Table of contents

Scalings of the central rotation in non-gettered, co-injected ISX-B discharges have been measured as a function of beam power, electron density and plasma current. Extensive studies are made possible by exploiting charge-exchange excitation (CXE) of 0 VIII lines to measure Doppler shifts. The rotation velocity, v_ϕ(0), tends to saturate at (1.0 − 1.2) × l0⁷ cm·s⁻¹ when P_b≅0.5 MW, showing little further increase up to the maximum input of 2 MW; v_ϕ(0) is independent of n_e and I_p. Momentum confinement times in quasi-steady plasmas are 10–16 ms for _e = 4.5 × 10¹³ cm⁻³. Counter-injection discharges always disrupt, but before this event v_ϕ(0) is the same as for co-injection plasmas. The addition of a third beam line, permitting injection of up to 2 MW of balanced neutral-beam power, has allowed comparisons of the energy and particle confinement in rotating and non-rotating plasmas with the same total neutral-beam input. In those cases where impurity buildup can be avoided, it is found that the ISX-B empirical scaling of energy confinement time is reproduced with balanced injection. Thus, the unfavourable dependence of is not the result of rotation. Studies of impurity behaviour under differing injection conditions have been extended to include fully stripped low-Z ions. The results are consistent with previous investigations of metallic elements which revealed strong dependences on the sense (co versus counter) of injection. The potentials calculated from momentum balance, using measured rotation profiles and typical plasma density and temperature profiles, are in qualitative agreement with the potentials measured directly for various combinations of co- and counter-injection.

The parameters and processes governing the low and high (L and H) confinement in ASDEX are investigated by computer modelling. The L-transport is shown to be connected with the non-Ohmic electron heating. After neutral injection, the L-phase is maintained by the ion-electron energy transfer rate for the significant time span of about 100 ms. It is found that the profile parameters η_e,i = d ln T_e,i/d ln n discriminate between the L-transport and the H-transport. Small η-values around unity characterize the H-regime, while larger η-values correspond to the L-regime. The operational limits of the H-mode are connected with variations of η_e and η_i due to changes in the injection heating profile. These results suggest that drift instabilities such as dissipative trapped-electron modes are involved in the anomalous L- and H-transport.

A computer code for solving the bounce-averaged Fokker-Planck equation appropriate to stellarator transport has been developed and its first applications have been made. The code is much faster than the bounce-averaged Monte-Carlo codes, which up to now have provided the most efficient numerical means for studying stellarator transport. Moreover, since the connection to analytic kinetic theory is more direct for the Fokker-Planck approach than for the Monte-Carlo approach, a comparison of theory and numerical experiment is now possible at a considerably more detailed level than previously.

The feasibility of steady-state spheromak operation using ECRH current drive is studied numerically. The model used comprises ray tracing and a linear relativistic treatment of current generation. It is assumed that the RF-generated current distribution will relax to that of the spheromak equilibrium. A reactor grade plasma with an average temperature of 18 keV, a density of 0.8 × 10²⁰ m⁻³, a maximum magnetic field of 4 T and a fusion power of 2100 MW can be sustained by launching less than 100 MW of ECRH power near the plasma surface.

Non-linear calculations of the free-boundary mode in a cylindrical tokamak plasma with finite conductivity are carried out by using the 'pseudo-vacuum' model. Applicability of the pseudo-vacuum model to the study of free-boundary kink modes is carefully checked by linear analysis both analytically and numerically. In the non-linear calculations, vacuum bubbles can be formed due to surface tearing mode as well as to free-boundary kink mode under the condition that the safety factor at the plasma surface, q_a, is around two. This bubble formation is a candidate for major disruptions in tokamak discharges with q_a approximately equal to two.

One key to achieving high beta in a tokamak may be the proper control of the plasma pressure and current density profiles. In principle, if steady-state RF current drive is used, the current density may be tailored by a careful selection of the wave frequency and the power spectrum. This study uses the selfconsistent theory of RF current drive in an axisymmetric torus to compute the MHD equilibrium generated by the fast (magnetosonic) wave current driver. This is accomplished by iterations in a calculation of the MHD equilibrium for a given diamagnetism function, F(ψ) = RB_t, and a ray-tracing calculation which determines F(ψ). The various wave and plasma characteristics which determine the equilibrium are examined, and it appears possible to create a large diversity of equilibria.

