8. Technological Developments
The physics produced by the Tevatron fixed target program depended on a whole sequence of technological developments in accelerators, detectors, and computing. Many experiments required one or more of these innovations to even be feasible in the first place, and to achieve their goals. The experimental collaborations themselves contributed significant advances in the technologies required. These advances have had far-reaching impact on subsequent work, both here at Fermilab and elsewhere in the world. Some of these advances transcend in importance the experiments for which they were developed. In this section we will hit a few of the highlights in this area. Where possible, we have tried to include a reference to each of these works.
8.2 Accelerator and Beam
lines
By far the most impressive, important, and obvious technical development is the Tevatron itself. From the experimentalist’s viewpoint, it not only provided twice the beam energy but also a much longer spill. The spill duty factor was increased from 1 second of beam every 15 second to 20 seconds of beam every 60 seconds; a factor of five improvement over the Main Ring. The longer spill length required several important innovations in the Tevatron itself. An extraction system, generically known as QXR, was extensively modified based upon microprocessor technology to handle the much longer beam spill.
The external beamlines required significant new developments as well. The longer spill implied much less intensity per second in the external beamlines. A new beam position monitoring system was developed to sense and control the external beams. These beams are as much as one million times less intense than the circulating beam in the Tevatron itself. The higher energy beam required twice the bending power in existing tunnels, which were originally designed for 200 GeV beams. The new requirement was to deliver the extracted 800 GeV protons to the Meson, Proton, and Muon Laboratories. This led to the development of the strings of superconducting magnets and their associated cryogenic plants, known as the left, right and muon bends. Many will remember channel 13 messages like “LEFT BEND QUENCH – NO ESTIMATE” - these things were far from trivial.
The higher beam energy and the sensitivity of the downstream cryogenic magnet strings required significant upgrades to the stations that split the extracted proton beams to each of the Proton, Neutrino, Muon, and Meson Laboratories. Motion controls and new pulsed magnets were installed to move the fast spill around the septa without burning them up, and the septa themselves were improved.
References
The energy saver test and commissioning history, Helen Edwards, Proceedings of the 12th International Conference On High-Energy Accelerators (August 11-16,1983).
Operation of the Tevatron extraction system, L. Chapman, et al., IEEE Transactions on Nuclear Science, NS-32, No. 5 (October 1985).
Tuned Beam Position Detector for the Fermilab Switchyard, Q. Kerns, et al., IEEE Particle Accelerator Conference – Accelerator Engineering and Technology (March 16-19, 1987).
Improvement of the high voltage properties of the Fermilab electrostatic septa, D. Trbojevic, et al., IEEE Transactions on Nuclear Science, NS-32, No. 5 (October 1985).
The most notable and noted advance in detector technology was the development of the silicon microvertex detectors by the Santa Barbara group for the charm photoproduction experiment E691. This development, built on the detector development work of European groups working at CERN, revolutionized heavy quark physics. This occurred first in charm production, and latter in studies of hadrons with beauty in both fixed target and collider experiments, as well as in the discovery of the top quark in the Tevatron collider . This development received significant recognition with the 1990 award of the Panofsky Prize to Mike Witherell for the advances in charm physics that it made possible.
The group from the Petersburg (then Leningrad) Nuclear Physics Institute made a major advance in precision electron identification with their development of a large transition radiation dectector (TRD) system for the precision S - beta decay experiment E715. This large system achieved electron/pion separation of several thousand, with an electron inefficiency of less than 1%. These techniques have been used and further extended at Fermilab in experiments E761, E781, and E799.
The KTeV experiment developed a CsI photon calorimeter with outstanding energy and position resolution, excellent linearity, and very high rate capabilities. They achieved better than 0.75 % energy resolution over virtually their entire photon energy range of 5-100 GeV. They developed a new digital readout system with 17 bits of dynamic range housed in the photomultiplier base of each of their 3100 CsI crystals.
The heart of The KTeV CsI readout was an application specific integrated circuit (ASIC) called the QIE chip, one of several ASIC’s developed on the 14th floor of Wilson Hall for use in the Tevatron fixed target program. Others included silicon microstrip and wire chamber electronic circuits. These chips, and the ability to develop new circuits in the latest technologies, have come into widespread use since these early efforts.
The technique of recording the position of Cerenkov photons at the focal plane of a large Cerenkov detector for particle identification over a broad angular range (the Ring Imaging Cerenkov Counter, RICH) was pioneered by the E605 experiment. A Fermilab group lead the effort in SELEX (E781) to develop a large ring imaging Cerenkov counter (RICH) based upon 2848 small phototubes as the photodetector. The SELEX RICH achieved useful p-K separation in full multi-hadronic events up to 165 GeV/c with 12 photons observed on a typical ring. The new CKM experiment planned for the Main Injector fixed target program is based, in part, on this detector technique that accommodates very high beam rates with excellent time resolution.
References
Test beam studies of a silicon microstrip vertex detector, P.E. Karchin, et al., IEEE Transactions on Nuclear Science, NS-32, 612 (1985).
Performance of the
E715 transition radiation detector, A. Denisov, et al., DPF
Conf.1984:358
Beam test of a prototype CsI calorimeter, R.S. Kessler, et al., Nucl. Inst. and Meth. A368, 653 (1996).
Identification of
Large Transverse Momentum Hadrons using a RICH, R.L. McCarthy et al.,
Nucl. Inst. and Meth. A248, 69 (1986).
The SELEX Phototube RICH Detector, J. Engelfried et al., Nucl. Inst. and Meth. A431, 53 (1999).
The use of “farms” of parallel computers bases upon commercially available processors is largely an invention of the Fermilab Advanced Compter Project (ACP). This technique has become widespread, both for on- and off-line computing - for selecting, acquiring, and processing the vast volumes of data that come with a full hardronic cross-section.
The original ACP I computer was developed at Fermilab in the mid 1980’s based on a Motorla 68020 processor and an ACP designed bus structure which allowed dozens of processors to analyze individual events in parallel. Much of the data from the early Tevatron fixed target runs were reconstructed off-line on farms of ACP I processors in the Feynman Computing Center. At its peak, we had 400 processors in use.
This technique was extended in the next generation to farms to the use of commercial UNIX workstations. Farms like these have been established in the Feynman Computing Center as well as collaborating universities and national laboratories. For example the 50 terabytes of data colleted by experiment E791 was reconstructed in parallel on farms at Fermilab, the University of Mississippi, Kansas State University and the CPBF in Rio de Janeiro.
This is an innovation which has become an industry standard in our field. Now there is hardly an experiment that does not have both an on-line computing farm for sophisticated software trigger decisions and an off-line farm for rapid parallel reconstruction of events.
As the Tagged Photon Laboratory charm group moved from experiment E769 to experiment E791, they pioneered the use of 8 mm magnetic tape for online data recording. As the before and after pictures from E769 and E791 show, they had good reason to do so. The change to a new recording medium was adopted as a new standard by the Computing Division (then the Computing Department in the Research Division). After considerable work (and pain), it became the de facto standard for data recording, both at Fermilab and in many other places in particle physics.
References
Use of new computer
technologies in elementary particle physics, I. Ganies and T. Nash., Ann.
Rev. Nucl. and Part. Sci. 27, 177
(1987).