Session TP10: Poster Session VII: Magnetic Fusion: DIII-D 2, Particle Acceleration, Beams, Relativistic Plasmas, Basic Plasma: Waves and Instabilities (9:30am-12:30pm)

Axial Magnetic Field Effects on Magnetically Insulated Line Oscillators (MILO)

The Magnetically Insulated Line Oscillator (MILO) is a crossed-field high power microwave device capable of multi-gigawatt operation. Intense axial currents generate a self-magnetic field strong enough that no external magnets are required, but contribute to low device efficiency (typically \textless 10{\%}). While MILOs have been studied for several decades, many fundamental questions remain. The University of Michigan is currently revisiting MILO theory and design, taking advantage of recent improvements in 3D particle-in-cell codes, pulsed power technology, and additive manufacturing techniques to study MILO. In particular, a MILO is being explored with an external, axial magnetic field to prevent electron losses before magnetic insulation is established. The operating condition, stability, and efficiency of MILOs will be investigated with versus without axial magnetic fields, as well as MILO variants capable of low voltage or multi-frequency operation. Simulations of axial magnetic field effects on MILO operation will be presented.   

An Exact Treatment of Helix Traveling Wave Tubes with Cold Tube Loss

Recent work on an exact treatment of a thin tape helix traveling wave tube (TWT) indicates that Pierce's classical linear theory requires revision at high beam currents to include space charge effects on the circuit mode [1]. These circuit mode space charge effects are characterized by a new parameter q, which acts in an analogous manner to the familiar space charge parameter Q that affects the beam mode. However, this approach has the crucial assumption that there are no ohmic losses in the tube, setting Pierce's cold tube loss parameter d $=$ 0. Here, we include these lossy effects by introducing an imaginary component of permittivity into the dielectric support structure and propose a modified dispersion relation that takes the effects of both q and d into account. We demonstrate the validity of our model by comparing our results to the classical theory in test cases with both uniform and non-uniform attenuation over the length of the tube. [1] P. Wong, D. Chernin, and Y. Y. Lau, IEEE Electron Device Lett. 39, 1238 (2018).   

Experimental Measurements and DFT-Based Modeling of Secondary Electron Yield of Materials of Interest to High Power Vacuum Electron Devices

Vacuum electron devices can experience degraded performance, including complete failure, due to multipactor breakdown (MPB). This is tied to the production and acceleration of secondary electrons due to electron impact and coupling to the RF fields. In order to better understand MPB with materials of interest, controlled measurements and density-functional theory (DFT)-based modeling of SEY are being conducted. SEY from electron bombardment in the low energy regime (10 eV to 1 keV) as a function of incident angle for Cu, Monel, Kovar, Invar, Al, and Fe (Cu Plated) has been measured. In addition, various surface cleaning protocols will be tested. The ability of DFT to reproduce the frequency dependent dielectric function of copper has been tested. The DFT data can be used to demonstrate that SEY can be predicted from first principles using Monte Carlo simulations. As a first step, two DFT codes Abinit and Vasp, both based on pseudopotentials and plane wave basis sets, have been used to calculate the frequency dependent dielectric function and energy loss function of copper. The best protocol for the calculation of these functions was obtained by comparing the computed and experimental values determined for copper in the optical limit.   

Highly-efficient terahertz radiation generation from laser-microplasma-waveguide interactions

When a relativistic laser pulse entersa micro-sized plasma waveguide, electrons are extracted from the inner surface and accelerated to hundreds of MeV. They emit a powerful THz pulse at the channel exit due to coherent diffraction radiation. In this work we investigate the dependence of THz generation on different target parameters via 3D particle-in-cell simulations, aiming at achieving a laser-to-THz conversion efficiency beyond the state-of-art. We find that when the transverse light pressure (TLP) of the fundamental waveguide mode is significantly strong (typically when the laser is intense and channel radius is small), the efficiency approaches its maximum on normal incidence. With appropriately choosing the channel length, one can control the divergence and charge of the electron bunch, maximizing the conversion efficiency. In the case that TLP is weak (due to practical issues with laser facilities or micro-engineering techniques), the laser incidence angle should be optimized in order to obtain a sufficient number of energetic electrons. Simulation results indicate that a conversion efficiency of over 1{\%} can be reached with the state-of-art laser facilities and micro-waveguide targets.   

Generation of pulsed THz signal by laser wake driven in a stratified leaky plasma channel

The laser wake, driven by a terawatt pulse in a radially uniform, longitudinally stratified plasma, is known to emit the broadband THz radiation in the forward hemisphere [L. M. Gorbunov and A. A. Frolov, JETP 83, 967 (1996).] We consider a more complex situation, when the plasma is pre-shaped as a leaky channel (still longitudinally stratified). The plasma density thus has a parabolic dip near axis, while rapidly falling off at the periphery. The wake-emitted THz radiation thus escapes into vacuum, which changes the radiation condition and emission pattern. An additional component to the THz rotational current is generated via coupling longitudinal electron velocity associated with the wake to the radial density gradient in the channel. Were the plasma longitudinally uniform, this pulsed current would be sub-luminal, and the THz signal radially evanescent. Longitudinal modulation of the background plasma density makes plasma wake locally superluminal, permitting generation of the outgoing THz radiation. Competition between the two mechanisms of producing THz radiation and emerging plasma diagnostic opportunities are explored.   

