-
TORAX: A Fast and Differentiable Tokamak Transport Simulator in JAX
Authors:
Jonathan Citrin,
Ian Goodfellow,
Akhil Raju,
Jeremy Chen,
Jonas Degrave,
Craig Donner,
Federico Felici,
Philippe Hamel,
Andrea Huber,
Dmitry Nikulin,
David Pfau,
Brendan Tracey,
Martin Riedmiller,
Pushmeet Kohli
Abstract:
We present TORAX, a new, open-source, differentiable tokamak core transport simulator implemented in Python using the JAX framework. TORAX solves the coupled equations for ion heat transport, electron heat transport, particle transport, and current diffusion, incorporating modular physics-based and ML models. JAX's just-in-time compilation ensures fast runtimes, while its automatic differentiation…
▽ More
We present TORAX, a new, open-source, differentiable tokamak core transport simulator implemented in Python using the JAX framework. TORAX solves the coupled equations for ion heat transport, electron heat transport, particle transport, and current diffusion, incorporating modular physics-based and ML models. JAX's just-in-time compilation ensures fast runtimes, while its automatic differentiation capability enables gradient-based optimization workflows and simplifies the use of Jacobian-based PDE solvers. Coupling to ML-surrogates of physics models is greatly facilitated by JAX's intrinsic support for neural network development and inference. TORAX is verified against the established RAPTOR code, demonstrating agreement in simulated plasma profiles. TORAX provides a powerful and versatile tool for accelerating research in tokamak scenario modeling, pulse design, and control.
△ Less
Submitted 12 June, 2024; v1 submitted 10 June, 2024;
originally announced June 2024.
-
Efficient training sets for surrogate models of tokamak turbulence with Active Deep Ensembles
Authors:
L. Zanisi,
A. Ho,
T. Madula,
J. Barr,
J. Citrin,
S. Pamela,
J. Buchanan,
F. Casson,
V. Gopakumar,
JET contributors
Abstract:
Model-based plasma scenario development lies at the heart of the design and operation of future fusion powerplants. Including turbulent transport in integrated models is essential for delivering a successful roadmap towards operation of ITER and the design of DEMO-class devices. Given the highly iterative nature of integrated models, fast machine-learning-based surrogates of turbulent transport ar…
▽ More
Model-based plasma scenario development lies at the heart of the design and operation of future fusion powerplants. Including turbulent transport in integrated models is essential for delivering a successful roadmap towards operation of ITER and the design of DEMO-class devices. Given the highly iterative nature of integrated models, fast machine-learning-based surrogates of turbulent transport are fundamental to fulfil the pressing need for faster simulations opening up pulse design, optimization, and flight simulator applications. A significant bottleneck is the generation of suitably large training datasets covering a large volume in parameter space, which can be prohibitively expensive to obtain for higher fidelity codes.
In this work, we propose ADEPT (Active Deep Ensembles for Plasma Turbulence), a physics-informed, two-stage Active Learning strategy to ease this challenge. Active Learning queries a given model by means of an acquisition function that identifies regions where additional data would improve the surrogate model. We provide a benchmark study using available data from the literature for the QuaLiKiz quasilinear transport model. We demonstrate quantitatively that the physics-informed nature of the proposed workflow reduces the need to perform simulations in stable regions of the parameter space, resulting in significantly improved data efficiency. We show an up to a factor of 20 reduction in training dataset size needed to achieve the same performance as random sampling. We then validate the surrogates on multichannel integrated modelling of ITG-dominated JET scenarios and demonstrate that they recover the performance of QuaLiKiz to better than 10\%. This matches the performance obtained in previous work, but with two orders of magnitude fewer training data points.
△ Less
Submitted 13 October, 2023;
originally announced October 2023.
-
Fast transport simulations with higher-fidelity surrogate models for ITER
Authors:
J. Citrin,
P. Trochim,
T. Goerler,
D. Pfau,
K. L. van de Plassche,
F. Jenko
Abstract:
A fast and accurate turbulence transport model based on quasilinear gyrokinetics is developed. The model consists of a set of neural networks trained on a bespoke quasilinear GENE dataset, with a saturation rule calibrated to dedicated nonlinear simulations. The resultant neural network is approximately eight orders of magnitude faster than the original GENE quasilinear calculations. ITER predicti…
▽ More
A fast and accurate turbulence transport model based on quasilinear gyrokinetics is developed. The model consists of a set of neural networks trained on a bespoke quasilinear GENE dataset, with a saturation rule calibrated to dedicated nonlinear simulations. The resultant neural network is approximately eight orders of magnitude faster than the original GENE quasilinear calculations. ITER predictions with the new model project a fusion gain in line with ITER targets. While the dataset is currently limited to the ITER baseline regime, this approach illustrates a pathway to develop reduced-order turbulence models both faster and more accurate than the current state-of-the-art.
△ Less
Submitted 1 June, 2023;
originally announced June 2023.
-
How accurate are flux-tube (local) gyrokinetic codes in modeling energetic particle effects on core turbulence?
