Table of contents

The propagation and absorption of second harmonic ion cyclotron range of frequency (ICRF) waves in NBI heated plasmas are investigated for the regime of plasma parameters in JT-60. The quasi-linear diffusion in the velocity space of ions is taken into account. The Fokker-Planck equation and the kinetic wave equation are numerically solved to obtain the spatial profile of the velocity distribution function and the power deposition profile in a self-consistent manner. The calculation of the power partition rate includes two processes: the direct absorption of the wave through wave-particle interactions and the power transfer due to Coulomb collisions. The power partition between electrons and ions is strongly affected by the plasma density and the RF power. The plasma temperature and the NBI power have some influence on power partition and tail formation in the experimental conditions of interest. The acceleration of the injected particles by ICRF heating is studied. The tail temperature, which is defined by the gradient of the velocity distribution function of tail ions, and the total stored energy of the plasma are calculated and the dependence on the plasma parameters is analysed.

The 0.8 MeV ³He ion created in the d(d,n) ³He fusion reaction is confined in TFTR and can undergo d(³He,p) α fusion reactions. The time evolution and profile of the 14.7 MeV proton emission indicate that collisional energy transfer with plasma electrons is the dominant process influencing the charged ³He fusion product for TFTR plasmas and that there is no observable anomalous transport. The magnitude of the burnup exceeds classical expectations by a factor of two to three, which is a little beyond experimental uncertainties.

The structures of scrape-off layer and divertor plasmas have been studied numerically with a neutral code and a two-dimensional fluid code. Doublet-III is taken as an example for an open divertor configuration. A decisive parameter is the distance between the plasma surface (determined by the magnetic separatrix) and the limiter, which is varied in order to assess the interaction of the plasma with the limiter as well as the effect of neutrals on the main plasma. The minimum value of the limiter clearance needed to prevent plasma-limiter interaction is determined. The scaling of the edge temperature and the dependence of the e-folding length of the scrape-off layer plasma on the heating power are obtained.

Neutral beam heated DIII-D expanded boundary divertor discharges have exhibited ASDEX-like H-mode behaviour over a wide parameter range. The deuterium H-mode energy confinement of 120 ms remained near the Ohmic value for up to 6 MW of neutral beam heating, where it was 2-2.5 times higher than the L-mode value at a plasma current of 1 MA. The hydrogen and helium H-mode energy confinement times were similar and substantially below the deuterium H-mode confinement time. The H-mode confinement times decreased with increasing neutral beam power and were only 30% better than the L-mode confinement times at 5 MW. In an H-mode with a mixture of hydrogen and deuterium ([H]/[H+D] ≅ 40%), the confinement time was in between the values obtained in the pure hydrogen and deuterium cases, increased linearly with plasma current for q₉₅ > 3.2, and decreased with increasing neutral beam power. The confinement quality in these plasmas was 85 ms per MA at a heating power of 5.6 MW. The lower energy confinement in the non-deuterium H-modes and the degradation of energy confinement with neutral beam power were both accompanied by an increase in the edge localized mode (ELM) amplitude and frequency. The changing ELM characteristics make a determination of the intrinsic isotopic and neutral beam effect on confinement difficult. For values of B_T < 0.9 T and q₉₅ < 3, the confinement quality in the deuterium and hydrogen/deuterium H-modes deteriorated to values near the L-mode level. This deterioration in energy confinement was not related to operation at high beta but instead appears to be due to a combined action between sawteeth and ELMs that becomes more pronounced at low q and low B_T. L-mode energy confinement was independent of ion species and in good agreement with Kaye-Goldston scaling. Odajima-Shimomura scaling disagreed with the present L-mode τ_E data in terms of isotopic mass dependence; their prediction for the hydrogen L-mode exceeds the present measurements by a factor of two.

