Superconducting Magnets

Contact: Program head Steve Gourlay.

End view of the experimental accelerator-style dipole HD1, which at 16 tesla is the latest in a series of record-setting magnets Ever-stronger magnets (which must be cost-effective as well) are a key to building tomorrow's high-energy accelerators and upgrading today’s. Our role—as not only a leading R&D group but also the administrators of the multi-institutional National Conductor Development Program—is to create both evolutionary improvements and paradigm shifts in the application of accelerator magnets, providing innovative technology that enables new science. Improvements in conductor, innovative structures to solve the challenges of high fields and brittle superconductors, and integration of computerized design and analysis tools are key.

The performance requirements of modern accelerators continue to press the limits of magnet technology. Ever-higher beam energy is a constant goal in high-energy physics. Stronger magnets mean that a given energy can be achieved in a smaller machine that needs less of every kind of infrastructure and perhaps even fits into an existing tunnel. With the cost of major accelerators in the hundreds of millions or even billions of dollars, not only these technical goals but also the cost-optimization of components are more important than ever.

Such matters are the focus of our base program here at LBNL— a program that is vertically integrated “from melt to magnet,” covering superconducting materials, cablemaking, and design and testing of the magnet. Our efforts, which are extensively collaborative with colleagues elsewhere and with industrial partners, are a coordinated part of an overall Department of Energy program.

In late 2004 we achieved fields of 16 tesla in an accelerator-style dipole magnet, at a time when other programs had reached no further than 11.5 T. Now we are considering how to push toward 20 T—an absurd number just a few years ago for practical accelerator-style magnets, but one which we have come to regard as a reasonable goal.

The implications go beyond our principal role, which involves present-day and next-generation projects in high-energy physics. Other applications that could benefit include magnetic resonance imaging; insertion devices for light sources; and the upcoming International Thermonuclear Experimental Reactor (shown), a large-scale fusion-energy experiment that will pose many interesting magnet challenges. Computer-Aided Design view of the International Thermonuclear Experimental Reactor

Background and Context

Our program was established in the early 1970s, when superconducting accelerator magnets were still a very young technology. We worked on dipoles and quadrupoles in support of such projects as ESCAR (Experimental Superconducting Accelerator Ring), ISABEL, and the Superconducting Super Collider, successfully industrializing the 6.6-tesla collider dipoles for the SSC (though that accelerator was never built) and achieving fields as high as 10 T in experimental accelerator-type magnets.

Ten tesla was about the limit for ductile superconductors such as niobium-titanium, so in the early 1990s we moved to niobium-tin and began developing new magnet designs and construction techniques more suited to these materials. Recently we successfully tested the world’s first dipole magnet constructed of niobium-tin materials (Nb3Sn) operating at 16 T. This magnet, HD1, is shown at top left of this page. It was the latest achievement in a continuing program to develop advanced high-field accelerator magnets.

We have also achieved a near-complete integration of design tools, leading to an unprecedented understanding of performance issues. Shown here is a Computer-Aided Design rendering of HD1. The proportion of actual superconductor (dark blue) relative to the mechanical structures and ancillary systems indicates how many considerations besides the magnetic design are important in a successful high-field superconducting magnet. Computer-Aided Design view of HD1

Having reached 16 T (which might be extended to 17 T without redesign simply by reducing the temperature to 1.8 rather than 4.3 K), we think 20 T is a reasonable goal for accelerator-type magnets made with Nb3Sn.

Activity timeline showing recent and projected achievements by year The overall goal of our endeavors in materials and in magnet design is simple: ever-stronger magnets of the type used in high-energy physics colliders. The details of getting there, of course, amount to many years’ work. Clicking this image will bring up an illustrated timeline (140 kB) summarizing our accomplishments and plans in various lines of inquiry.
While working to advance the state of the art in superconducting material, forming it into cable, and designing and building magnets, we emphasize cost-effective approaches to conductor and magnet fabrication techniques, for cost control too is a key aspect of the new and upgraded accelerators of the future.

Nb3Sn has been a cornerstone of the magnet program at LBNL for over a decade. Although we have used it in several magnets with progressively higher fields, reaching 13.5 T in 1996 and 14.3 T in 2001, the 2004 achievement of 16 T was a major confidence booster, solidly establishing the feasibility of the substance for accelerator applications. The program was given additional impetus by LARP, the US LHC Accelerator Research Program. One of LARP’s tasks is R&D for an eventual luminosity upgrade of the Large Hadron Collider, an endeavor so important that it was described as an “absolutely essential” medium-term goal by the DOE High-Energy Physics Advisory Panel—and so challenging that it is underway even before the initial version of the LHC is completed.

LBNL is managing the magnet portion of LARP. Although the initial version of the LHC will use NbTi, the upgrade will require larger apertures and high gradients combined with good field quality, as well as (in some positions) tolerance of radiation-induced heating; the job can only be done with Nb3Sn.

