B. Clavier, D. Zarzoso, D. Del-Castillo-Negrete, E. Frénod. Generative-machine-learning surrogate model of plasma turbulence. Physical Review E , 2025, 111 (1), pp.L013202. ⟨10.1103/PhysRevE.111.L013202⟩. ⟨hal-04966199⟩ Plus de détails...
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 autoencoder that encodes precomputed turbulence data into a reduced latent space, and a recurrent neural network and decoder that generate new turbulence states 400 times faster than the direct numerical integration. The model is applied to the Hasegawa-Wakatani (HW) plasma turbulence model, which 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 spatiotemporal 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.
B. Clavier, D. Zarzoso, D. Del-Castillo-Negrete, E. Frénod. Generative-machine-learning surrogate model of plasma turbulence. Physical Review E , 2025, 111 (1), pp.L013202. ⟨10.1103/PhysRevE.111.L013202⟩. ⟨hal-04966199⟩
Enrique de Dios Zapata Cornejo, David Zarzoso, S.D. Pinches, Andres Bustos, Alvaro Cappa, et al.. A novel unsupervised machine learning algorithm for automatic Alfvénic activity detection in the TJ-II stellarator. Nuclear Fusion, 2024, 64 (12), pp.126057. ⟨10.1088/1741-4326/ad85f4⟩. ⟨hal-04540368⟩ Plus de détails...
A novel sparse encoding algorithm is developed to detect and study plasma instabilities automatically. This algorithm, called Elastic Random Mode Decomposition, is applied to the Mirnov coil signals of a dataset of 1291 discharges of the TJ-II stellarator, enabling the identification of the Alfvénic activity. In the presented approach, each signal is encoded as a collection of basic waveforms called atoms, drawn from a signal’s dictionary. Then the modes are identified using clustering and correlations with other plasma signals. The performance of the proposed algorithm is dramatically increased by using elastic net regularization and taking advantage of GPU architectures, hence the signal size and the number of dictionary elements are no longer limiting factors for encoding complex signals. Once the modes are retrieved from the shots, they can be easily analyzed with standard clustering techniques, thereby describing the physical mode characteristics of this subset of TJ-II shots. The clustering features consider the relationship with the plasma current Ip, the diamagnetic energy W, and inverse squared root electronic density 1/√n, profiling different subtypes of Alfvénic activity. The proposed algorithm can potentially create large databases of labeled modes with unprecedented detail.
Enrique de Dios Zapata Cornejo, David Zarzoso, S.D. Pinches, Andres Bustos, Alvaro Cappa, et al.. A novel unsupervised machine learning algorithm for automatic Alfvénic activity detection in the TJ-II stellarator. Nuclear Fusion, 2024, 64 (12), pp.126057. ⟨10.1088/1741-4326/ad85f4⟩. ⟨hal-04540368⟩
H Betar, David Zarzoso, Jacobo Varela, Diego Del-Castillo-Negrete, Luis Garcia, et al.. Transport and losses of energetic particles in tokamaks in the presence of Alfvén activity using the new full orbit TAPaS code coupled to FAR3d. Nuclear Fusion, 2024, ⟨10.1088/1741-4326/ad7c66⟩. ⟨hal-04541528v2⟩ Plus de détails...
Recent developments and tools integrated into the TAPaS code are presented, enabling realistic scenario simulations of particle dynamics within experimental tokamak magnetic equilibria. In particular, the enhanced capabilities of TAPaS enable seamless coupling with external simulations, provided the metric and equilibrium magnetic field of the external code are known. Coupling TAPaS with the gyro-fluid code FAR3d, the transport and losses of energetic particles in the presence Alfvén eigenmodes (AEs) in DIII-D plasma discharge #159243 were investigated. Detailed analyses of prompt losses with and without collisions were performed. Then, further analysis was performed in the presence of electromagnetic perturbations resulting from AEs activity. The results indicate that, for the energies and the initial conditions considered here, the presence of AEs enhances the particle losses.
