Session TI2: Low Temperature Plasmas, Stellarator, and Disruptions

Cavitations induced by plasmas, plasmas induced by cavitations, and plasmas produced in cavitations

Cavitation bubbles are not static bubbles but have dynamics of expansion, shrinkage, and collapse. Since the collapse of a cavitation bubble is roughly an adiabatic process, the inside of the bubble at the collapse has a high temperature and a high pressure, resulting in the production of a plasma. This talk will be focused on cavitation-related plasma phenomena and the role of the cavitation bubble in the synthesis of nanoparticles. A method for inducing a cavitation bubble is laser ablation in liquid. After the disappearance of laser-produced plasma with optical emission, we have observed the formation of a cavitation bubble. We have found that the inside of the cavitation bubble is the reaction field for the synthesis of nanoparticles. The atomic and molecular species ejected from the ablation target toward the liquid are transported into the cavitation bubble, and they condense into nanoparticles inside it. It is important to note that nanoparticles are stored inside the cavitation bubble until its collapse. We have shown that the size and the structure of nanoparticles are controlled by controlling the dynamics of the cavitation bubbles. Another method for inducing cavitation bubbles is to use ultrasonic power. We have found a simple method for the efficient production of standing cavitation bubbles. The method is just inserting a punching metal plate into water irradiated by ultrasonic wave. The depth of water and the position of the punching plate should be tuned precisely. We have proposed the mechanism of the efficient production of cavitation bubbles by this method. Currently, we try to have electric discharges in cavitation bubbles with the intention of realizing nonequilibrium sonochemistry. In particular, the electric discharge in a laser-induced cavitation bubble shows interesting distortion of the bubble shape, which suggests the electrostatic characteristics of the cavitation bubble.   

Generation of anomalously energetic suprathermal electrons by an electron beam interacting with a nonuniform plasma

Electrons emitted by electrodes surrounding or immersed in the plasma are accelerated by the sheath electric field and become electron beams penetrating the plasma. In plasma applications where controlling the electron velocity distribution function (EVDF) is crucial, these beams are an important factor capable of modifying the EVDF and affecting the discharge properties. Recently, it was reported that an EVDF measured in a dc-rf discharge with 800 V dc voltage has not only a peak of 800 eV electrons emitted from the dc-biased electrode, but also a peak of suprathermal electrons with energy up to several hundred eV. Initial explanation of the suprathermal peak suggested that the fast long plasma waves excited by the beam decay parametrically into ion acoustic waves and short plasma waves with much lower phase velocity which accelerate bulk electrons to suprathermal energies. Particle-in-cell simulation of a dc beam-plasma system, however, reveals that the short waves appear not due to the parametric instability, but due to the plasma nonuniformity. Moreover, the acceleration may occur in two stages. Plasma waves excited by the beam in the middle of the system propagate towards the anode and enter the density gradient area where their wavelength and phase speed rapidly decrease. Acceleration of thermal electrons by these waves is the first stage. Some of the accelerated electrons reflect from the anode sheath, travel through the plasma, reflect near the cathode, and enter the accelerating area again but with the energy higher than before. The acceleration that occurs now is the second stage. The energy of a particle after the second acceleration exceeds the initial thermal energy by an order of magnitude. This two-stage mechanism plays a role in explaining previous observations of energetic suprathermal electrons in similar discharges. The study is performed in collaboration with I. D. Kaganovich (PPPL), P. L. G. Ventzek and L. Chen (Tokyo Electron America).   

Models, assumptions, and experimental tests of flows near magnetized boundaries

We present a history of research on the magnetized plasma boundary and recent first measurements of particle flows in such structures in laboratory plasmas using multi-dimensional laser-induced fluorescence (LIF). Our measurements show that the canonical model for this boundary proposed in 1982 [\textit{Chodura, Phys. Fluids (1982)}] is inaccurate for systems where the ion-neutral collision length is less than at least 4 times the ion gyro radius. Rather, our measurements validate more sophisticated plasma boundary fluid models that take neutral collisions into account [\textit{Riemann, Phys. Plasmas (1994); Ahedo, Phys. Plasmas (1997); Siddiqui et al., Phys. Plasmas (2014)}]. In light of these results, we show that both three-dimensional ion and neutral velocity distribution functions are strongly affected near the boundary. We discuss effects of these perturbed distributions on wall loading and erosion in experiments and applications such as divertor tokamak scrape-off layers and Hall thrusters. Finally, we propose modern definitions of the oft-used term, ``magnetic presheath.''   

