FUSION 233 the physical properties of magnetically confined plasmas; operational limits in high pressure regimes—in particular we plan to use the flexibility of TJ-II to exper­ imentally validate existing theories of instability development; and studies of confinement optimization and its relation­ ship to radial electric fields. Encouraged as we should be by all the progress made in the last decade in mag­ netic confinement, one of the greatest sci­ entific challenges confronting our commu­ nity is still open: the development of mod­ els to fully explain the mechanisms under­ lying the transport of energy in fusion plasmas. Our hope is that the experimen­ Fig 2 left the calculated confining magnetic surfaces of possible to its physical limits. tal flexibility of TJ-II could play an impor­ TJ-II and right the experimental ones measured at 5% of Main areas of research in TJ-II are: tant role in finding the solution to this the nominal field using the fluorescent rod method.The fundamental question. The facility is up excellent agreement between both Indicates that the studies of plasmas in low collisionality positioning and geom etry of all coils is within the regimes aimed at understanding the and running and at the full disposal of our specified construction tolerances (< 1 mm) nature of transport mechanisms caused by European scientific community.

Fig 1 below Photograph of a plasma in START. A central column carries the current to make the main toroidal MAST The Spherical Tokamak magnetic field, but it is so compact that the overall impression Is of a spherical plasma.The diameter of the 'sphere' is about 1 m and It carries up to 300 kA William Morris stability. In fact, Culham Science Centre, Abingdon, UK although these devices may be

When magnetic fusion has reached such built on a wealth of experience from tra­ an advanced state, it is fair to ask why ditional tokamaks. another variant on the tokamak is being The ST is largely based on theoretical discussed, as well as how it is possible for ideas initially developed in the late 1970s a torus to be spherical. Figure l addresses and 1980s. One of the major costs and both questions: the simplicity and com­ engineering challenges for the tokamak pactness of such a plasma have attrac­ is the production of an adequate mag­ tions as the core of a fusion device, and netic field to confine the plasma, so it the plasma is spherical in appearance. is important to make optimal use of There is a firm database in support of it. This means that the plasma pres­ a burning plasma device with a conven­ sure should be as high as possible for tional aspect ratio (ratio of major to a given field -the fusion performance minor radius of the torus), as embodied can be written in terms of the product in the ITER design for example. If, howev of the density, temperature and con­ er, there are advantages in pursuing a finement (or energy replacement) time. somewhat different plasma geometry, for Theoretical calculations using magne- instance with a larger or smaller aspect tohydrodynamic stability codes, tested ratio, then present confidence is limited against existing tokamak data, show that by the restricted range of aspect ratio the highest ratio of plasma pressure to data in the experimental database. A magnetic field pressure is achieved by major aim of MAST (Mega Amp Spherical reducing the aspect ratio as far as possi­ Tokamak) is to broaden the tokamak ble, and raising the plasma current database as well as to provide a novel (and/or reducing the magnetic field). environment in which to test many Furthermore, good plasma confinement thought of aspects of tokamak physics and opera­ time is found experimentally to be relat­ as low field, tion, and transfer results to more conven­ ed to high plasma current. A natural fea­ the field on the tional aspect ratio devices. Conversely, ture of ST geometry allows plasmas with inboard side can be as progress with the spherical tokamak (ST) very high plasma currents to be confined high as on larger aspect ratio devices. concept has been extremely rapid as it is at low magnetic field while maintaining One of the traditional concerns about a

