Session UO08: MFE: Low-Aspect Ratio Tokamaks

The New PEGASUS-III Experiment

Minimizing or eliminating the need for induction from a central solenoid during startup, ramp-up and sustainment of a tokamak plasma is a critical challenge in magnetic fusion. Solenoid-free startup potentially simplifies the cost and complexity of reactor-class devices by reducing the technical requirements of, or the need for, a central solenoid. The PEGASUS-III Experiment is a new solenoid-free, low aspect ratio ST (A ≥ 1.22, I_p ≤ 0.3 MA, B_T ≤ 0.6 T, pulse length ~ 100 ms) focused on studying innovative non-solenoidal tokamak startup techniques.  It is equipped with: a new local helicity injection (LHI) system capable of I_p ~ 0.3 MA; sustained and transient coaxial helicity injection (CHI) systems; and an RF system for initial electron Bernstein wave (EBW) and electron cyclotron (EC) heating. Initial experiments will establish high-I_p LHI scenarios, followed by deployment of transient CHI, modest sustained CHI, and low-power RF studies. PEGASUS-III will provide key enabling reactor relevant technology to directly test proposed plasma startup, ramp-up scenarios envisioned for larger scale STs, investigating methods to synergistically improve the target plasma for consequent bootstrap and NBI current sustainment.   

Predict-first startup scenario modelling supporting MAST-U commissioning

A reduced semi-empirical model for breakdown and the initial I_p ramp-up in spherical tokamaks (STs) was recently developed to support the preparation and execution of plasma startup in MAST-U.  The reduced model uses vacuum field calculations in order to achieve rapid solutions suitable for predict-first and between-shot analysis applications. The model was developed using a large database of startup discharges from NSTX, NSTX-U and MAST that use induction and different approaches to pre-ionization.  One key finding is that STs achieve a rapid breakdown (several milliseconds) by accessing a large E/P regime where ionized electrons are continually accelerating. Another finding is that the timescale of the initial increase in I_p is reproduced only if a mechanism for a self-generated electric field is included, similar to conventional-A tokamaks. The first plasma experiments on MAST-U are in good agreement with the model predictions and identified additional constraints that improve the predictive model.   

Core-Edge Coupled Modeling of NSTX in TRANSP with Reduced Scrape-Off-Layer Model

The dynamic interplay between the core and the edge plasma has important consequences in the confinement and heating of fusion plasma. The transport of the SOL plasma imposes boundary conditions on the core plasma, and the recycled neutrals in the SOL influence the core ionization sources and NBI deposition. In order to better study these effects in a self-consistent, time-dependent fashion, a new reduced model has been developed that calculates separatrix temperature and density given the core to edge particle and power fluxes and recycling coefficients. The reduced model is coupled with TRANSP to provide time-dependent boundary conditions for the core solver, which subsequently improves various components of core calculations. The reduced SOL plasma model will be supplemented by kinetic neutral simulations with the DEGAS2 code. We will demonstrate the improved data interpretation capabilities of TRANSP by comparing calculated quantities with measurements in NSTX discharges. Implications for future interpretive and predictive simulations will also be discussed.   

Energetic particle effect on neoclassical tearing mode stability in NSTX

The impact of energetic particles on neoclassical tearing mode (NTM) stability is modeled and compared to experiments in the spherical tokamak NSTX. It is shown that energetic particles can be an important destabilizing mechanism for neoclassical tearing modes, in that they allow small magnetic islands to overcome the polarization current stabilization effect, and the magnetic island growth may be damped with the loss of energetic particles due to orbit stochasticization. These results are obtained using the energetic particle and magnetic island parameters determined self-consistently by TRANSP simulations augmented by the “kick model” for energetic particle transport by instabilities [Podestà et al., Plasma Phys. Control. Fusion 59 095008 (2017)]. Inclusion of energetic particle effect improves the agreement between measured and predicted island width time evolution and may provide new insight on neoclassical tearing mode onset and growth. A new model is being implemented and tested in TRANSP for analysis and prediction of NTM stability in time-dependent simulations.   

Self-driven current generation in spherical tokamak plasmas with magnetic island perturbations

Magnetic islands, by altering the topological structure of the confining magnetic field, have varied and complex impacts on plasma transport and confinement in fusion experiments. One novel effect revealed by global gyrokinetic simulations results from island-induced three-dimensional ambipolar electric field. A magnetic island is shown to drive non-resonant electric potential islands centered at both the inner and outer edge of the magnetic island. Such potential islands may introduce a major change in plasma self-driven current through an efficient nonlinear parallel acceleration of electrons in conventional tokamak geometry, resulting in a significant global reduction of electron current with respect to the neoclassical bootstrap current. Remarkably, this potential island effect is weaker in low aspect ratio spherical tokamak regimes, and the overall island-induced bootstrap current reduction is significantly smaller than in large aspect tokamaks.  The reduction of the axisymmetric current scales with the square of island width, and in the meantime, a significant helical current is generated in the island region. On the other hand, turbulent fluctuations may drive a substantial anomalous bootstrap current in STs.   

