QUENCH PROTECTION OF SC QUADRUPOLE MAGNETS S. Feher, R. Bossert, J. DiMarco, D. Mitchell, M.J. Lamm, P.J. Limon, P. Mazur, F. Nobrega, D. Orris, J.P. Ozelis, J.B. Strait, J.C. Tompkins,A.V. Zlobin, Fermi National Accelerator Laboratory * , P.O. Box 500, Batavia, IL 60510, USA A.D. McInturff, Lawrence Berkeley National Laboratory * , Berkeley, CA USA Abstract The high gradient quadrupole (HGQ) being developed for the LHC interaction regions by the collaboration of FNAL/LBNL/BNL[1], relies on the use of quench protec- tion heaters. As part of the HGQ R&D program at Fer- milab, Tevatron low-β quadrupoles installed with quench protection heaters were tested in normal and superfluid he- lium. This paper focuses on heater operation time delay and quench propagation velocity measurements since these are important input parameters for designing the quench protection system of the HGQ. 1 INTRODUCTION The energy stored in a superconducting (SC) accelerator magnet is dissipated after a quench in the normal zones, heating the coil and generating a turn to turn and coil to ground voltage drop. The propagation velocity of the nor- mal zone is usually low relative to the heating rate of the cable and the cable temperature will rise so high that it will damage the cable. Quench heaters are used to protect the SC magnet by greatly increasing the coil normal zone thus allowing the energy to be dissipated over a larger conductor volume making the protection to be less dependent on the quench propagation velocity. Such heaters will be required for the HGQ. Without overheating the cable or developing too high voltages, the ellapsed time between the quench origin and the start time of the stored energy dissipation in the heater quenched part of the coil is usually quite short, for HGQ this value is in the order of a few millisecond[2]. This time depends mainly on the quench propagation velocity and the time delay of the heater operation. The study of these im- portant parameters in normal and superfluid helium as part of the HGQ R&D program has been started at Fermilab on low-β (LBQ) quadrupoles[3, 4]. This paper presents exper- imental results on a LBQ (R54002) heater operation time delay and quench propagation velocity. 2 MAGNET DESCRIPTION The magnet R54002 for this study is a modified 1.4 m long Tevatron low-β quadrupole. Details of the baseline design have been described elsewhere[5, 6]. This cold iron super- conducting quadrupole has two layer coils with a 76 mm diameter bore. There are copper wedges in the inner coils whose primary purpose is to minimize the geometric 12- * Supported by the U.S. Department of Energy and 20-pole harmonics. Four inner to outer coil splices are located in the magnet lead end radially beyond the outer coil and are made through pre-formed solder-filled cable originating from the lead end pole turn. The inner and outer coils are made from 36 strand Rutherford cable. The strands are 0.528 mm in diameter and contain 13 μm filaments. The cable insulation is made of 25μm thick and 9.53 mm wide Kapton tape covered both side with B-stage epoxy. Kapton tape is wrapped with 67% overlap forming three layer of insulation with total thick- ness of 75μm and 1-1.5 mm gaps in the outer insulation layer. The coils are supported in the body by aluminum col- lars. The coil lead and return ends are clamped with a 4 piece G-10 collet assembly enclosed in a tapered cylindri- cal can. Iron yoke laminations surround the coil in the body region, and stainless steel laminations surround the end region cylindrical can. A welded stainless steel skin surrounds the yoke. The quench protection heaters are 25μm thick and 12.5 mm wide stainless steel strips and are located radially beyond the outer coil, in the middle of four layers of 125 μm Kapton sheets. One heater covers approximately 12 turns of two midplane-adjacent outer coils. This is accom- plished by running the heater longitudinally along the body of the magnet and making appropriate folds on the heater in the magnet return end region. Two heaters oriented 180 degrees apart provide coverage for one side of each of the four outer coils. The resistance of the heater for coils A and B was 5.5 Ω, and that for C and D was 5.0 Ω. The system resistance (including cabling from the Strip Heater Firing Unit (SHFU) to the magnet) was 3.0 Ω , which means that ∼85.5 % of the SHFU voltage was deposited directly to the heaters. The 69 voltage taps that instrumented R54002 allowed for localization and determination of propagation veloc- ity for most quenches. Magnet was tested at the Fermilab Technical Division horizontal test facility[7]. 3 HEATER TIME DELAY The heater time delay (t fn ) is the time from protection heater current initiation to the presence of a detectable quench voltage in the outer coils. Figure 1 shows the time diagram of the heater and magnet voltage and an example of the t fn determination. The heater time delay as a function of voltage applied to heaters is shown in Figure 2. As one can see, at the lower heater energies (close to V min ) the t fn increases rapidly 3389 0-7803-4376-X/98/$10.00  1998 IEEE