H Betar, David Zarzoso, Jacobo Varela, Diego Del-Castillo-Negrete, Luis Garcia, et al.. Transport and losses of energetic particles in tokamaks in the presence of Alfvén activity using the new full orbit TAPaS code coupled to FAR3d. Nuclear Fusion, 2024, ⟨10.1088/1741-4326/ad7c66⟩. ⟨hal-04541528v2⟩
R Varennes, Guilhem Dif-Pradalier, P Ghendrih, V Grandgirard, O Panico, et al.. Synergy of turbulent momentum drive and magnetic braking. Plasma Physics and Controlled Fusion, 2024, 66 (10), pp.105008. ⟨10.1088/1361-6587/ad705c⟩. ⟨hal-03631224⟩ Plus de détails...
Abstract Relaxations of localized over-density in a plane transverse to the magnetic field are numerically investigated under the effect of drift-wave and interchange drives in SOL conditions. Such a controlled departure from thermodynamic equilibrium allows the investigation of fundamental processes at play in cross-field transport. Interchange instabilities generate ballistic outward radial flux with low amplitude zonal flow patterns, whereas drift-wave instabilities result in symmetric radial flux with large amplitude zonal flow patterns. When both instabilities are considered, the combined effects tend to favor drift-waves, leading to a weaker outward flux with larger zonal flow patterns.
R Varennes, Guilhem Dif-Pradalier, P Ghendrih, V Grandgirard, O Panico, et al.. Synergy of turbulent momentum drive and magnetic braking. Plasma Physics and Controlled Fusion, 2024, 66 (10), pp.105008. ⟨10.1088/1361-6587/ad705c⟩. ⟨hal-03631224⟩
E D D Zapata-Cornejo, D Zarzoso, S D Pinches, S E Sharapov, M Fitzgerald. Segmentation of MHD modes using Fourier transform, wavelets and computer vision algorithms. Plasma Physics and Controlled Fusion, 2024, 66 (9), pp.095016. ⟨10.1088/1361-6587/ad6a84⟩. ⟨hal-04599681v2⟩ Plus de détails...
MHD activity in fusion devices is typically analyzed by examining time-frequency spectrograms obtained from various diagnostics. MHD modes often co-exist with various types of noise and complex patterns generated by other events like pellet injection or active diagnostics. Traditionally, identifying MHD modes has been a manual task, making it labor-intensive. To overcome this issue, this study proposes the use of computer vision algorithms for noise removal and automatic feature extraction. First, the automatic detection of straight-line patterns is achieved by applying the Hough transform. Then, the discrete wavelet transform is proposed to break down spectrograms into sub-images of different scales, removing broadband noise and pellet injection signatures. The multiscale decomposition is subsequently extended to multiple directions using either 2D Fourier transforms or curvelets, achieving a high signal-to-noise ratio in spectrograms and eliminating undesired frequency sweeps of TAE antenna. Once MHD activity is successfully enhanced, a pipeline of algorithms for ridge detection, thresholding and labeling perform a segmentation of the image, automatically labeling individual modes. This study demonstrates the effectiveness of computer vision algorithms for the identification of MHD modes. The use of such algorithms may potentially help in the analysis process and the creation of large databases of modes.
E D D Zapata-Cornejo, D Zarzoso, S D Pinches, S E Sharapov, M Fitzgerald. Segmentation of MHD modes using Fourier transform, wavelets and computer vision algorithms. Plasma Physics and Controlled Fusion, 2024, 66 (9), pp.095016. ⟨10.1088/1361-6587/ad6a84⟩. ⟨hal-04599681v2⟩
R Varennes, G Dif-Pradalier, P Ghendrih, V Grandgirard, O Panico, et al.. Turbulent relaxation patterns in SOL plasma. Plasma Physics and Controlled Fusion, 2024, 66 (10), pp.105008. ⟨10.1088/1361-6587/ad705c⟩. ⟨hal-04599956⟩ Plus de détails...
