The Manhattan Project: Making the Atomic Bomb
Part IV: The Manhattan Engineer District in Operation
Pile Design: Changing Priorities
The fall 1942 planning sessions at the Met Lab led to the decision to build a second Fermi pile at Argonne as soon as his experiments on the first were completed and to proceed on design of the Mae West helium-cooled unit. When DuPont engineers assessed the Met Lab's plans in the late fall, they agreed that helium should be given first priority. They placed heavy water second and urged an all out effort to produce more of this highly effective moderator. Bismuth and water were ranked third and fourth in DuPont's analysis. Priorities changed when Fermi's calculations demonstrated a higher value for k than anyone had anticipated. Met Lab scientists concluded that a water-cooled pile was now feasible, while DuPont shifted its interest to air cooling. Since a helium-cooled unit shared important design characteristics with an air-cooled one, Greenewalt thought that an air-cooled semiworks at Oak Ridge would contribute significantly to designing the full-scale facilities at Hanford.
DuPont established the general specifications for the air-cooled semiworks and chemical separation facilities in early 1943. A massive graphite block, protected by several feet of concrete, would contain hundreds of horizontal channels filled with uranium slugs surrounded by cooling air. New slugs would be pushed into the channels on the face of the pile, forcing irradiated ones at the rear to fall into an underwater bucket. The buckets of irradiated slugs would undergo radioactive decay for several weeks, then be moved by underground canal into the chemical separation facility where the plutonium would be extracted with remote control equipment.
Met Lab activities focused on designing a water cooled pile for the full-scale plutonium plant. Taking their cue from the DuPont engineers, who utilized a horizontal design for the air-cooled semiworks, Met Lab scientists abandoned the vertical arrangement with water tanks, which had posed serious engineering difficulties. Instead they proposed to place uranium slugs sealed in aluminum cans inside aluminum tubes. The tubes, laid horizontally through a graphite block, would cool the pile with water injected into each tube. The pile, containing 200 tons of uranium and 1,200 tons of graphite, would need 75,000 gallons of water per minute for cooling.