arXiv subjects

D. Zarzoso

Publications and source records attributed to D. Zarzoso.

5 recordsLinked to original sources

arXiv2024

A generative machine learning surrogate model of plasma turbulence

Generative artificial intelligence methods are employed for the first time to construct a surrogate model for plasma turbulence that enables long time transport simulations. The proposed GAIT (Generative Artificial Intelligence Turbulence) model is based on the coupling of a convolutional variational auto-encoder, that encodes precomputed turbulence data into a reduced latent space, and a recurrent neural network and decoder that generates new turbulence states 400 times faster than the direct numerical integration. The model is applied to the Hasegawa-Wakatani (HW) plasma turbulence model, that is closely related to the quasigeostrophic model used in geophysical fluid dynamics. Very good agreement is found between the GAIT and the HW models in the spatio-temporal Fourier and Proper Orthogonal Decomposition spectra, and the flow topology characterized by the Okubo-Weiss decomposition. The GAIT model also reproduces Lagrangian transport including the probability distribution function of particle displacements and the effective turbulent diffusivity.

physics.plasm-ph↗

arXiv2022

Optimizing beam-ion confinement in ITER by adjusting the toroidal phase of the 3-D magnetic fields applied for ELM control

The confinement of Neutral Beam Injection (NBI) particles in the presence of n=3 Resonant Magnetic Perturbations (RMPs) in 15 MA ITER DT plasmas has been studied using full orbit ASCOT simulations. Realistic NBI distribution functions, and 3D wall and equilibria, including the plasma response to the externally applied 3D fields calculated with MARS-F, have been employed. The observed total fast-ion losses depend on the poloidal spectra of the applied n=3 RMP as well as on the absolute toroidal phase of the applied perturbation with respect to the NBI birth distribution. The absolute toroidal phase of the RMP perturbation does not affect the ELM control capabilities, which makes it a key parameter in the confinement optimization. The physics mechanisms underlying the observed fast-ion losses induced by the applied 3D fields have been studied in terms of the variation of the particle canonical angular momentum ($δP_ϕ$) induced by the applied 3D fields. The presented simulations indicate that the transport is located in an Edge Resonant Transport Layer (ERTL) as observed previously in ASDEX Upgrade studies. Similarly, our results indicate that an overlapping of several linear and nonlinear resonances at the edge of the plasma might be responsible for the observed fast-ion losses. The results presented here may help to optimize the RMP configuration with respect to the NBI confinement in future ITER discharges.

physics.plasm-ph↗

arXiv2020

Towards enhanced performance in fusion plasmas via turbulence suppression by MeV ions

Megaelectron volt (MeV) alpha particles will be the main source of plasma heating in magnetic confinement fusion reactors. Yet, instead of heating fuel ions, most of the energy of alpha particles is transferred to electrons. Furthermore, alpha particles can also excite Alfvenic instabilities, previously considered as detrimental. Contrary to expectations, we demonstrate efficient ion heating in the presence of MeV ions and strong fast-ion driven Alfvenic instabilities in recent experiments on the Joint European Torus (JET). Detailed transport analysis of these experiments with state-of-the-art modeling tools explains the observations. Here we show a novel type of turbulence suppression and improved energy insulation in plasmas with MeV ions and fully developed Alfvenic activities through a complex multi-scale mechanism that generates large-scale zonal flows. This mechanism holds promise for a more economical operation of fusion reactors with dominant alpha particle heating and, ultimately, cheaper fusion electricity

physics.plasm-ph↗

arXiv2017

Cross-code gyrokinetic verification and benchmark on the linear collisionless dynamics of the geodesic acoustic mode

The linear properties of the geodesic acoustic modes (GAM) in tokamaks are investigated by means of the comparison of analytical theory and gyrokinetic numerical simulations. The dependence on the value of the safety factor, finite-orbit-width of the ions in relation to the radial mode width, magnetic-flux-surface shaping, and electron/ion mass ratio are considered. Nonuniformities in the plasma profiles (such as density, temperature, or safety factor), electro-magnetic effects, collisions and presence of minority species are neglected. Also, only linear simulations are considered, focusing on the local dynamics. We use three different gyrokinetic codes: the lagrangian (particle-in-cell) code ORB5, the eulerian code GENE and semi-lagrangian code GYSELA. One of the main aims of this paper is to provide a detailed comparison of the numerical results and analytical theory, in the regimes where this is possible. This helps understanding better the behavior of the linear GAM dynamics in these different regimes, the behavior of the codes, which is crucial in the view of a future work where more physics is present, and the regimes of validity of each specific analytical dispersion relation.

physics.plasm-ph↗

arXiv2009

Observation of confined current ribbon in JET plasmas

we report the identification of a localised current structure inside the JET plasma. It is a field aligned closed helical ribbon, carrying current in the same direction as the background current profile (co-current), rotating toroidally with the ion velocity (co-rotating). It appears to be located at a flat spot in the plasma pressure profile, at the top of the pedestal. The structure appears spontaneously in low density, high rotation plasmas, and can last up to 1.4 s, a time comparable to a local resistive time. It considerably delays the appearance of the first ELM.

physics.plasm-ph↗