Relaxations of localized over-density in a plane transverse to the magnetic field are numerically investigated under the effect of drift-wave and interchange drives in SOL conditions. Such a controlled departure from thermodynamic equilibrium allows the investigation of fundamental processes at play in cross-field transport. Interchange instabilities generate ballistic outward radial flux with low amplitude zonal flow patterns, whereas drift-wave instabilities result in symmetric radial flux with large amplitude zonal flow patterns. When both instabilities are considered, the combined effects tend to favor drift-waves, leading to a weaker outward flux with larger zonal flow patterns.
R Varennes, G Dif-Pradalier, P Ghendrih, V Grandgirard, O Panico, et al.. Turbulent relaxation patterns in SOL plasma. Plasma Physics and Controlled Fusion, 2024, 66 (10), pp.105008. ⟨10.1088/1361-6587/ad705c⟩. ⟨hal-04599956⟩
Homam Betar, Daniele Del Sarto, A. Ghizzo, F. Brochard, David Zarzoso. A numerical study of electron-magnetohydrodynamics tearing modes in parameter ranges of experimental interest. Physics of Plasmas, 2024, 1st European Conference on Magnetic Reconnection in Plasmas, 31 (5), pp.052117. ⟨10.1063/5.0205061⟩. ⟨hal-04561813⟩ Plus de détails...
We perform a numerical study of the linear dynamics of tearing modes in slab incompressible electron- magnetohydrodynamics (EMHD) by considering some parameter ranges which can be of interest for laboratory plasmas (e.g., helicon devices) or for astrophysics (e.g., solar-wind turbulence). To this purpose several non-ideal effects are simultaneously retained (finite electron inertia, resistivity and electron viscosity) and we make distinction between the dissipation coefficients in the direction parallel and perpendicular to the guide field. We thus identify some new recon- nection regimes, characterized by a departure from the customary monotonic power-law scalings of the growth rates with respect to the non-ideal parameters. The results here presented can provide a useful indication for future studies of EMHD regimes relevant to experiments and for extensions of the EMHD tearing mode modelling to more complete regimes including kinetic effects (e.g., "electron-only" reconnection in kinetic regimes).
Homam Betar, Daniele Del Sarto, A. Ghizzo, F. Brochard, David Zarzoso. A numerical study of electron-magnetohydrodynamics tearing modes in parameter ranges of experimental interest. Physics of Plasmas, 2024, 1st European Conference on Magnetic Reconnection in Plasmas, 31 (5), pp.052117. ⟨10.1063/5.0205061⟩. ⟨hal-04561813⟩
Jeronimo Garcia, Yevgen Kazakov, Rui Coelho, Mykola Dreval, Elena de la Luna, et al.. Stable Deuterium-Tritium plasmas with improved confinement in the presence of energetic-ion instabilities. Nature Communications, 2024, 15 (1), pp.7846. ⟨10.1038/s41467-024-52182-z⟩. ⟨hal-04948995⟩ Plus de détails...
Providing stable and clean energy sources is a necessity for the increasing demands of humanity. Energy produced by Deuterium (D) and Tritium (T) fusion reactions, in particular in tokamaks, is a promising path towards that goal. However, there is little experience with plasmas formed by D-T mixtures, since most of the experiments are currently performed in pure D. After more than 20 years, the Joint European Torus (JET) has carried out new D-T experiments with the aim of exploring some of the unique characteristics expected in future fusion reactors, such as the presence of highly energetic ions in low plasma rotation conditions. A new stable, high confinement and impurity-free D-T regime, with reduction of energy losses with respect to D, has been found. Multiscale physics mechanisms critically determine the thermal confinement. These crucial achievements importantly contribute to the establishment of fusion energy generation as an alternative to fossil fuels.
