43 Chapter 3 TECHNICAL BASIS FOR THE SSC 3.1 INTRODUCTION The technical basis for the SSC is rooted in the experience and expertise gained in the construction and operation of large nuclear and high energy accelerator facilities throughout the world. The intricate and complex system of components of a large accelerator, described in Section 1.6, has been created and operated routinely in a number of locations, in ever increasing size. In principle, the extension of pre-Tevatron conventional accelerator technology to a machine of the capabilities of the SSC is straightforward, but in practice it is not. The use of superconductivity on a large scale is the key that makes the SSC practicable. Superconductivity is the property of conduction of electricity with absolutely no resistance whatsoever, a property exhibited by some elements, compounds, and alloys at temperatures very near absolute zero. Discovered by a Dutch scientist, Kamerlingh Onnes in 1911, superconductivity was a laboratory curiosity until recent times, when a combination of scientific and technological developments led to its practical application. The advantages of a ring of superconducting magnets are two-fold: (a) The magnetic field strength can be greatly increased over that attainable with conventional copper and iron magnets, increasing proportionately the maximum energy possible for a given sized ring; (b) the power consumption in resistive heat loss can be eliminated, with modest need of power elsewhere. Quantitatively, magnetic field strengths can be tripled with superconducting magnets, while power consumption in the magnet system can be reduced by factors of 10 or more. The conventional main ring of magnets at Fermilab dissipates approximately 130 MW at full energy. In the SSC, an instrument with a particle energy capability of 40 times the Fermilab main ring (20 TeV vs. 0.5 TeV), the losses in the magnets themselves are negligible, although there is roughly 30 MW needed for the refrigerator system to maintain the magnets in their cold superconducting state. Without superconductivity the SSC would need at least 4000 MW for its main rings! The development of a large synchrotron with superconducting bending and focusing magnets was pioneered by Fermilab. The successful operation of the Saver/Doubler ring in 1983 and the Tevatron Collider complex, already described in Sect. 1.6, provides a firm technological foundation and proof that a SSC-like accelerator is feasible. The success at Fermilab is being exploited in Hamburg, West Germany, where the Hadron-Electron-Ring Anlage (HERA), a proton-electron collider, is currently under construction, as mentioned in Sect. 1.5. The maximum proton energy will be 0.8 TeV, comparable to the Tevatron, with 4.7 Tesla superconducting magnets patterned closely on the Fermilab design. This facility is expected to be operational in 1990. 44 While the technical basis for the SSC is well established, it was realized early that R&D effort was desirable to optimize the design for a costeffective and reliable facility. The following sections describe briefly some of the issues and the associated R&D. A more detailed discussion of many of the subjects is given in the SSC Interim Report,* and also in the forthcoming Conceptual Design Report. 3.2 MAGNET DEVELOPMENT Our understanding of superconducting accelerator magnets has matured in the last few years; the Tevatron is operating, HERA is under construction in Europe. Many model magnets have operated successfully and superconducting materials are much better understood. To produce a minimum cost SSC, the magnet-bore diameter must be reduced to the limits allowed by beam dynamics (see Sect. 3.4), and long magnets must be built using mass-production techniques. The R&D efforts of the past two years have clarified the issues of field strength, field uniformity, predictability and reproductibility of performance, and have led to a clear choice of magnet type for the conceptual design of the SSC. (1) Magnetic Components Because reproducible, reliable, and long lasting bending magnets with good field characteristics are essential and are the single most costly technical component of the SSC, the magnet R&D effort has been considerable.