Search

Starting from recent JET experimental results that show a significant reduction of ion stiffness in the plasma core region due to plasma rotation in the presence of low magnetic shear, an experiment was carried out at JET in order to separate the role of rotation and rotation gradient in mitigating the ion stiffness level. Enhanced toroidal field ripple (up to 1.5%) and external resonant magnetic fields are the two mechanisms used to try and decouple the rotation value from its gradient. In addition, shots with reversed toroidal field and plasma current, yielding counter-current neutral beam injection, were compared with standard co-injection cases. These tools also allowed varying the rotation independently of the injected power. Shots with high rotation gradient are found to maintain their low stiffness level even when the absolute value of the rotation was significantly reduced. Conversely, high but flat rotation yields much less peaked ion temperature profiles than a peaked rotation profile with lower values. This behaviour suggests the rotation gradient as the main player in reducing the ion stiffness level. In addition, it is found that inverting the rotation gradient sign does not suppress its effect on ion stiffness.

A broad survey of the experimental database of neutral beam heated baseline H-modes and hybrid scenarios in the JET tokamak has established the ubiquity of non-diffusive momentum transport mechanisms in rotating plasmas. As a result of their presence, the normalized angular frequency gradient R∇ω/ω is higher than expected from momentum diffusion alone, by about unity in the core (r/a ∼ 0.3), rising to near 5 close to the edge, where its contribution to the total gradient is comparable to the gradient associated with the diffusive flux. The magnitude and parameter dependences of the non-diffusive contribution to the gradient are consistent with a theoretically expected pinch, which has its origin in the vertical particle drift resulting from the Coriolis force. Linear gyrokinetic calculations of the pinch number RV/χ φ and the Prandtl number χ φ /χ i are in good agreement with the experimental observations, with similar dependences on R/L n , q and ε = r/R. A contribution due to residual stresses may also be present, but could not be identified with certainty.

A broad survey of the experimental database of neutral beam heated plasmas in the JET tokamak has established the theoretically expected ubiquity, in rotating plasmas, of a convective transport mechanism which has its origin in the vertical particle drift resulting from the Coriolis force. This inward convection, or pinch, leads to inward transport of toroidal angular momentum and is characterized by pinch numbers RV/χϕ, which rise from near unity at r/a ≈ 0.25 to around 5 at r/a ≈ 0.85. Linear gyrokinetic calculations of the Coriolis pinch number and the Prandtl number χϕ/χi are in good agreement with the experimental observations, with similar dependences on plasma parameters. The data, however, do not rule out contributions from different processes, such as residual stresses.

The estimate of coefficients of the Convection-Diffusion Equation (CDE) from experimental measurements belongs in the category of inverse problems, which are known to come with issues of ill-conditioning or singularity. Here we concentrate on a particular class that can be reduced to a linear algebraic problem, with explicit solution. Ill-conditioning of the problem corresponds to the vanishing of one eigenvalue of the matrix to be inverted. The comparison with algorithms based upon matching experimental data against numerical integration of the CDE sheds light on the accuracy of the parameter estimation procedures, and suggests a path for a more precise assessment of the profiles and of the related uncertainty. Several instances of the implementation of the algorithm to real data are presented., Extended version of an invited talk presented at the 2012 EPS Conference. To appear in Plasma Physics and Controlled Fusion

The paper compares the transport properties of a set of dimensionless identity experiments performed between JET and JT-60U in the advanced tokamak regime with internal transport barrier, ITB. These International Tokamak Physics Activity, ITPA, joint experiments were carried out with the same plasma shape, toroidal magnetic field ripple and dimensionless profiles as close as possible during the ITB triggering phase in terms of safety factor, normalized Larmor radius, normalized collision frequency, thermal beta, ratio of ion to electron temperatures. Similarities in the ITB triggering mechanisms and sustainment were observed when a good match was achieved of the most relevant normalized profiles except the toroidal Mach number. Similar thermal ion transport levels in the two devices have been measured in either monotonic or non-monotonic q-profiles. In contrast, differences between JET and JT-60U were observed on the electron thermal and particle confinement in reversed magnetic shear configurations. It was found that the larger shear reversal in the very centre (inside normalized radius of 0.2) of JT-60U plasmas allowed the sustainment of stronger electron density ITBs compared with JET. As a consequence of peaked density profile, the core bootstrap current density is more than five times higher in JT-60U compared with JET. Thanks to the bootstrap effect and the slightly broader neutral beam deposition, reversed magnetic shear configurations are self-sustained in JT-60U scenarios. Analyses of similarities and differences between the two devices address key questions on the validity of the usual assumptions made in ITER steady scenario modelling, e.g. a flat density profile in the core with thermal transport barrier? Such assumptions have consequences on the prediction of fusion performance, bootstrap current and on the sustainment of the scenario.

