A Large Superconducting Detector Magnet without an Iron Return Path

This paper describes a detector magnet which retu:-n. \ix 'r ween the coils rather than through an iron return path. This actively si * ...~^ < niform field 2 T magnet can be fabricated in separate parts which can be manufactured on the SSC site. This magnet can be built so that central field Is uniform enough to permit a TPC detector to be used without iron poles. The field outside of the coil can be made to fall of as R' N power where N approaches 9. A major advantage of the magnet described in the paper is that there is no pole piece to block the particle jets emanating from the collision region in the forward and backward directions. Inexpensive materials such as earth and concrete can be used to provide the mass needed to analyze particles such as mu mesons. As a result, problems such as experimental hail subsidence can be reduced. Perhaps the cost of such an experiment can also be reduced. This type of magnet would require experimenters to rethink their experimental concepts.


BACKGROUND
The SSC, a large machine, will produce proton collisions of 20 TeV on 20 TeV. The collisions result in many fragments which must be detected in a large number of detectors around the collision point Large magnetic fields can be used to bend charged particles which will permit the mass and momentum of these particles to be calculated. The volume of the magnetic field must be large (at least 5 meters in diameter and 7 or 8 meters long).
In order to generate a large volume magnetic field, superconducting magnets provide an economical way of generating the magnetic field. A large volume of magnetic field implies that there must be a large ferromagnetic return path. If the magnetic field must be uniform, the iron return path and pole pieces must be machined in order to achieve the desired field uniformity. The conventional approach to building large detector magnets would require up to 30000 tons of iron to shape the magnetic flux and return it.
A large massive iron return path has a number of disadvantages which include: 1) The iron density is high enough to cause serious subsidence problems at the site of the experiment, 2) When the desired field is uniform, the iron blocks particles which travel at low angles in the direction from which the colliding beams come. The disadvantages given above can be modified scmawhat by tha proper design of iron return path. The result is a larger iron return path mass and an even mora complex design for the iron. This paper presents an alternative approach to detector magnets. This is a modest proposal for a large superconducting magnet which will produce a uniform field over a volume which is four meters in diameter and four meters in length. Tha proposed magnet has no iron return path, yet the magnetic induction 20 meters from tha collision point is lass than 10 Gauss. The no iron detector magnet requires that one rethink tha design of SSC physics detectors.

THE IRON FREE DETECTOR MAGNET AND ITS EFFECT ON PHYSICS
For soma types of SSC detectors, it may be desirable to eliminate the iron shielding and return yoke. Tha iron fraa detector magnet ia in principle like the actively shielded magnetic resonance imaging magnats (MRI), which have been built by Oxford 1 and other companies. 2 Two general designs ware looked at for magnats which have a fraa bora diameter of at least 7.5 maters.
Tha two designs can be described aa follows: 1) Tha first type of magnat ia the two solenoid type of actively shielded magnat, which ia shaped like a cylinder. This type of magnat is the state of tha art actively shielded MRI magnet. This type of magnet, when it has an inside bora of 7.5 m would have an overall length of about 28 to 30 maters and an outside diameter of about 16 maters. Magnets of this type have bean built with warm bore diameters of 1 meter. This type of magnet haa a peak field which ia 10 to 15 percent higher than tha central field of the magnet 3 This type of magnat ia wall suited for use as a high central induction magnet (>3T).
2. Tha second type of actively shielded magnet ia a spherical type of magnat which waa developed at Stanford University 4 for MRI imaging of tha heart and tha circulatory system. This mrgnet, when it haa a minimum warm bora diameter of 7.5 meters, will have an overall length of 14 meters and an outside diameter of about 19 meters. A magnet of thie type haa a peak induction at the conductor which ia 60-80 percent higher than tha central induction. This type of magnet is well suited for central inductions below 3.0T.
Tha advantages of the solenoRJal design are: 1} The peak field in the winding ia only 15 percent higher than the magnet central field. 2) The field is quite uniform over a region which ia up to 6 meters in diameter over a length of about 12 meters without an iron return path. Tha disadvantages of this type of magnat are as follows: 1) The magnet ia large, and it ia difficult to fabricate in pieces on tha site. The transport of magnets with a warm bora over 5 or 6 meters is difficult under the best conditions. 2) Tha free solid angle in tha forward and backwards directions from th% collision point is only about 20 degrees from the proton beam line. 3) Physics is difficult to do outside of the 3.75 mradiua over the full length of the solenoid magnet. 4) The 10 Gauss induction line is 38 meters from the collision point in tha radial direction and 48 meters from the collision point along tha beam axis.