Ultrashort Pulsed Laser Filamentation Microwave Emission Physics

The electrodynamics of microwave emission from the plasma left behind by an ultrashort pulsed laser optical pulse after it has self-focused in air via the Kerr effect is explored theoretically and experimentally. The mechanism for and distribution of electrical current responsible for electromagnetic radiation in the microwave regime is not well-understood. Our goal here is to establish a consistent picture of the physical processes that need to be modeled by an integrated, complete, and predictive simulation under development. To this end, measurements of the filament's radius, electrical conductivity, and axial electric current time integral are made along its length at a broad range of atmospheric pressures. The resultant microwave radiation pattern at great distance is also measured. The results are explained in terms of a consistent set of simulations and/or calculations of electron ionization and acceleration by the optical field, subsequent evolution by a nonisotropic form of Generalized Ohm's Law, and radiated electromagnetic field.   

Polarization effects on RF emission from ultrashort laser pulse generated plasma

Ultrashort pulse laser generated plasma has long been studied as a source of THz frequency electromagnetic emission. Recent work has shown that multiple centimeter long plasma columns formed by focusing an ultrashort pulse laser in air is also a source of broadband radio frequency (RF) emission. In order to better understand the current source of the electromagnetic pulse, we present a comparative study of RF generation for incident linearly and circularly polarized laser light. Circular polarization, which generates a slightly hotter electron population than linearly polarized light, is shown to increase the RF signal. Experimental results are compared to a temperature dependent electron diffusion model with multiphoton ionization.   

Plasma Length Dependence of Broadband Microwave Radiation from Laser Plasma in Air

A high power ultrashort laser pulse focused in air generates a plasma that radiates broadband electromagnetic waves. The transient current source responsible for radiation remains an open area of study. To better understand how the laser drives currents in plasma, we investigate the dependence of radiation at microwave frequencies on the plasma length. Radiation patterns of the microwaves are consistent with currents that flow longitudinally in the direction of laser propagation. We measured angular emission pattern of microwave frequency spectrum to determine microwave radiation source characteristics. At longer focal lengths and higher energies, the laser pulse intensity in the focal plane varies for a fixed beam diameter. After normalizing the pulse intensities, the emission patterns are compared to images of the visible plasma fluorescence. Relationships between frequency spectra and images indicate the dominant components of the radiation shift to lower frequencies with increased plasma lengths. Our research demonstrates the radiation mechanism is coupled with total variation in the current along the entire plasma length.   

Background gas species and pressure dependence of RF emissions generated by laser-produced filament plasmas

The Air Force Research Lab is investigating RF emissions generated by ultra-short pulse lasers. An 800 nm, terawatt class laser is used to propagate a plasma filament with various background gases under a range of pressures to study the RF emission from 2-40 GHz of the filament. Air pressure has been previously shown to have an inverse relationship with the amplitude of the electric field wave form, however there is limited data with specific background gas species. Nitrogen, helium, argon, and krypton gases will be used to gain a better understanding the of contributions election-neutral and electron-ion collisions to RF emission or if the RF is independent of these factors. Using microwave horns that measure from 2-13 GHz and 13-40 GHz, the effects of the background gas species and pressure will be quantified and their relation to the RF emission will be presented.   

Time-resolved conductivity measurements of air plasmas created by nonlinearly propagating ultrashort laser pulses

The properties of the short-lived plasma generated by a high peak power, ultrashort laser pulse propagating in the atmosphere depend sensitively on the evolution of the laser pulse intensity profile. The intensity is determined by nonlinear optical effects arising from the bound electrons in the air molecules, in addition to natural diffraction and plasma-induced defocusing. The simultaneous contribution of these effects to the laser intensity means that the plasma properties are difficult to predict. However, it is possible to experimentally measure the plasma conductivity by probing the plasma with weak external electric fields. We measure the time history of the conductivity by propagating the laser through a hole in a waveguide that contains a CW microwave field oscillating at a frequency of 40 GHz. The 25 picosecond period of the microwave field enables small temporal resolution of the conductivity relative to the plasma lifetime, which is on the order of a nanosecond. We present comparisons of the conductivity due to laser pulses with a wavelength of 800 nm for varying laser conditions, such as the pulse energy and polarization state.   

Dynamics of the plasma dipole field for highly efficient emission of the THz radiation

In our previous work, we suggested a new method to generate the plasma dipole oscillation, which can emit THz radiation from the narrow plasma strips. The idea was colliding two short, detuned laser pulses in a plasma. Over the overlapped region of the pulses, the in-phase oscillation of a bunch of electrons (plasma dipole oscillation, PDO) is generated, and a strong electromagnetic wave in the THz regime is emitted from the PDO. From the theoretical estimation, it was expected that the conversion efficiency of the driving laser pulse energy to the plasma dipole energy can be increased by optimizing the laser pulse duration and frequency detuning. We have investigated the effect of controlled laser pulse duration and frequency detuning by monitoring the dynamical evolution of plasma dipole field. The study shows that the efficiency tends to increase as those two parameters increase, while the applicable ranges of parameters are limited; too long pulse duration leads to the breaking of in-phase motion of the electrons. We show that such a problem can be remedied by obliquely shooting the laser pulses. We demonstrate how the efficiency and total power of the THz radiation evolve dynamically as the system parameters are changed.   

