Gustavo Grenfell

Understanding the performance of limiter plasmas in the ITER start-up phase is important for the whole pulse and a possible challenge if performed on tungsten, as will now be the case following the switch to a W first wall in the new 2024 ITER Baseline. Experiments were performed on ASDEX Upgrade and WEST to characterise limiter plasmas using boronizations with different degree of boron surface coverage and toroidal asymmetries. Non-boronized start-up is shown to be slow and laborious, and, on both machines, was aborted in favor of a non-homogeneous boronization, performed also in support of the ITER re-baseline to study the impact of spatially non-uniform boron coatings. This allows normal start-up and short (few 100 m s − 1 , ITER will run ≈ 10 s long limiter phases) limiter phases can be run without problems. Even with a full boronization, the limiter can de-condition and long limiter plasmas of several seconds suffer from high densities and radiation. Short limiter plasmas for plasma current ramp-up remain possible. The conditioning effect on limiter plasmas is documented, but also how the start-up is affected by an ageing boronization.

The quasi-coherent mode (QCM) observed in the type-I edgelocalized mode free quasi-continuous exhaust (QCE) regime of ASDEX Upgrade is studied through a detailed comparison of experimental measurements and linear gyrokinetic simulations using the GENE code. The QCM, detected by the thermal helium beam diagnostic near the separatrix, i.e. at ρ _pol ≈ 0.99, exhibits a frequency of approximately 30 kHz and a normalized poloidal wavenumber in the range 0.02 < k _θ ρ _s < 0.08. It propagates in the ion diamagnetic direction with roughly half the ion diamagnetic velocity in the plasma frame and shows electromagnetic characteristics. Cross-phase analyses reveal in-phase density and temperature fluctuations, i.e. α _ne,Te ≈ 0, while an interchange-like crossphase between density and potential fluctuation, i.e. α _ne,Te ≈ -0.7π is observed, indicating a pressure-driven nature. Linear local GENE simulations quantitatively reproduce its location at ρ _pol = 0.991, wavenumbers, propagation direction, and cross-phase behavior. Parameter scans demonstrate a strong dependence of the growth rate on the normalized electron pressure β _e , confirming the electromagnetic character of the instability, and negligible sensitivity to collisionality ν _c . The combined experimental and simulation results consistently identify the underlying QCM instability as a kinetic ballooning mode.

The 3D fluid edge code EMC3-EIRENE has been used to simulate three ASDEX Upgrade discharges with the same equilibrium but different upstream collisionality ν ∗ . In this scan, the plasma transits from an ELMy state at low collisionality to being in the quasi-continuous exhaust regime (QCE) at high ν ∗ . Particular attention was devoted to matching the experimental plasma profiles in front of the ion cyclotron resonance heating (ICRH) limiter, aiming to obtain a reliable estimation of the ICRH limiter peak heat flux in the QCE regime discharge. As a result, we observed that while moving from ELMy to QCE, the limiter peak heat flux increases, reaching values in the order of the MW m − 2 for a limiter-separatrix distance of 6 cm . These values were compared with those at the outer divertor, measured by infrared thermography, revealing that the heat flux perpendicular to the limiter surface can reach up to 90% of the one impinging onto the outer target. Finally, an attempt was made to extend the obtained results towards ITER. This simple analysis suggests that a large wall clearance and a small incidence angle of the magnetic field line, not only on the divertor target but also on ITER’s first wall, might have a crucial role in preventing wall damage during QCE discharges.

The intermediate phase (I-phase) is a plasma regime that exists at the transition from the low-confinement mode (L-mode) to the fully developed high-confinement mode (H-mode), or vice versa. This regime is characterized by improved particle and energy confinement compared to the L-mode, but not as pronounced as in the fully developed H-mode. Generally, type-I edge localized modes are absent in I-phase plasmas and instead smaller burst-like events occur. Using the thermal helium beam spectroscopy and ball-pen probe diagnostic at ASDEX Upgrade, we observe precursors to the burst, which are studied in detail in this work. We find that the I-phase precursor mode, a high-frequency edge fluctuation, is localized in the confined region and propagates in the electron diamagnetic direction in the plasma frame. It has a normalized poloidal wavenumber in the range 0.01 < k θ ρ s < 0.06 , indicating that it is in the mesoscale between classical magnetohydrodynamic and turbulent micro-instabilities. The precursor mode exists for several 100 µ s, with a rapid drop in frequency and change in cross-phase between potential and density fluctuations | α n e , ϕ | ⩾ 0 , before a burst sets in, causing a substantial particle ejection. Our results suggest that the precursor mode is an electromagnetic drift-wave (DW)-like mode, such as a micro-tearing or a kinetic peeling-ballooning mode. The microscopic analysis of single precursors reveals, that heat transport increases first, followed by particle transport. This is why we favor the precursor being a micro-tearing mode. In particular, our results lead to the hypothesis that the burst is the full development of the precursor. More specifically it undergoes a change in its underlying nature as indicated by the change in cross-phase, transitioning from a DW-like to a more interchange-like mode, which ultimately leads to the ejection of heat and particles characteristic of the burst.