Detailed experimental studies of ion heat transport have been carried out in JET exploiting the upgrade of active charge exchange spectroscopy and the availability of multi-frequency ion cyclotron resonance heating with 3He minority. The determination of ion temperature gradient (ITG) threshold and ion stiffness offers unique opportunities for validation of the well-established theory of ITG driven modes. Ion stiffness is observed to decrease strongly in the presence of toroidal rotation when the magnetic shear is sufficiently low. This effect is dominant with respect to the well-known ω E×B threshold up-shift and plays a major role in enhancing core confinement in hybrid regimes and ion internal transport barriers. The effects of T e/T i and s/q on ion threshold are found rather weak in the domain explored. Quasi-linear fluid/gyro-fluid and linear/non-linear gyro-kinetic simulations have been carried out. Whilst threshold predictions show good match with experimental observations, some significant discrepancies are found on the stiffness behaviour.

Experiments at DIII-D investigated the effects of magnetic error fields similar to those expected from proposed ITER test blanket modules (TBMs) containing ferromagnetic material. Studied were effects on: plasma rotation and locking, confinement, L–H transition, the H-mode pedestal, edge localized modes (ELMs) and ELM suppression by resonant magnetic perturbations, energetic particle losses, and more. The experiments used a purpose-built three-coil mock-up of two magnetized ITER TBMs in one ITER equatorial port. The largest effect was a reduction in plasma toroidal rotation velocity v across the entire radial profile by as much as Δv/v ∼ 60% via non-resonant braking. Changes to global Δn/n, Δβ/β and ΔH98/H98 were ∼3 times smaller. These effects are stronger at higher β. Other effects were smaller. The TBM field increased sensitivity to locking by an applied known n = 1 test field in both L- and H-mode plasmas. Locked mode tolerance was completely restored in L-mode by re-adjusting the DIII-D n = 1 error field compensation system. Numerical modelling by IPEC reproduces the rotation braking and locking semi-quantitatively, and identifies plasma amplification of a few n = 1 Fourier harmonics as the main cause of braking. IPEC predicts that TBM braking in H-mode may be reduced by n = 1 control. Although extrapolation from DIII-D to ITER is still an open issue, these experiments suggest that a TBM-like error field will produce only a few potentially troublesome problems, and that they might be made acceptably small.

Accurate and validated tools for calculating toroidal momentum sources are necessary to make reliable predictions of toroidal rotation for current and future experiments. In this work we present the first experimental validation of torque profile calculation from neutral beam injection (NBI) under toroidal field ripple. We use discharges from a dedicated experimental session on JET where neutral beam modulation technique is used together with time-dependent torque calculations from ASCOT code for making the benchmark. Good agreement between simulations and experimental results is found.

The scan of ion cyclotron resonant heating (ICRH) power has been used to systematically study the pump out effect of central electron heating on impurities such as Ni and Mo in H-mode low collisionality discharges in JET. The transport parameters of Ni and Mo have been measured by introducing a transient perturbation on their densities via the laser blow off technique. Without ICRH Ni and Mo density profiles are typically peaked. The application of ICRH induces on Ni and Mo in the plasma centre (at normalized poloidal flux ρ = 0.2) an outward drift approximately proportional to the amount of injected power. Above a threshold of ICRH power of about 3 MW in the specific case the radial flow of Ni and Mo changes from inwards to outwards and the impurity profiles, extrapolated to stationary conditions, become hollow. At mid-radius the impurity profiles become flat or only slightly hollow. In the plasma centre the variation of the convection-to-diffusivity ratio v/D of Ni is particularly well correlated with the change in the ion temperature gradient in qualitative agreement with the neoclassical theory. However, the experimental radial velocity is larger than the neoclassical one by up to one order of magnitude. Gyrokinetic simulations of the radial impurity fluxes induced by electrostatic turbulence do not foresee a flow reversal in the analysed discharges.

Several parametric scans have been performed to study momentum transport on JET. A neutral beam injection modulation technique has been applied to separate the diffusive and convective momentum transport terms. The magnitude of the inward momentum pinch depends strongly on the inverse density gradient length, with an experimental scaling for the pinch number being -Rv pinch/χϕ = 1.2R/L n + 1.4. There is no dependence of the pinch number on collisionality, whereas the pinch seems to depend weakly on q-profile, the pinch number decreasing with increasing q. The Prandtl number was not found to depend either on R/L n, collisionality or on q. The gyro-kinetic simulations show qualitatively similar dependence of the pinch number on R/L n, but the dependence is weaker in the simulations. Gyro-kinetic simulations do not find any clear parametric dependence in the Prandtl number, in agreement with experiments, but the experimental values are larger than the simulated ones, in particular in L-mode plasmas. The extrapolation of these results to ITER illustrates that at large enough R/L n > 2 the pinch number becomes large enough (>3–4) to make the rotation profile peaked, provided that the edge rotation is non-zero. And this rotation peaking can be achieved with small or even with no core torque source. The absolute value of the core rotation is still very challenging to predict partly due to the lack of the present knowledge of the rotation at the plasma edge, partly due to insufficient understanding of 3D effects like braking and partly due to the uncertainties in the extrapolation of the present momentum transport results to a larger device.

In this study, enhancement of the toroidal field (TF) ripple has been used as a tool in order to reveal the impact of the momentum pinch on the rotation profiles in H-mode JET discharges. The analysis showed that flatter rotation profiles were obtained in discharges with a high TF ripple, attributed to a smaller inward momentum convection. An average inward momentum pinch of approximately Vp ≈ 3.4 m s−1 and a normalized pinch value of RVp/χ ≈ 6.6 could explain the observation. The data show that the momentum at the edge affects the peaking of the rotation and momentum density profiles. Under the assumption that the heat and momentum diffusivities are equal, an estimate of the levels of the momentum pinch in all discharges in the JET rotation database was made. For H-mode discharge these ranged from 0.3 m s−1 < Vp < 17 m s−1, with 2 < RVp/χ < 10. A larger momentum pinch was found in discharges with a smaller density profile gradient length, i.e. a more peaked density profile.

