On Dec. 2 1942, under the stands at the University of Chicago’s Stagg Field football stadium, Nobel laureate Enrico Fermi led an experimental team that produced humankind’s first controlled nuclear chain reaction — an event that marked the dawn of the nuclear era, enabling the development of the first atomic bomb and the first nuclear power reactors.
To commemorate the first criticality of the Chicago Pile (CP-1), exactly 75 years later, MIT on Saturday restored a device similar the one used for that epochal event in Chicago. Researchers celebrated by restoring an MIT subcritical experimental facility, which is similar to those used during development of the CP-1 reactor and its landmark sustained nuclear chain reaction.
The historic experiment’s re-enactment was not merely a novelty. The researchers have revived a device, called a graphite exponential pile and originally built in 1957, that over the coming years will provide hands-on access to subcritical nuclear experiments for MIT’s students, and serve as a unique and valuable research tool that can be used to study new reactor designs for future nuclear power plants.
The device is essentially just a large cube-shaped pile of blocks made of pure graphite — the material used as the “lead” of a pencil — with holes drilled through to allow insertion of rods of uranium. These natural-uranium rods have such low radiation emissions that they could be safely handled with bare hands, as Fermi and his collaborators did in 1942 (though in this case they will be handled with protective gloves anyway).
In the decades following Fermi’s original experiment, more than two dozen similar graphite pile devices were built at universities and national laboratories around the country and used for basic research and teaching, but over the years most of those have been disposed of. The one at MIT, which though only half as big as Fermi’s original was the largest of these other installations, escaped that fate but had been unused and forgotten for many years, until being “rediscovered” last year by professor Michael Short of MIT’s Department of Nuclear Science and Engineering.
Kord Smith, the KEPCO Professor of the Practice of Nuclear Science and Engineering, was surprised to learn that the device was still intact. Covered in protective metal panels that made it look like a disused storage cabinet, it went unnoticed even by students and faculty working near it. Smith, working with colleagues in the Department of Nuclear Science and Engineering and David Moncton, director of the Nuclear Reactor Laboratory and his staff, quickly formulated a plan to restore the device for the 75th anniversary of the original groundbreaking experiment. MIT nuclear science and engineering student Richard Knapp made the design and construction of the system the subject of his BS thesis in 1957.
Now, with the device and its 30 tons of graphite and 2.5 tons of uranium fully cleaned and restored, the final slugs of uranium were ceremonially slid into place on Dec. 2 to complete the system. This took place before an invited group of 49 faculty, students, and guests — the same number who were present with Fermi in Chicago — at the precise time of the original experiment.
Smith explains that MIT’s subcritical graphite pile originally fell into disuse as the nuclear industry quickly shifted from graphite-based reactor designs to alternatives based on light water, heavy water, or liquid sodium. Experiments with the graphite system were thus seen as less relevant. In these devices, graphite (or water) serves as a moderator that slows down the speed of neutrons emanating from a radiation source, by a factor of more than a million, to get them to interact with other uranium atoms and initiate a self-sustaining chain reaction in which neutrons knock other neutrons out of an atom’s nucleus to create a cascade of collisions. Criticality of the much larger CP-1 graphite pile was controlled by inserting or withdrawing control rods, made of cadmium, to absorb the neutrons and interrupt the reaction.
Today, a wide variety of cutting-edge designs for proposed next-generation nuclear reactors, including designs that have passive cooling systems or continuous operation without requiring shutdowns for refueling, do once again make use of graphite, so the reactor is once again a useful research tool. Such a tool will permit students to actually handle nuclear fuel and be more accessible to students than full-scale nuclear reactors such as MIT’s own research reactor, which runs almost continuously and produces 6 megawatts of thermal power. Experiments done in that reactor, to study new kinds of fuel-rod cladding or new instruments for monitoring the reactions, for example, typically run for a year at a time.
Students will be able to install, run, and get results from experiments in the graphite exponential pile within a few hours or days, Smith says. Use of the graphite pile is anticipated to stimulate students’ interest in, and preparation for, performing cutting-edge experiments on the much more powerful MIT research reactor.
“Graphite as a medium for reactors has come and gone a few times over the years,” he says, but now, “we’re in the midst of a rebirth.” And even today, there are still significant aspects of exactly how neutrons from nuclear reactions scatter through the crystal lattice of graphite. In fact, Smith says, a new physics model to describe these interactions has recently been proposed, and using the graphite pile “we want to design experiments to test these new theoretical models.”
In addition to doing experiments that could help in the development of new reactor designs, fuel, and cladding types, or measurement systems, this device and the MIT reactor will be a valuable educational tools for nuclear engineers, Smith says. “We tend to get students who are very good at developing computational algorithms and models. But if you don’t have something to compare your calculations with, you start to think your simulations are perfect.” In the real world, though, the actual measurements usually don’t agree perfectly with predictions, and understanding such differences often lead to the development of improved theoretical models, he says.