Table of contents

The potentials of fuels that are predominantly deuterium, with a small tritium content (tritium-poor deuterium), are examined in the context of the inertial confinement fusion (ICF) scheme known as `fast ignitor', which is based on external heating of a portion of a precompressed target. For this purpose, the burn performance of compressed spheres of pure or nearly pure deuterium with an off-centre deuterium-tritium (DT) seed, ignited by an ultrapowerful external beam, has been studied using two dimensional (2-D) numerical simulations. The dependences of the fuel energy gain on the fuel mass and energy, and on the tritium fractional content F_T, have been studied; limiting gain curves have been obtained for fixed values of F_T and of the isentrope parameter alpha . It is found that assemblies with F_T=0.5-1%, fuel mass m approximately=20 mg, compressed at density rho approximately=1000 g/cm³, can achieve tritium self-sufficiency (i.e. net tritium production). At alpha approximately=1.5, the corresponding fuel energy at ignition is of the order of 1 MJ and the fuel energy gain can be about 1000. The potentials of tritium-poor configurations are also compared with those of equimolar DT fuels with different ignition configurations

Adjustments are described that should be made to a single filament description of a tokamak plasma if accurate modelling of a non-linear MHD simulation is to be achieved. It is shown that although the single filament model is excellent for purely radial, or purely vertical, fields, it does not describe the gradients of these fields. A simple scalar correction to the gradient terms in the model is found to facilitate the accurate prediction of the growth rate of the vertical instability. An extension to the modelling of X points is also presented. The corrected multi-pole model is shown to be appropriate for the design of robust controllers

The recent experiments on accumulative behaviour of heavy impurities in TEXTOR with test limiters of molybdenum and tungsten and puffing of xenon are briefly reviewed. The results of the reconstruction of the transport coefficients of high-Z ions in the accumulation stage are presented. They confirm the neoclassical nature of the convective particle transport, leading to peaking of the impurity density. A mechanism triggering accumulation, invoking the temperature dependence of neoclassical flows of impurities, is discussed and the threshold of accumulation obtained is compared with experimental data. Processes which can lead to a saturation of the accumulation, caused by a modification of the flux component proportional to the temperature gradient, are analysed. The results of numerical modelling for experiments in TEXTOR with different durations and intensities of xenon puffing are presented. The role of high-Z ions under reactor conditions is analysed, and it is shown that in a reactor, such as ITER, the discussed heavy impurity driven instabilities should be suppressed

A new fluctuation phenomenon is observed through Langmuir probe measurements at the edge plasma in the KT-5C tokamak by applying a -90 degrees phase shift feedback. Using a two point correlation technique, it is found that this fluctuation mode has a longer poloidal wavelength and a definite frequency when compared with the usual edge turbulence. It is also found through bispectral analysis that this mode is a spontaneously excited quasi-coherent mode, which has almost no contribution to the cross-field particle flux

The magnetic shear effect on thermal electron transport is studied in a large variety of non-inductive plasmas in Tore Supra. An improved confinement in the region of low and negative shear was observed and quantified with an exponential dependence on the magnetic shear (Litaudon, et al., Fusion Energy 1996 (Proc. 16th Int; Conf. Montreal, 1996), vol. 1, IAEA, Vienna (1997) 669). This is interpreted as a consequence of a decoupling of the global modes (Romanelli and Zonca, Phys. Fluids B 5 (1993) 4081) that are thought to be responsible for anomalous transport. This dependence is proposed in order to complete the Bohm-like L mode local electron thermal diffusivity so as to describe the transition from Bohm-like to gyroBohm transport in the plasma core. The good agreement between the predictive simulations of the different Tore Supra regimes (hot core lower hybrid enhanced performance, reversed shear plasmas and combined lower hybrid current drive and fast wave electron heating) and experimental data provides a basis for extrapolation of this magnetic shear dependence in the local transport coefficients to future machines. As an example, a scenario for non-inductive current profile optimization and control in ITER is presented