A new code for MHD equilibrium/current drive analysis has been developed which can be used in an analysis of high beta tokamak equilibria consistent with beam driven current profiles. With this new code, the critical beta equilibrium, which is stable against the ballooning mode, the kink mode and the Mercier mode, is determined by using first the MHD equilibrium and stability analysis codes EQLAUS/ERATO. Then, the current drive parameters and the plasma parameters required to sustain the critical beta equilibrium are determined by iterative calculations. The beam driven current profiles are evaluated by Fokker-Planck calculations on individual flux surfaces where the toroidal effects on the beam ion and plasma electron trajectories are considered. The pressure calculation takes into account the beam ion and fast alpha components. A peculiarity of this new method is that the solution obtained is consistent not only with the MHD equilibrium but also with the critical beta limit conditions for the current and pressure profiles. Using this new method, critical beta equilibria of β ∼ 21% (bean type) and β ∼ 6% (D-type) have been scanned for various parameters, such as major radius, magnetic field, temperature and injection energy. The achievable Q-value (Q = fusion power/beam power) for the bean type is found to be about 30% larger than that for the D-type. With strong beanness, Q ∼ 6 for DEMO type tokamaks (∼500 MW(th)) and Q ∼ 20 for power reactor size tokamaks (4.5 GW(th)) can be achieved. For D-type machines, however, the Q-value would not exceed 16.

The sustainment of plasma discharges (current drive) using helicity injection and relaxation is discussed. It is shown that toroidal plasmas may be sustained by applying only poloidal voltage and current. Some of the configurations which can be maintained in this way are calculated.

Intrinsic impurity behaviour and transport properties in neutral beam heated L- and H-mode PBX tokamak plasmas were studied with a variety of impurity diagnostics. Central impurity accumulation was most often observed in H-mode discharges; sometimes it resulted in a thermal collapse due to high central metallic radiation (∼1.5 W • cm⁻³). The impurity accumulation was evident from peaked Z_eff and radiated power profiles and was further substantiated by specific vacuum ultraviolet and X-ray spectroscopy measurements. It is shown that impurity accumulation was neither unique nor inevitable in H-mode discharges and it could be suppressed by sufficient gas puffing. Central impurity accumulation was also seen in L-mode plasmas even with co-injected neutral beams. This usually occurred at high beam power and relatively low density. While there was no significant difference in the degree of accumulation between L-mode and H-mode discharges, the Z_eff profile itself was more peaked in the H-mode because the electron density profiles are flatter in H-mode plasmas than in L-mode plasmas. The degree of accumulation increased as Z_eff(0) itself increased, which suggests that neoclassical convection and diffusion driven by both impurity-impurity and impurity-plasma ion collisions contribute to the central plasma particle transport.

The ion temperature gradient driven drift mode turbulence (η_i mode) has been studied with a fully toroidal fluid model. The saturation amplitudes and ion energy transport have been obtained by numerical simulations. An analytical model for mode coupling saturation in combination with a quasi-linear approximation has been used to derive a simple diffusion coefficient which is in good agreement with the simulations. The scaling of the diffusion coefficient χ_I with L_n/L_B indicates a considerably improved agreement with the experimental radial variation, and the scaling with T_e/T_j has been found to favour tokamak operation in the hot ion regime. The diffusion coefficient is consistent both with the improved confinement for pellet injection experiments and with the H-mode results for very large η_i.

By injection of high power lower hybrid (LH) waves of up to 6.0 MW into neutral beam heated (20 MW) plasmas in the medium electron density regime (_e ≤ 3.5 × 10¹⁹ m^-3) the plasma energy content is increased at the same incremental energy confinement time as it is increased with neutral beam heating alone. When LH heating is used in addition to thermal electron and ion heating, the beam ions are accelerated to energies considerably higher than the beam injection energy. In contrast to LH injection into ohmically heated plasmas in the same density regime, where substantial high energy electron production is observed, there are much less high energy electrons in the case of combined heating. The heating efficiency of LH waves during combined heating tends to decrease as the electron density is increased. Ray tracing analysis suggests that the accessibility condition prevents effective heating in a high density plasma. Calculation of wave damping, taking into account the beam component of the ion velocity distribution function, indicates that the waves are absorbed by beam ions before they are absorbed by electrons.

