Table of contents

The confinement properties of future magnetic fusion devices can be extrapolated from existing experiments using the scaling of heat transport with relative gyroradius ( rho _*). For H mode confinement regimes, this extrapolation to smaller rho _* is complicated by the requirement that the loss power from core transport remain above the H mode threshold power. If the increase in H mode threshold power is more rapid than the increase in loss power when rho _* becomes smaller, then the H mode threshold determines the scaling of global confinement for large devices rather than core transport. This can be avoided by choosing a scaling path at higher beta along which the loss power remains above the H mode threshold power to the point of ignition. The parameters for an ignition tokamak plasma that take full advantage of the gyro-Bohm-like scaling of heat transport in H mode plasmas are presented

Strong shaping, favourable for stability and improved energy confinement, together with a significant expansion of the central region of improved confinement in negative central magnetic shear target plasmas, increased the maximum fusion power produced in DIII-D by a factor of 3. Using deuterium plasmas, the highest fusion power gain, the ratio of fusion power to input power, Q, was 0.0015, corresponding to an equivalent Q of 0.32 in a deuterium-tritium plasma, which is similar to values achieved in tokamaks of larger size and magnetic field. A simple transformation relating Q to the stability parameters is presented

A magnetic determination of the dipole moment of the Pfirsch-Schluter current in a finite beta stellarator/heliotron plasma is described. The rotating cosine saddle diagnostic coils and the poloidal flux loops are used to identify the dipole current in a finite beta toroidal plasma. A typical dipole moment of the toroidal current in the standard Heliotron E configuration is 0.6 kA m, when the volume averaged beta is 0.50% at a magnetic field of 0.94 T. The measured equilibrium dipole current is practically proportional to the diamagnetic volume average beta available up to 0.95%. The observed dipole moment of the Pfirsch-Schluter current is compared with a magnetohydrodynamic (MHD) calculation for a finite beta stellarator plasma

A new code, Magnetic field solver for Finite-Beta Equilibria (MFBE), has been developed. It computes the magnetic field in a form suitable for field line tracing inside and outside the plasma boundary of finite-β Helias equilibria calculated with the NEMEC free boundary equilibrium code. Coupling of the NEMEC and MFBE codes allows a determination of the last closed magnetic surface. For various beta values up to (β) ≃ 0.04, the variations of the island sizes and positions outside the plasma boundary and the ergodization of the edge region are studied for a Helical Advanced Stellarator (Helias) configuration (Nuhrenberg and Zille, Phys. Lett. A 114 (1986) 129)

It is shown that when the resonant vertical chord omega = omega _BI (here omega is the applied frequency and omega _cI is the impurity cyclotron frequency) is located in the ripple well domain, strong convective outflux of the resonant impurity arises due to adiabatic trapping into the local magnetic well. This outflux could readily overcome the convective influx caused by parallel friction forces, thus preventing central impurity accumulation

A neutral particle analyser was deployed on JET for a measurement of the distribution function of megaelectronvolt energy protons driven by the high power ion cyclotron resonance frequency heating of deuterium plasmas in the hydrogen minority scheme. Unexpectedly, efficient neutralization of megaelectronvolt energy protons was observed in the plasma centre without recourse to injection of a beam of atoms to provide electron donors for charge exchange neutralization reactions. A model is presented that elucidates the role of carbon and beryllium, the main intrinsic impurities in JET plasmas, in this process and establishes charge exchange between hydrogen-like ions of the impurities and protons as the main neutralization process. A model calculation for deducing the proton energy distribution function f(E_p) from the measured hydrogen flux is described. The effects of uncertainties in neutralization cross-sections on the inferred f(E_p) are examined. The validity of this model of impurity induced neutralization (IIN) is tested by using it to describe the measured hydrogen flux in different conditions of plasma heating and fuelling. A crucial point in making these tests was to use measured local values of bare impurity ion densities; the required measurements were realized using charge exchange spectroscopy. Using IIN modelling and a procedure in which a known change in the density of deuterium atoms at the plasma centre was made by applying neutral beam injection, we have deduced the background thermal deuterium atom density at the plasma centre, which is an important new diagnostic result. Concerning future experiments, the model predicts that carbon and beryllium impurities will be major contributors to neutralization of hydrogenic ions (protons, deuterons and tritons) in ITER plasmas, for ion energies E<or=1 MeV/u. According to these estimates measurements of the distribution function, in the megaelectronvolt energy range, of ion cyclotron resonance frequency heating driven ions of hydrogen isotopes

An ion cyclotron range of frequencies (ICRF) heating experiment was performed on the Compact Helical System (CHS) device in order to identify the problems in establishing this as a reliable heating method. Radiofrequency heating was applied to plasmas with two ion species produced by ECH. Stored energy increased upto 2.2 kJ with the application of a 590 kW ICRF heating pulse, giving a heating efficiency comparable to that of NBI heating. More importantly, the `flat-top' of the stored energy was maintained during the pulse. Good heating was realized with 30% proton and 70% deuteron plasmas at high density (4*10¹⁹ m^-3). It was found that the best discharges are dominated by electron ICRF heating, which seems to be the reason for the successful results. On the contrary, a flat-top in stored energy has never been obtained in discharges where ion heating dominates. The loss of high energy ions with large pitch angles produced by ion heating appears to degrade the performance and limit the duration of the heating. A similar degradation was encountered in previous ICRF heating experiments in helical systems. A newly designed detector was used to study the loss of the helically trapped particles

