Search

In the present magnetically confined plasmas, the prediction of particle loading on material surfaces is a primary concern in view of the protection of plasma facing components for next step devices. Thus, an understanding of filament dynamics is needed. In this context, this work aims to develop an automatic detector for filaments arising in the MAST-U plasma. The identification of the filaments has been done starting from 2D images acquired with a fast visible camera. Therefore, it can be faced as an image object recognition problem. Currently, the object recognition is a key output of deep learning and machine learning algorithms. In this paper, a database of several thousands of images generated by a synthetic diagnostic, which reproduces the statistical properties of experimental filaments in terms of position, size and intensity has been used. The synthetic images are pre-processed by mapping them onto the toroidal midplane of the machine. Then a Faster R-CNN is customized to the problem of identifying the filaments. In particular, in order to enhance the performance of the detector, a suitable definition of the target-boxes defining the filament positions and sizes is adopted with good results.

There exists a large body of previous work using reduced two-dimensional models of the SOL, which model fluctuations in the drift-plane but approximate parallel transport with effective loss terms. Full size three-dimensional simulations of SOL turbulence in experimental geometries are now possible, but are far more computationally expensive than 2D models. We therefore use a flux-tube geometry model of the scrape-off layer to compare the results of 2D simulations to 3D simulations with a similar setup, looking for systematic differences. Overall there is good agreement in the basic radial profiles, probability distribution functions, and power spectra of fluctuations. However, the average temperature is over-predicted in 2D relative to 3D, and we explain the difference in terms of the effect of geometrical simplifications of devices at low power. Varying geometric parameters, we find that supersonic flow in the divertor leg, which occurs because our simulations do not include neutrals and so represent low-recycling conditions, means that the divertor leg length only has a weak effect on the output. Finally, we examine the effect of altering the magnitude of source and sink terms in 2D, concluding that they cannot easily be used to recreate both the density and temperature profiles observed in 3D simultaneously., 14 pages, 24 figures, submitted to Plasma Physics and Controlled Fusion

The modelling of JET corner configurations, in which the strike points are positioned deep in the corners of the divertor, is extremely challenging for edge plasma fluid modelling tools. To circumvent this technical limitation, a geometrical approximation has been proposed, consisting in considering an artificial minor modification of the geometry of the divertor targets plates. In this paper, we investigate how significantly this approximation impacts the output of transport simulations. Using the SOLEDGE2D-EIRENE transport code which has the unique capability to be able to cope with both the full and the approximated geometry, we have performed a density scan in H-mode for pulses in which the outer strike-point lies in the corner of the divertor. We report here how simulations in the artificial geometry differ from the ones in unaltered geometry. At the exception of low density cases, mid-plane profiles in the closed field lines region and the near scrape-off layer are little impacted. Further out however, in flux-surfaces that are concerned by the geometrical modification, we find that modifying the geometry leads to a strong overestimate of the plasma density. The density perturbation is not local and concerns the whole flux surfaces. Although the divertor geometry is modified only on the outer side, the largest impact is found at the inner divertor where densities are systematically overestimated by a factor that can exceed 10 in low density cases in the far Scrape-Off Layer (SOL) and temperature underestimated by 10 to 20 eV in most of the studied density range. The near SOL and strike point peak values are also impacted in the same direction with density changes by a factor of 2. As a consequence, the threshold to detachment of the inner divertor is found lower in the approximate geometry than in the unaltered one. Due to the large flux expansion between the outer and the inner target, the difference in plasma is especially sensitive at the top of the inner divertor baffle, which could have consequence on the evaluation of physical sputtering at that critical location.