___

Experiments have been carried out on the Joint European Torus tokamak to determine the diffusive and convective momentum transport. Torque, injected by neutral beams, was modulated to create a periodic perturbation in the toroidal rotation velocity. Novel transport analysis shows the magnitude and profile shape of the momentum diffusivity are similar to those of the ion heat diffusivity. A significant inward momentum pinch, up to 20 m/s, has been found. Both results are consistent with gyrokinetic simulations. This evidence is complemented in plasmas with internal transport barriers.

Experiments were carried out in the JET tokamak to determine the critical ion temperature inverse gradient length (R/L-Ti = R vertical bar del T-i vertical bar/T-i) for the onset of ion temperature gradient modes and the stiffness of Ti profiles with respect to deviations from the critical value. Threshold and stiffness have been compared with linear and nonlinear predictions of the gyrokinetic code GS2. Plasmas with higher values of toroidal rotation show a significant increase in R/L-Ti, which is found to be mainly due to a decrease of the stiffness level. This finding has implications on the extrapolation to future machines of present day results on the role of rotation on confinement.

A dedicated campaign has been run on JET to study the effect of toroidal field (TF) ripple on plasma performance. Radial electric field measurements from experiments on a series of plasmas with internal transport barriers (ITBs) and different levels of ripple amplitude are presented. They have been calculated from charge exchange measurements of impurity ion temperature, density and rotation velocity profiles, using the force balance equation. The ion temperature and the toroidal and poloidal rotation velocities are compared in plasmas with both reversed and optimized magnetic shear profiles. Poloidal rotation velocity (v(theta)) in the ITB region is measured to be of the order of a few tens of km s(-1), significantly larger than the neoclassical predictions. Increasing levels of the TF ripple are found to decrease the ion temperature gradient in the ITB region, a measure for the quality of the ITB, and the maximum value of v(theta) is reduced. The poloidal rotation term dominates in the calculations of the total radial electric field (E-r), with the largest gradient in E-r measured in the radial region coinciding with the ITB.

A new technique has been developed to produce plasmas with improved confinement relative to the H98,y2 scaling law (ITER Physics Expert Groups on Confinement and Transport and Confinement Modelling and Database ITER Physics Basics Editors and ITER EDA 1999 Nucl. Fusion 39 2175) on the JET tokamak. In the mid-size tokamaks ASDEX upgrade and DIII-D heating during the current formation is used to produce a flat q-profile with a minimum close to 1. On JET this technique leads to q-profiles with similar minimum q but opposite to the other tokamaks not to an improved confinement state. By changing the method utilizing a faster current ramp with temporary higher current than in the flattop (current overshoot) plasmas with improved confinement (H98,y2 = 1.35) and good stability (βN ≈ 3) have been produced and extended to many confinement times only limited by technical constraints. The increase in H98,y2-factor is stronger with more heating power as can be seen in a power scan. The q-profile development during the high power phase in JET is reproduced by current diffusion calculated by TRANSP and CRONOS. Therefore the modifications produced by the current overshoot disappear quickly from the edge but the confinement improvement lasts longer, in some cases up to the end of the heating phase.

In this work we apply the neutral beam modulation technique to a set of JET discharges in order to obtain information on the radial transport of the toroidal angular momentum. This perturbative technique allows one to separate possible pinch terms from the momentum diffusivity, otherwise impossible to decouple with a steady-state momentum balance analysis alone. In this paper a model based on the ion temperature gradient turbulence, which consists of a constant and uniform Prandtl number of the order unity combined with an inward Coriolis pinch term predicted by the gyro-kinetic theory, is validated against data from the JET experiment. The machine size and the good space and time resolution of the charge exchange system make JET particularly suitable for this study. The experimental data are matched with good accuracy, both the steady-state and the transient propagation of the momentum perturbation, choosing a Prandtl number slightly above one, provided the Coriolis pinch is included. Sensitivity studies assess the accuracy range of our predictive modelling.

An extensive database to study the scaling of rotation and momentum transport has been constructed at JET. The database contains information from various operational scenarios, amongst them H-mode discharges, and parameters that characterize the rotation, as well as those that describe the general plasma conditions. JET plasmas are predominantly heated by neutral beam injection which is also the main source for the observed toroidal rotation. Dimensionless Mach numbers are introduced to quantify rotation. The scaling of plasma rotation and the Mach numbers in particular has been studied. The thermal and Alfvén Mach numbers were found to scale inversely withqand with the ratio of torque and additional heating power. Although the momentum and energy confinement times were found to be of the same magnitude, the ratio was found to vary. Regression analyses showed a dependence of both the energy and momentum confinement times on plasma rotation. If rotation was included in the scaling model of energy and momentum confinement the quality of the fits substantially improved. Detailed analysis of the core and edge (pedestal) confinement showed that momentum confinement was improved in the core of the plasma compared with the energy confinement. However, the pedestal proved to be less confining for the momentum than for the energy.