Critical particle circulation caused by high-performance steady-state plasma discharge

Steady-state operation focused on the fusion reactor has been investigated in magnetic confined fusion devices, and plasma performance and duration time are steadily extended by the improvement of the quality of plasma heating and sophisticating plasma operation using the understanding of long-pulse plasma experiments. When higher-performance helium steady-state plasma discharges with duration time over 40 min, electron density of 1.2x10$^{19}$ m$^{-3}$, ion and electron temperatures over 2 keV and heating power of 1.2MW were repeatedly achieved in LHD, time-evolution of the wall-pumping and increasing frequency of impurity contaminations around the plasma edge clearly occurred. These are strongly related to the increasing mixed-material layer caused by continuous divertor erosion around geometrical dense divertor plates, which consists of carbon (\textgreater\ 90{\%}) and iron (\textless\ a few {\%}) with amorphous structure, that can retain the helium particles and affect the particle balance in long-pulse discharges. The mixed-material layer is easily exfoliated by the thermal stress and helium explosion in the layer, and small pieces of exfoliation enter the plasma edge in all toroidal sections. Uncontrolled flake contamination was one of the causes of plasma termination in long-pulse experiments. Increased plasma performance using higher heating power ($\sim$ 3.3 MW) with high quality makes robust plasma against impurity contaminations, and then a small amount of contamination of mixed-material does not terminate the helium plasma. Carbon impurity was circulated from the divertor plates and around the plates to the plasma edge in long-pulse plasma discharges, and the circulation was increased by the plasma duration and performance. The eroded material plays an important role in degrading the plasma performance as an impurity source and in the controllability of particle fueling in long-pulse discharges.   

Beta induced Alfven eigenmode excitation in the strongly shaped H-1NF stellarator

Recent advances in the modelling, analysis, and measurement of fluctuations in strongly shaped 3D magnetic configurations have significantly improved the diagnosis and understanding of Alfven eigenmodes in the H1 heliac. Visible light emissions from H1 reveal low frequency (10-40 kHz) electron density fluctuations correlated with magnetic measurements indicative of Alfven eigenmodes. Full 3D tomographic inversion of the emission was accomplished using a novel synchronous imaging technique, which achieves high SNR and aliases high frequency modes to frequencies that can be easily imaged. Excellent agreement is found between both the frequency and spatial structure of the mode and ideal MHD calculations of Beta-induced Alfven Eigenmodes using the CAS3D code. The resulting measurements and modelling demonstrates that the dominant low frequency Alfven eigenmodes in H1 are principally the axi-symmetric variety, similar to the modes in tokamaks. These results show experimentally that Alfven eigenmodes can exist in the acoustic range of frequencies even at low beta (approx. 10$^{-4})$ due to strong shaping in stellarators such as H1, unlike in tokamaks where significantly higher beta is required. Importantly, these modes are observed in the absence of confined ions near the Alfven velocity (V$_{A})$. Typical ion temperatures of 20 eV correspond to ion velocities several orders of magnitude below V$_{A}$. However, similar temperature thermal electrons closely match V$_{A}$ and provide a potential mechanism for mode excitation. Additionally, there is evidence that the complex harmonic structure of the H1 magnetic field may allow mode excitation by ion energies corresponding to velocities which are a small fraction of V$_{A}$.   

Optimization of Massive Impurity Injection Techniques for Thermal Quench Mitigation and Current Quench Control on DIII-D

Recent DIII-D experiments demonstrate the ability of massive impurity injection techniques to effectively control current quench (CQ) timescales and thermal quench (TQ) radiation fractions, which are essential design requirements for the ITER disruption mitigation system (DMS). Allowable CQ timescales for ITER are constrained by both a lower bound due to eddy current forces as well as an upper bound due to halo currents, and this must be achieved while maintaining sufficiently high radiated powers to minimize thermal loads. The DIII-D shattered pellet injection system has been modified to allow formation of mixed species pellets with variable quantities of high-Z radiating impurities (Ne) and main ions $(D_2)$, which is shown to provide control of TQ radiation fractions and the resulting post-TQ plasma resistivity which determines the CQ rate. By varying the pellet composition ranging from pure $D_2$ to pure Ne, TQ radiation fractions are observed to saturate with modest quantities of Ne, indicating that relatively small quantities of the radiating impurity provide effective thermal mitigation. Resulting CQ durations are found to remain within scaled eddy and halo current limits predicted for ITER, demonstrating that integrated control of TQ and CQ properties during disruption mitigation can be achieved. The effectiveness of these mitigation techniques in disruptive plasma scenarios with large MHD instabilities is also crucial for ITER. Such effects are observed to lead to increased assimilation of injected impurities during the TQ, implying the importance of MHD mixing during the initial assimilation. Longer timescale CQ metrics are relatively unaffected by the pre-existing MHD activity, allowing effective mitigation of electromagnetic loads. These results provide further confidence in the implementation of these injection techniques in the final design of the ITER DMS.