In the present work, the role of plasma facing components protection in driving the EU-DEMO design will be reviewed, focusing on steady-state and, especially, on transients. This work encompasses both the first wall (FW) as well as the divertor. In fact, while the ITER divertor heat removal technology has been adopted, the ITER FW concept has been shown in the past years to be inadequate for EU-DEMO. This is due to the higher foreseen irradiation damage level, which requires structural materials (like Eurofer) able to withstand more than 5 dpa of neutron damage. This solution, however, limits the tolerable steady-state heat flux to ~1 MW/m2, i.e. a factor 3–4 below the ITER specifications. For this reason, poloidally and toroidally discontinuous protection limiters are implemented in EU-DEMO. Their role consists in reducing the heat load on the FW due to charged particles, during steady state and, more importantly, during planned and off-normal plasma transients. Concerning the divertor configuration, EU-DEMO currently assumes an ITER-like, lower single null (LSN) divertor, with seeded impurities for the dissipation of the power. However, this concept has been shown by numerous simulations in the past years to be marginal during steady-state (where a detached divertor is necessary to maintain the heat flux below the technological limit and to avoid excessive erosion) and unable to withstand some relevant transients, such as large ELMs and accidental loss of detachment. Various concepts, deviating from the ITER design, are currently under investigation to mitigate such risks, for example in-vessel coils for strike point sweeping in case of reattachment, as well as alternative divertor configurations. Finally, a broader discussion on the impact of divertor protection on the overall machine design is presented.

Fluid models used to study the edge plasma region need to be benchmarked against similar conditions given that models can strongly differ in complexity and therefore the results they produce. Via this validation study undertaken through the framework of EUROfusion Enabling Research, four state-of-the art models—GBS, Hermes/BOUT++, hot-edge-sol-electrostatic and TOKAM3X—are compared to experimental plasma turbulence measurements on the ISTTOK tokamak. Statistical comparisons of simulation and experiment data show that fluid models used here can replicate most of the experiment in terms of I sat and V float fluctuations despite their differences. Furthermore, it is shown that without including more complex information (like core turbulence information and domain geometry details and magnetic topological aspects) in fluid models, the results recovered vary from their experimental counterparts. Via the simulations using these codes, it is demonstrated that fluid models continue to be a good cost-effective tool in recovering many global aspects of edge plasma behaviour.

A physics and engineering analysis of alternative divertor configurations is carried out by examining benefits and problems by comparing the baseline single null solution with a Snowflake, an X- and a Super-X divertor. It is observed that alternative configurations can provide margin and resilience against large power fluctuations, but their engineering has intrinsic difficulties, especially in the balance between structural solidity and accessibility of the components and when the specific poloidal field coil positioning poses further constraints. A hybrid between the X- and Super-X divertor is proposed as a possible solution to the integration challenge.

A double-null configuration is being considered for the EU-DEMO, due to its potential benefits for power exhaust arising from the use of two active divertors and magnetically disconnected low- and high-field sides. Using systematic parameter scans in fluid simulations, we have investigated the divertor power exhaust in the EU-DEMO in a connected double-null configuration, and compared the edge plasma properties to those obtained in a single-null configuration under detached conditions anticipated for reactor operation. Neglecting drift effects and kinetic behaviour of the neutrals, no clear benefits of the double-null configuration could yet be identified for the radiation pattern and power mitigation on open field lines. Future work should address the aforementioned physics as well as the effect of the additional X-point on core radiation.

We present an analysis of parallel electron energy transport in the scrape-off layer (SOL), considering the convective and conductive channels, as well as the radiation and neutral inelastic energy transfer channels involving atomic deuterium. Kinetic effects in both equilibria and conductive transients are explored by utilizing the capability of the SOL-KiT code to treat electrons as either a fluid or kinetically. We find kinetic effects in multiple channels, with an emphasis on those occurring during the investigated conductive transients. Energetic electron effects in the heat flux, as well as a modification of ionization rates of up to 40% compared to Maxwellian rates during perturbations in detached conditions, are reported and discussed.

