A SUPERCONDUCTING MAGNET SYSTEM FOR THE SPIRIT COSMIC RAY SPACE TELESCOPE

The SPIRIT (A Superconducting Passive Iron Isotope Telescope) experi­ ment requires a large volume (Ira-*) of 2T field in ordesr to achieve enough resolution to study heavy primary cosmic rays. It is proposed that the SPIRIT superconducting magnet system and its experimental package would be used in one of the space shuttles. The superconducting magnet design is based on Lawrence Berkeley Laboratory thin high current density solenoid technology. The superconducting magnet system consists of a number of coils which generate a 2T induction within the experiment, and at the same time allow free access to the package by cosmic rays. The superconducting magnet system uses high current density conductor which is protected by a shorted secondary circuit. The magnet coils are to be cooled by pumped two phase helium which is circulated through tubes. Refrigeration is supplied from a large liquid helium dewar.


IHTRODUCTION
The identity of the source of the cosmic radiation is one of the oldest and most interesting unanswered questions of 20th century physics.
While it has become increasingly clear that these energetic particles owe their existence to some of the most violent processes that occur in our galaxy (e.g. supernovae), a detailed understanding of the cosmic ray source and the conditions of galactic propagation has not been achieved. Of particular interest in this regard is the isotopic composition of the cosmic radiation since nuclear abundance anomalies would provide the most exciting clues as to their nuclear origin. The iron isotopes provide the most fruitful candidates for such a study since they are both abundant and are least modified by galactic transport. To date the most accurate isotopic studies of the iron group cosmic raysHJ have ruled out large deviations from solar system source composition. In order to achieve a convincing separation of the isotopes of iron it is necessary to design an instrument which can collect over ICP iron nuclei and achieve a mass resolution of a< 0.15 amu.
Three important developments have made the design of such an instru ment possible: 1) With considerable impetus from the new generation of accelerators, magnet and cryogenic technology has reached the stage where very large volumes can be filled with uniform fields in excess of 2T. 2) A track recording plastic detector made of CR-39 has been developed^] that is sensitive to minimum-ionizing particles of the so-called very heavy group (20 <_ Z <^ 30), has very good charge resolution, yields etched tracks of very high optical quality, and can be made in films thin enough that multiple Coulomb scattering can be neglected.
3) The advent of the space shuttle will make it possible to lift payloads weighing many tons into orbit for periods of several weeks. The measurement of magnetic rigidity in combination with the measurement of range will be used to determine particle mass.
In this paper we shall discuss the design of the superconducting magnet and cryostat systems. To achieve the resolution and collecting power necessary to meet the experimental objectives we need a superconduc ting magnet wi*-h an average field of 1.5 to 2 Tesla over a lin^ volume.
Spatial gradients need to be kept below 4Tm~l so that shifts in detector orientation expected during flight will not result in a degradation of resolution.
The whole apparatus must be contained in the space shuttle orbiter cargo bay which has a dynamic envelope diameter of 4.57m.

THE SUPERCONDUCTING MAGNET
In order to achieve the resolution and collecting power necessary to meet the experimental objectives or SPIRIT, a superconducting magnet with an average field of 2T over volume of lm^ is needed. The lm-* volume must be clearly accessible to cosmic radiation entering from a polar angle from 0 to about 45° (see Figure 1). The proposed magnet consists of six coils. The four inner coils generate a uniform high field over the lrp control volume. The two outer coils will generate a smaller field over a larger volume so that the entire assembly will have a zero net dipole moment, which results in a negligible distant field.  Nb-Ti * The inner coils which produce the 2T field within the experimental volume.
** The outer coil which cancel the dipole moment generated by the inner coils. The net dipole moment must be zero. The proposed operating current density and high stored energy result in a high EJ2 product; (E is stored energy, J is superconductor matrix current density). Thus, we propose that a well coupled secondary circuit made from very pure aluminum be used for quench protection. The shorted secondary quench protection system is used on the 2m diameter TPC solenoid which has an EJ^ product of 5.4 x lO^^Jk^m~^.

Im3. As a result, the use of
The shorted secondary concept affects the quench process in the following ways: 1) The shorted secondary causes the coil current to shift from the coil to the secondary circuit. As a result, there is less current in the coil to contribute to the conductor hot spot.
2) The shorted secondary circuit absorbs a substantial amount of the magnet stored energy. In the proposed magnet system, the shorted secondaries are expected to absorb about 70 percent of the magnetic -6-  // DNA means "does not apply". 3) The shorted secondary will cause "quench back" in the other coils when one of the six coils tvrns normal through ordinary quench propagation. Quench back is a key element in the protection of thin high current density solenoids whr'.ch have been built at LBL.
The sho.ted secondary circuit would be insulated from the supercon ductor. It is desirable that inductive coupling between the coil and the secondary circuit be maximized. We propose that the shorted secondary circuit be made from ultra pure aluminum (0.99999 pure or better) which has a residul resistance ratio at 4.2K and OT of about 2000. Aluminum has a much lower magnet resistance at 8T than copper. It also has one third the density. At full field, one can expect the shorted secondary circuit to have a RRR > 300. As a result, the shorted secondary circuit is expected to have a time constant in excess of 30 seconds. If the coupling between the coil and the shorted secondary circuits is good enough, effective shifting of the coil current will occur.
The proposed coils will have the superconductor, shorted secondary and a forced flow tubular cooling system combined into an integrated package, proposed that the superconducting magnet be located at the rear of the space shuttle bay.

The expected stray magnetic induction in a region
normally housing astronauts is expected to be around 10~^T.
It is proposed that coil 1 and 2 (in each half) be attached directly (see Figure 2). A compressive force 7J x 10% (77 metric tons) is ex pected between the two halves (between coils 1A, 2A, 3A and coils IB, 2B, 3B). We propose to carry this force with cold column struts between coll 1A, 2A and coil IB and 2B. The columns are arranged so that there is full access of the cosmic rays to the experiment. A tensile force of 5.8 x 10°N (560 tons) is expected between each of the two outer coils and their companion inner coils. We propose to carry this force with a continuous web of metal between the outer and inner coils. The six coils are expected to act as a rigid frame which will have a cold mass of about 4000kg (the helium tanks and the coil cryogenics will attach directly to this frame).

THE CRYOSTAT AND CRYOGENIC COOLING SYSTEM
The proposed superconducting magnet coils will be cooled using two phase helium pumped through tubes in the coil package (see Figure 3). We plan to circulate the two phase helium from a separate helium storage tank located at the end of the experiment (see Figure 1). Forced tubular cooling offers a number of advantages over the more conventional bath cooled systems: I'J 1) Tubular cooled systems can be cooled easily from room temperature by a refrigerator which is external to the cryostat system.
2) Only a small fraction of the liquid helium is in direct thermal contact with the superconductor at any one time during a quench.
The tubular cooling system can contain the pressure rise due to this small amount of helium. Helium boil off during a quench is orderly and well controlled.