This paper is the first in a series of three papers to summarize the recent work of an European-wide collaboration which is ongoing since about one decade using Particle-in-Cell (PIC) methods in low temperature plasma physics. In the present first paper the main aspects of this computational technique will be presented. In the second paper, an overview of applications in low-temperature plasma modelling will be given, whereas the third part will put emphasis on the specific results of modelling ion thrusters.
Edge Localised Modes (ELMs) are universally recognised as one of the greatest threats to the viability of ITER and future fusion power plants based on the tokamak concept. They are plasma relaxations driven by MHD modes and are thought to originate in the steep pressure gradient region of the edge transport barrier characteristic of H-mode plasmas. In ITER, extrapolations from JET predict that Type I ELMs in the Q DT = 10 baseline scenario will expel between 3-8% of the 350 MJ plasma stored energy, depositing energy fluxes of 0.6 -3.4 MJm -2 on the divertor targets [1]. Only at the lowest values of this energy range would the subsequent target erosion be tolerable. Of late, concerns are being raised not just for the divertor targets, where most of the ELM energy is intercepted, but also for the main chamber walls to where ELM power fluxes are now known to extend.The mechanisms governing the ELM origin location and non-linear evolution within the H-mode pedestal and the subsequent cross-field propagation within the scrape-off layer (SOL) remain the subjects of keen debate. Once in the SOL, however, the thermal energy within the filament is removed predominantly by parallel losses to divertor targets, a process which is better understood but which is nevertheless complex, comprising both kinetic and fluid effects. This contribution aims to demonstrate how experiments and modelling at JET are significantly advancing our understanding of the ELM SOL parallel transport, providing many of the key elements required for an integrated, quantitative treatment of the ELM energy fluxes and their subsequent consequences for plasma-wall interaction.Infra-red thermography is extensively employed for divertor target measurements at JET and has recently been complemented by a unique new wide angle view of the main chamber wall surface. Such measurements are technically challenging due to the fast transient nature of the ELM and the presence of thin surface layers, which can differ radically from surface to surface. Taking proper account of this reveals that ELMs deposit energy preferentially (~ factor 2) in the outer target at low pedestal collisionality (ν*) but that this ratio inverts in favour of inner target energy deposition (up to ~ factor 3) with rising ν*.These target plate measurements of the ELM heat flux transient, in combination with fast triple Langmuir probe data, have provided the first known experimental evidence for
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.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.