We present the first parallel electron transport results obtained using the newly developed 1D transport code SOL-KiT. With the capability to switch between consistent kinetic and fluid models for the electrons, we explore and report the differences in both equilibrium and transient simulations. Significant kinetic effects are found during transients, especially in the behaviour of the electron sheath heat transmission coefficient, which shows up to an eightfold increase. Equilibria are obtained for an input power scan with parameters relevant to medium size tokamaks. Detached equilibria are found to persist to higher input powers when electrons are treated kinetically. Furthermore, non-monotonic behaviour of the electron sheath heat transmission coefficient is observed in the power scan, with values being up to 40% above the classical value. We discuss the implications of the pr

The role of neutral and charged hydrogenic molecules in detached regimes of tokamak plasmas is investigated using a simplified 1D numerical model. Using MAST Upgrade like conditions, simulations are implemented to study the rollover of target flux Γ in upstream density scan. It is found that if H 2 and H 2 + are considered in simulations a lower target temperature and a larger upstream density will be required to trigger divertor detachment under the same input power and particle flux, and the critical detachment threshold (the critical ratio of upstream static pressure to the power entering the recycling region) is found to be p up P recl ∼ 8.1 N M W − 1 at rollover. Molecule–plasma interactions are found to be as crucial as atom–plasma interactions during divertor detachment, both of which account for the majority of plasma momentum loss in the cases studied here. Further analysis of the momentum loss decomposition shows molecule-plasma elastic collisions dominate molecule-plasma interactions, while molecular charge exchange cannot effectively reduce plasma momentum. In terms of H alpha emission, a strong rise of H alpha signal is found to be due to molecular excitation channels when the upstream density further increases after rollover.

Future tokamak reactors of conventional design will require high levels of exhaust power dissipation (more than 90% of the input power) if power densities at the divertor targets are to remain compatible with active cooling. Impurity seeded H-mode discharges in JET-ITER-like Wall (ILW) have reached a maximum radiative fraction (F-rad) of similar to 75%. Divertor Langmuir probe (LP) measurements in these discharges indicate, however, that less than similar to 3% of the thermal plasma power reaches the targets, suggesting a missing channel for power loss. This paper presents experimental evidence from limiter LP for enhanced cross-field particle fluxes on the main chamber walls at high F-rad. In H-mode nitrogen-seeded discharges with F-rad increasing from similar to 30% to up to similar to 75%, the main chamber wall particle fluence rises by a factor similar to 3 while the divertor plasma fluence drops by one order of magnitude. Contribution of main chamber wall particle losses to detachment, as suggested by EDGE2D-EIRENE modeling, is not sufficient to explain the magnitude of the observed divertor fluence reduction. An intermediate detached case obtained at F-rad similar to 60% with neon seeding is also presented. Heat loads were measured using the main chamber wall thermocouples. Comparison between thermocouple and bolometry measurements shows that the fraction of the input power transported to the main chamber wall remains below similar to 5%, whatever the divertor detachment state is. Main chamber sputtering of beryllium by deuterium is reduced in detached conditions only on the low field side. If the fraction of power exhaust dissipated to the main chamber wall by cross-field transport in future reactors is similar to the JET-ILW levels, wall plasma power loading should not be an issue. However, other contributions such as charge exchange may be a problem. For complete list of authors see http://dx.doi.org/10.1016/j.nme.2017.02.010

A novel approach using Bayesian inference has been implemented to interpret the filamentary dynamics measured by a Langmuir probe fixed to a reciprocating assembly on MAST. The model describes the system as a superposition of time-displaced filaments and a fixed background component. Each filament is parameterised in terms of a characteristic rise and fall time and maximum amplitude centred on local maxima in the measured data time-series. A distinctive feature of the approach is that no minimum threshold is set for the existence of filaments. It is observed that whereas large amplitude filaments are well characterised in terms of rise times, smaller amplitude filaments are often unconstrained by the data and are limited by the details of the prior. Based on these findings, a new definition for the plasma filaments is proposed based on the uncertainty in the filament rise times. The remaining filaments together with the constant background component forms a new time-dependent signal referred to as the computed background fluctuation signal. The characteristics of these signals (for the plasma filaments and for the background fluctuations) are reported in terms of their spatial variation as the probe moves through the SOL and into the core plasma.