The effect of a toroidal field (TF) ripple on the formation and performance of internal transport barriers (ITBs) has been studied in JET. It was found that the TF ripple had a profound effect on the toroidal plasma rotation. An increased TF ripple up to δ = 1% led to a lower rotation and reduced the rotational shear in the region where the ITBs were formed. ITB triggering events were observed in all cases and it is thought that the rotational shear may be less important for this process than, for example, the q-profile. However, the increase in the pressure gradient following the ITB trigger was reduced in discharges with a larger TF ripple and consequently a lower rotational shear. This suggests that toroidal rotation and its shear play a role in the growth of the ITB once it has been triggered.

One of the critical issues for ITER is access to an H-mode regime with good confinement, H98 = 1. The most basic scaling laws for power threshold for the L–H transition, Pth, take the variation with plasma density, magnetic field and plasma size into account. However, the large variations in the Pth data from the values estimated with such simple scaling laws indicate other underlying dependencies. Another important consideration for ITER is that H-modes with higher values of energy confinement factors are often obtained with input power values much greater than Pth. This paper presents results from recent studies on JET to assess possible hidden variables for H-mode access over a wide range of plasma conditions. Experimental results demonstrate that sensitivity to the magnetic shaping and divertor geometry could account for some of the scatter in the international power threshold database. Hysteresis in the L–H transition Pth has been studied in detail for the first time on JET by comparing values of Pth at the forward and back H-mode transitions over a range of densities. The impact of the edge plasma rotation on H-mode access has also been considered on JET with a toroidal field ripple scan across the L–H and H–L transitions. Finally, the total input power required relative to the measured value of Pth for access to a steady-state H-mode with H98 = 1 has been examined for a highly shaped magnetic configuration. The implications of these results for the attainment of H-mode with good confinement on ITER are discussed.

Currently, plasmas with internal transport barriers (ITBs) appear the most likely candidates for steady-state scenarios for future fusion reactors. In such plasmas, the broad hot and dense region in the plasma core leads to high fusion gain, while the cool edge protects the integrity of the first wall. Economically desirable large bootstrap current fraction and low inductive current drive may, however, lead to degraded fast ion confinement. In this work the confinement and heating profile of fusion alphas were compared between H-mode and ITB plasmas in realistic JET geometry. The work was carried out using the Monte Carlo-based guiding-center-following code ASCOT.For the same plasma current, the ITB discharges were found to produce four to eight times more fusion power than a comparable ELMy H-mode discharge. Unfortunately, also the alpha particle losses were larger (~16%) compared with the H-mode discharge (7%). In the H-mode discharges, alpha power was deposited to the plasma symmetrically around the magnetic axis, whereas in the current-hole discharge, the power was spread out to a larger volume in the plasma center. This was due to wider particle orbits, and the magnetic structure allowing for a broader hot region in the centre.

New results on electron heat wave propagation using ion cyclotron resonance heating power modulation in the Joint European Torus (JET) [P. H. Rebut , Nucl. Fusion 25, 1011 (1985)] plasmas characterized by internal transport barriers (ITBs) are presented. The heat wave generated outside the ITB, and traveling across it, always experiences a strong damping in the ITB layer, demonstrating a low level of transport and loss of stiffness. In some cases, however, the heat wave is strongly inflated in the region just outside the ITB, showing features of convective-like behavior. In other cases, a second maximum in the perturbation amplitude is generated close to the ITB foot. Such peculiar types of behavior can be explained on the basis of the existence of a critical temperature gradient length for the onset of turbulent transport. Convective-like features appear close to the threshold (i.e., just outside the ITB foot) when the value of the threshold is sufficiently high, with a good match with the theoretical predictions for the trapped electron mode threshold. The appearance of a second maximum is due to the oscillation of the temperature profile across the threshold in the case of a weak ITB. Simulations with an empirical critical gradient length model and with the theory based GLF23 [R. E. Waltz , Phys. Plasmas, 4, 2482 (1997)] model are presented. The difference with respect to previous results of cold pulse propagation across JET ITBs is also discussed. (C) 2007 American Institute of Physics.

In the 2006 experimental campaign, progress has been made on JET to operate non-inductive scenarios at higher applied powers (31 MW) and density (n(1) similar to 4 x 10(19) m(-3)), with ITER-relevant safety factor (q(95) similar to 5) and plasma shaping, taking advantage of the new divertor capabilities. The extrapolation of the performance using transport modelling benchmarked on the experimental database indicates that the foreseen power upgrade (similar to 45 MW) will allow the development of non-inductive scenarios where the bootstrap current is maximized together with the fusion yield and not, as in present-day experiments, at its expense. The tools for the long-term JET programme are the new ITER-like ICRH antenna (similar to 15 MW), an upgrade of the NB power (35 MW/20s or 17.5 MW/40s), a new ITER-like first wall, a new pellet injector for edge localized mode control together with improved diagnostic and control capability. Operation with the new wall will set new constraints on non-inductive scenarios that are already addressed experimentally and in the modelling. The fusion performance and driven current that could be reached at high density and power have been estimated using either 0D or 1-1/2D validated transport models. In the high power case (45 MW), the calculations indicate the potential for the operational space of the non-inductive regime to be extended in terms of current (similar to 2.5 MA) and density (n(1) > 5 x 10(19) m(-3)), with high beta(N) (beta(N) > 3.0) and a fraction of the bootstrap current within 60-70% at high toroidal field (similar to 3.5 T).

