The Curse of the Demon Core

Neatorama tells the story of two Manhattan Project scientist at Los Alamos who slipped up in deadly ways in performing criticality experiments - experiments to determine how chain reactions propagated in plutonium. Here's a sample:

Daghlian began surrounding the core with bricks of tungsten carbide, a very dense metal that reflects neutron radiation. The more enclosed in metal the core became, the more neutrons were reflected back into the core, rather than simply flying off. That meant that the rate of neutron bashing and atom splitting in the core increased as Daghlian added more and more bricks. (A geiger counter indicated whether the experiment was working, by clicking faster and faster.) Two very important notes:

* Daghlian wanted the chain reaction to increase to just below a critical state, meaning to a controlled chain reaction.

* He did not want the reaction to grow into a supercritical state, meaning one that was escalating completely out of control.

Using the bricks, Daghlian built walls, about ten inches on a side and ten inches high, around the plutonium. He then took a brick and slowly positioned it -he was simply holding it in his hand- over the opening at the top of the structure, right over the core. The geiger counter clicked wildly. Enough neutrons were now being reflected back into the core that it was headed toward a supercritical state.

Daghlian went to jerk the brick away - and dropped it.

The article explains more about the physical processes involved and describes another criticality experiment gone wrong.

One of the people present at the second accident was Raemer Schreiber, who became an executive at the Los Alamos Scientific Laboratory and who made sure that a facility was developed that allowed criticality experiments to be done remotely.

Cheryl Rofer