Search

This paper studies the effect of 3D magnetic perturbations (MPs) on fast-ion confinement, and its impact on the toroidal rotation velocity profile. Two low collisionality H-mode experiments carried out at the ASDEX Upgrade tokamak have been analysed. The two discharges feature different magnetic field helicity (q95), and differences in the velocity-space and level of fast-ion losses are observed. A new analysis technique has been developed that sheds light on the dependencies between fast-ion losses and toroidal rotation, providing for the first time correlation patterns resolved in radius and velocity space of the lost fast-ions. The correlation intensifies towards the plasma edge and is strongly dependent on the orbit topology of the lost fast-ions. The ASCOT orbit following code has been used to characterize the fast-ion resonant transport and beam driven torques, using the vacuum approach and including plasma response (PR). The change of the toroidal canonical momentum, which serves as figure of merit for resonant fast-ion transport, has been calculated with ASCOT. The beam geometry and q95 are found to have a strong impact on the fast-ion transport and losses. The fast-ion transport induced by the MPs affects the beam driven torques. The effect of the changes of the j × B and collisional torques on plasma rotation is analysed using the torques simulated by ASCOT and simple momentum balance calculations. For the low q95 = 3.8 discharge, which benefits from a resonant amplification, we find excellent agreement with the measured variation of the toroidal velocity. For the high q95 = 5.5 discharge, the inclusion of the PR improves the comparison with experimental data with respect to the vacuum estimation, but still some differences with experiments are observed. This suggests that other non-resonant effects could play a role for the determination of the toroidal rotation profile.

The first stage of a significant enhancement of the ASDEX Upgrade experiment with in-vessel coils for non-axisymmetric magnetic perturbations is now operational. First experiments have shown that ELM mitigation can be achieved using various perturbation field configurations with toroidal mode numbers n = 1, 2, 4. The main access criteria is the plasma edge pedestal density to exceed a threshold, which takes the lowest value of about 60% of the Greenwald density for resonant |n| = 1 perturbations. In H-mode plasmas, mode locking or error field-induced magnetic islands are generally not observed. Due to the low local shear of the plasma magnetic field in the vicinity of the perturbation coils around the outboard midplane, the magnetic perturbation is resonant simultaneously on several rational surfaces. It is hypothesised that the existence of image currents on these surfaces ensures good shielding of the error field in the confined plasma.

Previous work carried out in the ASDEX Upgrade tokamak on the role of the edge radial electric field and ion heat flux in the L-H transition physics in deuterium plasmas has been extended in hydrogen plasmas. Similar discharges were performed in the two gases providing a detailed comparison of the edge kinetic profiles and heat fluxes in L-mode up to the L-H transition, as the heating power is increased. At the L-H transition, the edge ion heat flux just inside the separatrix is about two times higher in hydrogen than in deuterium. However, the ion plasma parameters at the plasma edge, T i and , as well as the radial electric field well, are found to be very similar in the two gases. The transport analysis based on this data reveals that, at the L-H transition, the ion heat transport at the plasma edge is about two times higher in hydrogen than in deuterium, in agreement with the well-known isotope effect of confinement and transport. This indicates that the higher power threshold in hydrogen is mainly due to the higher ion heat transport in this gas.

We measure the deuterium density, the parallel drift velocity, and parallel and perpendicular temperatures (T∥, T⊥) in non-Maxwellian plasmas at ASDEX Upgrade. This is done by taking moments of the ion velocity distribution function measured by tomographic inversion of five simultaneously acquired spectra of Dα-light. Alternatively, we fit the spectra using a bi-Maxwellian distribution function. The measured kinetic temperatures (T∥ = 9 keV, T⊥ = 11 keV) reveal the anisotropy of the plasma and are substantially higher than the measured boron temperature (7 keV). The Maxwellian deuterium temperature computed with TRANSP (6 keV) is not uniquely measurable due to the fast ions. Nevertheless, simulated kinetic temperatures accounting for fast ions based on TRANSP (T∥= 8.3 keV, T⊥ = 10.4 keV) are in excellent agreement with the measurements. Similarly, the Maxwellian deuterium drift velocity computed with TRANSP (300 km s-1) is not uniquely measurable, but the simulated kinetic drift velocity accounting for fast ions agrees with the measurements (400 km s-1) and is substantially larger than the measured boron drift velocity (270 km s-1). We further find that ion cyclotron resonance heating elevates T∥ and T⊥ each by 2 keV without evidence for preferential heating in the Dα spectra. Lastly, we derive an expression for the 1D projection of an arbitrarily drifting bi-Maxwellian onto a diagnostic line-of-sight.