Plasma current density simulations of JET and JT-60U shots in reverse-q advanced scenarios based on the previous data analysis of the identity plasma experiments have been performed. The effects of the main differences between the shots (neutral beam current density (jnbi), electron density and geometry) have been studied. The reversed q-profile (which was the target in this identity experiment and is observed at the beginning of each shot) was sustained in JT-60U while it became flat in JET towards the end of the shot. In JET, jnbi is peaked on-axis whereas in JT-60U it is peaked off-axis (at ρ = 0.5) while the NBI fraction of the selected shots is the same (22-24%). A strong density ITB appeared (at ρ = 0.5) in JT-60U but not in JET. The plasma geometry was mainly set to match but it was not identical. In addition, the extrapolation to JET steady-state operation has been done by testing the sensitivity of q to different externally driven currents, electron density and geometry in predictive current diffusion simulations. Moreover, critical bootstrap current density has been analysed. The reasons for the different time evolution of q-profile have been studied with predictive current simulations with the 1.5D transport code JETTO. The current diffusion model was validated against reverse-q shots, and simulations were performed with experimental data profiles, jNBI given by ASCOT and neoclassical resistivity and bootstrap current calculated by NCLASS. Bootstrap current density was the most efficient way to sustain the beneficial reverse shape of q-profile. Replacing the JET ne profile with one from JT-60U leads to an increase of 0.2-0.3 MA in the bootstrap current (f bs increases from 15% to 30%). However, sustaining the stationary reverse q is not achieved in JET with bootstrap current induced by the density gradient of JT-60U. Even 10 times larger gradient than in JT-60U helps to sustain the shape of the q-profile longer than the experimentally observed density profile in JET but the minimum value of q moves closer to central plasma and the shape of q is not stationary in a 10 s simulation. The effect of different shape of NBI current density profile is negligible due to quite small fraction. Sustaining the reverse q requires 45% or larger added off-axis fraction. However, increasing the inverse aspect ratio increases bootstrap current and decreases the critical bootstrap current more effectively than increasing the density gradient: two times larger inverse aspect ratio produces almost three times larger bootstrap fraction but ten times larger density gradient only two times larger bootstrap fraction. The conclusions based on these simulations indicate that the need for bootstrap current is larger in JET and the same conditions cause smaller bootstrap fraction, which suggests that achieving steady-state operation in JET under these conditions is unlikely.

The simultaneous real-time control of the current and pressure profiles could lead to the steady state sustainment of an internal transport barrier (ITB), region where heat and particle transport are strongly reduced and so to a stationary optimized plasma regime. Recent experiments in JET have demonstrated significant progress in achieving such a control: different current and temperature gradient target profiles have been reached and sustained for several seconds using a controller based on a static linear model identification. Nevertheless, these experiments have shown that the controller was sensitive to rapid plasma events such as transient ITBs during the safety factor profile evolution or Magneto-Hydrodynamics (MHD) instabilities which modify the pressure profiles on the confinement time scale. The control technique has been improved by using a multiple-time-scale approximation in order to better respond to these rapid plasma events. The paper describes the theoretical analysis and closed-loop simulations using the controller that will be tested experimentally in JET during the forthcoming campaign.

Predictive, time-dependent transport simulations with a semi-empirical plasma model have been used in closed-loop simulations to control the q-profile and the strength and location of the internal transport barrier (ITB). Five transport equations (Te, Ti, q, ne, vΦ) are solved, and the power levels of lower hybrid current drive, NBI and ICRH are calculated in a feedback loop determined by the feedback controller matrix. The real-time control (RTC) technique and algorithms used in the transport simulations are identical to those implemented and used in JET experiments (Laborde L. et al 2005 Plasma Phys. Control. Fusion 47 155 and Moreau D. et al 2003 Nucl. Fusion 43 870). The closed-loop simulations with RTC demonstrate that varieties of q-profiles and pressure profiles in the ITB can be achieved and controlled simultaneously. The simulations also showed that with the same RTC technique as used in JET experiments, it is possible to sustain the q-profiles and pressure profiles close to their set-point profiles for longer than the current diffusion time. In addition, the importance of being able to handle the multiple time scales to control the location and strength of the ITB is pointed out. Several future improvements and perspectives of the RTC scheme are presented.

In 2003, the performance of the 'hybrid' regime was successfully validated in JET experiments up to βN = 2.8 at low toroidal field (1.7 T), with plasma triangularity and normalized Larmor radius (ρ*) corresponding to identical ASDEX Upgrade discharges. Stationary conditions have been achieved with the fusion figure of merit (H89.βN/q295) reaching 0.42 at q95 = 3.9. The JET discharges show similar MHD, edge and current profile behaviour, when compared with the ASDEX Upgrade. In addition, the JET experiments have extended the hybrid scenario operation at higher toroidal field of 2.4 T and lower ρ* towards the projected ITER values. Using this database, transport and confinement properties are characterized with respect to the standard H-mode regime. Moreover, trace tritium has been injected to assess the diffusion and convective coefficients of the fusion fuel. The maximization of confinement and stability properties provides, to this scenario, a good probability of achieving a high fusion gain at reduced plasma current for durations of up to 2000 s in ITER.