The paper summarizes the experiments performed with ion cyclotron resonance heating (ICRH) on ASDEX, from November 1984 until March 1986; the most interesting results are reported and discussed in detail. Heating and confinement studies using the hydrogen second harmonic scheme and the hydrogen minority scheme (P_IC < 2.6 MW, t_IC < 1.5 s) show a typical L-mode behaviour, i.e. a power dependent confinement degradation, which is rather similar to that found with neutral beam injection (NBI) heating. ICRH is accompanied by a slightly improved particle and energy confinement compared with that of NBI; this is also true for a combined ICRH + NBI scheme, up to P_tot ≈ 4.5 MW, absorbed in the plasma. Particular efforts have been devoted to investigations of the second harmonic regime in H/D plasmas with n_H/n_e ≈ 0.1 - 1, with a view to heating mixtures in reactor relevant plasmas. The achievement of H-mode transitions with ICRH alone in the hydrogen minority scheme at an absorbed RF power of about 1.1 MW supports the assumption of common confinement properties in auxiliary heated tokamaks, since they appear to be widely independent of the additional heating method. ICRH specific impurity problems, such as the strong release of iron from the vessel walls, have been overcome by applying extensive in situ wall carbonization. The mechanisms responsible for impurity generation have partly been identified and analysed; however, the problem still remains to be solved. Impurities preferentially released from the ICRH antenna do not pose problems.

Different target designs are discussed which may be used in an inertial confinement fusion (ICF) reactor system driven by lithium ions. These different designs are for various energies of the incident ions. For ion energies of the order of 30 MeV, targets without a tamper can be used; however, for higher energies, tamped targets are needed. The paper presents simulation results for the compression, ignition and burn of the latter type of target. These are single-shell multi-layered targets containing 3.2 mg of D-T fuel. The gain of these targets is sufficiently high that they can be used in an ICF reactor system.

Micrometre sized tantalum droplets were injected into a tokamak plasma by a controllable arcing gun located behind the wall. The trajectories of the ablating particles were photographed by a high speed camera. Various possible mechanisms which may explain the observed curvature of the particle paths are discussed. The migration of the ablated material in the tokamak was studied by post-mortem analysis of collector probes and limiters.

Light impurity density profiles following hydrogen gas puff and hydrogen pellet fuelling have been obtained via charge exchange recombination spectroscopy on TEXT. After hydrogen pellet fuelling, C⁺⁶ and O⁺⁸ profiles are more strongly peaked on axis than the electron density profile. In contrast, light impurity profiles measured during sawtoothing high density gas puff fuelled discharges peak to a lesser degree relative to the electrons, with lower central densities than in the pellet case. The shapes of the peaked pellet impurity profiles are well described by neoclassical theory. The measured transport coefficients are larger than neoclassical values. The inferred inward convective velocity is consistent with values measured by impurity injection during sawtoothing discharges, but the diffusion coefficient is an order of magnitude lower than that obtained with impurity injection.

In the STP-3(M) reversed field pinch, a thick conducting shell is surrounded coaxially by a one-turn poloidal coil, which also acts as a shell. Field errors at the shell gap in this double-shell structure are shown to be an order of magnitude smaller than those in a conventional shell structure. Reduction of the field errors resulted in a remarkable improvement of the plasma parameters through control of the average plasma position with a quasi-stationary vertical field, without any compensation of local field errors.

Resistive internal kink modes are the most difficult resistive instabilities to stabilize in tokamaks. Generally, the mode is unstable if q < 1. However, with special q-profiles, it is possible to stabilize resistive internal kink modes in tokamaks with non-circular cross-section. This result holds for large aspect ratio. Previously, similar stabilization was found for small aspect ratio circular cross-section tokamaks. The stabilizing effects of non-circular cross-section and small aspect ratio are expected to be additive. The conformal co-ordinates used in the numerical study are discussed.

The first IAEA Technical Committee Meeting on Pellet Injection and Toroidal Confinement, sponsored jointly by the IAEA and the Max-Planck-Institut für Plasmaphysik (IPP), Garching, was held at Gut Ising, Upper Bavaria. The local organizing committee was headed by M. Kaufmann of IPP. The Meeting followed the traditions of previous international workshops on pellet fuelling, the last of which was organized by ORNL at La Jolla in 1985.

Because of the substantial improvement in the plasma parameters of pellet fuelled tokamak discharges recently achieved in different laboratories, this Meeting was primarily devoted to the discussion and analysis of the confinement properties of pellet fuelled tokamak and stellarator discharges.