Table of contents

The main parameters of the beam required to ignite a precompressed DT fuel, as foreseen by the recently proposed scheme of fast ignition by laser-accelerated protons (Roth et al 2001 Phys. Rev. Lett.86 436), are studied by 2-D numerical simulations and a simple model. For simplicity, instantaneous proton generation at distance d from the compressed fuel and exponential proton energy spectrum, dn/dε∝exp (-ε/T_p), are assumed. An analytical expression and parametric numerical results are then given for the dependence of the minimum required beam energy on d, T_p and on the fuel density ρ. For the parameters of Roth et al (d≈4 mm; ρ≈400 g/cm³) the minimum total proton energy for ignition is about 40 kJ.

The B2.5 code is applied for modelling SOL plasma in the HT-7U superconductor tokamak in order to construct a database for the design of a divertor for the device. The plasma parameters at the core-SOL interface set as the boundary conditions for B2.5 modelling are varied extensively in order to survey the operation of the divertor over a wider range of plasma parameters. Detailed distributions of the plasma heat flux and the other plasma parameters in the SOL, especially at the target plate of the divertor and the outer wall, are found by modelling, and a detailed analysis for the power balance in the SOL is carried out. The modelling results show that most of the electron heat flux and a small part of the ion heat flux arrive at the target plate of the divertor, while a part of the ion heat flux and a small part of the electron heat flux arrive at the outer wall. For the same boundary conditions the SOL physics widths λ_Te, λ_Ti and λ_ne,i increase with SOL thickness D_SOL and the peak values of plasma heat flux and some other plasma parameters at the target plate of the divertor decrease with increasing D_SOL. For a boundary density N_edge of 2×10¹⁹ m^-3, a boundary temperature T_edge of 300 eV and D_SOL = 5 cm, the peak value of the electron heat flux at the target plate is 3.92 MW/m², with 52.81% of the electron heat flux and only 4.20% of the ion heat flux arriving at the target plate. The modelling results form a database for the preliminary design of the HT-7U tokamak divertor.

Internal transport barriers (ITBs) can be produced in JET by the application of strong additional heating during the current rise of plasma discharges. These `so-called' optimized shear experiments with low positive magnetic shear have revealed a strong dependence between the formation of the barrier and the integer q magnetic surfaces (q = 2 or q = 3). Further analysis also shows a correlation between the emergence of the ITB and the edge external MHD activity which is triggered when an integer q surface occurs at the edge of a strongly heated plasma (q = 4, q = 5 or q = 6). Mode coupling is the prime candidate to explain the link between the internal integer flux surfaces, where the ITB is triggered, and the plasma edge. Modelling of the MHD behaviour confirms the possibility of such a mechanism. Once coupled, this destabilized mode is thought to enhance locally the E×B shearing rate, either by magnetic braking or by the radial transport losses resulting from the modification of the field line topology. This could then trigger an ITB inside the internal integer surface at q = 2 or q = 3.

Mechanically attached blanket module insertion tests were carried out for various misalignments between the module and the back plate. In the insertion tests, the module was successfully inserted up to a misalignment of ±10 mm with a clearance of ±(0.16-0.18) mm between key and groove. This was achieved owing to the passive compliance due to the flexibility of the manipulator with the assistance of the chamfer configuration of the key for smooth insertion. In addition, the `correlation coefficient' based on the results obtained by the strain gauges located at the end effector was found to be useful for estimating the forces of the complex end effector during module insertion for the development of sensor based control.

The experimental and theoretical activity performed in the RFX device has allowed a deeper insight into the MHD properties of the reversed field pinch (RFP) configuration. A set of successful experiments has demonstrated the possibility of influencing both the amplitude and the spectrum of the magnetic fluctuations which characterize the RFP configuration. A new regime (quasi-single-helicity states) where the dynamo mechanism works in a nearly laminar way and a helical core plasma is produced has been investigated. With these studies a reduction of magnetic chaos has been obtained. The continuous rotation of wall locked resistive tearing modes has been obtained by an m = 0 rotating perturbation. This perturbation induces rotation of m = 1 non-linearly coupled modes.

Heat transport in steady state ASDEX Upgrade discharges is compared with the predictions of four theory based models: the IFS/PPPL, Weiland and GLF23 models, based on ion temperature gradient (ITG) and trapped electron mode physics, and the current diffusive ballooning mode (CDBM) model. The discharges selected provide scans over heating power, density and plasma current, by varying these quantities separately. The ion temperature profiles are found to be stiff, with the same linear relation between the core and edge temperatures for all the scans considered. The ITG based models reproduce this behaviour well qualitatively and quantitatively, while the CDBM model fails to predict profile stiffness. The electron temperature, however, does change shape when the density is lowered. All models reproduce this feature qualitatively. The stored energy is well predicted by the IFS/PPPL and Weiland models. The GLF23 model is less good, showing a clear trend to underpredict the energy at higher temperatures. The predictions made with the CDBM model are rather poor.