Authors:
A. Di Siena,
T. Hayward-Schneider,
P. Mantica,
J. Citrin,
F. Vannini,
A. Bottino,
T. Goerler,
E. Poli,
R. Bilato,
O. Sauter,
F. Jenko
Abstract:
Flux-tube gyrokinetic codes are widely used to simulate drift-wave turbulence in magnetic confinement devices. While a large number of studies show that flux-tube codes provide an excellent approximation for turbulent transport in medium-large devices, it still needs to be determined whether they are sufficient for modeling supra-thermal particle effects on core turbulence. This is called into que…
▽ More
Flux-tube gyrokinetic codes are widely used to simulate drift-wave turbulence in magnetic confinement devices. While a large number of studies show that flux-tube codes provide an excellent approximation for turbulent transport in medium-large devices, it still needs to be determined whether they are sufficient for modeling supra-thermal particle effects on core turbulence. This is called into question given the large temperature of energetic particles (EPs), which makes them hardly confined on a single flux-surface, but also due to the radially broad mode structure of energetic-particle-driven modes. The primary focus of this manuscript is to assess the range of validity of flux-tube codes in modeling fast ion effects by comparing radially global turbulence simulations with flux-tube results at different radial locations for realistic JET parameters using the gyrokinetic code GENE. To extend our study to a broad range of different plasma scenarios, this comparison is made for four different plasma regimes, which differ only by the profile of the ratio between the plasma kinetic and magnetic pressure. The latter is artificially rescaled to address the electrostatic limit and regimes with marginally stable, weakly unstable and strongly unstable fast ion modes. These energetic-particle-driven modes is identified as an AITG/KBAE via linear ORB5 and LIGKA simulations. It is found that the local flux-tube simulations can recover well the global results only in the electrostatic and marginally stable cases. When the AITG/KBAE becomes linearly unstable, the local approximation fails to correctly model the radially broad fast ion mode structure and the consequent global zonal patterns. According to this study, global turbulence simulations are likely required in regimes with linearly unstable AITG/KBAEs. In conditions with different fast ion-driven modes, these results might change.
△ Less
Submitted 5 May, 2023;
originally announced May 2023.
-
Integrated modelling and multiscale gyrokinetic validation study of ETG turbulence in a JET hybrid H-mode scenario
Authors:
J Citrin,
S Maeyama,
C Angioni,
N Bonanomi,
C Bourdelle,
F. J Casson,
E Fable,
T Goerler,
P Mantica,
A Mariani,
M Sertoli,
G Staebler,
T Watanabe,
JET Contributors
Abstract:
Previous studies with first-principle-based integrated modelling suggested that ETG turbulence may lead to an anti-GyroBohm isotope scaling in JET high-performance hybrid H-mode scenarios. A dedicated comparison study against higher-fidelity turbulence modelling invalidates this claim. Ion-scale turbulence with magnetic field perturbations included, can match the power balance fluxes within temper…
▽ More
Previous studies with first-principle-based integrated modelling suggested that ETG turbulence may lead to an anti-GyroBohm isotope scaling in JET high-performance hybrid H-mode scenarios. A dedicated comparison study against higher-fidelity turbulence modelling invalidates this claim. Ion-scale turbulence with magnetic field perturbations included, can match the power balance fluxes within temperature gradient error margins. Multiscale gyrokinetic simulations from two distinct codes produce no significant ETG heat flux, demonstrating that simple rules-of-thumb are insufficient criteria for its onset.
△ Less
Submitted 20 September, 2022;
originally announced September 2022.
-
The Problem of Marginality in Model Reductions of Turbulence
Authors:
C. Gillot,
G. Dif-Pradalier,
Y. Sarazin,
C. Bourdelle,
Y. Camenen,
J. Citrin,
X. Garbet,
Ph. Ghendrih,
V. Grandgirard,
P. Manas,
F. Widmer
Abstract:
Reduced quasilinear (QL) and nonlinear (gradient-driven) models with scale separations, commonly used to interpret experiments and to forecast turbulent transport levels in magnetised plasmas are tested against nonlinear models without scale separations (flux-driven). Two distinct regimes of turbulence -- either far above threshold or near marginal stability -- are investigated with Boltzmann elec…
▽ More
Reduced quasilinear (QL) and nonlinear (gradient-driven) models with scale separations, commonly used to interpret experiments and to forecast turbulent transport levels in magnetised plasmas are tested against nonlinear models without scale separations (flux-driven). Two distinct regimes of turbulence -- either far above threshold or near marginal stability -- are investigated with Boltzmann electrons. The success of reduced models especially hinges on the reproduction of nonlinear fluxes. Good agreement between models is found above threshold whilst reduced models would significantly underpredict fluxes near marginality, overlooking mesoscale flow organisation and turbulence self-advection. Constructive prescriptions whereby to improve reduced models is discussed.
△ Less
Submitted 8 September, 2022;
originally announced September 2022.
