Table of contents

Electron heating during ICRH in the JET tokamak is studied. Electron heating profiles are obtained by measuring the increase in electron temperature after sawtooth crashes. By studying the increase in electron heating during power ramp-up it is possible to distinguish direct electron heating from indirect electron heating via minority ions. For direct electron heating, power deposition profiles are obtained. For indirect electron heating, the time dependence is consistent with classical slowing-down of high energy ions on electrons through Coulomb collisions.

The stability properties of the toroidal ion temperature gradient (ITG) mode change dramatically for plasmas with flat density profiles, as found in H-mode plasmas. Even though η_i ≫ 1, the relevant stability parameter is is the ion temperature gradient scale length and R is the major radius). It is proposed that a significant element in the improved confinement observed in the L- to H-mode transient is the diminution of ITG mode turbulence, mediated by broader H-mode ion temperature profiles. Confinement results of L- and H-mode simulations of DIII-D, based on a simplified energy transport equation, are presented to support this hypothesis.

Soft X-ray measurements have been made with surface barrier diode (SBD) arrays in the REPUTE-1 RFP device (R/a = 82/22 cm). Arrays of seven and eleven SBDs are arranged so that they view a minor cross-section of the plasma from the vertical and horizontal diagnostic ports, respectively. Applying the tomography technique to the soft X-ray data, a two-dimensional emissivity profile can be reconstructed. In the sustainment phase of the RFP discharge, the soft X-ray emissivity profile exhibits a periodic peaking and a subsequent large crash across the minor cross-section. The fluctuation period is in the range of 60–150 μs, whereas the duration of the large crash is always about 30 μs (corresponding to ∼ 100 × τ_A, where τ_A is the Alfvén transit time). The soft X-ray emission is due to two modes: one is a dominantly m = 0 mode, which corresponds to the axisymmetric increase or decrease of the soft X-ray emissivity profile, and the other is an m = 1 mode. The toroidal correlation length of the m = 0 mode ranges from 20° to 80°, while the m = 1 mode has a high toroidal mode number. In the crash process, an m = 1 distortion of the emissivity profile, which indicates reconnection of the flux surfaces, can be observed. The magnetic fluctuations observed simultaneously are well correlated with the soft X-ray fluctuations. It is found that the soft X-ray fluctuations reflect the relaxation mechanism of the RFP plasma. This phenomenon is considered to be driven by global MHD activity.

A new method for modelling the sawtooth instability and other MHD activity in axisymmetric tokamak transport simulations is proposed. A hyper-resistivity (or current viscosity) term is included in the mean field Ohm's law to describe the effects of the three-dimensional fluctuating fields on the evolution of the inverse transform q characterizing the mean fields. This term has the effect of flattening the current profile while dissipating energy and conserving helicity. A fully implicit MHD transport and two-dimensional toroidal equilibrium code has been developed to calculate the evolution in time of the q-profile and the current profile using this new term. The results of this code are compared with the Kadomtsev reconnection model in the circular cylindrical limit.

The complete heavy impurity cycle of the divertor tokamak ASDEX is presented. The impurity sources, i.e. sputtering at the vessel wall and at the divertor plates, have been investigated spectroscopically or with special erosion probes. The measured erosion fluxes agree with data from sputtering calculations using measured incident charge exchange (CX) neutral and ion fluxes. The impurity transport in the plasma boundary has been simulated with a time dependent one-dimensional computer code. Central impurity confinement has been investigated with a laser blow-off technique. Investigations on impurity transport from the divertor plates back to the main plasma have shown a surprisingly low retention capability of the divertor for plate produced impurities at low densities, but no adequate code calculation is available. Finally, the deposition of heavy impurities has been investigated with time resolving probes and long term wall probes as well as through a post mortem analysis of the limiters and divertor plates. The results show clearly that CX sputtering is the main source of wall impurities, while the erosion of the divertor plates is mainly caused by ionic sputtering. Despite the fact that, in regimes of high confinement, impurity radiation often terminates the discharge, the present investigation shows a decrease in impurity production in ASDEX during these phases. The deposition pattern of heavy impurities indicates that impurities produced in the divertor are redeposited mainly on the divertor walls rather than on the divertor plates.