Numerical and experimental investigation of ELM filaments on NSTX

Edge localized modes (ELMs) are routinely observed in H-mode plasma regimes of the National Spherical Torus Experiment (NSTX). During the ELM’s explosive outburst significant amount of heat and particles are ejected from the plasma in the form of filaments. They carry enough energy to the plasma facing components to be able to permanently damage them, hence, it is important to mitigate these events. Studying the fast dynamics of the ELM filaments could lead to development of novel ELM mitigation techniques. Gas-puff imaging was utilized on NSTX to study the dynamics of the ELM filaments. To estimate their frame-by-frame velocity and structure evolution novel analysis techniques were developed.

 

Our initial analysis focused on a single ELM event which featured a distinct ELM filament with clear radial and poloidal velocity peaks. Investigation of the experimental data revealed that the structural parameters of the ELM filament are not significantly different from the intermittent blobs in the scrape-off layer. However, their non-linear triggering phase and propagation dynamics differ greatly. To find the underlying triggering and propelling mechanism, non-linear numerical simulations were performed with M3D-C1 and NIMROD, and they were compared to the experimental findings.   

Ion gyro orbit heat load simulations on real CAD using open source software

Accurate and precise heat load predictions are critical to ensuring that plasma facing component (PFC) designs meet performance targets in high power density tokamaks.  Recently, the open source Heat flux Engineering Analysis Toolkit (HEAT) was developed to compute time-evolving PFC heat loads incident upon fully featured 3D PFC CAD geometries.  Up until now, HEAT has simulated heat loads under the 'optical approximation' - the assumption that heat is transported directly along the magnetic field lines.  In reality, particles follow helical 'gyro orbit' trajectories as they precess about the magnetic field lines at finite Larmor radii.  Including the effects of gyro orbit trajectories in the heat load calculations can result in deviations from the optical approximation for the ion heat flux, as particles can access regions that are magnetically shadowed from straight field line fluxes.  This effect is increased during transients such as ELMs where free streaming of ions at the pedestal temperature can increase the radius.   A new HEAT module has been developed that enables users to calculate gyro orbit heat loads.  A side by side comparison between optical and gyro orbit results will be presented using castellated tile geometries for the NSTX-U divertor.     

Infrared Thermography Diagnostics Design and Heat Flux Calculations on NSTX-U

NSTX-U will operate with a plasma current up to 2MA for 5 seconds, and with NBI heating power up to 10MW. The increasing challenge for the safety of the plasma facing components (PFCs) requires a large coverage of IR diagnostics that view the PFCs. High spatial (1 mm) and time (several kHz) resolution IR cameras will be implemented to view the lower and upper divertors PFCs, with an 11° field of view. A wide-angle fast IR camera with a 108° field of view was designed to monitor the central stack, outer chamber, and upper and lower divertors simultaneously. A new 2-D heat flux calculation is being developed that allows temperature dependent PFC material properties, and is being compared with the THEODOR code [A. Herrmann et al., 1995 Plasma Phys. Control. Fusion 37 17]. The IR camera data will inform the new transient heat flux calculation, and in this presentation, we will discuss the effect of mesh size and meshing method on the calculated heat flux.   

Accessibility enhanced ECCD for plasma current start-up and ramp-up *

An efficient low-field-side launched X-mode ECH current start-up and ramp-up regime was identified. By choosing an appropriate operational density range, it is possible to access the EC resonance for the LFS X-mode. In this regime, the current drive is particularly efficient due to the strong cyclotron interaction with unidirectional passing electrons aided by the wave accessibility. The accessible density goes up with the square of frequency or magnetic field which favors high field devices including compact fusion pilot plant. This concept can now be tested in the higher field ST experimental devices becoming available such as NSTX-U, MAST, and ST-40.   

Particle balance analysis of RMP discharges in L- and H-mode at MAST

The application of resonant magnetic perturbations (RMPs) at MAST has been shown to cause a reduction in density, the so-called particle pump-out effect, during discharges which have a particular MHD response. This can occur in both L-mode and H-mode discharges. An analysis of the changes in fueling and exhaust using a 0-D single reservoir particle balance of the main ion species has shown that during RMP application there is an increase in total fueling to the plasma but also a significant drop in particle confinement time such that there is a net particle pump-out. To more accurately calculate the recycling and particle fueling, a detailed analysis of the measured D_α emission was performed. Synthetic diagnostics were created with the ray-tracing code CHERAB using plasma sources from modelling of these discharges done with the 3-D plasma and neutral code EMC3-EIRENE. The generated synthetic images were compared to the corresponding absolutely calibrated images from experiment to better constrain the experimental measurements and account for surface reflections. Results show there is a relatively greater increase in fueling flux during RMP application than shown in the initial analysis, causing a greater reduction in the particle confinement time in that period.   