Jeronimo Garcia, Yevgen Kazakov, Rui Coelho, Mykola Dreval, Elena de la Luna, et al.. Stable Deuterium-Tritium plasmas with improved confinement in the presence of energetic-ion instabilities. Nature Communications, 2024, 15 (1), pp.7846. ⟨10.1038/s41467-024-52182-z⟩. ⟨hal-04948995⟩
Ivan Kudashev, Anna Medvedeva, Nicolas Fedorszak, David Zarzoso, Manuel Scotto d’ Abusco, et al.. Development of a set of synthetic diagnostics for the WEST tokamak to confront 2D transport simulations and experimental data. Journal of Instrumentation, 2023, 18 (02), pp.C02058. ⟨10.1088/1748-0221/18/02/C02058⟩. ⟨hal-04010344⟩ Plus de détails...
Significant scientific effort has been focused on optimizing the scenarios and plasma parameters for tokamak operations. The lack of comprehensive understanding of underlying physical processes leads to simplifications used both in plasma simulation codes and for diagnostics, which is also complicated by the harsh plasma environment. One of the main tools to couple, check and verify these assumptions are the synthetic diagnostics. In this work we demonstrate current results of the development of the set of synthetic diagnostics for the WEST tokamak to couple experimental data with the SolEdge3X-HDG 2D transport code.
Ivan Kudashev, Anna Medvedeva, Nicolas Fedorszak, David Zarzoso, Manuel Scotto d’ Abusco, et al.. Development of a set of synthetic diagnostics for the WEST tokamak to confront 2D transport simulations and experimental data. Journal of Instrumentation, 2023, 18 (02), pp.C02058. ⟨10.1088/1748-0221/18/02/C02058⟩. ⟨hal-04010344⟩
J Varela, D Spong, L Garcia, Y Ghai, David Zarzoso, et al.. Effect of the neutral beam injector operational regime on the Alfven eigenmode saturation phase in DIII-D plasma. Plasma Physics and Controlled Fusion, 2023, 65 (12), pp.125004. ⟨10.1088/1361-6587/ad05d4⟩. ⟨hal-04541045⟩ Plus de détails...
The aim of this study is to analyze the effect of the neutral beam injector (NBI) operation regime on the saturation phase of the Alfven Eigenmodes (AEs) in DIII-D plasma. The analysis is done using the linear and nonlinear versions of the gyro-fluid code FAR3d. A set of parametric analyses are performed modifying the nonlinear simulation EP β (NBI injection power), EP energy (NBI voltage) and the radial location of the EP density profile gradient (NBI radial deposition). The analysis indicates a transition from the soft (local plasma relaxation) to the hard MHD (global plasma relaxation) limit if the simulation EP β ⩾ 0.02 , leading to bursting MHD activity caused by radial AEs overlapping. MHD bursts cause an enhancement of the EP transport showing ballistic-like features as avalanche-like events. Simulations in the soft MHD limit show an increment of the EP density gradient as the EP β increases. On the other hand, there is a gradient upper limit in the hard MHD limit, consistent with the critical-gradient behavior. AEs induce shear flows and zonal current leading to the deformation of the flux surfaces and the safety factor profile, respectively, particularly strong for the simulation in the hard MHD limit. Simulations in the hard MHD regime show a decrease of the AE frequency in the saturation phase; this is caused by the destabilization of a transitional mode between a 9 / 3 − 10 / 3 TAE and a 9 / 3 RSAE that may explain the AE frequency down-sweeping observed in some DIII-D discharges. Reducing the EP energy in the nonlinear simulations leads to a weakening of the plasma perturbation. On the other hand, increasing the EP energy causes the opposite effect. Nonlinear simulations of off-axis NBI profiles indicate a lower plasma perturbation as the EP density gradient is located further away from the magnetic axis.