5.2 Mast The Spherical Tokamak Morris Europhysics News November/December 1998 234 FUSION a scaled up START (a factor 2 in linear dimension) with a much longer pulse length (up to 5 seconds). The 4.4 m high 4 m diameter cylindrical vacuum vessel is as tall as the JET vessel. There have been, however, some engineering challenges. Perhaps the foremost of these is the design of the centre column (comprising 24 wedge-shaped conductors and a com­ pact high-stress solenoid) whose diameter has to be minimised subject to mechani­ cal, thermal and electrical stress con­ straints in order to minimise the aspect ratio of the plasma. Since the centre rod expands during a shot, and the vacuum vessel during baking, relative motion of up to 12 mm has to be accommodated to avoid fracture of the toroidal field coils or joints: a comprehensively-tested high cur­ rent sliding joint is used at the ends of the centre rod. Since the ST is a new concept, it pre­ sents an opportunity to combine radical design solutions with the large technolo­ gy database developed for the convention­ al aspect ratio tokamaks to produce com­ pact low-capital cost fusion devices, once the underlying physics has been demon­ Fig 2 Above Cutaway drawing of the MAST load early stage of development). strated experimentally. Potential applica­ assembly. Below the nominal parameters All such theory has to be tested. During tions include a compact volume neutron 1988 to 1990, a very low budget proof-of- source for testing materials and large Plasma current 1-2 MA principle device, Small Tight Aspect Ratio complex components (blanket modules, Toroidal field (0.7m) <0.63 T Tokamak (START) was built at Culham, welded/brazed assemblies etc) for any Major radius 0.7 m UK, almost entirely from spare parts. It fusion power plant, and a low capital cost Minor radius 0.5 m Aspect ratio >1.3 finally ceased operation in March 1998 core of a power plant as a long-term alter­ Elongation ~2 after producing such an astonishing set of native. Outline designs exist for a driven Pulse length <5 s results that it had become abundantly component test facility (Q~1) generating Auxiliary heating power, NBI 5 MW clear that the ST concept had to be pur­ some tens of megawatts of fusion power, Auxiliary heating power, ECRH <1.5 MW sued on a larger device. These results which would enable components with a included: the achievement of a β value total surface area of several square meters tight aspect ratio is the expectation that (ratio of average plasma pressure to to be exposed to neutron fluxes and flu- there will be many particles ‘trapped’ in toroidal magnetic field pressure) that ences at power-plant-level. There are, how­ the magnetic mirrors of the tight torus, exceeded the record on any other toka­ ever, some technological issues to be and these will have large excursions from mak by more than a factor of 3; energy resolved for a ST fusion device, as might the magnetic surfaces, leading to large confinement times in excess of that pre­ be expected with a new concept. The next losses. Also, it had been thought that the dicted for START using the latest scalings step is to proceed with the physics basis, bootstrap current (required for steady- from the ITER physics R&D activities; and and data from START, MAST and other state operation) would be small. Here the configurations that may, from experimen­ new STs around the world combined with answer lies in the detail, and in careful tal evidence on START, be free from ter­ accompanying theoretical studies should application of the relevant theory to accu­ minal plasma disruptions (which are a allow designs for fusion devices to be cre­ rately computed magnetic field distribu­ serious design issue for large tokamaks). ated based on proven physics. tions. When the theory of neoclassical The adaptable MAST experiment is the The first extended physics campaign on transport (originally a large aspect ratio next step in Europe and complements the MAST is planned for the first part of 1999. approximation) is modified for the spheri­ similarly sized NSTX facility under con­ The work described in this article is jointly cal tokamak it is found that the cross-field struction at Princeton in the US, these funded by the UK Department of Trade & transport is much reduced, while the boot­ being the largest ST devices in the world. Industry and EURATOM. strap current remains substantial. This is It will have strong additional heating Further reading an ongoing research field, and includes (5 MW of injected 70 keV deuterium Y-K.M. Peng, D.J. Strickler Features of spherical torus more complex particle orbits such as ‘pota­ atoms, and 1.5 MW of 60 GHz microwaves plasmas Nuclear Fusion 26 769 (1986) to’ orbits to supplement the well-known to match the electron cyclotron reso­ A. Sykes et al High-β performance of the START ‘banana’ orbits. There does indeed appear nance), and will be of a size (minor radius spherical tokamak Plasma Physics and Controlled to be self-consistent steady state scenarios ~50 cm) and current to allow comparison Fusion 39 B247 (1997) for spherical tokamaks in relation to with the existing medium-sized MA-level A.C. Darke et al The Mega Amp Spherical Tokamak Proc 16th Symposium of Fusion Engineering 2 1456 transport and current drive (solutions to conventional-aspect-ratio tokamaks in (1995)

Europhysics News November/December 1998 power and particle exhaust issues are at an Europe and elsewhere. MAST is essentially