-
2022 Review of Data-Driven Plasma Science
Authors:
Rushil Anirudh,
Rick Archibald,
M. Salman Asif,
Markus M. Becker,
Sadruddin Benkadda,
Peer-Timo Bremer,
Rick H. S. Budé,
C. S. Chang,
Lei Chen,
R. M. Churchill,
Jonathan Citrin,
Jim A Gaffney,
Ana Gainaru,
Walter Gekelman,
Tom Gibbs,
Satoshi Hamaguchi,
Christian Hill,
Kelli Humbird,
Sören Jalas,
Satoru Kawaguchi,
Gon-Ho Kim,
Manuel Kirchen,
Scott Klasky,
John L. Kline,
Karl Krushelnick
, et al. (38 additional authors not shown)
Abstract:
Data science and technology offer transformative tools and methods to science. This review article highlights latest development and progress in the interdisciplinary field of data-driven plasma science (DDPS). A large amount of data and machine learning algorithms go hand in hand. Most plasma data, whether experimental, observational or computational, are generated or collected by machines today.…
▽ More
Data science and technology offer transformative tools and methods to science. This review article highlights latest development and progress in the interdisciplinary field of data-driven plasma science (DDPS). A large amount of data and machine learning algorithms go hand in hand. Most plasma data, whether experimental, observational or computational, are generated or collected by machines today. It is now becoming impractical for humans to analyze all the data manually. Therefore, it is imperative to train machines to analyze and interpret (eventually) such data as intelligently as humans but far more efficiently in quantity. Despite the recent impressive progress in applications of data science to plasma science and technology, the emerging field of DDPS is still in its infancy. Fueled by some of the most challenging problems such as fusion energy, plasma processing of materials, and fundamental understanding of the universe through observable plasma phenomena, it is expected that DDPS continues to benefit significantly from the interdisciplinary marriage between plasma science and data science into the foreseeable future.
△ Less
Submitted 31 May, 2022;
originally announced May 2022.
-
Two step clustering for data reduction combining DBSCAN and k-means clustering
Authors:
Bart J. J. Kremers,
Aaron Ho,
Jonathan Citrin,
Karel L. van de Plassche
Abstract:
A novel combination of two widely-used clustering algorithms is proposed here for the detection and reduction of high data density regions. The Density Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm is used for the detection of high data density regions and the k-means algorithm for reduction. The proposed algorithm iterates while successively decrementing the DBSCAN search…
▽ More
A novel combination of two widely-used clustering algorithms is proposed here for the detection and reduction of high data density regions. The Density Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm is used for the detection of high data density regions and the k-means algorithm for reduction. The proposed algorithm iterates while successively decrementing the DBSCAN search radius, allowing for an adaptive reduction factor based on the effective data density. The algorithm is demonstrated for a physics simulation application, where a surrogate model for fusion reactor plasma turbulence is generated with neural networks. A training dataset for the surrogate model is created with a quasilinear gyrokinetics code for turbulent transport calculations in fusion plasmas. The training set consists of model inputs derived from a repository of experimental measurements, meaning there is a potential risk of over-representing specific regions of this input parameter space. By applying the proposed reduction algorithm to this dataset, this study demonstrates that the training dataset can be reduced by a factor ~20 using the proposed algorithm, without a noticeable loss in the surrogate model accuracy. This reduction provides a novel way of analyzing existing high-dimensional datasets for biases and consequently reducing them, which lowers the cost of re-populating that parameter space with higher quality data.
△ Less
Submitted 22 November, 2021;
originally announced November 2021.
-
The data-driven future of high energy density physics
Authors:
Peter W. Hatfield,
Jim A. Gaffney,
Gemma J. Anderson,
Suzanne Ali,
Luca Antonelli,
Suzan Başeğmez du Pree,
Jonathan Citrin,
Marta Fajardo,
Patrick Knapp,
Brendan Kettle,
Bogdan Kustowski,
Michael J. MacDonald,
Derek Mariscal,
Madison E. Martin,
Taisuke Nagayama,
Charlotte A. J. Palmer,
J. Luc Peterson,
Steven Rose,
J J Ruby,
Carl Shneider,
Matt J. V. Streeter,
Will Trickey,
Ben Williams
Abstract:
The study of plasma physics under conditions of extreme temperatures, densities and electromagnetic field strengths is significant for our understanding of astrophysics, nuclear fusion and fundamental physics. These extreme physical systems are strongly non-linear and very difficult to understand theoretically or optimize experimentally. Here, we argue that machine learning models and data-driven…
▽ More
The study of plasma physics under conditions of extreme temperatures, densities and electromagnetic field strengths is significant for our understanding of astrophysics, nuclear fusion and fundamental physics. These extreme physical systems are strongly non-linear and very difficult to understand theoretically or optimize experimentally. Here, we argue that machine learning models and data-driven methods are in the process of resha** our exploration of these extreme systems that have hitherto proven far too non-linear for human researchers. From a fundamental perspective, our understanding can be helped by the way in which machine learning models can rapidly discover complex interactions in large data sets. From a practical point of view, the newest generation of extreme physics facilities can perform experiments multiple times a second (as opposed to ~daily), moving away from human-based control towards automatic control based on real-time interpretation of diagnostic data and updates of the physics model. To make the most of these emerging opportunities, we advance proposals for the community in terms of research design, training, best practices, and support for synthetic diagnostics and data analysis.
△ Less
Submitted 22 November, 2021;
originally announced November 2021.