Characterization of Optical Field Ionized Plasmas Formed in a 20 cm Long Gas Jet

We present a new $20$ $cm$ gas jet capable of forming a highly uniform gas profile of molecular densities around $3x10^{18}$ $cm^{-3}$ and $1000$ $\mu m$ width. The jet utilizes a carefully optimized throat and reservoir to provide supersonic, steady-state gas flow with a rise time of about $3$ $ms$. The transverse profile of this steady-state flow is characterized using neutral gas interferometry while plasma fluorescence measurements demonstrate uniformity and steep density gradients at the edges of the gas flow. Since the jet provides a gas profile with sharp cutoffs at the edges, it is an optimal target for the formation of sub-critical density Optical Field Ionized (\textit{OFI}) plasmas. Plasmas are formed using the line focus of a $J_0$ Bessel beam (intensities on the order of $10^{15}$ $W/cm^2$), and hydrodynamically expand outward, forming a radial electron density gradient. Transverse interferometry of the plasma column is used to trace the dynamics of this channel formation.   

Dynamics of plasmas formed by higher order Bessel beams via optical field ionization

A hollow Bessel beam ($J_1$) with high intensity ($10^{15}W/cm^2$) and 50 fs pulse length is generated using a spiral phase plate and reflective axicon. The pulse has an on axis minimum, and initially generates a long, tubular plasma structure in hydrogen backfill (100 torr) through optical field ionization. The plasma expands hydrodynamically into the background neutral gas, both outside and inside the tube. The inward radial expansion results in on-axis plasma formation via collisional ionization. The temporal dynamics of this process are measured by transverse interferometry. We present a parametric study of laser pulse intensity and axicon angle (varies the plasma tube radius), and compare to measurements of a plasma formed with $J_0$ Bessel beam. The $J_0$ beam has an intensity maximum along the optical axis (also $10^{15} W/cm^2$) and produces a plasma column on axis which then expands radially outward. Despite the stark difference in initial plasma profiles, It is found that after few nanoseconds, the plasmas formed by $J_0$ and $J_1$ Bessel beams evolve into similar structures.   

Uphill acceleration in a spatially modulated electrostatic field particle accelerator

Spatially modulated electrostatic fields can be designed to efficiently accelerate particles by exploring the relations between the amplitude, the phase velocity, the shape of the potential and the initial velocity of the particle. The acceleration process occurs when the value of the velocity excursions of the particle surpass the phase velocity of the carrier, as a resonant mechanism. The ponderomotive approximation based on the Lagrangian average is usually applied in this kind of system. The mean dynamics of the particle is well described by this approximation far from resonance. However, the approximation fails to predict some interesting features of the model near resonance, such as the uphill acceleration phenomenon. Canonical perturbation theory is more accurate in these conditions and may be applied in different systems. We compare the results from the Lagrangian average and from canonical perturbation theory, focusing in regions where these two approaches differ from each other.   

Ponderomotive and resonant effects in the acceleration of particles by electromagnetic modes

In the present analysis, we study the dynamics of charged particles under the action of slowly modulated electromagnetic carrier waves. With the use of a high-frequency laser mode along with a modulated static magnetic wiggler, we show that the ensuing total field effectively acts as a slowly modulated high-frequency beat-wave field typical of inverse free-electron laser schemes. This effective resulting field is capable of accelerating particles in much the same way as space-charge wake fields do in plasma accelerators, with the advantage of being more stable than plasma related methods. Acceleration occurs as particles transition from ponderomotive to resonant regimes, so we develop the ponderomotive formalism needed to examine this problem. The ponderomotive formalism includes terms that, although not discussed in the usual applications of the approximation, are nevertheless of crucial importance in the vicinity of resonant capture. The role of these terms is also briefly discussed in the context of generic laser-plasma interactions.   

Laser-Wakefield Application to Endoscopic Oncology

Recent developments in fiber laser and nanomaterials have opened the possibility of using laser wakefield acceleration (LWFA) as the source of low-energy electron radiation for endoscopic and intraoperative cancer therapy, a scheme in which sources of radiation for cancer treatment are brought directly to the affected tissues, avoiding collateral damage to intervening tissues. To this end, the electron dynamics of LWFA is examined in the high-density regime. In the near-critical density regime, electrons are accelerated by the ponderomotive force followed by an electron sheath formation, resulting in a flow of bulk electrons. These low-energy electrons penetrate tissue to depths on the order of millimeters. First a typical resonant laser pulse is used, followed by lower-intensity, longer-pulse schemes, which are more amenable to a fiber-laser application. [1]   

Chromatic matching in a plasma undulator

The principle of color tuning, using two laser modes of different geometric modes numbers and of different colors such that they copropagate at the same group velocity, is proposed in order to realize the plasma undulator concept. By using color tuning it is possible to overcome the limit of group velocity slippage, whereby lower order modes outrun higher order ones, allowing for extended interaction lengths. The dephasing limit can be overcome by using a particular tapering of the plasma channel such that the electron bunch propagates in phase with the laser mode as well as maintain constant undulator frequency. In addition, controlled dephasing is proposed as a means to induce a chirp in the generated x-ray spectrum.   

Laser and Beam Hybrid Wakefield Accelerator

We report a new wakefield acceleration scheme where a laser pulse and an electron driver bunch are overlapped to drive a wakefield in the blowout regime. The laser pulse can provide direct laser acceleration (DLA) to the driver bunch. The driver bunch is then accelerated through the DLA mechanism instead of being decelerated by its wakefield [1], and its propagation distance is extended by several times. Additionally, the plasma channel sustained by the driver bunch also helps to guide the laser pulse by creating a plasma bubble.~ This combination can achieve 10GeV energy gain of the electron at a single stage. We also describe the simulation tool we used: an efficient full 3D quasi-static PIC code with the ability to capture sophisticated particle-laser resonance interactions over distances exceeding tens of centimeters.~ This code solves simple and advanced quasi-static equations[2] without complicate predict-correct algorithm. An in-house parallel multigrid algorithm is developed which shows good scalabilities on the machine with thousands of cores. Comparison between the new code and the first-principle PIC code will be presented.~ [1] V. N. Khudik, et al. Physics of Plasmas~25.8 (2018): 083101. [2] T. Wang, et al. Physics of Plasmas 24.10 (2017): 103117.   