In plasmas exhibiting an internal transport barrier (ITB), locally very high pressure gradient (∇P) is obtained. It induces high values of the magnetohydrodynamic α parameter (α = −q 2 βR∇P/P, with R the major radius, q the safety factor, P the pressure, ∇ the radial gradient and β the ratio between kinetic and magnetic pressure). Similarly to low or negative magnetic shear (s), high α reduces the curvature and ∇ B drifts driving curvature-type microinstabilities. Therefore, high values of α can stabilize part of the microturbulence, which leads to higher pressure gradient and to even higher α. This possibility for entering a positive feedback loop is very attractive to sustain ITBs in high performance plasmas. Indeed, α scales favourably with higher pressure and does not require any external momentum input. In this paper, after having discussed the α stabilization mechanism in detail, we report the experimental microstability analyses of ITBs from an international multi-machine database—the International Tokamak Physics Activity database, accessible on the webb. We show that α is indeed a relevant parameter of ITB physics that should be taken into account in interpretative and predictive one-dimensional transport codes.

Predictive transport modelling and gyrokinetic stability analyses of demonstration hybrid (HYBRID) and advanced tokamak (AT) discharges from the International Tokamak Physics Activity (ITPA) profile database are presented. Both regimes have exhibited enhanced core confinement (above the conventional ITER reference H-mode scenario) but differ in their current density profiles. Recent contributions to the ITPA database have facilitated an effort to study the underlying physics governing confinement in these advanced scenarios. In this paper, we assess the level of commonality of the turbulent transport physics and the relative roles of the transport suppression mechanisms (i.e. E × B shear and Shafranov shift (α) stabilization) using data for select HYBRID and AT discharges from the DIII-D, JET and AUG tokamaks. GLF23 transport modelling and gyrokinetic stability analysis indicate that E × B shear and Shafranov shift stabilization play essential roles in producing the improved core confinement in both HYBRID and AT discharges. Shafranov shift stabilization is found to be more important in AT discharges than in HYBRID discharges. We have also examined the competition between the stabilizing effects of E × B shear and Shafranov shift stabilization and the destabilizing effects of higher safety factors and parallel velocity shear. Linear and nonlinear gyrokinetic simulations of idealized low and high safety factor cases reveal some interesting consequences. A low safety factor (i.e. HYBRID relevant) is directly beneficial in reducing the transport, and E × B shear stabilization can dominate parallel velocity shear destabilization allowing the turbulence to be quenched. However, at low-q/high current, Shafranov shift stabilization plays less of a role. Higher safety factors (as found in AT discharges), on the other hand, have larger amounts of Shafranov shift stabilization, but parallel velocity shear destabilization can prevent E × B shear quenching of the turbulent transport, and only E × B suppression is achieved.

Internal Transport Barriers (ITB's) in tokamak plasmas are a promising way to achieve steady-state plasmas with good confinement properties in a fusion reactor. Despite a large activity in this field, the formation and self-sustainment of barriers are still not well understood. The physics of ITB's in JET has been investigated with various techniques, namely micro-stability analysis, profile modelling and turbulence simulations. The calculation of linear growth rates show that the magnetic shear plays an important role in the formation of the ITB. The Shafranov shift, ratio of the ion to electron temperature, and impurity content further improve the stability. However the ExB velocity shear is important for the sustainment and the motion of the barrier. This picture is consistent with profile modelling and global fluid simulations of electrostatic drift waves. Turbulence simulations also show that the formation mechanisms are different for electron and ion barriers. Ion barriers are mainly due to the formation of a gap in the density of low wavenumber resonant surfaces whereas electron barriers are rather due to a reversal of the precession drift of trapped electrons.

New algorithms using a truncated singular value decomposition of a linearised model have been implemented in the JET control system. Using three heating and current drive actuators (LHCD, NBI, ICRH), successful control of the safety factor profile has been achieved in quasi steady state conditions where the loop voltage was small and a large fraction of the plasma current was carried by the bootstrap current.

The recent development of real-time measurements and control tools in JET has enhanced the reliability and reproducibility of the relevant ITER scenarios. Diagnostics such as charge exchange, interfero-polarimetry, electron cyclotron emission have been upgraded for real-time measurements. In addition, real-time processes like magnetic equilibrium and q profile reconstruction have been developed and applied successfully in real-time q profile control experiments using model based control techniques. Plasma operation and control against magnetohydrodynamic instabilities are also benefiting from these new systems. The experience gained at JET in the field of real-time measurement and control experiments operation constitutes a very useful basis for the future operation of ITER scenarios.

Injection of Lower Hybrid Heating and Current Drive (LHCD) into the current ramp-up phase of JET [Joint European Torus] discharges can produce extremely reversed q-profiles characterized by a core region of near zero current density (within Motional Stark Effect diagnostic measurement errors). Non-inductive, off-axis co-current drive induces a back electromotive force inside the non-inductive current radius that drives a negative current in the plasma core. The core current density does not go negative, although current diffusion calculations indicate that there is sufficient LHCD to cause this. The clamping of the core current density near zero is consistent with n=0 reconnection events redistributing the core current soon after it goes negative. This is seen in reduced MHD simulations and in nonlinear resistive MHD simulations which predict that these discharges undergo n=0 reconnection events that clamp the core current near zero.