-
Application of Gaussian process regression to plasma turbulent transport model validation via integrated modelling
Authors:
Aaron Ho,
Jonathan Citrin,
Fulvio Auriemma,
Clarisse Bourdelle,
Francis J. Casson,
Hyun-Tae Kim,
Pierre Manas,
Gabor Szepesi,
Henri Weisen,
JET Contributors
Abstract:
This paper outlines an approach towards improved rigour in tokamak turbulence transport model validation within integrated modelling. Gaussian process regression (GPR) techniques were applied for profile fitting during the preparation of integrated modelling simulations allowing for rigourous sensitivity tests of prescribed initial and boundary conditions as both fit and derivative uncertainties a…
▽ More
This paper outlines an approach towards improved rigour in tokamak turbulence transport model validation within integrated modelling. Gaussian process regression (GPR) techniques were applied for profile fitting during the preparation of integrated modelling simulations allowing for rigourous sensitivity tests of prescribed initial and boundary conditions as both fit and derivative uncertainties are provided. This was demonstrated by a JETTO integrated modelling simulation of the JET ITER-like-wall H-mode baseline discharge #92436 with the QuaLiKiz quasilinear turbulent transport model, which is the subject of extrapolation towards a deuterium-tritium plasma. The simulation simultaneously evaluates the time evolution of heat, particle, and momentum fluxes over $\sim10$ confinement times, with a simulation boundary condition at $ρ_{tor} = 0.85$. Routine inclusion of momentum transport prediction in multi-channel flux-driven transport modelling is not standard and is facilitated here by recent developments within the QuaLiKiz model. Excellent agreement was achieved between the fitted and simulated profiles for $n_e$, $T_e$, $T_i$, and $Ω_{tor}$ within $2σ$, but the simulation underpredicts the mid-radius $T_i$ and overpredicts the core $n_e$ and $T_e$ profiles for this discharge. Despite this, it was shown that this approach is capable of deriving reasonable inputs, including derivative quantities, to tokamak models from experimental data. Furthermore, multiple figures-of-merit were defined to quantitatively assess the agreement of integrated modelling predictions to experimental data within the GPR profile fitting framework.
△ Less
Submitted 3 May, 2021;
originally announced May 2021.
-
Neural network surrogate of QuaLiKiz using JET experimental data to populate training space
Authors:
Aaron Ho,
Jonathan Citrin,
Clarisse Bourdelle,
Yann Camenen,
Francis J. Casson,
Karel L. van de Plassche,
Henri Weisen,
JET Contributors
Abstract:
Within integrated tokamak plasma modelling, turbulent transport codes are typically the computational bottleneck limiting their routine use outside of post-discharge analysis. Neural network (NN) surrogates have been used to accelerate these calculations while retaining the desired accuracy of the physics-based models. This paper extends a previous NN model, known as QLKNN-hyper-10D, by incorporat…
▽ More
Within integrated tokamak plasma modelling, turbulent transport codes are typically the computational bottleneck limiting their routine use outside of post-discharge analysis. Neural network (NN) surrogates have been used to accelerate these calculations while retaining the desired accuracy of the physics-based models. This paper extends a previous NN model, known as QLKNN-hyper-10D, by incorporating the impact of impurities, plasma rotation and magnetic equilibrium effects. This is achieved by adding a light impurity fractional density ($n_{imp,light} / n_e$) and its normalized gradient, the normalized pressure gradient ($α$), the toroidal Mach number ($M_{tor}$) and the normalized toroidal flow velocity gradient. The input space was sampled based on experimental data from the JET tokamak to avoid the curse of dimensionality. The resulting networks, named QLKNN-jetexp-15D, show good agreement with the original QuaLiKiz model, both by comparing individual transport quantity predictions as well as comparing its impact within the integrated model, JINTRAC. The profile-averaged RMS of the integrated modelling simulations is <10% for each of the 5 scenarios tested. This is non-trivial given the potential numerical instabilities present within the highly nonlinear system of equations governing plasma transport, especially considering the novel addition of momentum flux predictions to the model proposed here. An evaluation of all 25 NN output quantities at one radial location takes $\sim$0.1 ms, $10^4$ times faster than the original QuaLiKiz model. Within the JINTRAC integrated modelling tests performed in this study, using QLKNN-jetexp-15D resulted in a speed increase of only 60 - 100 as other physics modules outside of turbulent transport become the bottleneck.
△ Less
Submitted 3 May, 2021;
originally announced May 2021.
-
Quasilinear gyrokinetic theory: A derivation of QuaLiKiz
Authors:
Cole Darin Stephens,
Xavier Garbet,
Jonathan Citrin,
Clarisse Bourdelle,
Karel Lucas van de Plassche,
Frank Jenko
Abstract:
In order to predict and analyze turbulent transport in tokamaks, it is important to model transport that arises from microinstabilities. For this task, quasilinear codes have been developed that seek to calculate particle, angular momentum, and heat fluxes both quickly and accurately. In this tutorial, we present a derivation of one such code known as QuaLiKiz, a quasilinear gyrokinetic transport…
▽ More
In order to predict and analyze turbulent transport in tokamaks, it is important to model transport that arises from microinstabilities. For this task, quasilinear codes have been developed that seek to calculate particle, angular momentum, and heat fluxes both quickly and accurately. In this tutorial, we present a derivation of one such code known as QuaLiKiz, a quasilinear gyrokinetic transport code. The goal of this derivation is to provide a self-contained and complete description of the underlying physics and mathematics of QuaLiKiz from first principles. This work serves both as a comprehensive overview of QuaLiKiz specifically as well as an illustration for deriving quasilinear models in general.