Tokamak scenario development requires an understanding of the properties that determine the kinetic profiles in non-steady plasma phases and of the self-consistent evolution of the magnetic equilibrium. Current ramps are of particular interest since many transport-relevant parameters explore a large range of values and their impact on transport mechanisms has to be assessed. To this purpose, a novel full-discharge modelling tool has been developed, which couples the transport code ASTRA (Pereverzev et al 1991 IPP Report 5/42) and the free boundary equilibrium code SPIDER (Ivanov et al 2005 32nd EPS Conf. on Plasma Physics vol 29C (ECA) P-5.063 and http://epsppd.epfl.ch/Tarragona/pdf/P5_063.pdf), utilizing a specifically designed coupling scheme. The current ramp-up phase can be accurately and reliably simulated using this scheme, where a plasma shape, position and current controller is applied, which mimics the one of ASDEX Upgrade. Transport of energy is provided by theory-based models (e.g. TGLF (Staebler et al 2007 Phys. Plasmas 14 055909)). A recipe based on edge-relevant parameters (Scott 2000 Phys. Plasmas 7 1845) is proposed to resolve the low current phase of the current ramps, where the impact of the safety factor on micro-instabilities could make quasi-linear approaches questionable in the plasma outer region. Current ramp scenarios, selected from ASDEX Upgrade discharges, are then simulated to validate both the coupling with the free-boundary evolution and the prediction of profiles. Analysis of the underlying transport mechanisms is presented, to clarify the possible physics origin of the observed L-mode empirical energy confinement scaling. The role of toroidal micro-instabilities (ITG, TEM) and of non-linear effects is discussed.

The installation of a new poloidal charge exchange recombination spectroscopy (CXRS) diagnostic at ASDEX Upgrade (AUG) has enabled the determination of the radial electric field, Er, using the radial force balance of impurity ions. Er has been derived from charge exchange (CX) spectra measured on different impurity species, such as He2+, B5+, C6+ and Ne10+. The resulting Er profiles are found to be identical within the uncertainties regardless of the impurity species used, thus, demonstrating the validity of the diagnostic technique. The Er profile has been compared to the main ion pressure gradient term, which is found to be the dominant contribution at the plasma edge, thus, supporting that the Er well is created by the main ion species. The Er profile has been measured in different confinement regimes including L-, I- and H-mode. The depth of the Er well and the magnitude of the Er shear are correlated with the ion pressure at the pedestal top. The temporal evolution of the measured CX profiles and the resulting Er have been studied during an edge-localized mode (ELM) cycle. At the ELM crash, the Er minimum is less deep resulting in a reduction of the E???B shear. Within 2?ms after the ELM crash, the edge kinetic profiles have nearly recovered and the Er well is observed to recover simultaneously. In high density type-I ELM mitigated H-mode plasmas, obtained via externally applied magnetic perturbations (MPs) with toroidal mode number n?=?2, no clear effect on Er due to the MPs has been observed.

The results from ASDEX Upgrade discharges dedicated specifically to the investigation of low-density L-to-H transitions are presented. The plasmas were heated by electron cyclotron resonance heating to achieve a separation of electron and ion heat channels. Under such conditions, the ratio of electron to ion temperature at the plasma edge increases with decreasing density at the L–H transition and can be as high as 3.5. Our results strongly support the essential role of the ion channel in the L–H transition, via the diamagnetic E r provided by the ion pressure gradient.

