In FY 98, the emphasis of our high-field magnet work shifted from "proof of principle" to a broad-based search for a cost-effective magnet solution for the next generation collider beyond the LHC. This shift is consistent with the recommendations of High-Energy Physics Advisory Panel (HEPAP) subpanel on Planning for the Future of U.S. High Energy Physics, sometimes known as the "Gilman Panel." These recommendations include a statement that "an expanded program of R&D on cost reduction strategies, enabling technologies, and accelerator physics issues for a VLHC" is desirable. This slightly more focused approach fits within the overall goal of the program to "write the book on magnets".

Our R&D approach is centered on a simple racetrack coil geometries. At present, accelerator magnet technology is dominated by the use of NbTi superconductor. To achieve fields above 10 Tesla requires the use of A15 compounds, the most practical and available of which is Nb3Sn. In a practical geometry, magnets based on Nb3Sn technology should be able to exceed fields of 14 -15 Tesla at 4.2 K. The challenge lies in incorporating the intrinsically brittle, strain sensitive material, into a realistic magnet where it is subjected to stresses approaching 150 MPa. Advances in fabrication techniques and materials have allowed us to reinvestigate the advantages of simple racetrack coil geometries lending advantages in support structure design and fabrication. In particular, we are concentrating on the common coil configuration for its potential simplicity of construction and consequent cost effectiveness. The design concept consists of a pair of racetrack coils shared between two apertures, producing fields in opposite directions. This geometry is intrinsically suited for a collider, but modifications of this design can be used for single-aperture applications as well.

Materials and magnet designs for use above the field range of the ductile alloys are being developed. Niobium-tin conductors are being optimized for use in accelerator magnets, where maximum current density is essential. Using these conductors as a prototype, the techniques for fabricating accelerator magnets from brittle superconductors are being explored. These techniques can then be applied to other brittle materials, such as niobium-aluminum, Chevrel-phase, or high-temperature superconductors, once these materials are available in adequate lengths and with critical current densities comparable or superior to that of niobium-tin.

The current phase of the program began with a 6 Tesla common coil magnet, RD-2 (Racetrack Dipole), that was tested in September of 1998 as a demonstration of the feasibility of the design. It was designed, built, and tested in less than one year, much faster than with previous designs. Our goals with this first magnet were to gain experience in the fabrication of this type of coil, to evaluate training behavior as compared with that of cosine theta magnets, and to establish some design parameters before taking the next step, which will be a 14 T common coil magnet.

Though the field was limited by the available conductor, it performed extremely well. RD-2-01 required no training and achieved a thermally dependent 4.4 K plateau current of 8.29 kA on the first ramp. Several modifications were made to the structure to study the effects on performance. The mechanically modified versions of the magnet (RD-2-02 - RD-2-05) performed identically to the original configuration, despite the sizable differences in loading and loading histories.

Recent Activities

The group is now in the process of constructing a 14 Tesla magnet (RD-3). The design consists of one inner and two outer coil modules. The outer coil modules were tested using a support structure similar to that designed for the 6 Tesla magnet, in March of 2000. Designated RT-1 for Racetrack Test, the coil configuration achieved a field of approximately 12.2 Tesla (12.36 Tesla peak field on the winding), more than twice that of RD-2. This was made possible by the availability of state-of-the-art superconductor with a non-Cu Jc of over 2,000 A/mm2. This represents a factor of three increase compared to the conductor used for RD-2.

The construction of magnets at higher fields requires careful consideration of mechanical support structure issues. In fact, aside from the conductor, the support structure is the most important factor in developing a robust and cost effective high field magnet. Relative to the 6 Tesla magnet (RD-2) the 14 Tesla magnet will have forces 5 times higher. The integrated horizontal Lorentz forces total about 9 MN (1,000 tons), acting to push the windings apart. The success of the interim test using the outer coil modules prompted development of a coil support design that makes use of inflatable bladders which are used as a temporary internal "press" to load the coil modules inside an aluminum shell, using a support structure geometry similar to that of RT-1. The operational parameters of the bladders have been determined to be more than sufficient, reliably operating in excess of 12 kpsi with a stroke of 1 mm and up to 10 kpsi with a 3 mm stroke. A 1/3 scale, mechanical model test has been completed. The yoke and shell were assembled with a "dummy" coil module and cooled to liquid helium temperature. Resistive strain gauges provided data that were compared with analytical and FEM model predictions. The model was intended to demonstrate that the fundamental design provided the required preload and behaved in a manner consistent with that predicted by analysis. It also served to verify that the assembly procedure was as simple as anticipated and that the iron and aluminum alloys used for the structure were adequate for this application. The results of the mechanical test were consistent with model predictions, the only exceptions being differences that can be understood in terms of frictional effects. All necessary criteria for pursuit of this design have been met and have resulted in the adoption of the PBS support structure as the new baseline. With the critical questions answered, it is believed that this scheme will be simpler and faster to implement than the wire wind method previously proposed.

The inner coil module was combined with the two outer coil modules using the new support scheme and the final assembly of RD-3 was completed in August. Testing commenced in early September, but during the first ramp an insulation failure occurred, which resulted in arc damage to two modules. The damaged coils are now being rebuilt and we anticipate re-testing the magnet in April of 2001.

Continuing Program

Efficient magnetic field generation requires that the conductor be as close to the bore as possible. Our initial goal is to choose a coil geometry and support structure which will adequately support the coil with little or no training. Once this step is successfully accomplished, we will begin considering cross section designs that provide field quality. Racetrack cross sections which have good magnetic field quality are intrinsically more complex and the simple cross section will need to be augmented. The more evolved designs will also need to take into account the operational consequences of high field magnets and the issues associated with the conductors we use to achieve high fields. For example, magnetization effects are much greater for Nb3Sn due to the higher current density and large filament diameter. Ways to minimize this effect are being investigated both in terms of conductor development and in magnetic design. Other consequences of high fields are the large heat loads generated by synchrotron radiation. The muon collider magnets will also have a heat load problem due to the electrons from muon decay. Solutions to these problems will require optimization of the coil geometry, field requirements and cryogenic design.

The next magnet, RD-4, will be a modification of RD-3 to achieve better field quality. Racetrack cross sections that have good magnetic field quality are intrinsically more complex and the simple cross section will need to be augmented. This will involve rebuilding the inner coil with spacers and a different conductor placement, together with auxiliary coils. The outer layer coils and support structure can be reused with small modifications. As a consequence of the constraint to re-use the RD-3 outer coils, the field of RD-4 will be substantially lower. However, a successful test of this magnet will be the first example of a large bore, common coil magnet with good field quality.

While we develop designs for today's applications, such as cost effective magnets for the VLHC, we are continuing our primary goal of pushing technology to the limit in order to construct the highest field dipole magnets

pursuit of the ultimate in high field dipole magnets. What is the ultimate field limit for a dipole? What conductor properties are required? Our goal is to push the field to the mechanical limit of materials and available superconductors and to define the properties required to go even further.