Table of contents

A new method of measuring the fast confined alpha particle distribution in a reacting plasma is proposed. The presence of alpha particles in a DT plasma will create high energy tails in the deuterium and tritium ion energy distributions. A 3.5 MeV alpha can transfer 3.4 MeV to a tritium ion in a single elastic scattering interaction. Calculations of the size of these knock-on tails in tokamaks such as TFTR, JET and ITER show that it may be possible to measure these tails and provide information on the fast confined alphas. The knock-on tail ions will produce DT neutrons with energies up to 20.6 MeV, so that DT neutron spectroscopy can be used to monitor the alpha population. Neutron spectroscopy looks especially attractive for ITER. A collimated array of threshold neutron activation detectors could be used to measure the confined alpha density profile. Tests of this diagnostic can also be done on TFTR and JET. Existing high energy neutral particle analysers may allow observation of the ion tails directly via passive and/or active charge exchange

The effects of wall stabilization and differential plasma rotation on resistive tearing modes are discussed for configurations with multiple rational surfaces. Results are presented for purely rigid plasma rotation as well as for the case when two rational surfaces are in differential rotation

Supershots in TFTR often suffer a performance deterioration characterized by a gradual decrease of the DD fusion neutron yield and plasma stored energy after several hundred milliseconds of auxiliary heating. The correlation between this performance deterioration and the development of low m (the poloidal mode number), n (the toroidal mode number) MHD modes is studied through shot-to-shot comparisons and statistical data analyses. A good correlation is observed between performance deterioration and the appearance of strong 3/2 and 4/3 macroscopic modes. The magnetic island structures are observed using Mirnov and ECE diagnostics. The measured T_e, T_i and n_e profiles show that development of the islands corresponds to a nearly constant decrement of these quantities over the core region r < r_s, where r_s is the mode rational surface, on a transport time-scale (t > τ_E). The observed energy deterioration scaling, δW/W varies as w/a, where w is the magnetic island width and a is the plasma minor radius, agrees with a local transport model. Numerical simulations based on the local transport model reveal many features consistent with the experiments. Besides the MHD effect, it is found that a continuous increase of edge recycling rate during the neutral beam injection phase also has a large effect on the performance deterioration

A non-linear analysis is presented of localized regions of strong divergence of parallel heat flux-thermal fronts-in the edge and scrape-off layers of tokamaks. The phenomena of divertor detachment and MARFEs are discussed in terms of one dimensional thermal conduction in the parallel co-ordinate, but retaining cross-field transport power divergence and radiation as source terms. A full finite aspect-ratio geometry is retained and has important effects. Thermal fronts are shown to require either localization of the source terms or density control, to be stable against parallel motion. The edge density range over which the front is naturally localized to the divertor leg is shown to be rather small. The size, stability and preferred position of MARFEs are predicted and are in agreement with experiment

A technique using a single hidden layer backpropagation neural network is described to establish a nonlinear mapping between a set of magnetic flux measurements and some shaping parameters of a noncircular plasma. The technique has been applied for the identification of limiter and X point equilibria in the ASDEX Upgrade geometry; the dataset of equilibria required for training and testing the neural network has been generated by means of an integrated use of a fixed and a free boundary MHD code. The average accuracy of the identification procedure is quite good, with a further improvement if a linear connection between the input and output layers is introduced. A procedure is also proposed for the selection of the optimum location of a limited number of sensors. The relationship existing between the behaviour of the neural network and some statistical parameters of the dataset is analysed and discussed

In many of the medium and high beta discharges in PBX-M, low n modes (n=1-4) with different n numbers are observed. The prevalence of low n modes decreases with increasing n number. Furthermore, if two modes of different frequency and n number (ω₁, and ω₂; k₁ and k₂) are simultaneously present in the plasma, these modes interact non-linearly and create sidebands in frequency (ω₂ ± ω₁) and wavenumber (k₂ ± k₁ or n₂ ± n₁ and m₂ ± m₁). It has also been observed that if these fundamental modes, ω₁/k₁ and ω₂/k₂, contain strong harmonics, the harmonics also interact non-linearly, creating more non-linear products: kω₂ ± lω₁ and kk₂ ± lk₁, where k and l are integers describing the harmonics. These multiple three wave couplings on a given rational surface can create several modes of the same m and n numbers but of different phase velocities. The multiple perturbations of the rational surfaces can increase the plasma losses. In fact, a decrease in neutron rate and an enhanced loss of medium energy ions are observed during this three wave coupling interaction. Although the three wave coupling has been observed primarily in medium or high beta discharges, it is a more general plasma phenomenon because it also occurs occasionally at low β

A new computational model for neutral particle transport in the outer regions of a diverted tokamak plasma chamber is presented. The model is based on the calculation of transmission and escape probabilities using first-flight integral transport theory and the balancing of fluxes across the surfaces bounding the various regions. Both long and short mean free path regions can be handled. The complex geometry of the outer regions of a diverted tokamak can be represented accurately and economically using precomputed probabilities which depend only on the mean free path of the region. Benchmark comparisons with Monte Carlo illustrate that quite accurate results may be expected with the method

Experiments that decoupled H mode plasma current and plasma density have been performed for the first time on the DIII-D tokamak by the application of divertor cryopumping. Deuterium ELMing H mode discharges were operated in the single null configuration with 6 MW of neutral beam injection. These steady state discharges were operated at 1.5 and 0.75 MA with density ranges of (4.0-8.3) × 10¹⁹ and (2.7-4.0) × 10¹⁹ m^-3, respectively. Therefore, for the first time in a steady state DIII-D H mode, a factor of 2 range in I_p (n_e) has been obtained at fixed n_e (I_p). A power law dependence of ELMing thermal confinement was assumed, with the result that for a given magnetic field and a fixed power dependence, τ_th = 0.18I_p^0.91±0.08n_e^0.18±0.09P_L^-0.5. If a linear dependence of τ _th on internal inductance is assumed the density dependence disappears. A local power balance analysis employing the 1½-D ONETWO transport code found that the one fluid effective diffusivity did not change with a factor of 2 change in density and was strongly reduced with increasing current. The reduction with increasing current occurs most dramatically for ρ > 0.5. Utilizing ONETWO for particle transport analysis demonstrated that the core particle transport was more rapid at lower density

Parametrization of the decay index n_decay of the equilibrium magnetic field in a tokamak with respect to plasma geometry, β_p value and profile peakedness is investigated numerically for a fundamental understanding of vertical displacement instability (VDI). A D shaped plasma with higher triangular deformation and broader profiles is more stable against VDI. For the parameters referred to, as well as those of both DIII-D and ITER, pressure can play either a stabilizing or a destabilizing role, depending on both the geometry and the profile peakedness of the plasma. A saturation of n_decay with respect to the elongation and the profile peakedness exists