Plasma Subgroup Question #2 Comments

Martin Peng Sun 25 April 1999 12:51 Subject: Normal Magnets for High Beta Confinement Systems Dear PQ2 of Plasma Technology Topic Leaders, I am very impressed with the rapid preparation of the nice WebPage information on the Technology WG. The prospectus on magnet issues is already clearly written. My comment addresses the question: under what conditions will normal conductor magnets be appropriate for steady-state fusion energy development or producing devices? I would also like to find out if there are good reasons to include this type of question in the key questions and issues list within KQ2. The basic reasons are that normal conducting magnets can be: 1) Acceptable for energy development devices such as the VNS, where economy is not a critical issue whereas the operating cost may not be a go-no-go issue 2) Economic for power plants where the magnetic confinement system provides very high beta, such as projected for the ST (~50% toroidal average beta) and the FRC (~100% average beta) 3) Indispensable for ST, where a single turn center leg of the TFC without using insulator in the high dose region, to permit compact devices in the case of VNS and achieve economy in the case of ST power plants. I look forward to receiving your comments on whether this or a similar question could be included in the KQ2's list of key questions and issues. Thanks. Martin

Joel Schultz Wed 30 June 1999 11:19 The following is a draft of answers to the Common Cross-Cutting Questions for Snowmass. A. What are the most important contributions that technology can make over the next 10 years: 1) to improve the vision for an attractive and competitive fusion product, and 2) reduce the cost of R&D for fusion? 1) Attractive and competitive fusion product The purpose of magnet technology is to produce the fields needed by magnetic and inertial fusion. The magnets should be inexpensive, compact, simple to assemble and maintain, and high field, where appropriate. Over the next 10 years, we imagine a "mixed-magnet" portfolio, as there has been in the past 10 years, including normal copper and superconductors, wound and plate magnets, and air, water, nitrogen, and helium cooling systems. The major difference is that we will begin to introduce high-temperature superconductors in fusion experiments, during this period. Low-temperature superconductors are now sufficiently mature that they should be used only when they save total program cost or enhance mission, when compared with normal magnets. A recent example would be the selection of a Nb3Sn Floating-Coil and NbTi Charging Coil for LDX, with all other magnets air or water-cooled. Another interesting example may be NCSX, where normal magnets have marginal performance for the needed current density and pulse length, while superconducting magnets provide long- pulse performance margin, if pulsed performance is adequate. A third might be an upgrade of the Electra Helmholtz coils with superconducting magnet pairs of similar overall size, but 50 times less conductor and losses. A brief survey of upcoming experiments indicates that the program should deliberately begin the fabrication of small, inexpensive high-temperature superconductors for near-term experiments and moderate size magnet systems for medium-term (2-5 years from now) experiments. The reason for this is that several orders of magnitude in scale and two orders of magnitude of current scaleup have to be done, in order to be prepared for a large experiment in twenty years. The adoption of new manufacturing technologies that would lower magnet system cost can also be accelerated by the fusion program. These include the development of fully parameterized coil winding robotics, which are particularly useful for the large number of small quadrupole magnets in Heavy Ion Fusion Drivers. The use of "additive manufacturing", in which laser or electron-melted or electromechanically deposited material builds up a component may be used to lower the cost of ST central columns or superconducting conduits. 2) Reduced cost of R&D Because of the diversity of the fusion program at this point, the only way to reduce magnet costs with a small development program is to reduce the cost and improve the performance of magnet components. We have introduced metrics for the superconductor, matrix, structure, insulation, joints, quench detection system, thermal isolation, and electrical isolation components for this purpose. In all cases, it appears that inexpensive programs (< $1 M/year/component) can double the performance/cost ratio of each component with the significant exception of structures, which require the multiyear level of effort that was, for example, needed for the development of Incoloy 908. Uncertainty as to which fusion topologies are the most important can be minimized by concentrating on the improvement of "cross-cutting" technologies, which would benefit almost all possible topologies. Examples would be the development of radiation-resistant, high shear-strength insulations with broad operating temperature range; high-current density, high-current, high-temperature