practice during the weeks before the attack on Nagasaki The size of the bomb was constrained by the available aircraft, which were investigated for suitability by
Norman Foster Ramsey. The only Allied bombers considered capable of carrying the Fat Man without major modification were the British
Avro Lancaster and the American
Boeing B-29 Superfortress. British scientist
James Chadwick advocated the Lancaster which had a limited range but had a larger single bomb bay; but this was less of a problem when the Fat Man replaced the long (
Thin Man. At the time, the B-29 represented the epitome of bomber technology with significant advantages in
maximum takeoff weight, range, speed, flight ceiling, and survivability. Without the availability of the B-29, dropping the bomb would likely have been impossible. However, this still constrained the bomb to a maximum length of , width of and weight of . Removing the bomb rails allowed a maximum width of . Drop tests began in March 1944 and resulted in modifications to the Silverplate aircraft due to the weight of the bomb. High-speed photographs revealed that the tail fins folded under the pressure, resulting in an erratic descent. Various combinations of stabilizer boxes and fins were tested on the Fat Man shape to eliminate its persistent wobble until an arrangement dubbed a "California Parachute" was approved, a cubical open-rear tail box outer surface with eight radial fins inside of it, four angled at 45 degrees and four perpendicular to the line of fall holding the outer square-fin box to the bomb's rear end. In drop tests in early weeks, the Fat Man missed its target by an average of , but this was halved by June as the bombardiers became more proficient with it. The early Y-1222 model Fat Man was assembled with some 1,500 bolts. This was superseded by the Y-1291 design in December 1944. This redesign work was substantial, and only the Y-1222 tail design was retained. Later versions included the Y-1560, which had 72 detonators; the Y-1561, which had 32; and the Y-1562, which had 132. There were also the Y-1563 and Y-1564, which were practice bombs with no detonators at all. The final wartime Y-1561 design was assembled with just 90 bolts. On 16 July 1945, a Y-1561 model Fat Man, known as the Gadget, was detonated in a
test explosion at a remote site in
New Mexico, known as the "
Trinity" test. It gave a yield of about . Some minor changes were made to the design as a result of the Trinity test.
Philip Morrison recalled that "There were some changes of importance... The fundamental thing was, of course, very much the same." For certain non-atomic parts the manufacturer had not met the delivery schedule, so many tests had to be done twice over, one with all components except for the missing item, and at a critically late date with the complete assembly minus the nuclear components. So the first live tests of the missing part were conducted only a few days before it was dropped on Nagasaki. ==Interior==