Table of contents

For the investigation of turbulent transport in magnetized plasmas, the cross-phase between density and plasma potential fluctuations is a key parameter. In the majority of the experimental studies, the floating potential of a conventional Langmuir probe is used as an estimate for the plasma potential. It is well known, however, that plasma and floating potential fluctuations can have different phases if temperature fluctuations are present in the plasma. A diagnostic giving a direct estimate of the plasma potential fluctuations is the emissive probe. In this work, potential measurements from both emissive and Langmuir probes are directly compared. The experiments are carried out in the magnetically confined low-temperature plasma of the TJ-K torsatron. Potential, electric field and density fluctuations are measured with a three-tip probe array in the entire plasma volume. Profiles of mean values and fluctuations of the parameters are compared. The deviation of the mean plasma potential from the floating potential is larger than expected from simple probe theory. The measurements of fluctuations, however, which are relevant for turbulence studies, do not show a significant difference. This justifies the use of Langmuir probes for turbulence studies at least in toroidal low-temperature plasmas.

In this work three techniques that refine the magnetic reconstruction of the mega-ampere spherical tokamak (MAST) equilibrium are detailed: kinetic reconstruction, in which the thermal pressure profile is fitted to thermal electron and ion data; bootstrap (BS) reconstruction, in which the edge current profile is modified to be self-consistent with the BS fraction (in the limit that edge current is BS dominated); and fast-particle reconstruction, in which an effective fast-ion pressure component is added, representing ions driven by charge exchange of the thermal ions with injected neutrals. Kinetic reconstructions for some high performance shots suggest normalized beta, β_n, up to 4.95 and BS fractions up to 30%, with internal inductance l_i ≈ 1 and pressure peaking factor p(0)/⟨p⟩ ≈ 2.2. Full-orbit simulations suggest that up to 25% of the total stored energy in these high performance discharges is in the fast-ion population: addition of effective fast-particle pressures boosts β_n to 5.56. Ideal MHD pressure driven stability thresholds of n = 1, 2 and ∞ displacements are examined for kinetic and BS reconstructions of four high β_n MAST discharges. Based on kinetic reconstructions it is found that the no-wall instability threshold to external n = 1 displacements is β_n ≈ 5−6, and the with-wall β_n limit 10% higher than the no-wall limit. In comparison, the n = 1 no-wall instability threshold based on BS reconstructions is slightly below (≈95%) that determined using kinetic reconstructions. Comparison to the MAST database suggests that MAST is approaching a regime in which passive stabilization is required to prevent ideal disruptions at higher β_n. Finally, vertical stability of an earlier set of MAST discharges is also examined, an estimate of the MAST effective wall for n = 0 modes provided, and the wall radius for marginal stability parameterized as a function of l_i and κ. Together, these provide a measure of proximity to marginal stability.

In the TEXTOR tokamak, experiments were performed to simultaneously determine the molecular, atomic and total particle flux of deuterium in front of a graphite limiter, the temperature of which can be controlled independently of the plasma conditions. With rising limiter temperatures, T_TL, but constant plasma conditions an increase in Balmer emission and a decrease in Fulcher-band emission were observed. This variation is associated with a change in the type of released species: molecules dominate at low temperatures (550 K < T_TL < 1100 K), whereas at temperatures T_TL ⩾ 1100 K the direct atomic release starts to become important. The total number of deuterons remains constant for all temperatures. Since not all molecules dissociate into two potentially radiating atoms, it is necessary to take into account the ratio of atoms to molecules when deducing the total particle flux from the Balmer emission. We present a spectroscopic method which allows the determination of the atomic, molecular and total deuterium particle flux and which also gives effective conversion factors, (S/XB)_eff, to deduce the total deuterium flux from Balmer-α emission alone.

Analysis of the spectroscopic data of both species can be performed to determine the rotational and vibrational populations for the molecules by means of Fulcher-α spectroscopy, and the penetration depth and energy for the atoms using Balmer spectroscopy. This further analysis gives additional information about the release mechanism, showing that both species, atoms and molecules, are released predominantly as thermalized particles.

We describe experiments with a biased electrode inserted into the scrape-off layer (SOL) of the CASTOR tokamak. The resulting radial and poloidal electric field and plasma density modification are measured by means of Langmuir probe arrays with high temporal and spatial resolutions. Poloidally and radially localized stationary structures of the electric field (convective cells) are identified and a related significant modification of the particle transport in the SOL is observed.

The work shows that a single elongated island immersed in the quasi-neutral current sheet makes it MHD unstable. Typical values of growth rate are found to be several per cent of inverse Alfven time for broad sheets. Hall dynamics greatly enhance instability and growth rate reaches one-tenth of ion cyclotron frequency when sheet width is comparable to ion inertia length. A weak square-root dependence of increment on island width and length is derived both from numerical simulation and analytical analysis. At the non-linear phase of evolution a phenomenon of fast impulsive intensification of current is found. After a finite time of about 10 ion-cyclotron periods it ends by a collapse of the sheet. It is shown that Hall dynamics is responsible for such a behaviour. Possible implications for magnetotail substorms are discussed.

A novel new kinetic simulation model has been developed to investigate dynamics in collisionless plasmas. In this model, the electrons are treated as gyrokinetic (GK) particles and ions are treated as fully kinetic (FK) particles. In the GK-electron and FK-ion (GKe/FKi) plasma simulation model, the rapid electron cyclotron motion is removed, while keeping finite electron Larmor radii, realistic electron-to-ion mass ratio, wave–particle interactions, and off-diagonal components of the electron pressure tensor. The model is particularly suitable for plasma dynamics with wave frequencies lower than the electron gyrofrequency, and for problems in which the wave modes ranging from Alfvén waves to lower-hybrid/whistler waves need to be handled on an equal footing. Using this model, the computation power can be significantly improved over that of the existing full-particle codes. The GKe/FKi model, furthermore, can also handle physics with realistic electron-to-ion mass ratio and dynamic processes on the global Alfvén time/spatial scales. With respect to the hybrid (i.e. FK ion and fluid electron) model, the GKe/FKi model has the advantage that important electron kinetic physics, such as wave–particle resonances and finite electron Larmor radius effects, are included. The simulation model has been successfully benchmarked for linear waves in uniform plasmas against analytic dispersion relation.