Core toroidal rotation behavior and momentum transport have been examined in neutral beam injection (NBI) heated plasmas with and without electron cyclotron resonance heating (ECRH) in ASDEX Upgrade (AUG). The impurity ion temperature and rotation are measured by means of charge exchange recombination spectroscopy (CXRS) and the main ion rotation is calculated neoclassically based on the measured impurity ion temperature and rotation profiles. In purely NBI heated discharges the plasma spins up in the co-current direction and forms peaked rotation profiles. However, when ECRH power is added to these discharges the rotation decreases significantly leading to flat and occasionally slightly hollow profiles. The rotation at the edge of the plasma (ρ > 0.65) is largely unaffected by the application of the ECRH. During ECRH phases the electron temperature in the core increases dramatically concomitant with a flattening of the ion temperature profile. In these phases peaking of the impurity ion and electron density profiles is also observed. The change in toroidal rotation is sensitive to both the amount of ECRH power deposited as well as to the deposition location. The measured rotation changes cannot be explained by a modification to the NBI torque deposition profile, a preferential loss of fast ions from the plasma core, or by a simple increase in momentum diffusivity. In addition, the inward directed Coriolis momentum pinch is predicted to increase, not decrease, during the ECRH phases. Altogether, the data suggest the presence of either an outward convection of momentum or an intrinsic, counter-current directed torque. Although the physics behind these possibilities remains unclear, they are presumably related to either a convective or residual stress driven momentum flux, which responds to the ECRH-induced changes in the plasma profiles and turbulence.

A fast-ion D-alpha (FIDA) diagnostic has been developed for the fully tungsten coated ASDEX Upgrade (AUG) tokamak using 25 toroidally viewing lines of sight and featuring a temporal resolution of 10 ms. The diagnostic’s toroidal geometry determines a well-defined region in velocity space which significantly overlaps with the typical fast-ion distribution in AUG plasmas. Background subtraction without beam modulation is possible because relevant parts of the FIDA spectra are free from impurity line contamination. Thus, the temporal evolution of the confined fast-ion distribution function can be monitored continuously. FIDA profiles during on- and off-axis neutral beam injection (NBI) heating are presented which show changes in the radial fast-ion distribution with the different NBI geometries. Good agreement has been obtained between measured and simulated FIDA radial profiles in MHD-quiescent plasmas using fast-ion distribution functions provided by TRANSP. In addition, a large fast-ion redistribution with a drop of about 50% in the central fast-ion population has been observed in the presence of a q = 2 sawtooth-like crash, demonstrating the capabilities of the diagnostic.

The ASDEX Upgrade tokamak is currently being enhanced with a set of in-vessel saddle coils for non-axisymmetric perturbations aiming at mitigation or suppression of edge localized modes (ELMs). Results obtained during the first experimental campaign are reported. With n = 2 magnetic perturbations, it is observed that type-I ELMs can be replaced by benign small ELM activity with strongly reduced energy loss from the confined plasma and power load to the divertor. During these phases with ELM mitigation, no density reduction (density 'pump-out') is observed. ELM mitigation has, so far, been observed in plasmas with different shape and different heating mixes and, therefore, different momentum input. The ELM mitigation regime can be accessed with resonant and non-resonant perturbation field configurations. The main threshold requirement appears to be a critical minimum plasma edge density which depends on plasma current. So far it is not possible to distinguish whether this is an edge collisionality threshold or a critical fraction of the Greenwald density limit.

First experiments with nonaxisymmetric magnetic perturbations, toroidal mode number n=2, produced by newly installed in-vessel saddle coils in the ASDEX Upgrade tokamak show significant reduction of plasma energy loss and peak divertor power load associated with type-I edge localized modes (ELMs) in high-confinement mode plasmas. ELM mitigation is observed above an edge density threshold and is obtained both with magnetic perturbations that are resonant and not resonant with the edge safety factor profile. Compared with unperturbed type-I ELMy reference plasmas, plasmas with mitigated ELMs show similar confinement, similar plasma density, and lower tungsten impurity concentration.

The influence of pure electron heating versus combined electron and ion heating on a high collisionality H-mode plasma has been investigated in ASDEX Upgrade. This was done by replacing both neutral beam injection (NBI) and ion cyclotron resonance frequency (ICRF) heating in small steps with electron cyclotron resonance heating (ECRH) while keeping the total heating power constant. The stability of the global plasma parameters and the response of the kinetic profiles on the changed heating mix and torque input is shown. The differences in edge localized mode behaviour between the different heating mixes is reported. The data were analysed with an interpretative transport model and linear gyrokinetic simulations were performed to evaluate the underlying transport mechanism.