The physics of ITB formation in JET has been investigated using micro-stability analysis, profile modelling and turbulence simulations. The calculation of linear growth rates show that the magnetic shear plays a crucial role in the formation of the ITB. The Shafranov shift, ratio of the ion to electron temperature, and impurity content further improve the stability. This picture is consistent with profile modelling and global fluid simulations of electrostatic drift waves. Turbulence simulations also show that rational q values may play a special role in triggering an ITB. The same physics also explains how double internal barriers can be formed.

In JET discharges where lower hybrid heating and current drive (LHCD) is applied early during the current ramp, a region of the plasma with zero current density is formed near the axis. At the boundary of this region the current density is large and B? increases rapidly over a small distance. In the central region the safety factor, q, is effectively infinite, but this falls steeply in the boundary region. Outside the boundary region q reaches a minimum, where the magnetic shear s?r/q (dq/dr) becomes zero. The formation of this region of zero current is dependent on both the heating and the current drive effects of the LHCD. When LHCD is switched off the current profile begins to relax towards the resistive peaked current distribution of fully inductive tokamak operation. If LHCD is not used in the current rise then these current profiles are not established. Although the physical mechanism exists to drive the central plasma current below zero, in most cases it appears to be prevented from going negative. At least one MHD mechanism has been identified which could be responsible for this. The presence of the zero central current is closely linked to the periodic relaxation events seen in these discharges. In these discharges, internal transport barriers have been observed with additional heating powers substantially below the values required to obtain barriers in monotonic q profile cases.

An overview is presented of the results obtained on JET by the Heating and Current Drive Task Force (TF-H) in the period May 2000 - March 2001. A strongly improved Lower Hybrid (LH) coupling was achieved by optimising the plasma shape and by controlling the local edge density via the injection of CD4. Up to 4 MW have been coupled in type III ELMy H-mode and/or on Internal Transport Barrier (ITB) plasmas with reflection coefficients as low as 4%. Long lasting quasi steady-state ITBs have been obtained by adding the LH current to the bootstrap and beam driven components. Furthermore the use of LH in the pre-heat phase results in electron temperature in excess of 10 keV, deep negative magnetic shear and strongly reduced power threshold for ITB formation. Preliminary results on ICRF coupling are reported including the effect of CD4 injection and the commissioning of the wide band matching system on ELMy plasmas. 1C CD scenarios have been studied in H and 3He minority and used to modify the stability of the sawtooth to influence the formation of seed islands for the appearance of NTM. Up to 3 MW of 1C power was coupled in the high magnetic field fast wave CD scenario. Preliminary MSE measurements indicate differences in the current profiles between -90° and +90° phasing. Careful measurements of the toroidal rotation, in plasmas heated by ICRF only show some dependence on the position of the resonance layer. Finally the use of ICRF minority heating under real-time control, in response to measured plasma parameters to simulate the effect of alpha particles, is presented. ICRF heating results in ITER non-activated scenarios are reported in a companion paper.

RF heating and current drive with ion cyclotron waves and Lower Hybrid waves have been crucial for the development of the Optimized Shear scenario on JET to high performance. Peaked electron temperature profiles and improved energy confinement could be obtained with electron heating both from LHCD and ICRH during plasma current ramp up. ICRH and NBI comparisons allow to separate heating and fueling and suggest a dominant role of core heating in the formation of an Internal Transport Barrier (ITB). ICRH and NBI powers are there equivalent. Pressure profile control by varying the composition of centrally peaked ICRF and broader NBI deposition improves MHD stability. Current profile modifications in a wide range have been obtained with LHCD and in combination with NBI during current ramp up. During the high performance phase, however, LH coupling degrades strongly due to the steep edge density gradient resulting in a drop of the density in front of the LH antenna to the cut-off density. High fusion performance achieving simultaneously high beta values and bootstrap currents is predicted in scenario modeling using pressure and current profile control with ICRF and LHCD.

Transport calculations with relevant lower hybrid current drive control have been performed with the JETTO transport code. The heat transport model and various particle transport models reproducing the experimental JET data have been used in JETTO for predictive high performance modelling. Application of 3.5 MW LHCD power provides a slightly inverted or flat q-profile across 70% of the plasma radius whereas, without LHCD the q-profile is monotonic during the flat-top phase. The results predict a fusion power up to 30 MW for the high performance DT plasmas in the optimised shear scenario at Bt=3.4 T and Ip=3.9 MA. Large uncertainties, however, still persist in particular on the particle transport which strongly influences on the modelling calculations. The presence of not well understood MHD instabilities is also likely to reduce the performance. The most optimistic model, still consistent with experimental results, predicts a fusion gain approaching Q=1.

