The twentyone collisions collected in the course of that short engineering run were an injection of optimism in the veins of the CDF collaborators, who had worked very hard to put together their detector, under the pressure of the excellent results that the UA1 and UA2 experiments at CERN were achieving. It was evident that the experiment would soon be ready to compete with the other particle physics experiments around the world, and this was a quite strong motivation to get the detector ready for a real data collection run.
Everybody wanted data -lots of data, at the unprecedented energy that the Tevatron could provide. But CDF still needed to be finalized in a number of ways: by the integration of end-cap sections which would allow the measurement of jets emitted at low angle with respect to the beam direction; by inserting in its core the CTC - the "central tracking chamber", a marvelous 3-meter-wide cylinder filled with a argon-methane mixture and instrumented with thousands of sensitive wires which would enable an efficient reconstruction of the tracks of charged particles in the messy environment; and by moving the main ring away from the central detector, which until then it still traversed. Further, the electronics for data acquisition were not in place yet, and a proper triggering system existed only in blueprint. It would take two years to complete the detector and get ready for a real physics run.
Besides the completion of the detector and its electronics, the detector elements already in place and working, which had provided the proof of principle with the twentyone 1.6 TeV collisions, were also in need of a precise calibration; among other things, the response of the hadron calorimeter had to be "equalized". This was done by exposing each of its modules to a reference signal: the possible differences in the total light yield of the stacks of scintillating sheets invested by fixed-energy ionizing particles could be recorded and later used to correct their output. This would become an important ingredient in the measurement of hadronic jets and missing transverse energy.
To determine the response map of a calorimeter element as a function of the position of a particle hit, which varied due to geometrical factors, one could use a beam of particles of fixed energy; once that map was known, subtler differences between different modules of nominally equal geometry were gauged with the use of radioactive sources.
The main procedure consisted in inserting a long thin rod of steel inside a small dedicated hole which ran through the length of each calorimeter wedge. The rod was hollow, and was filled with a radioactive powder, a cesium isotope (137Cs) which released energetic gamma rays. The photons converted to electron-positron pairs in the iron slabs of the calorimeter, and yielded a reference signal in its scintillating sheets: the collection of that signal and the comparison of its intensity across the various detector elements allowed to extract the required intercalibration constants. In principle such a procedure was quite effective, but its practical execution was technically challenging, due to the difficulty of moving around the flexible steel rod.
The cesium source was indeed a rather odd thing, and it duly deserved the name it had been given by the late Aldo Menzione, "la spada sacra": the sacred sword. As any other radioactive source it required careful handling, a detailed bookkeeping of the duration of its use, and the deployment of clear signs to delimit the area where the radiation field produced by the source was significantly above the natural background. Also, one had to conform to strictly controlled procedures for its storage, and the enforcing of a practice called "ALARA": radiological operators had to carefully plan in advance the use of the source, in order to carry out their work by minimizing their own exposure to the radiation, achieving the goal of receiving a dose "As Low As Reasonably Achievable".
All the above were common requirements for the usage of radioactive sources at Fermilab; but a long and flimsy steel rod which emitted gamma rays along its whole length was quite a bit more complex to handle than the typical pocked-sized sources in use throughout the lab. Working with it was fishing for trouble.
The Italians Franco Cervelli and Arnaldo Stefanini, who worked at the calibration of that detector in the summer of 1986, were the protagonists of the worst radiological incident in many years for the whole laboratory; fortunately one which can be now remembered for its rather funny twists rather than for its gravity. Standing on top of a forklift in the bottom floor of the CDF assembly hall, the two scientists together held the "sacred sword", which was over two meters long and was thus quite flexible and hard to manouver; it was also lubricated with a graphite coating.
They directed it into the hole of each of the modules of the EndWall calorimeter, pushing it through the full length of the detector; then they measured the current produced in the photomultiplier tubes by the light yield of the plastic scintillator sheets, extracted the rod, and moved to a different module.
Admittedly, the setup was uncommon, and it is surprising that the Fermilab security allowed the calibration procedure to be carried out that way. Already the simple moving around on a forklift standing several meters above the ground was a peculiar sight; add to that the handling of a flimsy rod, and the dangerous nature of the latter, and you get something closer to a movie scene by Laurel and Hardy than a particle physics experiment.
At some point Cervelli and Stefanini, who were concentrating their attention on the end of the rod which they tried to insert in the appropriate hole, inadvertently caused an elastic movement of the other end, which ran into the poles of the electrical battery powering the forklift, which was lodged at its base. The ensuing short-circuit ignited the solder cap of the hollow rod; the cesium powder released from the interior partly vaporized and partly dispersed on the ground of the CDF assembly hall. Horrified, the Frascati physicists immediately extinguished the fire and rushed to the restrooms of the assembly hall to wash themselves -thereby spreading the contamination!, and called the Fermilab security.
Witnesses of the incident swear that the scene outside CDF soon became one similar to the aftermath of a terrorist attack: in the matter of a few minutes a couple dozen vehicles among fire brigade teams, police patrols, and ambulances crowded the area around the big orange industrial building and filled the air with loud sirens and bright coloured flashes. The decontamination of the area where the incident had taken place would take quite some time, and would include the breaking of the walls of the bathroom where the two had unwittingly washed themselves, to dig out the sink tubes and recover the contaminated fluids; for a few days the assembly hall remained off limits. As for the two terrified researchers, they were taken to the Argonne National Laboratories, 20 miles southeast .
At Argonne, in a low-background underground facility shielded with non-radioactive steel obtained from old warships, they could be screened to measure the radiation dose they had been exposed to. But they were not the only ones subjected to that unpleasant treatment: their colleague Andrea Sansoni, a young Frascati researcher who had not participated in any way to the incident, was taken hostage and withstood the same procedure. The radiologists in fact needed a "control sample": a few months before, prior to their trip to the US, the Italians had been exposed to the radioactive fallout from the Russian Chernobyl accident of May 1986; because of that some radioactive contamination -in particular from the same cesium isotope!- was expected in their body. In order to determine how much radioactive contamination had been absorbed by Cervelli and Stefanini from the vaporized source, a subtraction of the Chernobyl fallout background to which they had previously been exposed had to be performed.
The screening of the three Italians, performed by scanning their bodies with a sodium iodide detector, showed a clear peak of potassium-40, which all of us have in some measure in our bodies, as well as the more ominous one due to the cesium-137 contamination. Fortunately, there was no large difference between the intensity of the latter signal in the test samples (Cervelli and Stefanini) and the reference sample (the unguilty Sansoni): the incident proved inconsequential.