This paper presents the effect of self-consistent plasma backgrounds including plasma-neutral interactions, on the dynamics of filament propagation. The principle focus is on the influence of the neutrals on the filament through both direct interactions and through their influence on the plasma background. Both direct and indirect interactions influence the motion of filaments. A monotonic increase of filament peak velocity with upstream electron temperature is observed, while a decrease with increasing electron density is observed. If ordered by the target temperature, the density dependence disappears and the filament velocity is only a function of the target temperature. Smaller filaments keep a density dependence, as a result of the density dependence of the plasma viscosity. The critical size $\delta^*$, where filaments are fastest, is shifted to larger sizes for higher densities, due to the plasma viscosity. If the density dependence of the plasma viscosity is removed, $\delta^*$ has no temperature dependence, but rather a density dependence.

Intrinsic Scrape Off Layer (SOL) instabilities are studied using flute approximation and incorporating the appropriate sheath boundary conditions at the target. The linear growth rate and the structure of the modes are obtained. The associated diffusion is estimated using a γ / k ⊥ 2 approach for the fastest growing modes. The model used includes curvature and sheath drives, finite Larmor radius effects and partial line tying at the target. The magnetic geometry is obtained using current carrying wires, representing the plasma current and the divertor coils, and naturally generates X-point geometry and magnetic shear effects. The calculation is performed for ITER relevant parameters and scans in SOL width and distance from the separatrix are presented. In addition to a standard Lower Single Null, Super-X and Snowflake configurations are examined in order to assess the importance of the geometry on the stability of the boundary plasma.

Using a kinetic model for the ions and adiabatic electrons, we solve a steady state, electron-repelling magnetic presheath in which a uniform magnetic field makes a small angle $\alpha \ll 1$ (in radians) with the wall. The presheath characteristic thickness is the typical ion gyroradius $\rho_{\text{i}}$. The Debye length $\lambda_{\text{D}}$ and the collisional mean free path of an ion $\lambda_{\text{mfp}}$ satisfy the ordering $\lambda_{\text{D}} \ll \rho_{\text{i}} \ll \alpha \lambda_{\text{mfp}}$, so a quasineutral and collisionless model is used. We assume that the electrostatic potential is a function only of distance from the wall, and it varies over the scale $\rho_{\text{i}}$. Using the expansion in $\alpha \ll 1$, we derive an analytical expression for the ion density that only depends on the ion distribution function at the entrance of the magnetic presheath and the electrostatic potential profile. Importantly, we have added the crucial contribution of the orbits in the region near the wall. By imposing the quasineutrality equation, we derive a condition that the ion distribution function must satisfy at the magnetic presheath entrance --- the kinetic equivalent of the Chodura condition. Using an ion distribution function at the entrance of the magnetic presheath that satisfies the kinetic Chodura condition, we find numerical solutions for the self-consistent electrostatic potential, ion density and flow across the magnetic presheath for several values of $\alpha$. Our numerical results also include the distribution of ion velocities at the Debye sheath entrance. We find that at small values of $\alpha$ there are substantially fewer ions travelling with a large normal component of the velocity into the wall., Comment: 69 pages, 20 figures

This paper investigates the effect of the plasma background, including neutrals in a self-consistent way, on filaments in the scrape-off layer (SOL) of fusion devices. A strong dependency of filament motion on background density and temperature is observed. The radial filament motion shows an increase in velocity with decreasing background density and increasing background temperature. In the simulations presented here, three neutral-filament interaction models have been compared, one with a static neutral background, one with no interaction between filaments and neutrals, and one co-evolving the neutrals self consistently with the filaments. With the background conditions employed here, which do not show detachment, there are no significant effects of neutrals on filaments, as by the time the filament reaches maximum velocity, the neutral density has not changed significantly.