The maximum achievable density in a tokamak is restricted by the so-called density limit. In this contribution, we observe that a MARFE precedes the final chain of events leading to disruption in ramp up L-mode discharges, resulting in strong power dissipation and plasma cooling, which likely makes the plasma unstable to MHD modes. However, this phase is preceded by an initial cooling governed by enhanced radial transport due to higher fluctuations and a more favorable cross-phase for transport, which likely impacts the MARFE formation. The increase in radial turbulent transport near the separatrix is primarily observed following the detachment of the outer divertor. This occurs when an X-point radiator forms and the edge collisionality rises, indicating that the increased collisionality influences the turbulence characteristics in the plasma boundary. The causal connection between the events that precede the disruption is discussed.

We report on developments in the edge and scrape-off layer turbulence code GRILLIX to self-consistently simulate plasma turbulence coupled to a fluid neutral gas model. The neutrals model is extended from a single fluid moment (neutrals density) to three moments, describing additionally the dynamics of neutrals parallel momentum and neutrals pressure. GRILLIX employs the flux-coordinate-independent (FCI) approach, wherein the mesh is neither conformal to the wall nor the magnetic field. A new flux evaluation method consistent with FCI allows for implementing plasma recycling at the boundaries that preserves particles to high precision. The improved plasma-neutrals model is applied to simulate an ASDEX Upgrade attached L-mode discharge. Diffusion and viscosity of neutrals parallel momentum introduce a loss channel for ion parallel momentum. This increases the plasma-neutrals interaction time, which is expected to facilitate detachment. Evolution of neutrals pressure leads to a smoother neutrals distribution. Through the charge-exchange coupling, fluctuation amplitudes of ion temperature filaments in the divertor region are reduced. When additionally applying self-consistent recycling boundary conditions, neutrals density and ionization rates at the strike-lines further increase, which impacts the heat load distribution on the target plates.

Reliable estimation of equilibrium flow shear from reciprocating probe measurements is challenging since the quantity of interest corresponds to the second derivative of the observable plasma potential. In addition, a time series of the plasma potential obtained by plunging a probe is affected by both the probe head position and plasma fluctuations, complicating the estimation of equilibrium components and their errors. We tackle this problem by employing Gaussian process regression that is able to infer even the derivatives of a spatial or temporal profile in the form of a probability distribution function. The proposed inference framework is validated by using synthetic data generated by gyrofluid simulations. While the inference result based on a single plunge is unstable in certain spatial locations, we have obtained reasonable agreement between the inference result and the true flow shear profile by combining data sets taken from several plunges.

Understanding the transport processes that determine the plasma profile widths in the scrape-off layer (SOL) and divertor region of tokamaks is crucial for successful power and particle exhaust management in future devices. Plasma transport from the SOL into the private flux region (PFR) broadens the profiles and could mitigate the power exhaust challenge. Analysis of ion current profiles, measured by Langmuir probes in the ASDEX upgrade (AUG) tokamak, shows that the ion current width in the PFR, normalized by the flux expansion between outer target and midplane, is about 1.2 mm in L-mode and 0.9 mm in inter-ELM H-mode plasmas. The measured widths agree within a factor of two with predictions from an analytical model, based on Pfirsch-Schlüter flows. According to the model, the ion PFR width increases with the distance ΔR between the outer target and X-point major radii, and scales inversely with the poloidal magnetic field B p . For the tokamaks ITER and SPARC the model predicts PFR widths below 0.3 mm.

The ICRF slow wave is a potential carrier for parallel RF electric fields known to cause unwanted plasma-wall interactions in magnetic confinement fusion experiments. In nowadays machines the slow wave is usually confined to the far scrape-off layer or the limiter shadow, but conditions in future experiments and reactors may allow the slow wave to be propagative in a larger region. Simulations with RAPLICASOL for various geometries show that the ICRF slow waves appear as the so-called resonance cones (RCs) characterized by large localized electric fields. The RCs emerge from the points along the plasma-antenna interface where (in the cold plasma approximation) the radio frequency electric field diverges. We demonstrate that in the parameter range of interest, the propagation of the RCs in a plasma with a density gradient is defined by a simple geometric model, using the local plasma density and the frequency as input parameters. In the context of experiments at ASDEX Upgrade, simulations illustrate that the RCs can emerge from a single tile of the ICRF antenna limiter. Experiments at the Ion-cyclotron System Hardware Test ARrangement (ISHTAR) were conducted to test the detection principle in a simple environment. In agreement with simulations and with predicted characteristics which depend on operation parameters, the RCs are excited by an RF antenna and propagate through the relatively homogeneous plasma in ISHTAR. The cones are detected at a distance from the antenna using two probes scanning through the plasma. In ASDEX Upgrade, a single tile of an antenna limiter was modified to launch RF power into a specially tailored low-density scrape-off layer. Probes at the mid-plane manipulator were then used to detect the wave electric fields at a distance from the RF source. The detected RF signals show that the signal maxima are located close to the lower hybrid resonance density and are highest when the source and the probes are connected along magnetic field lines. These observations agrees with the model for RCs from the simulations.