** Initially, parallel efforts were pursued to explore a number of alternative styles of magnet. The various ideas (two magnetically coupled beam tubes and coils in a single thermally insulating cryostat; an iron-dominated, low field design; an improved design based on the Tevatron magnets) were explored to determine whether potential cost savings in construction and/or in operation of the accelerator were indeed realizable. In the course of these efforts at the national laboratories (Brookhaven National Laboratory, Fermilab, Lawrence Berkeley Laboratory) and the Texas Accelerator Center, many short (1 meter) and long (4.5 meters and more) model magnels were built and tested. Figure 3-1 shows a sampling of test results on the peak field for six 4.5 m bending (dipole) magnets built at Brookhaven National Laboratory and eight 1.0 m dipoles from Lawrence Berkeley Laboratory. The data show that all the R&D magnets repeatedly achieved peak magnetic fields in excess of 6 tesla at 4.2 K after only a few cycles of bringing the magnet up to full power. Some small variation from magnet to magnet can be seen as a consequence of different superconducting cables being used. The trend towards higher peak field with successive LBL models is a reflection of the improvement of the quality of the superconductor, as is discussed in Sect. 3.3. A significant feature of these results is the evidence for higher peak fields at lower operating temperatures. The BNL models achieved fields close * SSC Central Design Group, "Interim Report," SSC-SR-1011 (June 1985). Figure 3-1. Peak magnetic fields attained at 4.2K in model magnets upon successive slow rampings up in current. At top are data from six 4.5 meter "type D" model magnets made at Brookhaven National Laboratory. Below are the results for eight 1.0 meter "type D" models built at Lawrence Berkeley Laboratory. 46 to 8 tesla when operated at 2.5 K rather than 4.2 K, while the LBL models exceeded 8 tesla at 1.8 K. Lower operating temperatures would permit an increase in the maximum possible field and energy of the protons for the same sized ring and the same magnets, although such operation would require increased refrigeration capacity. The uniformity of the field in the bending magnets is as important as the maximum attainable field strength. Small systematic departures from uniformity can be compensated by correction coils, but random departures because of inevitable small errors in fabrication cause nonuniformity of the field that varies randomly from one magnet to the next. Such variations must be kept within acceptable limits to have a successful accelerator. The departures from the ideal of a uniform field are described in terms of multipole coefficients (an, bn) which give the relative contributions of successive multipole fields (quadrupole, sextupole, etc.). A convenient unit of measurement of the multipole coefficients is 10-480, that is, one part in 10,000 of the main dipole field. If multipole coefficients are kept to the order of magnitude of a few or less, in these units, the proton beams can be successfully guided through the confinement system without loss. Data from a sample of six model magnets (type D, 4 cm aperture, 4.5 m long, built at Brookhaven National Laboratory) are shown in Fig. 3-2. The particular design has certain relatively small systematic multipoles that need not concern us here (since they are easily corrected for). The important question is reproducibility, in other words, the random errors or deviations found in a set of actual magnets. On the basis of the details of the magnet fabrication process and its various tolerances, a number of models were developed to permit estimation of the random errors. Model parameters were adjusted to fit the measured random errors of the Tevatron and CBA magnets and then make predictions for the SSC magnets. The rms deviations of the various multipole coefficients expected from this analysis are shown by the open circles and triangles in Fig. 3-2. The actual rms deviations measured for the integral field in the six BNL models are given by the solid circles and triangles. For the lowest 12 coefficients the measured random errors are all less than the predictions. For the highest order multipoles studied, the coefficients are zero within measurement errors. Since the expectations of the models were the basis of the aperture studies described in Sect. 3.4, and the measurements indicate even smaller random errors, one can conclude that the manufacturing techniques are capable of producing many identical magnets of completely acceptable field quality, within tight tolerances. The conclusions of the extensive R&D effort on model magnets are that the magnet performance can be predicted reliably, that model SSC magnets achieve the necessary peak field strengths and have adequate field quality for accelerator operation, that the magnet fabrication techniques assure reproducibility from magnet to magnet, and that these techniques are ready to be transferred to industry. On the latter point, it should be noted that, although the six Brookhaven 4.5 m model magnets were produced as prototypes in the laboratory, the fabrication methods and tooling were designed with largescale industrial production in mind. |