△ Less
Submitted 18 March, 2021;
originally announced March 2021.
-
First-principles-based multiple-isotope particle transport modelling at JET
Authors:
M. Marin,
J. Citrin,
C. Bourdelle,
Y. Camenen,
F. J. Casson,
A. Ho,
F. Koechl,
M. Maslov,
JET Contributors
Abstract:
Core turbulent particle transport with multiple isotopes can display observable differences in behaviour between the electron and ion particle channels. Experimental observations at JET with mixed H-D plasmas and varying NBI and gas-puff sources [M. Maslov et al., Nucl. Fusion 7 076022 (2018)] inferred source dominated electron peaking, but transport dominated isotope peaking. In this work, we app…
▽ More
Core turbulent particle transport with multiple isotopes can display observable differences in behaviour between the electron and ion particle channels. Experimental observations at JET with mixed H-D plasmas and varying NBI and gas-puff sources [M. Maslov et al., Nucl. Fusion 7 076022 (2018)] inferred source dominated electron peaking, but transport dominated isotope peaking. In this work, we apply the QuaLiKiz quasilinear gyrokinetic transport model within JINTRAC flux-driven integrated modelling, for core transport validation in this multiple-isotope regime. The experiments are successfully reproduced, predicting self consistently $ j $, $ n_{e} $, $ n_{Be} $, $ T_{e} $, $ T_{i} $, $ω_{tor}$ and the isotope composition. As seen in the experiments, both H and D profiles are predicted to be peaked regardless of the core isotope source. An extensive sensitivity study confirmed that this result does not depend on the specific choices made for the boundary conditions and physics settings. While kinetic profiles and electron density peaking did vary depending on the simulation parameters, the isotope ratio remained nearly invariant, and tied to the electron density profile. These findings have positive ramifications for multiple-isotope fuelling, burn control, and helium ash removal.
△ Less
Submitted 16 March, 2021;
originally announced March 2021.
-
Multiple-isotope pellet cycles captured by turbulent transport modelling in the JET tokamak
Authors:
M. Marin,
J. Citrin,
L. Garzotti,
M. Valovic,
C. Bourdelle,
Y. Camenen,
F. J. Casson,
A. Ho,
F. Koechl,
M. Maslov,
JET Contributors
Abstract:
For the first time the pellet cycle of a multiple-isotope plasma is successfully reproduced with reduced turbulent transport modelling, within an integrated simulation framework. Future nuclear fusion reactors are likely to be fuelled by cryogenic pellet injection, due to higher penetration and faster response times. Accurate pellet cycle modelling is crucial to assess fuelling efficiency and burn…
▽ More
For the first time the pellet cycle of a multiple-isotope plasma is successfully reproduced with reduced turbulent transport modelling, within an integrated simulation framework. Future nuclear fusion reactors are likely to be fuelled by cryogenic pellet injection, due to higher penetration and faster response times. Accurate pellet cycle modelling is crucial to assess fuelling efficiency and burn control. In recent JET tokamak experiments, deuterium pellets with reactor-relevant deposition characteristics were injected into a pure hydrogen plasma. Measurements of the isotope ratio profile inferred a Deuterium penetration time comparable to the energy confinement time. The modelling successfully reproduces the plasma thermodynamic profiles and the fast deuterium penetration timescale. The predictions of the reduced turbulence model QuaLiKiz in the presence of a negative density gradient following pellet deposition are compared with GENE linear and nonlinear higher fidelity modelling. The results are encouraging with regard to reactor fuelling capability and burn control.
△ Less
Submitted 16 March, 2021;
originally announced March 2021.
-
Flux-driven integrated modelling of main ion pressure and trace tungsten transport in ASDEX Upgrade
Authors:
O. Linder,
J. Citrin,
G. M. D. Hogeweij,
C. Angioni,
C. Bourdelle,
F. J. Casson,
E. Fable,
A. Ho,
F. Koechl,
M. Sertoli,
the EUROfusion MST1 Team,
the ASDEX Upgrade Team
Abstract:
Neoclassical and turbulent heavy impurity transport in tokamak core plasmas are determined by main ion temperature, density and toroidal rotation profiles. Thus, in order to understand and prevent experimental behaviour of W accumulation, flux-driven integrated modelling of main ion heat and particle transport over multiple confinement times is a vital prerequisite. For the first time, the quasili…
▽ More
Neoclassical and turbulent heavy impurity transport in tokamak core plasmas are determined by main ion temperature, density and toroidal rotation profiles. Thus, in order to understand and prevent experimental behaviour of W accumulation, flux-driven integrated modelling of main ion heat and particle transport over multiple confinement times is a vital prerequisite. For the first time, the quasilinear gyrokinetic code QuaLiKiz has been applied for successful predictions of core kinetic profiles in an ASDEX Upgrade H-mode discharge in the turbulence dominated region within the integrated modelling suite JETTO. Neoclassical contributions are calculated by NCLASS; auxiliary heat and particle deposition profiles due to NBI and ECRH prescribed from previous analysis with TRANSP. Turbulent and neoclassical contributions are insufficient in explaining main ion heat and particle transport inside the $q=1$ surface, necessitating the prescription of further transport coefficients to mimic the impact of MHD activity on central transport. The ion to electron temperature ratio at the simulation boundary at $ρ_\mathrm{tor} = 0.85$ stabilizes ion scale modes while destabilizing ETG modes when significantly exceeding unity. Careful analysis of experimental measurements using Gaussian process regression techniques is carried out to explore reasonable uncertainties. In following trace W impurity transport simulations performed with additionally NEO, neoclassical transport under consideration of poloidal asymmetries alone is found to be insufficient to establish hollow central W density profiles. Reproduction of these conditions measured experimentally is found possible only when assuming the direct impact of a saturated $(m,n)=(1,1)$ MHD mode on heavy impurity transport.