An extended impurity transport theory, which includes NBI momentum input and radial momentum transfer effects, inertial strong-rotation (v_ø ∼ v_th) effects and radial electric field effects, has been applied to analyse the impurity accumulation experiments in ISX-B and PLT. Theory and experiment are found to be in good agreement. Counter-injection causes a central impurity accumulation that is several times larger than for co-injection, and this difference can be understood in terms of differences in transport properties between co-injection and counter-injection.

Self-consistent plasma modelling, on the assumption that all particle species have the same confinement time, is applied to a field reversed configuration (FRC) to show that plausible designs of an ignited advanced fuel reactor are possible within the framework of present day technology. FRCs in the ⟨β⟩ ∼ 1 state are predicted to be microscopically and macroscopically stable; here, any ignited state must be of this kind since the plasma is heated up by a vast amount of nuclear power which even exceeds the electric output. In the ⟨β⟩ ∼ 1 state, the steep pressure gradient appears to concentrate dissipation into a thin sheath formed on the plasma surface; hence, the problem can be solved independently of the detailed local dissipative properties, in analogy to the case of a supersonic flow decelerated through an Hugoniot adiabatic shock. It was found out that, in the ignited state, self-pinching of the electrons takes place in the scrape-off layer towards the X-point so that the size of the loss hole from the scrape-off layer is reduced, which raises the edge plasma pressure to a high level; thereby, serious impurity concentration is eliminated. In the paper, possible D-based steady state fuel cycles are studied systematically by using a formulation which predicts each important parameter of a simple toroidal FRC reactor if only the electric power output and the plasma radius are specified. In optimizing the burning temperature, care was taken to satisfy the criteria for thermal stability. A study has also been pursued on a complex closed reactor system which is fed by D alone and exhausts the final fusion products only. It is found that flexible design is possible for such a complex system to minimize construction cost or environmental effects. Advanced fuel reactors are demonstrated to be very compact in size so that a field strength and a plasma volume approximately equal to those of JT-60 may be sufficient for a fusion reactor constituting the basis of a 1000 MW(e) power plant

An analysis has been performed to investigate the limits of high field commercial tokamak reactors. Variations of the maximum mechanical stress, the major radius and the performance parameter (where B₀ is the magnetic field on axis and a is the minor radius) have been investigated to evaluate the potential impact of higher strength structural materials and higher field superconductors. For fixed values of major radius and stress, the tokamak parameters have been chosen such as to optimize machine performance. Increasing values of allowable stress correspond to decreasing capital cost and may be an effective approach to economic commercial fusion. The paper discusses prospects of using very high strength structural materials to provide a sufficiently high field for Ohmic dominated heating to ignition in power reactors with a major radius of about 6 m.

The paper presents two-dimensional simulations of the locking of tearing modes to resistive walls in tokamaks during the early phase of density limit disruptions. Two different trigger mechanisms for wall locking have been considered. In the scenario believed to be most relevant for density limit disruptions, the m=2/n=l tearing mode becomes unstable while still rotating and develops a rotating magnetic island. When the mode lock occurs, wall stabilization is lost and the magnetic island grows, leading to disruption. Factors determining the increase in island size resulting from the mode locking are discussed. Another trigger mechanism is slowing down of the rotation through changes in the equilibrium, which decreases the wall stabilization and makes the tearing mode linearly unstable.

Using a modified version of the 1½-D BALDUR transport code, important aspects of the time evolution of D-T plasmas in a compact, high field tokamak of the Ignitor type with strong Ohmic heating are investigated. For the canonical electron heat transport, a generalization of the Coppi-Mazzucato-Gruber diffusion for axisymmetric toroidal geometry is assumed as well as alpha particle heating contributions; this reproduces the power dependent degradation of confinement observed in present-day auxiliary heated experiments. Anomalous ion heat conductivity is based on the existence of the η_i mode halfway out in the discharge, with an assumed spatial dependence of the transport magnitude superimposed on a parameter scaling related to wave breaking in the direction perpendicular to the magnetic field. The particle density profiles are held fixed, and convective energy fluxes are ignored. Results from three model examples associated with the set of transport coefficients are considered, and the impact of variations in density, conductivity magnitudes, resistivity, plasma impurity level and flux surface equilibrium configuration on the approach to ignition is discussed.