J Varela, D Spong, L Garcia, Y Ghai, David Zarzoso, et al.. Effect of the neutral beam injector operational regime on the Alfven eigenmode saturation phase in DIII-D plasma. Plasma Physics and Controlled Fusion, 2023, 65 (12), pp.125004. ⟨10.1088/1361-6587/ad05d4⟩. ⟨hal-04541045⟩
Samuele Mazzi, David Zarzoso. Parametric Validation of the Reservoir Computing–Based Machine Learning Algorithm Applied to Lorenz System Reconstructed Dynamics. Complex Systems , 2022, 31 (3), pp.311-339. ⟨10.25088/ComplexSystems.31.3.311⟩. ⟨hal-03838327⟩ Plus de détails...
A detailed parametric analysis is presented, where the recent method based on the reservoir computing paradigm, including its statistical robustness, is studied. It is observed that the prediction capabilities of the reservoir computing approach strongly depend on the random initialization of both the input and the reservoir layers. Special emphasis is put on finding the region in the hyperparameter space where the ensemble-averaged training and generalization errors together with their variance are minimized. The statistical analysis presented here is based on the projection on proper elements method.
Samuele Mazzi, David Zarzoso. Parametric Validation of the Reservoir Computing–Based Machine Learning Algorithm Applied to Lorenz System Reconstructed Dynamics. Complex Systems , 2022, 31 (3), pp.311-339. ⟨10.25088/ComplexSystems.31.3.311⟩. ⟨hal-03838327⟩
S Mazzi, J Garcia, David Zarzoso, Ye Kazakov, J Ongena, et al.. Gyrokinetic study of transport suppression in JET plasmas with MeV-ions and toroidal Alfvén eigenmodes. Plasma Physics and Controlled Fusion, 2022, 64 (11), pp.114001. ⟨10.1088/1361-6587/ac91f3⟩. ⟨hal-03838290⟩ Plus de détails...
The impact of fast ions, generated in the MeV-range through the efficient application of the three-ion scheme in JET plasmas, on the turbulence properties is presented through complex numerical simulations. The suppression of the ion-scale turbulent transport is studied by means of in-depth gyrokinetic numerical analyses. Such a suppression is demonstrated to be achieved in the presence of toroidal Alfvén eigenmodes (TAEs) destabilized by the highly energetic ions. Details on the TAE excitation are also provided with a multi-code analysis. The inherently nonlinear and multi-scale mechanism triggered by the fast ions, also involving the high-frequency modes and the large-scale zonal flows, is deeply analyzed. Such mechanism is thus demonstrated, with experimental validating studies, to be the main cause of turbulence suppression and improvement of ion thermal confinement. Additional simulations address the implications of reversed shear magnetic equilibrium on the turbulent transport.
S Mazzi, J Garcia, David Zarzoso, Ye Kazakov, J Ongena, et al.. Gyrokinetic study of transport suppression in JET plasmas with MeV-ions and toroidal Alfvén eigenmodes. Plasma Physics and Controlled Fusion, 2022, 64 (11), pp.114001. ⟨10.1088/1361-6587/ac91f3⟩. ⟨hal-03838290⟩
Samuele Mazzi, Yann Camenen, Jeronimo Garcia, David Zarzoso, D. Frigione, et al.. Effects of the parallel flow shear on the ITG-driven turbulent transport in tokamak plasmas. Nuclear Fusion, 2022, 62 (9), pp.096024. ⟨10.1088/1741-4326/ac7ac2⟩. ⟨hal-03838283⟩ Plus de détails...
Abstract The impact of the parallel flow shear on the tokamak plasma stability and turbulent transport driven by the ion temperature gradient (ITG) modes is analyzed by means of local gyrokinetic numerical analyses. It is shown that the parallel flow shear increases the ITG growth rate in the linear regime, and induces a broadening and shift of the radial spectrum. Then, the different effects of the finite parallel shear on the ITG turbulence characteristics are deeply analyzed in the nonlinear regime. These studies highlight that a reduction of the thermal-ion turbulent heat flux is induced by a complex mechanism involving the nonlinear generation of an enhanced zonal flow activity. Indeed, the turbulent sources of the zonal flows are increased by the introduction of the finite parallel flow shear in the system, beneficially acting on the saturation level of the ITG turbulence. The study has been carried out for the Waltz standard case below the critical threshold of the destabilization of the parallel velocity gradient instability, and then generalized to a selected pulse of a recent JET scenario with substantial toroidal rotation in the edge plasma region. It is, thus, suggested that the investigated complex mechanism triggered by the finite parallel flow shear reducing the ITG turbulent heat fluxes could be complementary to the well-established perpendicular flow shear in a region with sufficiently large plasma toroidal rotation.