△ Less
Submitted 3 February, 2020;
originally announced February 2020.
-
Fast modeling of turbulent transport in fusion plasmas using neural networks
Authors:
Karel Lucas van de Plassche,
Jonathan Citrin,
Clarisse Bourdelle,
Yann Camenen,
Francis J. Casson,
Victor I. Dagnelie,
Federico Felici,
Aaron Ho,
Simon Van Mulders,
JET Contributors
Abstract:
We present an ultrafast neural network (NN) model, QLKNN, which predicts core tokamak transport heat and particle fluxes. QLKNN is a surrogate model based on a database of 300 million flux calculations of the quasilinear gyrokinetic transport model QuaLiKiz. The database covers a wide range of realistic tokamak core parameters. Physical features such as the existence of a critical gradient for the…
▽ More
We present an ultrafast neural network (NN) model, QLKNN, which predicts core tokamak transport heat and particle fluxes. QLKNN is a surrogate model based on a database of 300 million flux calculations of the quasilinear gyrokinetic transport model QuaLiKiz. The database covers a wide range of realistic tokamak core parameters. Physical features such as the existence of a critical gradient for the onset of turbulent transport were integrated into the neural network training methodology. We have coupled QLKNN to the tokamak modelling framework JINTRAC and rapid control-oriented tokamak transport solver RAPTOR. The coupled frameworks are demonstrated and validated through application to three JET shots covering a representative spread of H-mode operating space, predicting turbulent transport of energy and particles in the plasma core. JINTRAC-QLKNN and RAPTOR-QLKNN are able to accurately reproduce JINTRAC-QuaLiKiz T i,e and n e profiles, but 3 to 5 orders of magnitude faster. Simulations which take hours are reduced down to only a few tens of seconds. The discrepancy in the final source-driven predicted profiles between QLKNN and QuaLiKiz is on the order 1%-15%. Also the dynamic behaviour was well captured by QLKNN, with differences of only 4%-10% compared to JINTRAC-QuaLiKiz observed at mid-radius, for a study of density buildup following the L-H transition. Deployment of neural network surrogate models in multi-physics integrated tokamak modelling is a promising route towards enabling accurate and fast tokamak scenario optimization, Uncertainty Quantification, and control applications.
△ Less
Submitted 1 April, 2020; v1 submitted 13 November, 2019;
originally announced November 2019.
-
Non-Maxwellian fast particle effects in gyrokinetic GENE simulations
Authors:
A. Di Siena,
T. Görler,
H. Doerk,
R. Bilato,
J. Citrin,
T. Johnson,
M. Schneider,
E. Poli,
JET Contributors
Abstract:
Fast ions have recently been found to significantly impact and partially suppress plasma turbulence both in experimental and numerical studies in a number of scenarios. Understanding the underlying physics and identifying the range of their beneficial effect is an essential task for future fusion reactors, where highly energetic ions are generated through fusion reactions and external heating sche…
▽ More
Fast ions have recently been found to significantly impact and partially suppress plasma turbulence both in experimental and numerical studies in a number of scenarios. Understanding the underlying physics and identifying the range of their beneficial effect is an essential task for future fusion reactors, where highly energetic ions are generated through fusion reactions and external heating schemes. However, in many of the gyrokinetic codes fast ions are, for simplicity, treated as equivalent-Maxwellian-distributed particle species, although it is well known that to rigorously model highly non-thermalised particles, a non-Maxwellian background distribution function is needed. To study the impact of this assumption, the gyrokinetic code GENE has recently been extended to support arbitrary background distribution functions which might be either analytic, e.g. slowing down and bi-Maxwellian, or obtained from numerical fast ion models. A particular JET plasma with strong fast-ion related turbulence suppression is revised with these new code capabilities both with linear and nonlinear gyrokinetic simulations. It appears that the fast ion stabilization tends to be less strong but still substantial with more realistic distributions, and this improves the quantitative power balance agreement with experiments.
△ Less
Submitted 13 February, 2018;
originally announced February 2018.