Materials R&D and the DOE Conductor Development Program

In addition to the goal of high field, any viable candidate for a future collider must be very cost effective. Making magnets both stronger and more affordable starts with better materials and cable, and the goals of the National Conductor Development Program are quite demanding: a critical current density of 3000 A/mm2 (at 4.2 K and 12 T) and a net cost of $1.5 per kA-m. (For comparison, the NbTi superconductor for the Superconducting Super Collider in the early 1990s had the same cost per kA-m, but the field was only 6.6 T.) Now in its sixth year and supported at $500k per year, the National Conductor Development Program has made impressive progress towards the goals of high field and low cost. The Jc (critical current density) of Nb3Sn conductor now exceeds that 3000 A/mm2 goal, and we are projecting that the cost goals can be met when the manufacturing processes, now at the research stage, are properly industrialized.

With those goals achieved or in sight, the emphasis of the program is turning toward reducing filament diameter while maintaining the current density. Related studies include:

  • Cabling work to reduce critical current (Ic) degradation.
  • Techniques for making cable out of new conductor designs such as powder-in-tube.
  • Heat treatment studies to optimize the residual resistivity ratio and Jc.
  • The effect of transverse strain on degradation and the relationship of strain degradation to the conductor substructure.
Besides our work in the National Conductor Development Program per se, we continue to provide significant cabling support for other magnet and cable development programs.

This work is essential to the success of the magnet program because, ultimately, magnet performance is determined by the conductor. There are many critical conductor parameters that contribute to the success of a magnet design, and they must be optimized as a whole, with consideration given to the trade-offs between individual parameters.

Magnets: Recent Achievements and Future Directions

For several years it has been clear that progress much beyond 10 T requires the use of the “A15” compounds, the most practical and available of which is Nb3Sn. One of the challenges as we move away from ductile materials is to find magnet geometries that support the brittle, strain-sensitive A15 conductors against the powerful Lorentz forces inside a high-field magnet while keeping the stress to below 200 MPa or lower. Another key change was the need for "wind-and-react" fabrication techniques; the heat treatment that turns these materials into superconductors also makes them brittle, so the coils must be fabricated and then heated, in contrast to the older materials, which remained flexible after their heat treatment. Therefore the successful push toward 16 T over the last few years required not only new materials but also new magnet and support-structure designs, fabrication techniques, and instrumentation.

The route to these very-high field accelerator magnets is being pursued along parallel research paths. One, of course, is the construction of accelerator magnets about one meter long and otherwise dimensionally realistic. Another is the use of sub-scale magnets for focused technology development. This sub-scale approach i gives us a cost-efficient way to test support-structure designs, conductors and cables, quench-protection schemes, and fabrication techniques. If new materials such as MgB2 or Bi-2212 become available in sufficient quantity and with good properties, coils of those materials will be fabricated and tested in the sub-sized models.

It also became clear that sophisticated analysis is the key to building successful magnets, so we have placed more and more emphasis on this area, making significant progress. Our design team has combined a number of engineering tasks into a single streamlined process. Integration of mechanical, magnetic, structural, thermal and electrical design has now become routine.

This level of integration of multiple disciplines, which appears to be unique to our program, is used to investigate new and more complex magnet designs. It will also help us unravel complex mechanics-related performance issues, such as the near universally observed but still incompletely understood “training” process by which a magnet reaches its full potential in a series of ramps to successively higher fields, interspersed with quenches.

Industrial Synergy

Much of our magnet R&D—especially the development of advanced superconductors and improved manufacturing methods—involves close collaboration with industry. This area of R&D includes superconducting materials with high critical current density at high field, as well as very fine filaments (to improve stability and reduce losses). New cable manufacturing methods and insulation materials are also required. These collaborations are carried out through the DOE/HEP National Conductor Development Program as well as the Small Business Innovation Research (SBIR) program. Each year we contribute to the SBIR program by providing technical support to the companies and advice on the DOE programmatic goals.
Photo of ex-situ magnet prototype Development of a superconducting magnet for ex-situ nuclear magnetic resonance is a Laboratory-Directed R&D project in progress here at LBNL. The resulting magnet, shown here at the prototype stage, will be used by the research group led by Alex Pines of the Materials Sciences Division and UC-Berkeley. Superconducting undulators for synchrotron-light sources have also been the subject of a recent LDRD project.

Exploring Further

For additional technical details on our progress in both materials and magnets, where we plan to go in 2005 and 2006, a closer look at our LDRD projects and other notable spinoff benefits, and links to papers we published, download the Superconducting Magnets chapter of the AFRD Research Highlights in Portable Document Format.
The Superconducting Magnet Program has a website of its own with information on its people, projects, and technologies. Here are some other relevant pages on other servers at Berkeley Lab and beyond:
Administrative and legal information on the AFRD homepage is applicable to this page.