Using footage from high speed movies taken of the boundary plasma in the Mega Amp Spherical Tokamak (MAST) general properties of filaments are inferred through statistical moments. Filaments are observed up to and beyond the ψ N = 1.5 flux surface which, in single null configurations, lies well beyond the secondary separatrix and leads to filaments observed > 30 cm from the top of the plasma. In the divertor filaments are observed to connect through to the target, however a quiescent region is observed close to the X-point where no coherent filaments are identified. This region coincides with a sharp rise in the integrated magnetic shear which may change the nature of the filament cross-section.

This paper presents the use of a novel modelling technique based around intermittent transport due to filament motion, to interpret experimental profile and fluctuation data in the scrape-off layer (SOL) of JET during the onset and evolution of a density profile shoulder. A baseline case is established, prior to shoulder formation, and the stochastic model is shown to be capable of simultaneously matching the time averaged profile measurement as well as the PDF shape and autocorrelation function from the ion-saturation current time series at the outer wall. Aspects of the stochastic model are then varied with the aim of producing a profile shoulder with statistical measurements consistent with experiment. This is achieved through a strong localised reduction in the density sink acting on the filaments within the model. The required reduction of the density sink occurs over a highly localised region with the timescale of the density sink increased by a factor of 25. This alone is found to be insufficient to model the expansion and flattening of the shoulder region as the density increases, which requires additional changes within the stochastic model. An example is found which includes both a reduction in the density sink and filament acceleration and provides a consistent match to the experimental data as the shoulder expands, though the uniqueness of this solution can not be guaranteed. Within the context of the stochastic model, this implies that the localised reduction in the density sink can trigger shoulder formation, but additional physics is required to explain the subsequent evolution of the profile.

International audience; Three-dimensional and two-dimensional seeded blob simulations are performed with five different fluid models, all based on the drift-reduced Braginskii equations, and the numerical results are compared among themselves and validated against experimental measurements provided by the TORPEX device ( Fasoli et al 2006 Phys. Plasmas 13 055902). The five models are implemented in four simulation codes, typically used to simulate the plasma dynamics in the tokamak scrape-off layer, namely BOUT++ (Dudson et al 2009 Comput. Phys. Commun. 180 1467), GBS (Ricci et al 2012 Plasma Phys. Control. Fusion 54 124047), HESEL (Nielsen et al 2015 Phys. Lett. A 379 3097), and TOKAM3X (Tamain et al 2014 Contrib. Plasma Phys. 54 555). Three blobs with different velocities and different stability properties are simulated. The differences observed among the simulation results and the different levels of agreement with experimental measurements are investigated, increasing our confidence in our simulation tools and shedding light on the blob dynamics. The comparisons demonstrate that the radial blob dynamics observed in the three-dimensional simulations is in good agreement with experimental measurements and that, in the present experimental scenario, the two-dimensional model derived under the assumption of k(vertical bar vertical bar) = 0 is able to recover the blob dynamics observed in the three-dimensional simulations. Moreover, it is found that an accurate measurement of the blob temperature is important to perform reliable seeded blob simulations.

Reciprocating Mach probe data is used to estimate the radial wave number of oscillatory zonal flows in Ohmic MAST plasma. An intermittent ~10 kHz mode, previously identified as a geodesic acoustic mode (GAM), is detected in the wavelet decomposition and windowed spectra of plasma potential fluctuations of the MAST tokamak edge plasma. Two-points phase differencing technique is then applied to probe pins with radial and poloidal separations giving an estimate of the radial wave number at the desired range of frequencies. The phase velocity of propagation and an estimate of the shearing rate of the GAM is obtained. We measure the radial mode number range k r ≈ 0.3–1.0 cm−1 and a radial propagation speed of up to ~1 km s−1. The GAM shearing rate is an order of magnitude smaller than the growth rate of drift-like turbulence. These results are consistent with the estimates obtained previously from multi-fluid numerical simulations of GAM in MAST.