Ion motion and hosing suppression in plasma-based accelerators

Plasma accelerators have been proposed as drivers for the next generation of colliders. Achieving high efficiency while preserving excellent beam quality is critical to realizing this application. High efficiency requires large longitudinal wakefield excitation by the witness beam, and this has an associated large transverse wake that will drive the hosing instability. Furtheremore, for high-energy beams with low emittance, the focusing forces in the plasma will pinch the witness beam and increase the beam density, orders of magnitude above the background ion density, leading to ion motion. This results in nonlinear focusing and emittance growth. We present a solution to mitigate the hosing instability in plasma accelerators that relies on ion motion. The response of the ions to a high-density beam is described, including the coupling to the hosing instability. It is shown that the ion-motion-induced head-to-tail variation in the focusing experienced by the beam suppresses hosing. A class of initial beam distributions are identified that are equilibrium solutions in the plasma wake, including ion motion. Using these beam distributions enables ion motion without emittance growth. Hence, stable and quality-preserving acceleration in plasma-based accelerators is possible.   

Effect of carrier envelope phase on betatron oscillations and direct laser acceleration of electrons in ion channels and plasma bubbles

When the laser pulse length is of the order several wavelengths [1] or has a very sharp front due to etching [2], the ponderomotive approximation breaks down. Such laser waves propagating inside a plasma cause controllable asymmetric plasma electron expulsion from laser according to the carrier envelope phase (CEP) and form a periodically oscillating plasma bubble. This can result in periodic transverse kicking of the electrons in the plasma bubble, which may significantly affect their dynamics, energy gain, and the excitation of betatron oscillations. We find that under certain conditions, the transverse CEP force can constructively interfere with direct laser acceleration (DLA) and enhance the betatron oscillations. Using first-principles 3D Particle-In-Cell (PIC) simulations and analytic calculations, we estimate the effect of CEP-induced oscillations on the energy gain of electrons via DLA in ion channels and plasma bubbles. [1] E.N.Nerush and I.Yu.Kostyukov Phys.Rev.Lett. 103,035001(2009) [2] Ma, Yong, et al. Scientific Reports. 5,30491(2016)   

Enhancement of synchrotron photon emission in a strongly nonuniform magnetic field

It has been previously shown that efficient multi-MeV photon emission can be achieved by accelerating an electron beam in a laser-generated magnetic field [PRL116, 185003]. However, it is not feasible to simulate the emission process using a conventional field solver of a particle-in-cell code because of the extremely short wavelength of the electromagnetic radiation representing multi-MeV photons. A Monte-Carlo algorithm [PPCF 53(2011)015009] has been developed to address this problem, where photons are emitted as individual particles based on a synchrotron spectrum for a uniform magnetic field. We found that this approach is not adequate for the entire photon spectrum, with the lower frequency part of the spectrum being significantly enhanced in the presence of a strong magnetic field gradient that is characteristic of laser-driven magnetic fields. We have derived the corresponding frequency and the enhancement factor.   

Soft X-ray spectral measurements from Laser Wakefield Acceleration

We have performed soft X-ray measurements from LWFA experiments using the HERCULES laser system at University of Michigan. A high-resolution spectrometer captured radiation emitted from LWFA interactions with a gas cell target and 30 fs pulses with powers of 100 TW. Spectral lines and broadband emission measurements were made alongside electron beam characterizations, and an inference to the electron acceleration mechanism is discussed. Simulations were also performed showcasing radiation generation.   

Optimization of high repetition-rate laser wakefield accelerators using machine-learning techniques

Many potential applications of laser accelerator sources require operation at high repetition rate. Here, 20 milliJoule pulses are generated at kilohertz repetition rate for pulse self-compression and laser wakefield acceleration experiments. A genetic algorithm is implemented using a Dazzler acousto-optic programmable dispersive filter with the laser pulse characteristics from FROG measurements or wakefield electron beam signal optimized onto several different masks used as feedback. This procedure allows a heuristic search for the optimal laser pulse phase characteristics up to 4th order to produce a desired arbitrary wakefield electron beam or a well self-compressed pulse. Additionally, in progress is the implementation of a spiral phase plate in order to produce a $\text{Laguerre-Gaussian}_{01}$ laser pulse with optical angular momentum. We’re investigating the use of this exotic beam for laser wakefield acceleration experiments.   

External injection in laser wakefield acceleration at the CLARA accelerator facility: a preparatory study

External injection of an electron beam into a laser wakefield accelerating structure is a very attractive scheme because in principle allows for better controlling the characteristics of the electron bunch. Among few in the world, the CLARA accelerator facility at Daresbury Laboratory in the UK offers the right set of infrastructure to conduct proof-of-principle external injection experiments. In this poster, we are going to show preliminary results of one-to-one particle-in-cell simulations modelling the injection of the 50 MeV CLARA electron beam into the wakefield. A detailed parameter scan is performed to analyze the effect of different laser and plasma parameters. With the aim of providing indications useful for a future experiment, we study different injection angles and explore the possibility to accommodate the beam in one bucket or to spread it among a few buckets.   