Samuele Mazzi, Yann Camenen, Jeronimo Garcia, David Zarzoso, D. Frigione, et al.. Effects of the parallel flow shear on the ITG-driven turbulent transport in tokamak plasmas. Nuclear Fusion, 2022, 62 (9), pp.096024. ⟨10.1088/1741-4326/ac7ac2⟩. ⟨hal-03838283⟩
David Zarzoso, Diego Del-Castillo-Negrete, Rémi Lacroix, Pierre-Eric Bernard, Stanislas Touzet. Transport and losses of fusion-born alpha particles in the presence of tearing modes using the new Toroidal Accelerated Particle Simulator (TAPaS). Plasma Physics and Controlled Fusion, 2022, 64 (4), pp.044003. ⟨10.1088/1361-6587/ac493a⟩. ⟨hal-03843243⟩ Plus de détails...
Abstract The transport and losses of fusion-born alpha particles are studied in the presence of a single-helicity tearing mode, characterized by ( m = 2 , n = 1 ). The analysis is performed by means of the recently developed Toroidal Accelerated P a rticle Simulator ( TAPaS ). Although such modes have usually been believed to result only in a local flattening of the radial profiles, it is shown that the density profile can exhibit a global modification leading to significant losses of alpha particles. This is due to the fact that, although the magnetic field does not exhibit any chaotic behaviour, the trajectories of alpha particles do, as revealed by their Poincaré maps. Such results are in qualitative agreement with past observations and simulations of energetic particles generated by neutral beam injection in TFTR, DIII-D and AUG tokamaks. In-depth analysis is carried out to characterize the impact of the tearing mode on the transport and losses of fusion-born alpha-particles with a realistic density profile. The impact of the amplitude is evidenced. Moreover, the effect of the island rotation frequency is assessed based on a detailed analysis of the linear resonances in phase-space, in agreement with the simulation results. Finally, the probability density function of the exit time has been computed and the transport of alpha particles has been found to be anomalous.
David Zarzoso, Diego Del-Castillo-Negrete, Rémi Lacroix, Pierre-Eric Bernard, Stanislas Touzet. Transport and losses of fusion-born alpha particles in the presence of tearing modes using the new Toroidal Accelerated Particle Simulator (TAPaS). Plasma Physics and Controlled Fusion, 2022, 64 (4), pp.044003. ⟨10.1088/1361-6587/ac493a⟩. ⟨hal-03843243⟩
Samuele Mazzi, David Zarzoso. Parametric Validation of the Reservoir-Computing-Based Machine Learning Algorithm Applied to Lorenz System Reconstructed Dynamics. Complex Systems , In press. ⟨hal-03200715v2⟩ Plus de détails...
A detailed parametric analysis is presented, where the recent method based on the Reservoir Computing paradigm, including its statistical robustness, is studied. It is observed that the prediction capabilities of the Reservoir Computing approach strongly depend on the random initialisation of both the input and the reservoir layers. Special emphasis is put on finding the region in the hyperparameter space where the ensemble-averaged training and generalization errors together with their variance are minimized. The statistical analysis presented here is based on the Projection on Proper Elements (PoPe) method [T. Cartier-Michaud et al., Phys. Plasmas 23, 020702 (2016)].
Samuele Mazzi, David Zarzoso. Parametric Validation of the Reservoir-Computing-Based Machine Learning Algorithm Applied to Lorenz System Reconstructed Dynamics. Complex Systems , In press. ⟨hal-03200715v2⟩