Clear filamentary structures are observed at the edge of tokamak plasmas. These filaments are ejected out radially and carry plasma in the far Scrape Off Layer (SOL) region, where they are responsible for producing most of the transport. A study has been performed of the characteristics of the filaments observed in L-mode plasma on MAST, using visible imaging. A comparison has then been made with the observed particle and power profiles obtained at the divertor as a function of the plasma current. The radial velocity and to a lesser extent the radial size of the filaments are found to decrease as the plasma current is increased at constant density and input power. The results obtained in this paper on the dependence of the average filament dynamics on plasma current are consistent with the idea that the filaments are responsible for determining the particle profiles at the divertor., 24 pages, 17 figures. This is an author-created, un-copyedited version of an article submitted for publication in Plasma Physics and Controlled Fusion. IoP Publishing Ltd is not responsible for any errors or omissions in this version of the manuscript or any version derived from it

Four numerical codes are employed to investigate the dynamics of scrape-off layer filaments in tokamak relevant conditions. Experimental measurements were taken in the MAST device using visual camera imaging, which allows the evaluation of the perpendicular size and velocity of the filaments, as well as the combination of density and temperature associated with the perturbation. A new algorithm based on the light emission integrated along the field lines associated with the position of the filament is developed to ensure that it is properly detected and tracked. The filaments are found to have velocities of the order of 1 km s(-1), a perpendicular diameter of around 2-3 cm and a density amplitude 2-3.5 times the background plasma. 3D and 2D numerical codes (the STORM module of BOUT++, GBS, HESEL and TOKAM3X) are used to reproduce the motion of the observed filaments with the purpose of validating the codes and of better understanding the experimental data. Good agreement is found between the 3D codes. The seeded filament simulations are also able to reproduce the dynamics observed in experiments with accuracy up to the experimental errorbar levels. In addition, the numerical results showed that filaments characterised by similar size and light emission intensity can have quite different dynamics if the pressure perturbation is distributed differently between density and temperature components. As an additional benefit, several observations on the dynamics of the filaments in the presence of evolving temperature fields were made and led to a better understanding of the behaviour of these coherent structures.

The interaction of L-mode and inter-ELM Scrape-Off Layer (SOL) filaments is investigated by simulating the dynamics of two seeded density perturbations with different initial separation. A numerical campaign of 2D and 3D simulations is presented to explore the effect that different filaments’ perpendicular size and relative distance have on their reciprocal action. It is shown that the interaction is weak unless the filaments are in close proximity, i.e. with a separation of the order of the filaments’ diameter in the drift plane. The interaction occurs when the dipolar electrostatic fields associated with the filaments merge or cancel each other and it leads to changes in the radial velocity in the centre of mass of the system that are modest (≤ 50%) and quickly decaying with separation. These results show that filaments travelling in the SOL can be considered to a good approximation as independent entities.

We develop a gyrokinetic treatment for ions in the magnetic presheath, close to the plasma-wall boundary. We focus on magnetic presheaths with a small magnetic field to wall angle, (in radians). Characteristic lengths perpendicular to the wall in such a magnetic presheath scale with the typical ion Larmor orbit size, . The smallest scale length associated with variations parallel to the wall is taken to be across the magnetic field, and ordered , where is assumed. The scale lengths along the magnetic field line are assumed so long that variations associated with this direction are neglected. These orderings are consistent with what we expect close to the divertor target of a tokamak. We allow for a strong component of the electric field in the direction normal to the electron repelling wall, with strong variation in the same direction. The large change of the electric field over an ion Larmor radius distorts the orbit so that it is not circular. We solve for the lowest order orbits by identifying coordinates, which consist of constants of integration, an adiabatic invariant and a gyrophase, associated with periodic ion motion in the system with . By using these new coordinates as variables in the limit , we obtain a generalised ion gyrokinetic equation. We find another quantity that is conserved to first order and use this to simplify the gyrokinetic equation, solving it in the case of a collisionless magnetic presheath. Assuming a Boltzmann response for the electrons, a form of the quasineutrality equation that exploits the change of variables is derived. The gyrokinetic and quasineutrality equations give the ion distribution function and electrostatic potential in the magnetic presheath if the entrance boundary condition is specified.