Narrow bandwidth Laser-Plasma Accelerator driven Thomson photon sources

Compact, high-quality photon sources at MeV energies for nuclear nonproliferation and other applications can be provided by Thomson scattering of a laser from the electron beam of a Laser-Plasma Accelerator (LPA). Recent experiments and simulations demonstrate controllable LPAs in the energy range appropriate to MeV sources and indicate that high flux photon beams with narrow energy spread can be achieved via control of the accelerator, scattering laser pulse shape and laser guiding. Undesired background bremsstrahlung can be mitigated by plasma based deceleration of the electron beam after photon production. The path from current experiments towards a compact photon source system will be presented.   

Measurements of mid-infrared radiation from a laser wakefield accelerator

In the bubble regime of Laser Wakefield Acceleration (LWFA), a density up-ramp at the leading edge of the plasma bubble creates a region where the driving laser pulse sees a negative gradient in refractive index. This gradient generates frequency shifts in the laser, extending its spectral content to the mid-infrared. Experiments performed using the HERCULES laser system at the University of Michigan measured the spectrum, energy and beam profile of mid-infrared radiation produced during LWFA. Spectra with wavelengths extending to 2.5 micrometers and containing up to 15 mJ of energy were obtained. Enhanced spectral broadening using tailored density targets was experimentally demonstrated. Under certain conditions, quasi-monochromatic spectral features were produced. A sensitivity analysis of the spectral features with varying laser and plasma conditions was performed to guide optimization of this process for the realization of tunable infrared sources. Supporting PIC simulations indicated that slow-moving long-wavelength radiation, which slips backward relative to the driving laser pulse, can interact with the accelerated electron bunch, and may serve as a diagnostic of bunch formation and dynamics.   

Developing shaped bunches to improve beam loading in laser wake-field accelerators

Important to successful achievement and integration of monoenergetic laser wake-field acceleration into mainstream use in science is optimization of the injection control process. In addition to low transverse emittance, a significant challenge is accelerating low energy spread beams. One such technique to improve the final energy spread of the electron beam is achievement of beam loading by way of electron bunch shaping. Optimization of bunch parameters can have a significant effect on other bunch parameters, leading to significant increases in quality of monoenergetic Wakefield produced electron beams. Presented here are studies on the production of shaped bunches by tailoring the plasma profile.   

Imaging of small-scale dynamic phenomena with a laser-wakefield acceleration driven x-ray source

Laser wakefield accelerators are a promising alternative for generation high-brightness and coherent x-rays at a fraction of the cost and facility size of conventional synchroton-like electron accelerators. The radiation source is mediated by the oscillatory motion of electrons inside the plasma wakefield, also known as betatron oscillations. Characteristics of the x-ray source include low beam divergence (few miliradians), ultrafast femtosecond pulse duration, and micrometer size resolution. These properties make a wakefield accelerator radiation source an excellent candidate for imaging of ultrafast events with high-temporal and high-spatial resolution. In this work we present imaging of small scale dynamic phenomena using a betatron x-ray source generated by the interaction of a high-intensity laser pulse with a gas cell. Some of the preliminary results of interest include properties of the betatron spectrum and imaging resolution.   

Modeling of Capillary Discharge Plasmas for Electron Beam Transport and Acceleration

Next generation accelerators demand sophisticated beam sources to produce ultra-low emittances at large accelerating gradients. Furthermore, the transport of these beams between accelerating stages requires similarly capable beamline optics. Capillary discharge plasmas may address each of these challenges. As sources, capillaries have been shown to increase the energy and quality of laser wakefield accelerators, and as active plasma lenses they provide orders-of-magnitude increases in peak magnetic field. Capillaries are sensitive to energy deposition, heat transfer, ionization dynamics, and magnetic field penetration; therefore, advances in capillary design requires careful modeling. We present simulations of capillary discharge plasmas using FLASH, a publicly-available multi-physics code developed at the University of Chicago. We report on the implementation of a 2D, cylindrically symmetric capillary model for capturing plasma density and temperature evolution with realistic conductivities and magnetic fields. We then illustrate the use of laser energy deposition to model channel formation for the guiding of intense laser pulses. Lastly, we present simulations of active capillary plasmas with varying fill species and comparisons to experiment.   

VPIC on GPU

Efficient operation of Particle-in-Cell codes on Graphics Processing Units (GPUs) has been a coveted goal since they were adopted by High Performance Computing platforms years ago. While a variety of research exist on this topic, many of the world’s highest performing PIC codes have yet to demonstrate their ability to effectively use large GPU machines at extreme scale. In this work we demonstrate the effort of Los Alamos National Laboratory to port VPIC to run on large-scale DoE GPU super computers. We demonstrate the codes ability to scale to thousands of GPUs, directly compare code performance to previous systems, and present lessons learnt from porting to the performance-portable code framework Kokkos. We document our strategy for data management in the context of the limited memory regime presented by GPUs, and also demonstrate the strategies we employ to minimize data copies between host and device.   

Guiding Center Theory for Large Electric Field Gradients

It is important to understand the physics of edge transport barriers, yet the steep gradient region of the pedestal and scrape-off layer of a tokamak challenge the validity of standard guiding center and gyrokinetic theory. In this work, the guiding center equations for magnetized particles are extended to the regime of large electric field gradients perpendicular to the magnetic field. Shear in the electric field modifies the oscillation frequency and causes the particle orbits to deform from circular to elliptical trajectories. In turn, the elliptical orbits modify the polarization and magnetization of the particle. In order to retain a good adiabatic invariant, there can only be strong dependence on a single coordinate at lowest order, so that resonances do not generate chaotic motion that destroys the invariant. In this case, the drift equations are modified, but retain a mathematical form that is similar to the unsheared case and can be used to develop a more accurate gyrokinetic theory in a relatively straightforward manner. It is of great interest to continue exploring the physical implications of the extended ordering and its convergence for strong field variations.   

