Table of contents

Alpha particle loss was measured during the TFTR DT experiments with a scintillator detector located at the vessel bottom in the ion Del B drift direction. The DT alpha particle loss to this detector was consistent with the calculated first orbit loss over the whole range of plasma current I=0.6-2.7 MA. In particular, the alpha particle loss rate per DT neutron at a given plasma current did not increase significantly with fusion power up to 10.7 MW, indicating the absence of any new 'collective' alpha particle loss processes in these experiments

The behaviour of the material ablated from the wall in a hard disruption is analysed, in order mainly to ascertain its capability of protecting the wall from more severe damage. The investigation is limited to the fast first phase of the disruption, during which most of the bulk plasma thermal energy is given off. Within the ablated material an interaction region and a transmission region are defined. The first region (the 'virtual limiter') is where the bulk plasma thermal energy is dumped and the associated kinetic energy flux changed to radiation flux. The second region (the 'plasma shield') is where the radiative energy flux is attenuated through interaction with ablation matter, thus preventing more serious injuries to the wall. The whole problem is investigated in terms of a simplified mathematical model, which is shown to have a similarity solution. By making further approximations, an explicit solution is derived in full analytical terms. Applications are made to ITER type and IGNITOR type tokamaks

Experimental data from the reversed field experiment, ZT-40M, have been re-examined in an attempt to determine the scaling behaviour of the physical plasma quantities and their fluctuations. A subset of the data is defined, allowing a reduced number of independent variables to describe the behaviour. For flat-top ZT-40M discharges the independent variables are chosen as being the toroidal current, I_ϕ, and the dimensionless pinch parameter, Theta , which is proportional to the ratio of the toroidal current to the toroidal magnetic flux. The amplitudes of the dependent variables, including the electron temperature, plasma resistance, toroidal flux, the ratio of I_ϕ to the mean electron density and their fluctuation amplitudes, exhibit minima as functions of Θ for constant I_ϕ. These minima move towards lower Theta values with increasing I_ϕ. Over the range of conditions for acceptable operation, the scaling of variables with I_ϕ is not unique but depends on the variation of Θ as I_ϕ increases. The Θ variation is governed by the specific conditions (such as constant poloidal β, β_p) chosen to set the desired RFP operational constraints. Contour plots of the dependent variables versus the two independent variables, I_ϕ and Θ, allow the determination of the I_ϕ-Θ trajectory that corresponds to discharges that meet the chosen condition. The analysis shows that the amplitude of the low frequency fluctuations correlates with the mean β_p and energy confinement time of ZT-40M. By modifying the external circuits on ZT-40M, low frequency fluctuations were reduced. Comparing the designs of different RFP experiments and their operating behaviour, these modifications suggest design changes for present and future RFP experiments that will benefit their performance

The toroidicity induced shear Alfven eigenmode (TAE) instability driven by the fusion alpha particle pressure gradient is found to be stabilized in the presence of a resonant RF field with Ω = k_0⊥. v_α⊥, where Ω is the frequency, k₀ the wavenumber and v_α⊥ is the alpha particle velocity component perpendicular to the magnetic field. This kind of resonance could occur during lower hybrid (LH) current drive in an ignited tokamak plasma. The RF electric field strength that is needed for stabilization is available from modern RF power sources

Special versions of the 1.5-D BALDUR predictive transport code are used to explore the confinement in the ignited ITER EDA by self-consistent calculations. The code computes 2-D equilibria and solves 1-D transport equations in the bulk and scrape-off layer with empirical transport coefficients for the ohmic, L and ELMy H mode regimes. The emphasis is on scenarios with flat density profiles and high, fixed radiative power in the main chamber due to the seeded impurities argon and neon. It is shown that self-sustained steady state thermonuclear burn is achieved for 370 s and is compatible with the flat density profiles and strong radiative cooling. The necessary local energy and particle transport are presented. In the argon and neon scenarios, the required radiation corrected energy confinement times are 4.1 and 3.5 s, respectively, which are achievable according to the ITER ELMy H mode scaling. The advantage of neon originates from its smaller radiative loss within the separatrix of 37% of the total radiation in the main chamber, compared with 60% in the case of argon. A significant radiative loss from the confinement zone, mainly by bremsstrahlung, cannot be avoided. It raises the required energy confinement time and is the price to be paid for reduction of the divertor heat load by radiative cooling in the main chamber. In steady state, a helium fraction of 5% is computed. The fractions of helium, argon and neon and the resulting fuel dilution are considerably lower than with peaked density profiles

A reversed field pinch configuration with a quadra-null toroidal divertor has been sustained for a much longer time than the magnetic diffusion time on the TPE-2M device. By applying toroidal divertor fields, the plasma resistance has been reduced by 40% in low plasma current operations (40-55 kA), along with a decrease in the line emission of light impurities and limiter materials (molybdenum). Direct measurement of internal magnetic fields by inserted magnetic probes has shown that a separatrix exists within the limiter radius. Though the behaviour of the slow magnetic fluctuations has not deteriorated, fast and spiky m=0 mode fluctuations have been excited, especially near the reversal surface

The leading hypothesis for the cause of the very high confinement observed during VH mode in DIII-D is that a spontaneous increase in the core flow shear suppresses microturbulence which reduces transport loss and, for example, further increases flow, in a positive feedback cycle. This is confirmed by using 'magnetic braking' as an independent control of the toroidal rotation contribution to the flow; when the toroidal rotation shear of a VH mode discharge is reduced, the turbulence and transport increase back to the level observed during 'standard' H mode

The density perturbation induced by electron cyclotron resonance heating was examined on the JFT-2M tokamak using two different heating methods, a pulse injection and a modulated injection. The measured temporal behaviours of the density perturbations were compared with numerical simulations, in which the ECRH induced flux was modelled as local outward flux. It was found that this model can explain the measured density perturbation satisfactorily

Results are presented from an experiment in which a large amount of deuterium gas was continuously puffed at the scrape-off layer and exhausted from the DIII-D pump divertor ('puff and pump' technique). The experiments were conducted with a single null divertor configuration during edge localized mode (ELM)ing H mode. Steady state density conditions were attained during the strong deuterium puff. The deuterium puff reduced the central plasma density of argon, puffed continuously as a trace impurity, by as much as a factor of 20, and the central argon radiated power density by as much as a factor of 10

The usual demonstration that neoclassical transport is automatically ambipolar, and independent of the radial electric field (E_r), is shown to be valid only in an equilibrium axisymmetric plasma in which there is no other source or damping of toroidal momentum. If it were generally true, it would be difficult to explain the measured current drawn from a biased internal probe, as used to induce H modes. However, this current and its variation with applied voltage are consistent with the alternative analysis by Galeev and Sagdeev (1968), in which the ion and electron fluxes depend on E_r - B_θŪ_||. The behaviour of E_r and impurity pump-out during enhanced MHD activity can also be explained using this analysis

The parallel electron viscosity is shown to play an important role in sawtooth collapse. A new sawtooth collapse model is established with a Sweet-Parker type analysis. The reconnection time based on this parallel electron viscosity model is less than that of the electron inertia model for a high temperature tokamak plasma

Report on the Workshop held before the 11th PSI Conference, Nagoya, Japan, 19-20 May 1994

Report on the 11th International Conference, held at Mito, Japan, 23-21 May 1994.