Numerical simulations of L-mode plasma turbulence in the Scrape Off Layer of MAST are presented. Relevant features of the boundary plasma, such as the thickness of the density layer or the statistics of the fluctuations are related to the edge density and temperature, plasma current and parallel connection length. It is found that the density profile is weakly affected by the edge density, it broadens when the current or the temperature are decreased while the connection length has the opposite effect. The statistics of the turbulence is relatively insensitive to variations of all the edge parameters and show a certain degree of universality. Effective transport coefficients are calculated for several plasma conditions and display a strong nonlinear dependence on the parameters and on the radial variable. Finally, it is shown how the perpendicular particle fluxes in the Scrape Off Layer are related to the edge parameters., Comment: 21 pages, 6 figures

A theoretical framework is developed to clarify the relation between the profiles of density and temperature in the Scrape Off Layer (SOL) with the fluctuations (filaments) that generate them. The framework is based on the dynamics of independent filaments and on their statistical behaviour and can be used to rigorously understand the mechanisms that lead to the non-exponential nature of the radial SOL profiles as well as the increase of the relative fluctuation amplitude in the far SOL. Several models for the dynamics of the filaments, which can be applied to the framework, are derived and discussed for the purpose of identifying how different assumptions lead to the emergence of features in the profiles. It is found that multiple alternative models can explain the observations, thus motivating more stringent and focused experimental analysis. In particular, radially accelerating filaments, less efficient parallel exhaust and also a statistical distribution of the velocity of the filaments can all contribute to induce flatter profiles in the far SOL. A quite general result is the resiliency of the non-exponential nature of the profiles. At the same time, several of the models discussed can also capture the increase of the relative fluctuation amplitude observed in the far SOL. It is also shown that several scenarios are compatible with the broadening of the SOL, which could be caused by charge exchange interactions with neutral particles or by a significant radial acceleration of the filaments., Comment: 28 pages, 5 figures. This is an author-created, un-copyedited version of an article submitted for publication in Plasma Physics and Controlled Fusion. IoP Publishing Ltd is not responsible for any errors or omissions in this version of the manuscript or any version derived from it. arXiv admin note: text overlap with arXiv:1604.08790

Simulations have been carried out to establish how electron thermal physics, introduced in the form of a dynamic electron temperature, affects isolated filament motion and dynamics in 3D. It is found that thermal effects impact filament motion in two major ways when the filament has a significant temperature perturbation compared to its density perturbation: They lead to a strong increase in filament propagation in the bi-normal direction and a significant decrease in net radial propagation. Both effects arise from the temperature dependence of the sheath current which leads to a non-uniform floating potential, with the latter effect supplemented by faster pressure loss. The reduction in radial velocity can only occur when the filament cross-section loses angular symmetry. The behaviour is observed across different filament sizes and suggests that filaments with much larger temperature perturbations than density perturbations are more strongly confined to the near SOL region., Comment: 24 pages, 15 figures. This is an author-created, un-copyedited version of an article submitted for publication in Plasma Physics and Controlled Fusion. IoP Publishing Ltd is not responsible for any errors or omissions in this version of the manuscript or any version derived from it

The nonlinear saturation of the tearing mode is revisited in slab geometry by taking into account higher-order harmonics in the outer solution. The general formalism for tackling this problem in the case of a vanishing current gradient at the resonant surface is derived. It is shown that, although the higher-order harmonics lead to corrections in the final saturation equation, they are of higher order in the perturbation parameter, which provides a formal proof that the standard one-harmonic approach is asymptotically correct., Comment: Accepted to Plasma Physics and Controlled Fusion

Recent theoretical work on magnetic reconnection in hot plasma confinement devices is reviewed. The presentation highlights the common aspects of reconnection phenomena, and current research trends are emphasized. Progress in understanding the dynamics of slowly evolving modes of the tearing family, based on advanced analytic techniques and numerical simulation, as well as of faster modes that lead to internal disruptions, is reported.