Numerical finite particle effects and wave breaking limits

Seminal works of Dawson, Akhiezer and Polovin, and Coffey provide theoretical cold, relativistic, and warm wave breaking limits, respectively. There is renewed interest in understanding wave breaking limits to either achieve or avoid them, depending on the injection and acceleration mechanism. Numerical simulations are often used to investigate breaking limits. When particle methods are used, they commonly employ finite numerical particle size or, equivalently, a smoothed Green's function for the electric interaction. In this work we present studies of finite numerical particle effects on wave breaking limits. We discuss phase mixing times and breaking limits in relativistic and non relativistic cases when using finite sized numerical particles in simulation. We compare PIC models with the Dawson sheet model and other particle methods, using a 1d grid-free particle method employing a smoothed kernel.   

New developments in the OSIRIS 4.0 framework

The OSIRIS [1] Electromagnetic particle-in-cell (EM-PIC) code is widely used in the numerical modeling of many kinetic plasma laboratory and astrophysical scenarios. In this work, we report on the new developments recently introduced into the framework. In particular, we will describe our implementation of a tile-based dynamic load balancing algorithm, and the support for the latest hardware (new GPU architectures / ARM Neon). We will focus on the use of a customized field solver used to mitigate NCI, and provide better accuracy under extreme fields. We will also address the new particle beam initialization capabilities and new EMF sources available. Finally, we will discuss our new diagnostic i/o subsystem that provides order of magnitude performance improvements over the existing HDF5 implementation. [1] R. A. Fonseca et al., Lecture Notes in Computer Science \textbf{2331}, 342-351 (2002)   

WarpX: efficient modeling of plasma-based accelerators with mesh refinement

Plasma-based accelerators are being developed to provide a more compact and economical alternative to standard accelerator technology. High accelerating gradients were demonstrated in centimeter long plasmas. Recent studies focus on mitigating multiple non-linear, fast processes and instabilities that deteriorate the quality of plasma-based accelerated beams. High-fidelity numerical codes that can model beam propagation in plasma fields are necessary to study those nonlinear processes. Simulations are typically computationally demanding because they resolve small structures over large distances. The Adaptive Mesh Refinement (AMR) technique, where selected regions are modeled with higher resolution, can make simulations more efficient. For the Exascale Computing Project, we have been developing the WarpX tool that incorporates AMR through the AMReX framework in the Particle-In-Cell (PIC) code Warp. We present recent studies of beam evolution in consecutive plasma stages, done with and without using mesh-refinement.   

Optimizations for a semi-implicit, energy- and charge-conserving particle-in-cell algorithm with iVPIC

A semi-implicit, energy- and charge-conserving PIC algorithm has recently been developed for solving the relativistic Vlasov-Maxwell system. \footnote{Chen et al, arXiv:1903.01565v2, submitted to J. Comput. Phys.} The algorithm employs the leap-frog scheme for Maxwell's equations, and a Crank-Nicolson scheme for the particle equations. The implicit field-particle integration ensures exact accounting of energy transfer between the field and particles. A new particle pusher is used to be exactly energy- and charge-conserving. We have designed a simple and effective Picard iteration algorithm that only requires a single orbit computation per outer iteration, thereby minimizing wall-clock time impact vs. the explicit VPIC algorithm. The Picard algorithm requires only a few iterations (3-5) to converge to single precision round-off levels. With further code optimizations we have obtained speedups of a factor of 3 vs. a naive implementation, resulting in a cost per implicit iteration comparable to a single explicit update of the baseline VPIC implementation. As a result, the semi-implicit algorithm is only a few times slower than the explicit baseline. We present numerical results that demonstrate the speedups of the algorithmic and code optimizations with sample test problems.   

Computing Wake Functions in Plasma Accelerators Using Particle-in-Cell Simulations

Plasma accelerators in the blowout regime are a possible candidate for a future TeV lepton collider due to their high gradients. These high gradients come from the small scale of the accelerating plasma wave, typically hundreds of microns, compared to conventional rf-based structures, with scales in the range of tens of centimeters. This small scale can lead to very strong dipole wakes, which could drive a beam break-up instability and spoil the bunch for any collider applications. Understanding the beam break-up instability in plasma accelerators requires computing the wake functions in a plasma accelerator to characterize the linear response of the plasma wave to a perturbing charge. We present an approach to computing these wake functions from simulations in the FBPIC electromagnetic particle-in-cell code. We apply this approach to a beam-driven plasma wakefield accelerator using a drive and witness bunch with parameters similar to what is expected at the proposed FACET-II facility.   

SPACE : a relativistic, 3D Particle-in-Cell code with atomic physics support

A parallel, relativistic, three-dimensional Particle-in-Cell code SPACE has been developed for the simulation of electromagnetic fields, relativistic particle beams, and plasmas. In addition to the standard PIC algorithm, SPACE includes efficient, novel algorithms to resolve atomic physics processes such as the generation and evolution of plasma, recombination, and electron attachment on dopants in dense neutral gases. SPACE also contains a highly adaptive and artifact-free particle method, called AP-Cloud, for solving the Vlasov-Poisson problems. The code's structure, capabilities, parallelization strategy and performances has been discussed. Applications of the code in modeling various processes in the eRHIC program at Brookhaven National Laboratory (BNL), high pressure RF cavity experiments at Fermilab and $ \text{CO}_2$ laser driven wakefield accelerator experiments at BNL's Accelerator Test Facility is presented.   