An eight shot pellet injector is used to fuel the HL-1M tokamak from the low field side. The pellet ablation process is investigated with a CCD camera and an array of H_α emission detectors. It is observed that the velocity of the pellet ablation cloud is obviously slowed down after it enters the plasma. The ablation clouds are elongated along the magnetic field direction. By means of multi-exposures with the CCD camera during pellet ablation, the safety factor profile is estimated with the ablation cloud inclination angles with respect to the torus on HL-1M. The pellet penetration depth is analysed with Abel inversion of the H_α emissions. The result is compared with the prediction of ablation models. Kuteev's model is more suitable for describing pellet ablation processes in the HL-1M plasma than the neutral gas shielding model.

Recent experimental results show that the core electron temperature in the TJ-II stellarator almost doubles previously obtained values for the same heating power. These plasmas, heated with electron cyclotron waves, are characterized by their low density, and by having highly peaked electron temperature profiles and flat, or even hollow, density profiles. The conditions for obtaining these high electron temperature discharges regarding their density, injected power and dependence on plasma species are described. Neoclassical and experimental transport analyses are performed for these discharges, showing a reduction in the electron heat conductivity at the plasma core. The relations of this observed confinement enhancement to the CHS internal transport barrier and the W7-AS neoclassical electron root feature are discussed.

JET experiments using deuterium and tritium have made it possible to study sawtooth stability in plasmas approaching thermonuclear conditions. Record fusion yields were obtained in discharges where the sawtooth was delayed. In a sequence of discharges designed to study alpha heating, a new phenomenon was observed: the sawtooth period increases with tritium concentration. The internal kink stability for both high performance and alpha heating experiment plasmas was studied. Calculations for the record fusion discharge at the time of maximum fusion power (P_fus = 16 MW) showed that alpha particles make a significant stabilizing contribution to the potential energy of the m = 1 internal kink instability. The scaling of sawtooth period with tritium concentration implies a dependence on mean ion mass: possible reasons for such a dependence are considered. Calculations of the kinetic fast particle contribution to the kink potential energy indicate that the rise in sawtooth period with tritium concentration is likely to have arisen from two effects: an increase in the slowing down time of the beam ions, which in these plasmas is proportional to the mean ion mass, and an increase in the proportion of beam ions at the full injection energy.

Termination of the runaway electron current generated during plasma disruptions is found in JT-60U during simulated vertical plasma displacement events where the safety factor at the plasma surface q_s decreases. For all discharges with runaway electron generation, the runaway current disappears for q_s⩾2 with the appearance of spikes in the magnetic fluctuations. The growth rate of the spikes in the magnetic fluctuations decreases by an order of magnitude during the termination of runaway current. Corresponding to the loss of runaway electrons by magnetic fluctuations, heat flux pulses are measured at the inner divertor plates, which indicates interaction of the wall with the runaway electrons. The halo current during runaway termination is small and increases after runaway termination with a dominant toroidal mode of n = 1.

A theoretical framework for understanding the feedback mechanism for stabilization of external MHD modes has been formulated. Efficient computational tools - the GATO stability code coupled with a substantially modified VACUUM code - have been developed to effectively design viable feedback systems against these modes. The analysis assumed a thin resistive shell and a feedback coil structure accurately modelled in θ and ϕ, albeit with only a single harmonic variation in ϕ. Time constants and induced currents in the enclosing resistive shell are calculated. An optimized configuration based on an idealized model has been computed for the DIII-D device. Up to 90% of the effectiveness of an ideal wall can be achieved.

The ablated material of a frozen hydrogen isotope pellet which is injected into a hot tokamak plasma forms a high β plasmoid. This diamagnetic plasmoid is accelerated to the magnetic low field side of the torus. The high β plasmoid drift was directly observed by an optical diagnostic with high space and time resolution. Spectroscopic measurements of the emitted light allowed the density and temperature of the ablation cloud, and for the first time also of the drifting plasmoids, to be determined. The experiments give a new insight into the dynamics of the formation of striations during the pellet ablation; these striations cause the separation of the ablated material into a sequence of separated, drifting plasmoids. The influence of the drift on the mass deposition profile for high field side pellet injection is discussed. The plasmoid dynamics even influences the radial pellet motion, most probably owing to a rocket effect. The physical principles of the high β plasmoid drift are discussed and compared with the experimental observations.

Various ICRH scenarios for ITER-FEAT are evaluated. A wave propagation and damping study confirms the potential of the second harmonic tritium heating scenario for both heating and current drive purposes. The fundamental deuterium heating scheme is dominated by beryllium and alpha particle absorption. Owing to the reduced ITER size, the low frequency current drive window is lost in practice. A ³He minority greatly enhances the performance. In the early stage of the discharge, the power absorbed by the ³He is transferred to the background ions but, later on, the power of the fast particles (from both the ³He and the tritium tails) is roughly equally distributed between electrons and ions. Using the experimentally established expression for the L to H mode threshold, it is found that the H mode regime can always be reached using RF heating. To achieve Q = 10, high density operation is required. Minority current drive competes with heating and with electron current drive. The scenarios foreseen for the non-activated ITER-FEAT phase are also discussed.

