The UVa group was responsible for the design, fabrication and monitoring of the hadronic calorimeter. All of the components for The hadronic calorimeter was designed to provide a `muon blind' component to the standard left-right trigger and to reduce the trigger rate due to interactions in the spectrometer material. The calorimeter had to: 1) be exceptionally fast due to the high rate of interactions in the spectrometer, 2) have a sharp energy threshold, and 3) have to ability detect muons. The figure below shows how well the calorimeter performed. On the left hand side is a plot of the trigger acceptance as a function of energy for charged particles entering the calorimeter for a trigger threshold of about 50 GeV. Note the sharp turn on in acceptance. The right hand plot shows the total energy entering the calorimeter for charged particles of momentum between 100 and 120 GeV. Our preliminary (uncalibrated) estimate of the calorimeter resolution is: sigma/E = 80/sqrt(E) + 2.5%.
The calorimeter is 1 m square in area, weighs about 14 tons, and is subdivided into eight cells. All of the protons from Lambda's decaying in the evacuated decay volume lie within a fiducial area of about half the calorimeter area. It is a sampling calorimeter composed of 64 layers of 0.5 cm thick scintillator sandwiched between 2.41 cm of iron. Iron absorber was chosen over lead, which would allow a compensating calorimeter to be built, because the critical energy for muons in iron (350 GeV) is much greater than that for lead (120 GeV), minimizing muon bremsstrahlung. The readout of the scintillator (Kuraray SCSN-38) is via 2,048 2 mm waveshifting fibers embedded in keyhole shaped grooves. Use of fibers rather than waveshifting plates allowed for a more hermetic calorimeter, minimized the needed photocathode area, and greatly simplified the mechanical design. The fibers are the new Bicron BCF-92 G2 waveshifting fibers, which are very bright and fast (tau = 2.7 ns).
The fibers are read out at the bottom end, with the other end polished and sputtered with 2,000 angstroms of Al. The average fiber reflectivity was measured to be 86% with an average attenuation length (the long component) of 346 cm. This corresponds to a difference in light output of 6% from top to bottom, well within specifications.
The readout is via Hamamatsu R329 photomultipliers equipped with green extended photocathodes. Extensive studies of short and long-term rate effects in these photomultipliers were conducted at UVa and are the subject of Rajaram's Masters thesis and resulted in the design of a transistor stabilized base of exceptional rate stability. The summed outputs of all the photomultiplier tubes was discriminated to provide an energy cut. The absolute calibration of the calorimeter was relatively easy due to the fact that there was a large flux of protons with well-measured momenta impacting it. The cell-to-cell gains were adjusted using muons which produced about 1.5 photoelectrons per plane of scintillator. A light pulser system using a Nitrogen laser and PIN diodes was also used to monitor the time dependence of the calibration.
A fast (1 microsecond conversion time) 14-bit ADC was designed and fabricated at LBNL and used to read out the calorimeter.