Development of Novel Computational Methods for Plasma-Based Particle Accelerators Modeling

High-fidelity modeling and high-performance computing are essential components for advancing state-of-the-art experimental plasma-based particle accelerators. WarpX is a new tool for exascale Particle-In-Cell (PIC) simulation of plasma-based accelerators. It is being developed by a collaboration of researchers from LBNL, SLAC and LLNL within the U.S. Department of Energy's Exascale Computing Project. WarpX combines the most advanced numerical algorithms that include ultrahigher-order Pseudo-Spectral Analytical Time-Domain (PSATD) Maxwell solvers. The spectral solver offers arbitrary order with low numerical dispersion at any wavelength and angle and becomes dispersion-free at infinite order. It also enables efficient algorithms for mesh refinement and PIC algorithms with large time steps. However, it required the development of a novel algorithm for the application of Perfectly Matched Layers (PML) for open boundary problems. We present a novel “two-step” PML formulation that can be applied to any type of Maxwell solvers, including the PSATD solver. We will discuss the efficiency and performance of the algorithm, and provide examples of its application. We will also discuss a novel algorithm for PIC algorithms with large time steps.   

Recent development on open source QuickPIC

QuickPIC is a 3D parallel quasi-static PIC code for efficiently simulating the plasma based accelerator (PBA). It is developed based on the framework UPIC. QuickPIC has been widely used and played an important role in studying PBA problems. In 2017, we made QuickPIC an open source code on Github (https://github.com/UCLA-Plasma-Simulation-Group/QuickPIC-OpenSource). In this work, we will present the recent development of the code including the field ionization module, the laser module and 2D QuickPIC in r-z cylindrical coordinates.   

Comparison of Numerical Methods for the Calculation of Synchrotron Radiation From Electrons

The coherent synchrotron emission of electrons in a beam bunch can lead to micro-instabilities causing emittance degradation and beam disruption detrimental to advanced light source development. To understand how the beam interacts with synchrotron radiation for a design and its optimization, accurate and efficient numerical methods are essential. We investigate several existing methods for the calculation of the radiation in near fields, including the finite difference method, the Lieìnard-Wiechert method and a near-field method recently validated [1]. We focus on the accuracy and efficiency of these methods in computing the radiation fields (coherent and incoherent) in both steady state and dynamical beam trajectories. In our presentation, we will also discuss a self-similarity feature in the synchrotron radiation that can be exploited to improve the calculations. [1] F. Y. Li, et al, paper MOPGW116, Proceedings of 10th Int. Particle Accelerator Conf., (2019).   

The Effect of Quantum Radiation Emission in High-Energy Wakefield Stages

An electron beam passing through an undulator will experience radiation emission such that the high energy part of the beam will radiate more energy than the low energy part, decreasing its energy spread. In plasma accelerator stages with an injected electron beam at above one hundred GeV, stochastic radiation emission can cause a broadening of the energy spread as well. We use Particle-in-Cell simulations to study how quantum radiation emission would influence the energy spread and emittance of external injected beams inside laser wakefields. The effect of nonlinear focusing forces, beam energy spread and laser beam mismatch, however, can distort the phase space distribution and cause significant emittance growth. Theoretical analysis and numerical simulation were performed to find optimal conditions to minimize phase space distortions.   

Optimized laser-driven electron acceleration inside a hollow core target

When a hollow core plasma target is irradiated by a short laser pulse with intensity $I_{0}\sim {10}^{20}\thinspace \mathrm{W/}{\mathrm{cm}}^{\mathrm{2}}$, electrons are extracted from the channel walls and injected into the initially empty area by the transverse laser electric field. The accumulated net negative charge background provides an optimized structure for quasi-static electric and magnetic fields which allows the longitudinal laser electric field to facilitate collimated electron acceleration. When electrons move with the laser pulse inside the channel, the electron density exerts a non-negligible feedback on the dispersion relation of the laser pulse which raises its phase velocity $v_{ph}$. Leveraging on theoretical analyses and 2D particle-in-cell (PIC) simulations, we demonstrate that $v_{ph}$ is the key factor which dominates the electron acceleration process and determines both saturation time and maximum attainable energy. The correctional term introduced by the electron density fits our PIC simulation results more closely in comparison to the traditional waveguide dispersion relation. The dependence of the superluminal phase velocity on the laser amplitude $a_{0}$ and channel inner radius $R$ is also confirmed.   

Radiation Rebound and Quantum Splash in Electron Laser Collision

The radiation reaction (RR) is expected to play a significant role in light-matter interactions at extreme intensity. Utilizing the theoretical analyses and numerical simulations, we demonstrate that electron reflection, induced by the RR in a head-on collision with an intense laser pulse, can provide pronounced signatures to discern the classical and quantum RR. In classical regime, there is a precipitous threshold of laser intensity to achieve the whole electron bunch rebound. However, this threshold becomes a gradual transition in the quantum regime, where the electron bunch is quasi-isotropically scattered by the laser pulse and this process resembles a water splash. Based on the derived dependence of classical radiation rebound on the parameters of laser pulses and electron bunches, a practical detecting method is proposed to distinguish the quantum discrete recoil and classical continuous RR force.   