Core momentum and particle transport in ASDEX Upgrade (AUG) have been examined in a wide variety of plasma discharges and via several different methods. Experiments were performed in which ECRH power was added to NBI heated H-modes causing the electron and impurity ion density profiles to peak and the core toroidal rotation to flatten. Turbulence calculations of these plasmas show a change in the dominant regime from ITG to TEM due to the ECRH induced changes in the electron and ion temperature profiles. The impurity and electron density behavior can be fully explained by the changes in the turbulent particle transport. Momentum transport analyses demonstrate that in the TEM regime there is a core localized, counter-current directed, residual stress momentum flux of the same order of magnitude as the applied NBI torque. The initial results from momentum modulation experiments performed on AUG confirm that the Prandtl number in AUG NBI heated H-modes is close to 1 and that there exists a significant inward momentum pinch. Lastly, an intrinsic toroidal rotation database has been developed at AUG which can be used to test theoretically predicted dependences of residual stress momentum fluxes. Initial results show a linear correlation between the gradient of the toroidal rotation and both the electron density gradient scale length and the frequency of the dominant turbulent mode.

Local gyrokinetic calculations of the logarithmic gradients at mid-radius of both electron and boron densities in ASDEX Upgrade H-mode plasmas are presented and compared with the experimental observations. The experimental results show that both the electron and the boron density profiles increase their peaking in response to the addition of central electron cyclotron heating over a background of neutral beam injection (NBI) heating. The boron density profiles are always less peaked than the electron density profiles in the confinement region, and are flat or even slightly hollow in the presence of NBI heating only. The experimental behaviours are well reproduced by the theoretical predictions. The agreement allows the identification, through theoretical modelling, of the transport mechanisms responsible for the observed dependences. In particular, the observed increase in the logarithmic electron density gradient with increasing central electron heating is explained by a concurrent reduction of the outward pure convection and an increase in the inward thermodiffusion. In addition, it is found that the plasma toroidal rotation velocity and its radial gradient play a non-negligible role in the turbulent boron transport, and allow the prediction of a decrease in boron peaking with increasing rotation velocity, which is consistent with the experimental observations.

This paper describes joint ITPA studies of the I-mode regime, which features an edge thermal barrier together with L-mode-like particle and impurity transport and no edge localized modes (ELMs). The regime has been demonstrated on the Alcator C-Mod, ASDEX Upgrade and DIII-D tokamaks, over a wide range of device parameters and pedestal conditions. Dimensionless parameters at the pedestal show overlap across devices and extend to low collisionality. When they are matched, pedestal temperature profiles are also similar. Pedestals are stable to peeling-ballooning modes, consistent with lack of ELMs. Access to I-mode is independent of heating method (neutral beam injection, ion cyclotron and/ or electron cyclotron resonance heating). Normalized energy confinement H-98,H-y2 >= 1 has been achieved for a range of 3 5 T. Issues for extrapolation to ITER and other future fusion devices are discussed.

Momentum transport and plasma rotation have been studied extensively on many tokamaks in recent years. Both experiments and theory have shown that sheared plasma rotation can stabilise turbulence while the rotation itself has beneficial effects on MHD instabilities, such as resistive wall modes. Future tokamaks, like ITER, will have larger moments of inertia and, therefore, less NBI torque per volume and less deep deposition profiles (NBI torque 'dominates' in present large tokamaks over other torque sources). Therefore, intrinsic torque will play a more important role in determining rotation for example in ITER. Numerous experimental results, both on momentum transport and intrinsic torque, have been reported on individual devices - yet only few experimental multi-machine momentum transport or intrinsic torque comparisons have been performed. This paper reports dedicated scans to study momentum transport and intrinsic torque, carried out as dedicated joint experiments on JET, DIII-D, AUG, NSTX and C-Mod within the ITPA framework.

Direction reversals of intrinsic toroidal rotation have been observed in Alcator C-Mod ohmic L-mode plasmas following modest electron density or toroidal magnetic field ramps. The reversal process occurs in the plasma interior, inside of the q = 3/2 surface. For low density plasmas, the rotation is in the co-current direction, and can reverse to the counter-current direction following an increase in the electron density above a certain threshold. Reversals from the co- to counter-current direction are correlated with a sharp decrease in density fluctuations with k(R) >= 2 cm(-1) and with frequencies above 70 kHz. The density at which the rotation reverses increases linearly with plasma current, and decreases with increasing magnetic field. There is a strong correlation between the reversal density and the density at which the global ohmic L-mode energy confinement changes from the linear to the saturated regime.