Paul Grannis and Dmitri Denisov on the DØ Detector
On June 11, 2019, Paul Grannis and Dmitri Denisov, co-spokespersons of the DØ Experiment, shared their memories of the DØ detector.
Central tracking system
The central tracking system had two subdetectors for measuring charged particle track positions and a magnetic field to cause tracks to bend, thereby allowing a measurement of their momenta.
The silicon microstrip tracker was located just outside the Tevatron beam pipes. Five barrels concentric with the beams and 16 disks with strips perpendicular to the beams provided precision measurements of charged track coordinates. These helped to determine particle momenta and to distinguish those particles that emerged from the primary collision point from those, like tau leptons and hadrons containing bottom quarks, that traveled a finite distance before decaying. It consisted of about 800,000 silicon strips of 50 micron width, capable of measuring track location to about 10 microns. The outer radius of the silicon detectors was limited to 10 cm due to their high cost.
The cylindrical scintillating fiber tracker occupied the radial region between 20 and 52 cm and 2.5 m along the beam line, outside the silicon tracker. Particles traversed eight layers of 835 micron diameter scintillating fibers. Light from each of the more than 75,000 fibers was transmitted to solid state sensors that created electronic signals that were digitized and logged. The fiber tracker spatial precision was about 100 microns.
A superconducting solenoid magnet was located just outside the fiber tracker created a 2 T magnetic field in the silicon and fiber tracker volume.
Calorimeter
The job of the calorimeters and associated subdetectors was the measurement of the energies of electrons, photons and charged and neutral hadrons. This was achieved by letting incident particles traverse multiple layers of dense inert material in which they interacted and created secondary particles. The collection of all such secondary particles is called a shower. Ultimately the energy of the progenitor particle was shared among many shower particles of much lower energy that ultimately stopped, at which point the shower ended. Between the layers of the inert material there were detectors in which the ionization of the particles was measured. The total ionization signal summed over the shower is proportional to the energy of the progenitor particle.
A cylindrical layer of scintillator based preshower strips was placed immediately outside the solenoid and read out with the fiber tracker sensors. Similar preshower detectors capped the ends of the tracking region. The material in the solenoid augmented with lead sheets caused primary electrons and photons to begin a shower of secondary particles. The preshower detector was thus the first stage of the calorimetry and gave a precise location of the particle impact point.
A central calorimeter outside and two end calorimeters capping the solenoid contained separate sections for measuring electromagnetic particles and hadrons. Uranium was chosen for the inert absorber plates owing to its very high density. The active gaps contained liquid argon with a strong electric field applied to collect the ionization of traversing particles on finely segmented planes of copper electrodes. These signals were ganged into 50,000 signals that measured the particle energies and the transverse and longitudinal shower shapes which helped identify the particle type. Each calorimeter contained about sixty uranium-liquid argon modules with a total weight of 240 - 300 metric tons. The total thickness of a calorimeter was about 175 cm so as to fully absorb the showers of the most energetic particles from a collision. The stainless steel vessels needed to contain the modules at liquid argon temperature (-190 C) were relatively thick, so scintillation detectors were inserted between central and end calorimeters to correct for energy lost in the cryostat walls.
A primary task for the calorimetry is identification of jets, the sprays of particles created as quarks and gluons escape from their collision point. Jet identification and measurement of their directions and energies allow analyses to recreate the momenta of the underlying quarks and gluons in the primary collision.
Muon Detector
The outermost shell was for muon detection. High energy muons are quite rare and are thus a telltale of interesting collisions. Unlike most particles, they did not get absorbed in the calorimeters, so tracks observed beyond the calorimeters are most likely muons. Scintillator planes provide a fast signature used to flag interesting events. One station of tracking chambers before and two stations after solid iron magnets record the muon tracks. The iron of the large central magnet was reclaimed from a NASA cyclotron built to simulate radiation damage in space. The muon detectors are large and thus there is a premium on the use of cost effective technologies.
Trigger and DAQ
The trigger system used the electronic signals to identify events of interest, such as those containing electrons, muons, photons, high energy jets or particles that travel some distance before decaying. The first trigger level used the fast electronic signals from each subdetector to make a decision within a few microseconds on whether to pause data-taking and digitize the signals. About 10,000 such Level 1 triggers were accepted. A second trigger level refined the selection using the digitized signals from several subdetectors in combination to form a more nuanced event profile, and reduced the candidate event pool to 1000 events per second. In the third level, a farm of computers analyzed the digital information in a stripped down version of the full offline computer code to yield up to 100 events per second to be permanently recorded and subsequently analyzed on large offline computer farms. The operation of the trigger system is a delicate balance between maximizing the number of events saved and minimizing the dead time incurred while collecting them. It must be robust and reliable, as the millions of events not selected by the trigger are lost forever.