Tha advantages of tha spherical design are: 1) The field is quite uniform over a diameter of 4 to 5 meters and a length of 6 to 8 meters without an iron return path. 2) The free solid angle in the forward and backwards directions from the collision point is about 35 degrees from the proton beam line. 3) At high solid
angles from 60 to 90 degrees from the proton beam around the collision point, physics can be dona out to a radius of 5.5 meters. 4) The 10 Gauss line is 18 maters from tha collision point in tha axial direction. 5) Tha magnat can be built in pieces on tha site and it can be assembled on site. The disadvantages of this type of magnat ara aa follows: 1) Tha field rise at tha conductor is 65 parcant highar than tha central field. This is accaptabia in a 2.0 Tasla magnat, but it is not accaptabia in a 4 or 5 Tasla magnat. 2) Tha mambars which carry tha forcas between tha magnat coils must be cold. These forcas ara much larger than in the solenoid design.
Tha spherical solenoid design waa selected aa tha candidate for an iron free detector magnat. Tha use of such a magnet configuration requires one to rethink how one might do tha physics. If tha magnet has no iron return yoke, one must use other materials such as concrete, heavy concrete with barrites, earth and other relatively low density materials to moderate the particles generated at the collision point around which the experiment is being done. This may require one to look at different types of detectors to do the physics.
The advantages of the iron free magnet are: 1) The up to 30000 tone of iron return path can be eliminated. If iron is used in tha experiment, the coH design must be altered. 2) Up to 35 degrees in the forward and backwards directions is completely free for looking at tha particle Jets created at tha intersection point. 3) Cheap materials such as concrete and earth can be used to moderate the particles. There is a saving in foundation cost and subsidence is greatly reduced. Tha disadvantages ara: 1) The magnat coils block particles at a radius of 4 maters from tha proton beam Una at solid angles from 35 to 60 degrees from tha proton beam line. This radius goes out to 5.5 meters for an angle of 60 to 90 degrees from tha proton beam line. 2) Soma kinds of experiments require dense materials to moderate the particle produced at the collision point. Iron is one of the least expensive dense materials to use aa a moderator. 3) Tha stray field from the magnet can have an adverse effect on some types of detectors. The colliding proton beams may have to be carried in superconducting shielded beam pipes.  The magnet design process uses Legendre polynomials of the first kind to expand the field inside of the coil.s Legendre polynomial of the second kind are used for the field outside of the coils (when one looks at the problem, the use of Legendre polynomials of the second kind are not needed). The magnetic field is designed to be axially symmetric. The net magnetic Legendre dipole, sextupole and decapole moments are zero by design (for both the inner and outer fields) and the quadrupoie, octupole, 12 pole and so on are zero by symmetry. Since the magnetic moments up to N • 7 are zero, the field outside of the magnet coils falls off rapidly (as radius to the minus nine power).  Figure 2 produces an axially symmetric dipole of 2 Tesla, with no axially symmetric quadrupoie, sextupole, octupole, decapole or 12 pole. The first higher term to appear in either the internal or external field is the axially symmetric 14 pole. A magnetic induction contour map for the magnet is shown in Figure 3 and the magnetic flux line contour map is shown in Figure 4 Table 1 presents parameters for the spherical no iron detector magnet shown in Figure 1 and 2. The eight coil magnet system has an overall coil length of 13.6 m and an overall diameter of 18.5 m. When a 2.0 Tesla central field is generated, the peak induction in con 2 is 3.3 Tesla. The external induction out to a distance of 25 meters from the collision point is shown in Figure 5. The 1 gauss line is about 24 meters from the collision point in the radial direction, and it is about 28 meters from collision point in the axial direction. The shielding achieved can be controlled by thin windings mounted on the outer coils (coils 3 and 4). Table 1 assumes that the inner and outer coils are operated off of a single 12kA power supply. Quench protection can be provided by a single 0.16 ohm external dump resistor. Table 2     Tha proposed detector magnet should ba built on tha site and assembled along with tha experiment. A no-iron detector magnet requires one to rethink about physics options for the SSC. This type of magnet will not achieve all the physics goals for tha SSC, but it will be good for analyzing particle Jets up to a solid angle of 35 degrees from tha proton beam line.