It might be desirable, before it is too late, for the surviving remnants, including me, who took part in the Manhattan Project and the engineering design and initial operation of the first Hanford reactor, to add some footnotes to Alvin Weinberg’s excellent summary of Eugene Wigner’s (and his own) contribution to the design and start-up of the first production reactor (see Physics Today, Physics Today 0031-9228 55 10 2002 42 https://doi.org/10.1063/1.1522166 October 2002, page 42 ). In common with the Richard Rhodes volume The Making of the Atomic Bomb (Simon & Schuster, 1986), Weinberg’s article largely skipped over or minimized the role of the DuPont company engineers in making the process viable.
Hood Worthington, the head of the DuPont design effort, actually saved the day. The design calculations employed the measured absorption cross sections of all the materials used in the reactor to calculate the critical mass. That calculation gave a minimum of about 1500 fully loaded fuel–coolant channels. Of course, more than the minimum number of channels were incorporated into the design to allow some safety margin. Somewhat more than the estimated critical number were loaded with fuel before start-up. Remarkably, that quantity turned out to be nearly correct when the control rods were displaced. Such accuracy was a triumph, considering that the measurements had all been taken at or close to zero power. However, as is well known, the reactor began to shut down spontaneously within a few hours. Adding fuel obtained the same result, although the shutdown happened more slowly. It then became a race to see which effect would dominate: the additional fuel mass or the buildup of some unknown poison. If the poison buildup had been dominant even up to the core capacity, the reactor, with 10-foot thick shielding, would have been inoperable. The course of the war in the Pacific would have been very different. An American invasion of Japan probably would have been necessary.
Fortunately, following normal chemical engineering practice at the time, Worthington had decreed a one-third margin of overcapacity. There were about 2000 coolant channels, and the reactor leveled out to a steady power output around 1950 channels. The poison culprit, which had not been observable at prior energy levels, was identified as xenon-135.
Another interesting sidelight on the story of the early graphite reactors was the design and procurement of the coolant tubes that contained the fuel slugs. The tubes were made from pure aluminum, except for two longitudinal aluminum-alloy ribs along the bottom of each horizontal tube. Ribs made of pure aluminum would have stuck to the aluminum-clad slugs by diffusion and reaction over several months of exposure to flowing water. The slugs would have been impossible to push out of the reactor without the possibility of loss of integrity.
The production process for those composite tubes was proprietary to Aluminum Co of America (now Alcoa). The company’s draw bench was the only one in the world long enough to produce 50-foot-long tubes. At the time it was being used to draw the main wing spar for the B-59 bombers of the Pacific campaign. The spar consisted of a strong aluminum-alloy body with a thin, pure aluminum covering to resist seawater corrosion. It took considerable effort to convince company officials that three academics from the University of Chicago’s Metallurgical Laboratory understood the problem and furthermore, that they had a directive from the federal government to shut down the wing spar production to make about 2000 tubes, each about 50 feet long, according to their design. Suffice it to say that the rapid reversal of body and thin covering materials was accomplished after some quick experimentation, and things progressed satisfactorily after that.