superconductor cables; and low volume, low DC loss, low AC loss joint topologies. B. What are the new technology issues that must be solved to allow the continued exploration of "Advanced Tokamaks" and to enable full development of the recently initiated or planned innovative confinement concepts and next step devices? "Solve" and "enable" are overly strong terms to apply to magnet development with regard to Advanced Tokamaks and some of the planned innovative confinement concepts for the very reason that these concepts were invented in order to reduce the magnet requirements through the use of high-beta or simpler magnet systems. Thus, in the cases of Spherical Tokamaks or Spheromaks, questions A and B have the same answer that magnet development can further reduce the size and cost of magnet systems. A possible "enabling" technology is the improvement of control magnet system design in order to extend elongation or aspect ratio, and either eliminate internal and field error control magnets or design them to be easily maintainable. Some of the innovative confinement concepts and next-step devices do present magnet design issues that will determine concept viability. The size and cost of Heavy Ion Fusion Drivers has a strong dependence on radical improvements in magnet design and technology, because the beam current density itself is inversely proportional to magnet size. Furthermore, the cost of the linear induction accelerator and the accelerator buildings and services are also directly proportional to the quadrupole array size. Stellarators, with their combination of complex geometry and comparatively low beta are challenges for the magnet system. In the near-term, the NCSX is challenging, because of the combined high current-density, low field-error requirements. Stellarator reactors will be challenging, because of the need for blanket and first-wall accessibility. Levitated Dipole reactors and experiments also have unique difficulties. In the near-term, the problem is guaranteeing safe operation of a levitated cryostat in a large tank, while the reactor has the nearly insoluble problem of designing a floating magnet system with onboard refrigeration, persistent current, and no mechanical links to the outside world. C. What contributions will Technology make to advancing science? What research areas will be pushing the frontiers of science? What constitutes concept exploration, engineering proof of principle and engineering performance extension for fusion energy systems? The answers to A and B apply also to C. Improvements in magnet size, cost and simplicity lowers the cost of all new scientific experiments involving magnets. 1) Advancing Science Magnet science is also an integral part of the fusion program. The theory of superconducting cables and twisted-filament strands has uncanny similarities to that of toroidal plasmas, including the helical current paths, a multiplicity of circulating current modes, coupled thermal-magnetic-mechanical behavior, and the existence of disastrous current quench/disruption modes. In both cases, the complexity of the problem has defied brute-force physical simulation; but, since the physical mechanisms are known, persistence should ultimately permit the design of superconducting magnets on a fully scientific basis. Success in physical modeling of superconducting cable transients will also improve the performance of all pulsed applications, including Superconducting Magnetic Energy Storage (SMES), transformers, generators, and motors, including magnetic levitation. 2) Research Areas Materials science is integral to the improvement of magnet systems. Magnets benefit from 1) the development of superconducting materials with higher critical current, field, and temperature that are also more strain-tolerant, 2) Stronger, tougher, and more weldable superalloys; as well as alloys that are less expensive and more tolerant of processing conditions, 3) compact insulations and isolators that withstand radiation, higher voltages and a shift to fiber optic, higher bandwith and nonconducting instrumentation systems. Materials science is also reaching a breakpoint in which new alloys can be designed less empirically by modeling and designing improved properties, before development. 3) Concept Exploration, Proof of Principle and Engineering Performance Extension A strong program is essential for data interpretation from component development experiments for concept exploration and for the advancement of codes for performance prediction. This effort should be coupled with fusion requirements for particular systems, for example HIFD quadrupoles, to progressively produce and test elements, then arrays, then incorporate more effective cost reduction manufacturing techniques. Extending engineering performance can be done by operating components beyond their design allowables in order to compare predicted and achieved margins in order to gain the knowledge base necessary to reduce margins. Active participation in international programs involving experiments on large scale magnet systems to measure performance against design is also essential.