The behaviour of the turbulence in magnetic fusion plasmas is discussed in the framework of dimensional analysis. Three main dimensionless parameters are identified, which directly affect the adiabatic response of the plasma and hence control the electron response. They were originally introduced by Scott (1997 Plasma Phys. Control. Fusion 39 1635) and Rogers and Drake (1997 Phys. Rev. Lett. 79 229), and represent the effect of the electromagnetic induction, collisions and electron inertia. The dimensionless parameters are then evaluated at different radial positions in four existing divertor machines (ALCATOR C-MOD, ASDEX-Upgrade, JET and MAST) and in three future experiments (MAST-Upgrade, ITER and DEMO). Clear multi-machine trends, suggesting a quasi-universal behaviour of the plasma, are identified and discussed. Exploiting this quasi-universality, useful insight into the physics of the machines is obtained without complex calculations or expensive numerical simulations. In particular, the trends show a clear separation of the plasma in four distinct regions: core, edge, scrape-off layer and divertor, each of them characterized by a different set of parameters which, in turn, lead to significant variation of the dominant physical mechanisms across the machine.

A numerical analysis of the interaction of resistive drift wave and interchange turbulence with a magnetic island in a two-dimensional slab is presented. The time-scale for the evolution of the island is assumed to be much longer than that for the turbulence, allowing the use of an electrostatic model. The effects of the turbulence are isolated by choosing the parameters such that only even modes are unstable. This makes it possible to compare turbulent states with quiescent states in which turbulence is suppressed by enforcing odd parity. The turbulence is found to reduce the propagation velocity of the island. Its effect is destabilizing for thin islands but becomes stabilizing for islands greater than a few times the Larmor radius. Analysis of the quiescent solutions reveals the possibility of oscillations of the island amplitude and frequency through hysteretic transitions between bistable states.

Numerical simulations of L-mode turbulence in the scrape-off layer (SOL) are used to construct power scaling laws for the characteristic decay lengths of the temperature, density and heat flux at the outer mid-plane. Most of the results obtained are in qualitative agreement with the experimental observations despite the known limitation of the model. Quantitative agreement is also obtained for some exponents. In particular, an almost linear inverse dependence of the heat flux decay length with the plasma current is recovered. The relative simplicity of the theoretical model used allows one to gain insight into the mechanisms determining the width of the SOL.

A complete model of the dynamics of scrape-off layer filaments will be rather complex, including temperature evolution, three dimensional geometry and finite Larmor radius effects. However, the basic mechanism of $\boldsymbol{E}\times\boldsymbol{B}$ advection due to electrostatic potential driven by the diamagnetic current can be captured in a much simpler model; a complete understanding of the physics in the simpler model will then aid interpretation of more complex simulations, by allowing the new effects to be disentangled. Here we consider such a simple model, which assumes cold ions and isothermal electrons and is reduced to two dimensions. We derive the scaling with width and amplitude of the velocity of isolated scrape-off layer filaments, allowing for arbitrary elliptical cross-sections, where previously only circular cross-sections have been considered analytically. We also put the scaling with amplitude in a new and more satisfactory form. The analytical results are extensively validated with two dimensional simulations and also compared, with reasonable agreement, to three dimensional simulations having minimal variation parallel to the magnetic field., Comment: This is an author-created, un-copyedited version of an article accepted for publication/published in Plasma Physics and Controlled Fusion. IOP Publishing Ltd is not responsible for any errors or omissions in this version of the manuscript or any version derived from it. The Version of Record is available online at http://dx.doi.org/10.1088/0741-3335/58/1/014030