Understanding the origin of rotation in ion cyclotron resonance frequency (ICRF) heated plasmas is important for predictions for burning plasmas sustained by alpha particles, being characterized by a large population of fast ions and no external momentum input. The angular velocity of the plasma column has been measured in JET H-mode plasmas with pure ICRF heating both for the standard low toroidal magnetic ripple configuration, of about similar to 0.08% and, for increased ripple values up to 1.5% (Nave et al 2010 Phys. Rev. Lett. 105 105005). These new JET rotation data were compared with the multi-machine scaling of Rice et al (2007 Nucl. Fusion 47 1618) for the Alfven-Mach number and with the scaling for the velocity change from L-mode into H-mode. The JET data do not fit well any of these scalings that were derived for plasmas that are co-rotating with respect to the plasma current. With the standard low ripple configuration, JET plasmas with large ICRF heating power and normalized beta, beta(N) approximate to 1.3, have a very small co-current rotation, with Alfven-Mach numbers significantly below those given by the rotation scaling of Rice et al (2007 Nucl. Fusion 47 1618). In some cases the plasmas are actually counter-rotating. No significant difference between the H-mode and L-mode rotation is observed. Typically the H-mode velocities near the edge are lower than those in L-modes. With ripple values larger than the standard JET value, between 1% and 1.5%, H-mode plasmas were obtained where both the edge and the core counter-rotated.

Perturbative experiments are essential to understand the complex transport phenomena in fusion plasmas. The perturbative methods used for transport studies are summarized and the main properties discussed. Based on this approach, transport of particles, heat and momentum has been intensively investigated. The main results obtained for the different channels are described and illustrated with selected examples.

A strong toroidal rotation braking has been observed in plasmas with application of an n = 1 magnetic perturbation field on the JET tokamak. Calculation results from the momentum transport analysis show that the torque induced by the n = 1 perturbation field has a global profile. The maximal value of this torque is at the plasma core region (ρ < 0.4) and it is about half of the neutral beam injection torque. The calculation shows that the plasma is mainly in the ν-√ν regime in the plasma core, but it is close to the transition between the 1/ν and ν-√ν regimes. The neoclassical toroidal viscosity (NTV) torque in the 1/ν and ν-√ν regimes is calculated. The observed torque is of a magnitude in between that of the NTV torque in the 1/ν and ν-√ν regimes. The NTV torque in the ν-√ν regimes is enhanced using the Lagrangian variation of the magnetic field strength. However, it is still smaller than the observed torque by one order of magnitude.

Recent experiments on DIII-D [J. L. Luxon, Nucl. Fusion 42, 614 (2002)] and National Spherical Torus Experiment (NSTX) [M. Ono, Nucl. Fusion 40, 557 (2000)] have focused on investigating mechanisms of driving rotation in fusion plasmas. The so-called intrinsic rotation is generated by an effective torque, driven by residual stresses in the plasma, which appears to originate in the plasma edge. A clear scaling of this intrinsic drive with the H-mode pressure gradient is observed. Coupled with the experimentally inferred pinch of angular momentum, such an edge source is capable of producing sheared rotation profiles. Intrinsic drive is also possible directly in the core, although the physics mechanisms are much more complex. Another option which is being explored is the use of nonresonant magnetic fields for spinning the plasma. It is found beneficially that the torque from these fields can be enhanced at low rotation, which assists in spinning the plasma from rest, and offers increased resistance against plasma slowing. © 2010 American Institute of Physics.

The loss of plasma toroidal angular momentum and thermal energy by edge localized modes (ELMs) has been studied in JET. The analysis shows a consistently larger drop in momentum in comparison with the energy loss associated with the ELMs. This difference originates from the large reduction in angular frequency at the plasma edge, observed to penetrate into the plasma up to r/a ~ 0.65 during large type-I ELMs. As a result, the time averaged angular frequency is lowered near the top of the pedestal with increasing ELM frequency, resulting in a significant drop in thermal Mach number at the edge. An increase in profile peaking of ion temperature and angular frequency is observed. At the same time the plasma confinement is reduced while the ratio of confinement times (Rτ = τE/τ) increases noticeably with ELM frequency. This change could be explained by the relatively larger ELM induced losses for momentum in combination with the observed longer build-up time for the momentum density at the plasma edge.

In recent years, JET diagnostic capability has been significantly improved to widen the range of physical phenomena that can be studied and thus contribute to the understanding of some ITER relevant issues. The most significant results reported in this paper refer to the plasma wall interactions, the interplay between core and edge physics and fast particles. A synergy between new infrared cameras, visible cameras and spectroscopy diagnostics has allowed investigating a series of new aspects of the plasma wall interactions. The power loads on the plasma facing components of JET main chambers have been assessed at steady state and during transient events like ELMs and disruptions. Evidence of filaments in the edge region of the plasma has been collected with a new fast visible camera and high resolution Thomson scattering. The physics of detached plasmas and some new aspects of dust formation have also been devoted particular attention. The influence of the edge plasma on the core has been investigated with upgraded active spectroscopy, providing new information on momentum transport and the effects of impurity injection on ELMs and ITBs and their interdependence. Given the fact that JET is the only machine with a plasma volume big enough to confine the alphas, a coherent programme of diagnostic developments for the energetic particles has been undertaken. With upgraded γ-ray spectroscopy and a new scintillator probe, it is now possible to study both the redistribution and the losses of the fast particles in various plasma conditions.