-
Tractable flux-driven temperature, density, and rotation profile evolution with the quasilinear gyrokinetic transport model QuaLiKiz
Authors:
J Citrin,
C Bourdelle,
F J Casson,
C Angioni,
N Bonanomi,
Y Camenen,
X Garbet,
L Garzotti,
T Görler,
O Gürcan,
F Koechl,
F Imbeaux,
O Linder,
K van de Plassche,
P Strand,
G Szepesi,
JET Contributors
Abstract:
Quasilinear turbulent transport models are a successful tool for prediction of core tokamak plasma profiles in many regimes. Their success hinges on the reproduction of local nonlinear gyrokinetic fluxes. We focus on significant progress in the quasilinear gyrokinetic transport model QuaLiKiz [C. Bourdelle et al. 2016 Plasma Phys. Control. Fusion 58 014036], which employs an approximated solution…
▽ More
Quasilinear turbulent transport models are a successful tool for prediction of core tokamak plasma profiles in many regimes. Their success hinges on the reproduction of local nonlinear gyrokinetic fluxes. We focus on significant progress in the quasilinear gyrokinetic transport model QuaLiKiz [C. Bourdelle et al. 2016 Plasma Phys. Control. Fusion 58 014036], which employs an approximated solution of the mode structures to significantly speed up computation time compared to full linear gyrokinetic solvers. Optimization of the dispersion relation solution algorithm within integrated modelling applications leads to flux calculations $\times10^{6-7}$ faster than local nonlinear simulations. This allows tractable simulation of flux-driven dynamic profile evolution including all transport channels: ion and electron heat, main particles, impurities, and momentum. Furthermore, QuaLiKiz now includes the impact of rotation and temperature anisotropy induced poloidal asymmetry on heavy impurity transport, important for W-transport applications. Application within the JETTO integrated modelling code results in 1s of JET plasma simulation within 10 hours using 10 CPUs. Simultaneous predictions of core density, temperature, and toroidal rotation profiles for both JET hybrid and baseline experiments are presented, covering both ion and electron turbulence scales. The simulations are successfully compared to measured profiles, with agreement mostly in the 5-25% range according to standard figures of merit.
△ Less
Submitted 7 August, 2017; v1 submitted 3 August, 2017;
originally announced August 2017.
-
Comparison between measured and predicted turbulence frequency spectra in ITG and TEM regimes
Authors:
J. Citrin,
H. Arnichand,
J. Bernardo,
C. Bourdelle,
X. Garbet,
F. Jenko,
S. Hacquin,
M. J. Pueschel,
R. Sabot
Abstract:
The observation of distinct peaks in tokamak core reflectometry measurements - named quasi-coherent-modes (QCMs) - are identified as a signature of Trapped-Electron-Mode (TEM) turbulence [H. Arnichand et al. 2016 Plasma Phys. Control. Fusion 58 014037]. This phenomenon is investigated with detailed linear and nonlinear gyrokinetic simulations using the \gene code. A Tore-Supra density scan is stud…
▽ More
The observation of distinct peaks in tokamak core reflectometry measurements - named quasi-coherent-modes (QCMs) - are identified as a signature of Trapped-Electron-Mode (TEM) turbulence [H. Arnichand et al. 2016 Plasma Phys. Control. Fusion 58 014037]. This phenomenon is investigated with detailed linear and nonlinear gyrokinetic simulations using the \gene code. A Tore-Supra density scan is studied, which traverses through a Linear (LOC) to Saturated (SOC) Ohmic Confinement transition. The LOC and SOC phases are both simulated separately. In the LOC phase, where QCMs are observed, TEMs are robustly predicted unstable in linear studies. In the later SOC phase, where QCMs are no longer observed, ITG modes are identified. In nonlinear simulations, in the ITG (SOC) phase, a broadband spectrum is seen. In the TEM (LOC) phase, a clear emergence of a peak at the TEM frequencies is seen. This is due to reduced nonlinear frequency broadening of the underlying linear modes in the TEM regime compared with the ITG regime. A synthetic diagnostic of the nonlinearly simulated frequency spectra reproduces the features observed in the reflectometry measurements. These results support the identification of core QCMs as an experimental marker for TEM turbulence
△ Less
Submitted 12 July, 2017;
originally announced July 2017.
-
Real-time capable first principle based modelling of tokamak turbulent transport
Authors:
J. Citrin,
S. Breton,
F. Felici,
F. Imbeaux,
T. Aniel,
J. F. Artaud,
B. Baiocchi,
C. Bourdelle,
Y. Camenen,
J. Garcia
Abstract:
A real-time capable core turbulence tokamak transport model is developed. This model is constructed from the regularized nonlinear regression of quasilinear gyrokinetic transport code output. The regression is performed with a multilayer perceptron neural network. The transport code input for the neural network training set consists of five dimensions, and is limited to adiabatic electrons. The ne…
▽ More
A real-time capable core turbulence tokamak transport model is developed. This model is constructed from the regularized nonlinear regression of quasilinear gyrokinetic transport code output. The regression is performed with a multilayer perceptron neural network. The transport code input for the neural network training set consists of five dimensions, and is limited to adiabatic electrons. The neural network model successfully reproduces transport fluxes predicted by the original quasilinear model, while gaining five orders of magnitude in computation time. The model is implemented in a real-time capable tokamak simulator, and simulates a 300s ITER discharge in 10s. This proof-of-principle for regression based transport models anticipates a significant widening of input space dimensionality and physics realism for future training sets. This aims to provide unprecedented computational speed coupled with first-principle based physics for real-time control and integrated modelling applications.
△ Less
Submitted 15 June, 2015; v1 submitted 25 February, 2015;
originally announced February 2015.