A non-linear Fokker-Planck code is applied to the study of a JT-60U hot ion plasma in which the experimentally measured carbon impurity temperature reached up to 45 keV with 90 keV deuterium beam injection. A non-Maxwellian deuteron distribution function is obtained numerically and the deuteron bulk temperature, which has not been determined experimentally, is evaluated from the slope of the energy spectrum. It is found that the deuteron bulk temperature can exceed the carbon temperature, indicating that the impurity temperature measurement does not lead to overestimation of the ion temperature. The deuteron effective temperature based on the average energy is, however, found to be almost the same as the carbon temperature. The DD fusion reactivity is also around a value given by the Maxwellian distribution with its temperature equal to the carbon temperature. Consequently, the carbon temperature may possibly be regarded as an equivalent ion temperature

The energy confinement time of TCV ohmic L mode discharges depends strongly on plasma shape. For fixed average current and electron densities, confinement times increase with plasma elongation and decrease with (positive) plasma triangularity. This dependence can be explained by the geometrical effects of flux surface expansion and compression on the temperature gradients together with the effect of power degradation, without the need to invoke a shape dependence of the transport coefficients. A global factor of merit, the shape enhancement factor H_s, is introduced to quantify this geometrical effect. The shape enhancement factor also has the potential to improve the description of the shape dependence in existing interdevice scaling laws. Modified versions of Neo-Alcator scaling and of Rebut-Lallia-Watkins scaling provide successful descriptions of ohmic L mode confinement for a large variety of plasma shapes in TCV by making use of H_s. Magnetohydrodynamic activity is also strongly dependent on plasma shape. Sawtooth amplitudes are largest at positive triangularity and sometimes vanish at negative triangularity, where the amplitude of MHD modes is highest. It is shown that the changes in MHD behaviour are to a large extent consequences of the confinement changes produced in these shaping experiments

Magnetic measurements alone on spherical tori should allow a very good separation of the poloidal beta beta _p, from the internal self-inductance l_i/2 and should even permit an accurate estimate of the current density j_phi profile. However, the reduced space allowed for magnetic sensors near the central conductor in a spherical tokamak, and the possibility of producing flux core spheromak configurations without a central conductor, could imply that magnetic probes are not present in the cavity of the spherical torus. The fluxes and fields of a variety of calculated spherical torus configurations, all endowed with a single or double null separatrix, are analysed in terms of spherical multipolar moments obtained from simulated magnetic measurements located only upon a sphere surrounding the spherical plasma. The solution to the problem of the absence of magnetic measurements in the cavity of the spherical torus is to fix from non-magnetic measurements (e.g., spectroscopy) the plasma inboard boundary r_in on the equatorial plane. This constraint is added to the constraints of matching the spherical multipolar expansion in an iterative solution of the Grad-Shafranov equation, on the basis of a spherical geometry. The convergence of the spherical reconstructive equilibrium code is extremely fast and gives an error on the total plasma current I_p of less than 1% at an aspect ratio A=1.2, an error on the position of the plasma boundary of less than 2% of the radius of the plasma sphere, an error on beta _p of at most 15% and, finally, the j_phi profile is extremely well reconstructed in peaked, flat and even hollow cases. The effect of an uncertainty +or- delta r_in upon the spectroscopic identification of the plasma inboard boundary on the equatorial plane r_in is assessed

The particle confinement time of boron, tau _B, is studied in the TPE-1RM20 reversed field pinch plasma by means of a laser ablation technique. The experimental conditions were as follows: pinch parameter, Theta , and plasma current, I_p, are scanned, and tau _B obtained from a zero dimensional approximation is compared with the energy confinement time, tau _E. The results show that tau _B increases almost linearly with tau _E, and the implication of this is discussed in terms of energy loss channels. The increase of tau _B with Theta supports the recently obtained improved confinement at high Theta . Laser ablation results are compared with the particle confinement time of deuterium, tau _p, from the D_alpha line spectrum measurement. It is shown that tau _b approximately= tau _p, which indicates a weak mass dependence in particle transport. Numerical simulation is also conducted and compared on a one dimensional basis with the experimentally obtained spatial distribution of the line emission intensity of the boron impurity injected by the laser ablation. The result shows reasonable agreement with experiment