A linearized non-rigid MHD consistent displacement model is shown to be adequate to the study of the vertical stability of a plasma in an air core or iron core tokamak. This simple but accurate approach, which can be extended to the study of shape and radial position control, can be used in determining the limits of validity of the linear approximation and simulating the behaviour of a detection system based on a set of flux measurements. The method provides a basis for the optimization of position measurements and feedback parameters. The features of the model are illustrated, with reference to the study of the vertical stability of a NET configuration, showing the significance of the choice of the input signal.

The angular distribution of a ripple induced flux of high energy alpha particles on the first wall of a circular tokamak reactor is considered. The authors show that the distribution is highly non-uniform in the poloidal angle and that its peaking factor reaches values of 10 to 15 for typical tokamak reactor parameters. The distribution does not depend on the energy of the escaping particles and is therefore practically equivalent to the distribution of the energy load produced by these particles on the surface of the chamber. Furthermore, it is only weakly dependent on the system parameters over the relevant range of practical interest.

The paper presents a theoretical model of the H-mode in tokamaks which is based on the bifurcation of the radial electric field at the plasma edge. The electric field is determined by the balance of the non-ambipolar fluxes of ions and electrons at the edge. It is found that bifurcations of the radial electric field, the particle flux and the convective energy loss occur when the edge gradient reaches a critical value. This is attributed to L-H or H-L transition. The critical conditions are examined and the role of neutral particles as well as the effect of impurities are incorporated in the model. Combining the confinement scaling laws for both the core plasma and the scrape-off layer plasma, the threshold power for the transition is derived. The temporal evolution associated with the transition is studied.

It is shown that non-equilibrium radiative cooling can give rise to a condensational instability of the drift waves in an inhomogeneous magnetized plasma. The corresponding particle and heat transport coefficients are estimated by means of the conventional quasi-linear theory. The appearance of electrostatic edge turbulence and marfes on the inboard edge of tokamak plasmas may be attributed to such a radiation driven drift wave instability.

Lower hybrid current drive (LHCD) experiments on JT-60 have been carried out in parameter ;_e = (0.6−1.6)×10¹⁹ m⁻³, I_p = 0.5−1.5 MA, B_T = 2.7−4.5 T, P_LH < 2.5 MW, f = 1.75 and 2.0 GHz, with and without NB heating (P_NB < 6 MW), limiter and lower side single null divertor configurations, and narrow and wide spectra. The current drive efficiency increases with the volume averaged electron temperature ⟨T_e⟩ and with decreasing Z_eff and a high efficiency of η_CD ≡ _eR_PI_RF/P_LH ≤ 2.65 × 10¹⁹ m⁻²·A·W⁻¹ was achieved by LHCD alone. The experimental data set has shown that the current drive efficiency is consistent with the scaling of η_CD ∼ 12 ⟨T_e⟩/(5+Z_eff)×l0¹⁹ m⁻²·A·W⁻¹. Low N₁ waves with a narrow spectrum have a higher current drive efficiency than higher N₁ waves with a wide spectrum insofar as launched waves satisfy the accessibility

An analysis of sawtooth postcursor oscillations in JET is presented. From electron cyclotron emission measurements, the radial T_e profile on the equatorial plane is obtained with high time resolution. The plasma rotation is used to reconstruct the constant temperature contours in the poloidal plane from the T_e profile. The postcursor oscillation observed after a partial collapse shows a cold region with a structure very much like that of an m = 1 magnetic island, while the hot plasma core is displaced but still has concentric T_e contours. After a full collapse, a cold region of similar structure is seen, while the hot core is displaced over a much larger distance and is squeezed into a crescent shape. Also a large amount of energy is lost from the hot core. It is concluded that partial and full collapses are likely to be due to the same instability and that reconnection is involved in the non-linear evolution of the instability. It is suggested that the amplitude at which the instability saturates determines the type of sawtooth collapse and the amplitude of the postcursor oscillation.