An ion and electron energy balance analysis has been made of the MHD stable high beta plasma produced by neutral beam injection in Heliotron E. The currentless target plasma was generated by 53.2 GHz second-harmonic ECH (∼ 160 kW) at a magnetic field of B_h = 0.94 T. The linear density rise due to intense gas puffing during neutral beam injection increased the volume averaged beta value up to ≈ 1.5 − 2.0%. The results of 1-D transport codes show that the global energy balance is dominated by the electron loss channels in which radiation is predominant. The particle balance in the beam-assisted density rise phase cannot be accounted for without an inward convection of bulk particles, indicating a similarity to the transport of impurities. So far, the observed beta values are not restricted by MHD activity.

A simple model for the interaction between a solid microsphere and a surrounding dense and hot plasma shows that plasma columns such as those produced in plasma focus devices may drive the implosion of the microsphere, thus creating situations of interest for inertial confinement research. The method can be used for microspheres structured in such a way that the implosion does not require a critical temporal profile of the pressure pulse applied on the surface, i.e. for microspheres having thin shells. In this case, it is shown that the plasma column of a plasma focus fed by a condenser bank of a few hundred kilojoules is, potentially, a driver comparable to a multibeam kilojoule Nd laser facility. It is concluded that attention should be paid to plasma focus research in the context of inertial confinement.

The theory of tokamak ripple transport in the banana-drift and ripple-plateau regimes is extended in a number of directions. The theory is valid for small values of the toroidal periodicity number n of the perturbation as well as for the moderate values (n ∼ 10–20) previously assumed. It is shown that low-n perturbations can produce much greater transport than the larger-n perturbations usually studied. In addition, the ripple perturbation is allowed arbitrary values of poloidal mode number m and frequency ω, making it applicable to the transport induced by MHD modes. Bounce averaging is avoided, so the theory includes the contributions to transport from all harmonics of the bounce frequency, providing a continuous description of the transition from the banana-drift regime to the ripple-plateau regime. The implications of the theory for toroidal rotation in tokamaks are considered.

An MHD code is used to find a sequence of equilibria where negative skin current penetrates progressively toward the magnetic axis, on which the toroidal current density (positive) and kinetic pressure are maintained constant. The initial equilibrium with no skin current has a central safety factor close to unity and an edge safety factor about three. Equilibria have been found for currents down to 5 per cent of the initial value (total negative current about half total positive current). For total currents less than a third of the initial current, the plasma becomes smaller because of the creation of a poloidal separatrix on the small major radius side. The vertical field necessary for equilibrium decreases more slowly than the total plasma current. This is explained by a large increase in the poloidal beta accompanied by a smaller increase in the internal inductance. The use of such a fast rampdown sequence to provide a controlled termination for an energetic tokamak plasma is discussed.

Reversed-field pinch configurations produced in the OHTE device using a resistive shell were maintained for up to 7.5 times the shell resistive time. The plasma density, resistance, and magnetic fluctuations do not evolve significantly while the plasma current is sustained, and the plasma behaviour is similar to that previously observed in conducting shell OHTE/RFP experiments. The experimental observations suggest that a highly conducting shell may not be required for the stability of a reversed-field pinch.

Invariance arguments are used for determining the form of the anomalous transport caused by rippling mode turbulence. It is found that the thermal diffusivity increases with the parallel thermal conduction. Rippling modes driven by impurity gradients are also considered.

The IAEA Technical Committee Meeting on Mirror Fusion Research was held in Tsukuba, 8-12 July 1985. The Meeting was the fifth in a series on this subject, the preceding one having been held in Moscow, September 1984. The purpose of the Meeting was to exchange the latest information on mirror-based fusion research. Eighty-six physicists from five countries attended the Meeting, among whom 61 from Japan, 21 from the USA, two from the USSR, one from Austria and one from China. The presentation of 52 papers and a visit to the GAMMA 10 experimental site took three days. Because of the. large number of papers, 34 papers had to be presented in parallel sessions. After presentation of the papers, two half days were spent on panel discussions.