Electron cyclotron (EC) power has technological and physics advantages for heating and current drive (CD) in a tokamak reactor, and advances in source development make it credible for applications in the International Thermonuclear Experimental Reactor (ITER). Strong single pass absorption makes heating to ignition in ITER particularly simple. At densities to 3.6*10²⁰ m^-3, with ohmic temperatures and wave frequency 170 GHz, heating in the plasma core is readily obtained. For outside launch of an ordinary mode (O mode) near the fundamental EC frequency, the optimized EC current drive (ECCD) efficiency ((n)IR/P) shows a linear temperature scaling at temperatures up to ~15 keV. For temperatures above 30 keV, the efficiency saturates at approximately 0.3*10²⁰ A/W.m² for a frequency of 220 GHz in an ITER target plasma with a toroidal field of 6 T, due primarily to harmonic overlap and to a lesser extent due to limitations arising from relativistic effects. The same efficiency can also be obtained at 170 GHz for the same plasma equilibrium and q profile except that the magnetic field is reduced to (170/220)*6 T=4.6 T. The ECCD efficiencies are obtained with the comprehensive 3-D, bounce averaged Fokker-Planck codes CQL3D and BANDIT3D

Naturally occurring electric current spatial profiles were measured at high spatial resolution by a probe array installed in the JET Mark 1 divertor target. Very narrow current layers near the separatrix strike point on the divertor target were observed. The current flow directions in the narrow layers changed with toroidal field direction. The currents do not behave like previously identified thermoelectric currents. A plasma plus neutral quasi-single-fluid MHD model is developed to adequately represent the scrape-off layer (SOL) equilibrium all the way through the divertor recycle zone and sheath to the target. The model shows that SOL Pfirsch-Schluter current passes through these zones relatively unattenuated and is measurable by target mounted probes. The observed new current features are consistent with being SOL Pfirsch-Schluter currents. Some possible consequences to the equilibrium when the Pfirsch-Schluter current is reduced by divertor plasma detachment from the target are briefly discussed

The temperature and density profiles of the Joint European Torus (JET) are parametrized using log additive models in the control variables. Predictive error criteria are used to determine which terms in the log linear model to include. The density and temperature profiles are normalized to their line averages (n and T). The normalized ohmic density shape depends primarily on the parameter n/B_t, where B_t is the toroidal magnetic field. Both the low mode (L mode) and the edge localized mode-free (ELM-free) high mode (H mode) temperature profile shapes depend strongly on the type of heating power, with ion cyclotron resonant heating (ICRH) producing a more peaked profile than neutral beam injection (NBI). Given the heating type dependence, the L mode temperature shape is nearly independent of the other control variables. The H mode temperature shape broadens as the effective charge, Z_eff, increases. The line average L mode temperature scales as B_t^0.96 (power per particle)^0.385. The L mode normalized density shape depends primarily on the ratio of line average density, n, to the edge safety factor, q₉₅. As n/q₉₅ increases, the profile shape broadens. The current, I_p, is the most important control variable for the normalized H mode density. As the current increases, the profile broadens and the gradient at the edge sharpens. Increasing the heating power, especially ICRH, or decreasing the average density, peaks the H mode density profile slightly

Tearing mode formation is investigated for ohmic density limit discharges on the ASDEX Upgrade tokamak at medium and high safety factors (q₉₅=3.8, 4.9, 6.0). Low electron temperatures inside magnetic islands and the observation of localized C III impurity radiation suggest that a thermal instability, as proposed by Rebut and Hugon, destabilizes the (m, n)=(3,1) and (4,1) islands, which grow during the current profile contraction phase. In contrast, the (2,1) islands appear to be thermally stable. Minor disruptions lead to step-wise loss of confinement, first localized at the 4 2 surface, then after a time delay comparable to the resistive time-scale, at the q=3 surface and after a second time delay, at the q=4 surface. It is found that (3,1) islands, unlike (2,1) islands, are quenched by the high heat flux during minor disruptions

Discharges with strongly peaked density profiles, density peaking factor >4, have been produced on the Tokamak Fusion Test Reactor (TFTR) tokamak by deuterium pellet injection. Following the pellet perturbation, these plasmas have been heated using ³He or hydrogen minority ion cyclotron resonance heating (ICRH) at power levels up to 6 MW. As seen in experiments on other devices, a strong transient enhancement of the central plasma parameters, in particular of the fusion reactivity and central electron pressure, has been observed in these discharges. The transport characteristics of these plasmas during the period of enhanced core confinement have been analysed using the TRANSP time dependent interpretive code. The time evolution of the q profile is calculated and compared with measurements. The influence of plasma current and heating power is examined. The local effective thermal and electron particle diffusivities are inferred and compared with those from similar non-enhanced discharges produced by shallow pellet fuelling. The core diffusivities during the enhanced phase are found to be more than a factor of 2 lower than those in the non-peaked discharges, with improvement localized within the plasma core. Enhancement in global confinement is generally modest but operation at low edge q appears to increase the volume of the enhanced region and the magnitude of global confinement. The enhanced phase is terminated by a magnetohydromagnetic (MHD) mode localized off-axis. Operation at higher central q by earlier injection of the pellet and higher current delays the onset of MHD and extends the duration of the enhanced phase. Operation at low q increases the radial extent of the core region