Boronization of NSTX has allowed access to higher density higher confinement plasmas. A glow discharge with 4 mtorr helium and 10% deuterated trimethyl boron deposited 1.7 g of boron on the plasma facing surfaces. Ion beam analysis of witness coupons showed a B + C areal density of 10¹⁸ cm^-2 corresponding to a film thickness of 100 nm. Subsequent ohmic discharges showed oxygen emission lines reduced by a factor of 15, carbon emission reduced by a factor of two and copper reduced to undetectable levels. After boronization, the plasma current flat-top time increased by 70% enabling access to higher density higher confinement plasmas.

Substantial progress has been made towards both understanding and control of internal transport barriers (ITBs) on DIII-D, resulting in the discovery of a new sustained high performance operating mode termed the quiescent double barrier (QDB) regime. The QDB regime combines core transport barriers with a quiescent ELM-free H mode edge (termed QH mode), giving rise to separate (double) core and edge transport barriers. The core and edge barriers are mutually compatible and do not merge, resulting in broad core profiles with an edge pedestal. The QH mode edge is characterized by ELM-free behaviour with continuous multiharmonic MHD activity in the pedestal region and has provided density and radiated power control for longer than 3.5 s (25τ_E) with divertor pumping. QDB plasmas are long pulse high performance candidates, having maintained a β_NH₈₉ product of 7 for five energy confinement times (T_i⩽16 keV, β_N⩽2.9, H₈₉⩽2.4, τ_E⩽150 ms, DD neutron rate S_n⩽4×10¹⁵ s^-1). The QDB regime has only been obtained in counter-NBI discharges (injection antiparallel to the plasma current) with divertor pumping. Other results include successful expansion of the ITB radius using (separately) both impurity injection and counter-NBI, and the formation of ITBs in the electron thermal channel using both ECH and strong negative central shear (NCS) at high power. These results are interpreted within a theoretical framework in which turbulence suppression is the key to ITB formation and control, and a decrease in core turbulence is observed in all cases of ITB formation.

Controllability of current profiles in long duration discharges is studied by two kinds of combination of wave spectra. First, by superposition of a higher N_∥ spectrum wave at 2.45 GHz on a target plasma driven by a lower N_∥ one at 8.2 GHz, a hollow j(r) profile has been achieved and sustained for 20 s. The current profile can be well controlled below a threshold power of 14 kW by varying the power at 2.45 GHz. A spontaneous transition phenomenon to a peaked j(r) profile, however, has occurred above the threshold power even after a reasonable steady state condition had been obtained. Second, using two oppositely travelling waves at 8.2 GHz the total current is clearly reduced, and the j(r) profile becomes peaked when the backward travelling lower hybrid waves (BWs) are superposed on the forward lower hybrid wave at a low power of 20 kW. When the BW power is increased above 27 kW, further current reduction is suppressed, and above 34 kW the direction of the current driven by the BWs is completely reversed and the j(r) profile becomes broad. Thus, current modification by BWs shows a highly non-linear behaviour with respect to the BW power.

A new method of particle motion control in toroidal magnetic traps with a rotational transform using an estafette of drift resonances and stochasticity of particle trajectories is proposed. The use of the word estafette here means that particles pass through a set of resonances in consecutive order from one to another during their motion. The overlapping of adjacent resonances can be moved radially from the centre to the edge of the plasma by switching on the corresponding perturbations at times in accordance with a particular rule. In this way, particles (e.g., cold alpha particles) can be removed from the centre of the confinement volume to the plasma periphery. For an analytical treatment of the stochastic behaviour of particle motion, the equation for the drift surface function Ψ* (r,ϑ,φ,V_∥,V_⊥,t) = const is reduced to a diffusion type equation in the case when some adjacent resonances take place and the stochastic diffusion coefficients D_r,r, D_r,ϑ and D_ϑ,ϑ in the real space variables are introduced. This new approach is demonstrated by numerical computations of test helium particle trajectories in the LHD toroidal trap.

A detailed systems analysis code has been used to compare the projected cost of electricity (COE) for inertial fusion energy for direct drive (DD) and indirect drive (ID) target scenarios, based on a diode pumped solid state laser driver with Yb:S-FAP (Yb doped Sr₅(PO₄)₃F) gain media. Previously published target gain curves which resulted in a target gain at the optimal DD operating point that is 30% higher than that for the ID scenario have been used. This gain advantage for DD is offset by a requirement for improved beam smoothing, which was obtained via smoothing by spectral dispersion (SSD) with a 1 THz bandwidth at 349 nm. Such a large SSD bandwidth has a number of effects on laser performance, including greater risk of optics damage from non-linear effects, lowered harmonic conversion efficiency, altered extraction parameters and higher front-end costs. The non-linear effects, which contribute to optical component damage by amplification of intensity non-uniformities, were parameterized through a constraint on the maximum allowable B integral (i.e. the total average phase retardation due to the non-linear indices of all materials traversed by the beam). If we constrain B to be no larger than 1.8 rad, which is the presently accepted safe value based on observations in single shot glass laser facilities, the COEs for DD and ID are predicted to be the same within the uncertainties. If technology permits the B limit to be raised, the optimized COE for DD is predicted to decrease relative to that for ID.