Overview of recent results from the ST40 high-field spherical tokamak

ST40 is a high-field spherical tokamak (ST) built and operated by Tokamak Energy Ltd., a privately funded company based in the UK. The goal of ST40 is to extend the high-field ST physics basis and test engineering solutions for future ST reactors.

ST40 Programme 2.2 (P2.2) operations began in March 2021 and are scheduled to run until January 2022. The target parameters for P2.2 are: BT=3T (with LN2 cooling of the Cu magnet system), IP=1MA, PNBI=1.5MW (0.8MW/50kV, 0.7MW/25kV), flat-top duration =200ms, A=1.6-1.9, and RGeo=0.4-0.5m. Operations are in hydrogen and both double null diverted (DND) and centre column limited configurations are possible. Plasma start-up is achieved via merging-compression (MC), where two high voltage in-vessel coils are used to initiate the plasma. ST40 diagnostics include: Interferometers, X-ray Crystal Spectrometer, Diagnostic Neutral Beam and Visible Spectrometer (for Charge Exchange Recombination Spectroscopy), Neutral Particle Analyser, Soft X-ray Camera/Spectrometer, Neutron and Hard X-ray Spectrometers, Divertor/Main Chamber IR Cameras, Magnetics and ECE Radiometer. A Thomson Scattering system is planned for later in 2021.

Some of the goals of ST40 P2.2 are to maximise central ion temperature, study hot-ion modes, achieve H-mode operation and investigate confinement in a high-field ST. To date, a plasma current of 800kA has been reached at a toroidal field of 1.9T. At 450kA, an ion temperature of around 2keV has been obtained with one neutral beam at 50kV.

This presentation will provide an overview of the ST40 high-field spherical tokamak and latest results from P2.2 operations.   

Integrated Modelling of Plasmas in the ST40 High-Field Spherical Tokamak

The ST40 tokamak [1], built and operated by Tokamak Energy, has recently been upgraded with upper and lower divertors to enable double null diverted operations with up to 1MA of plasma current and 1.5MW of neutral beam heating. ST40 is a high field spherical tokamak (ST), B_T=3T at R₀=0.4-0.5m, A=1.6 – 1.9 with a goal to extend the high field ST physics basis. Crucially, transport and confinement in high field, high temperature STs will be explored in support of the design of future ST power plants [2]. Extensive modelling activities have been undertaken to support ST40 operations. Integrated data-analysis and modelling has been carried out on measurements from the recent experimental campaign for data and model validation. A range of plasma equilibrium in double-null configuration have been designed along with detailed scenario modelling, including 1.5D transport simulations and 2D SOL modelling. Gyrokinetic analysis has been performed to assess the level of expected turbulent transport. Building upon the NSTX pedestal database the pedestal width and height in the high performance ST40 scenarios have been predicted. MHD stability analysis and beta limit have been assessed. ST40 will be initially operated in hydrogen with up to 1.5 MW of NBI (0.8MW at 55kV and 0.7MW at 25kV). ST40 will be upgraded in view of the follow up Programme in deuterium, with 2MW of 55kV NBI and around 1.6MW 105/140GHz ECRH, and a new inner vacuum vessel and PFCs. Careful analysis of the power deposited in the divertor during high performance operation has also been carried out.

[1] M. Gryaznevich, O. Asunta (2017) Fusion Eng. Des. 123 177-180

[2] P F Buxton et al (2019) Plasma Phys. Control. Fusion 61 035006   

Linear micro-stability properties of a high β 1GW spherical tokamak

Spherical tokamaks (STs) offer an attractive route towards a compact high performance fusion reactor as historically they have achieved high plasma beta and strong shaping that facilitate plasmas with high bootstrap current. While scaling laws can be used to estimate confinement in a reactor relevant ST, their value is limited if the plasma regime is a substantial extrapolation from the parameter space used to generate these laws. In reality the confinement will be set by the turbulence that would arise in such a device, and any reduced transport model needs to describe the relevant characteristics of the dominant turbulence.

The local initial value gyrokinetic solvers, GS2 and CGYRO are used to analyse the nature of the modes that arise in a conceptual 1 GW ST. The dominant linear instabilities are found to be a collisional micro-tearing mode (MTM) and a kinetic ballooning mode (KBM) at the ion scale below k_yρ_s = 1. Towards the electron scale around k_yρ_s = 5, a collisionless MTM is found. Furthermore, when k_yρ_s > 10 the equilibrium examined is found to be completely stable, with no electron temperature gradient (ETG) mode. The KBMs and high k_yρ_s collisionless MTMs are suppressed by the anticipated level of flow shear, while the low k_yρ_s collisional MTMs are robust and are therefore expected to be the dominant source of transport nonlinearly. These low k_yρ_s MTMs have extended eigenfunctions in ballooning space indicating that nonlinear simulations may require very high radial resolution approaching the electron scale.

The dependence of these linear modes on different equilibrium parameters are examined to determine possible routes to stabilise each mode. With this information, potential options for generating a more stable equilibrium are discussed and a marginally stable plasma equilibrium is found linearly.