-
Electromagnetic stabilization of tokamak microturbulence in a high-$β$ regime
Authors:
J. Citrin,
J. Garcia,
T. Goerler,
F. Jenko,
P. Mantica,
D. Told,
C. Bourdelle,
D. R. Hatch,
G. M. D. Hogeweij,
T. Johnson,
M. J. Pueschel,
M. Schneider
Abstract:
The impact of electromagnetic stabilization and flow shear stabilization on ITG turbulence is investigated. Analysis of a low-$β$ JET L-mode discharge illustrates the relation between ITG stabilization, and proximity to the electromagnetic instability threshold. This threshold is reduced by suprathermal pressure gradients, highlighting the effectiveness of fast ions in ITG stabilization. Extensive…
▽ More
The impact of electromagnetic stabilization and flow shear stabilization on ITG turbulence is investigated. Analysis of a low-$β$ JET L-mode discharge illustrates the relation between ITG stabilization, and proximity to the electromagnetic instability threshold. This threshold is reduced by suprathermal pressure gradients, highlighting the effectiveness of fast ions in ITG stabilization. Extensive linear and nonlinear gyrokinetic simulations are then carried out for the high-$β$ JET hybrid discharge 75225, at two separate locations at inner and outer radii. It is found that at the inner radius, nonlinear electromagnetic stabilization is dominant, and is critical for achieving simulated heat fluxes in agreement with the experiment. The enhancement of this effect by suprathermal pressure also remains significant. It is also found that flow shear stabilization is not effective at the inner radii. However, at outer radii the situation is reversed. Electromagnetic stabilization is negligible while the flow shear stabilization is significant. These results constitute the high-$β$ generalization of comparable observations found at low-$β$ at JET. This is encouraging for the extrapolation of electromagnetic ITG stabilization to future devices. An estimation of the impact of this effect on the ITER hybrid scenario leads to a 20% fusion power improvement.
△ Less
Submitted 5 September, 2014;
originally announced September 2014.
-
Nonlinear stabilization of tokamak microturbulence by fast ions
Authors:
J. Citrin,
C. Bourdelle,
J. Garcia,
J. W. Haverkort,
G. M. D. Hogeweij,
F. Jenko,
T. Johnson,
P. Mantica,
M. J. Pueschel,
D. Told,
JET-EFDA contributors
Abstract:
Nonlinear electromagnetic stabilization by suprathermal pressure gradients found in specific regimes is shown to be a key factor in reducing tokamak microturbulence, augmenting significantly the thermal pressure electromagnetic stabilization. Based on nonlinear gyrokinetic simulations investigating a set of ion heat transport experiments on the JET tokamak, described by Mantica et al. [Phys. Rev.…
▽ More
Nonlinear electromagnetic stabilization by suprathermal pressure gradients found in specific regimes is shown to be a key factor in reducing tokamak microturbulence, augmenting significantly the thermal pressure electromagnetic stabilization. Based on nonlinear gyrokinetic simulations investigating a set of ion heat transport experiments on the JET tokamak, described by Mantica et al. [Phys. Rev. Lett. 107 135004 (2011)], this result explains the experimentally observed ion heat flux and stiffness reduction. These findings are expected to improve the extrapolation of advanced tokamak scenarios to reactor relevant regimes.
△ Less
Submitted 26 June, 2013;
originally announced June 2013.
-
Ion temperature profile stiffness: non-linear gyrokinetic simulations and comparison with experiment
Authors:
J. Citrin,
F. Jenko,
P. Mantica,
D. Told,
C. Bourdelle,
R. Dumont,
J. Garcia,
J. W. Haverkort,
G. M. D. Hogeweij,
T. Johnson,
M. J. Pueschel,
JET-EFDA contributors
Abstract:
Recent experimental observations at JET show evidence of reduced ion temperature profile stiffness, hypothesised to be due to concomitant low magnetic shear (s) and significant toroidal rotational flow shear. Non-linear gyrokinetic simulations are performed, aiming to investigate the physical mechanism behind the observations. A comprehensive set of simulations are carried out, comparing the impac…
▽ More
Recent experimental observations at JET show evidence of reduced ion temperature profile stiffness, hypothesised to be due to concomitant low magnetic shear (s) and significant toroidal rotational flow shear. Non-linear gyrokinetic simulations are performed, aiming to investigate the physical mechanism behind the observations. A comprehensive set of simulations are carried out, comparing the impact on the ion heat flux of various parameters that differ within the data-set. These parameters include q, s, rotation, effect of rotation on the magnetohydrodynamic (MHD) equilibrium, R/L_n, beta_e, Z_eff, and the fast particle content. The effect of toroidal flow shear itself is not predicted by the simulations to lead to a significant reduction in ion heat flux, due both to an insufficient magnitude of flow shear and significant parallel velocity gradient destabilisation. It is however found that non-linear electromagnetic effects due to both thermal and fast-particle pressure gradients, even at low beta_e, can significantly reduce the profile stiffness. A total of five discharges are examined, at both inner and outer radii. For all cases studied, the simulated and experimental ion heat flux values agree within reasonable variations of input parameters around the experimental uncertainties.
△ Less
Submitted 16 July, 2013; v1 submitted 9 March, 2013;
originally announced March 2013.
