What Caused The Demon Core Criticality In Each Incident?

I've always kept a short mental checklist when I tell people about the demon core: the common thread is reflectors and a near‑critical plutonium mass. In both incidents the core was already close to critical, and someone changed the neutron leakage — once by dropping tungsten carbide bricks around it, once by letting two beryllium halves close together when a screwdriver slipped. That drop in leakage raised the neutron multiplication factor above 1 and produced a prompt burst of neutrons and gamma radiation.

Both events were human‑error driven experiments with manual control of a dangerous configuration; the physics cause is the same, the handling was reckless by today’s standards, and both resulted in fatal acute radiation doses. It’s a sobering example of how a small mechanical change around fissile material can have catastrophic radiological consequences, and why remote procedures and strict protocols are non-negotiable now.

When I first dug into the demon core stories I felt like I was reading a tragic science thriller — and the science behind each incident is actually pretty straightforward once you strip away the drama. In the August 1945 accident, the experimenter was stacking tungsten-carbide bricks around a plutonium core to act as a neutron reflector. Reflectors bounce escaping neutrons back into the fissile material, which reduces the critical mass needed for a sustained chain reaction. One misplaced brick completed a reflective geometry that pushed the neutron multiplication factor (k) above 1, creating a prompt supercritical event. The burst of neutrons was intense but brief; the technician disassembled the stack and walked away, but had received a lethal acute dose and died weeks later.

The May 1946 incident had a similar causal root but a different hands-on mechanism. Two beryllium hemispheres were being used as a variable reflector around the same core, and an operator was holding them slightly apart with a screwdriver as a sort of improvised shim. The screwdriver slipped, the halves snapped together, and the reflectors instantly made the configuration supercritical. That pulse was even more violent — witnesses reported a blue flash and a rush of heat and radiation — and the operator got a fatal dose in a fraction of a second. In both cases the immediate cause was increasing neutron reflection (reducing leakage) around an already near-critical plutonium mass, but human error and unsafe, manual experimental practice turned what could have been a controlled test into a deadly one. I always think about how those mistakes reshaped lab culture — after that, remote handling and strict protocols became the baseline rather than the exception.

I tend to think about the demon core incidents in clinical, almost mechanical terms, because the root cause in both cases was the same physical lever: change neutron leakage. In nuclear terminology, the core was marginally subcritical under normal conditions. Add a reflector — whether tungsten-carbide bricks or beryllium hemishells — and you suppress neutron leakage; the effective multiplication factor rises. If it crosses unity you get a self-sustaining chain reaction; if it goes significantly above unity in a short time you get a prompt critical excursion, which produces an intense neutron and gamma burst.

What differed between the two accidents was the method of introducing that reflector and the human error involved. In the first event a single dropped brick completed a symmetrical reflective geometry and held long enough for a dangerous pulse. In the second, the reflective halves were being mechanically propped apart and then accidentally allowed to close when a simple tool slipped. Both were essentially shortcuts: manual, makeshift control of a variable reflector rather than remote, engineered interlocks. The physics — neutron economy, prompt versus delayed neutrons, and the role of reflectors — explains the mechanism; the tragedy was compounded by the lack of established safety barriers. Reading about it makes me appreciate how much modern criticality safety borrows from those painful lessons, insisting on distance, shielding, and fail-safe automation.