Enhanced gamma-ray emission in structured targets irradiated by counter-propagating laser pulses

Previous research has shown that a structured over-dense target irradiated by a high-intensity laser pulse ($I\approx5\times10^{22}$W/cm$^2$) is an efficient source for collimated beams of multi-MeV gamma-rays [Stark et al. PRL 116, 185003 (2016)]. The newly-constructed laser facilities, such as ELI Beamlines, will enable experiments with multiple laser pulses instead of just one. Motivated by this experimental capability, we consider a setup where a structured target with an embedded relativistically transparent channel is irradiated by two counter-propagating laser pulses. Using 2D particle-in-cell simulations, we show that the laser energy conversion rate into multi-MeV photons is significantly increased due to head-on collision between laser-accelerated electrons and a counter-propagating laser. The main advantage of the setup is that the target channel provides automatic alignment between the electrons and the beam.   

High power gamma flare generation in multi-petawatt laser interaction with tailored targets

Using quantum electrodynamics particle-in-cell simulations, we optimize the gamma flare ($\gamma$-flare) generation scheme from the interaction of high power petawatt-class laser pulse with tailored cryogenic hydrogen target having extended pre-plasma corona. We show that it is possible to generate an energetic flare of photons with energies in the GeV range and total flare energy being on a kilojoule level with the efficient conversion of the laser pulse energy to $\gamma$-photons. We discuss how the target engineering and laser pulse parameters influence the $\gamma$-flare generation efficiency. High-Z targets, the oblique incidence of the laser pulse, and the effects of realistic pre-plasma obtained by hydrodynamical simulations are also discussed. This type of experimental setup for laser-based $\gamma$ source would be feasible for the upcoming high power laser facilities.   

Radiative cooling effect on the stability of magnetic islands

A magnetic island is a well-known concept of plasma physics composed of a magnetic flux tube bounded by a separatrix (in 2D). Magnetic reconnection is a typical plasma phenomenon where magnetic islands emerge either separated by X points or Y points. One of the simplest ways to model the equilibrium of a cylindrical flux tube (the magnetic island being the cross-section) is given by the Bennett equilibrium where total plasma pressure and the magnetic pressure balances out the magnetic tension. We study analytically the stability of a slowly time-evolving Bennett equilibrium under the influence of radiative cooling that could occur due to synchrotron emission. If the temperature of the plasma is not sustained because of radiation losses, the equilibrium pinch is broken due to a lower pressure which eventually leads to the collapse of the island [1]. The analytical model, confirmed by particle-in-cell simulations, suggests that the collapse can only occur for a special hierarchy of the characteristic times of the system. These results have profound implications for magnetic reconnection in ultra-intense fields.[1] Kevin Schoeffler, Thomas Grismayer, Dmitri Uzdensky, Ricardo Fonseca, Luís Silva, ApJ 870 (2019).   

Characterizing extreme laser intensities by ponderomotive acceleration of protons from rarefied hydroge

A new method to diagnose extreme laser intensities through measurement of angular and spectral distributions of protons directly accelerated by the laser beam focused into a rarefied gas is proposed. This work addresses a range of laser parameters with intensities from 10$^{\mathrm{21}}$ to 10$^{\mathrm{24}}$ Wcm$^{\mathrm{-2}}$, pulse durations from 15 to 80 fs and focal spot diameters from 1$\lambda $ to 4$\lambda $, where $\lambda =$0.8 $\mu $m is the laser wavelength. The Stratton-Chu diffraction integrals are used to describe focusing of a laser pulse with different spatio-temporal parameters and the test particle method with relativistic ponderomotive force is used to simulate laser-proton interactions. This approach has allowed us to construct analytical formulas estimating spectral and angular characteristics of the accelerated protons as a function of laser intensity. The proposed method is relaxed in terms of the experimental realization and can be used to accurately compare the results of different extreme intensity experiments with theoretical predictions and with each other. It can be, therefore, valuable for the commissioning of new laser facilities.   

Scaling laws for direct laser acceleration with radiation reaction

Electrons can be directly accelerated by a laser within a hollow (or low density) plasma channel. If betatron resonance is achieved, we can expect to obtain multi-GeV electron energies. However, this acceleration process is very sensitive to the local initial conditions. By increasing the laser intensity, it is possible to enter a regime where radiation reaction becomes important in the particle dynamics, which further complicates the analysis. In this case, the particles can radiate away a considerable fraction of their energy, and their acceleration is limited by this emission. However, radiation reaction can also be beneficial for the acceleration process, as it allows for radiative electron trapping and changes the onset of the betatron resonance by altering the local conditions. Here, we study the direct laser acceleration within a plasma channel using the parameters of the upcoming generation of 10 PW-class lasers. We show that even with radiation reaction, it is possible to obtain multi-GeV electrons in a single-stage acceleration within a 0.5 mm-long channel provided that the optimal initial conditions are satisfied. We present those conditions in a form of explicit analytical scaling laws that can be applied to guide the future electron acceleration experiments.   

Electron Bunch Width Limitations on the Spectrum of the Coherent Synchrotron Emission from Ultrathin Foils

High-order harmonic generation from the interaction of intense lasers with solid targets offers a compact solution for generating brilliant and broadband extreme-ultraviolet and soft x-ray radiation. Detailed numerical simulations and theoretical work have revealed the origin of this radiation to be from the sub-cycle bunching and acceleration of electrons, which undergo synchrotron-like trajectories, emitting bursts of attosecond pulsed radiation. The use of ultra-thin targets, as opposed to thick targets, has been shown to result in higher conversion efficiencies. In this ultra-thin foil regime, attosecond pulses are emitted in both the specular and transmitted directions. In this work, we use particle-in-cell simulations to investigate the effects of the electron bunch width on the spectral power-law and attosecond pulse formation, in particular, showing how the measured bunch width limits the efficiency of the emission process.   

Scaling relativistic laser-solid interaction using 30fs laser pulses

There has been growing interest in relativistic laser-solid interaction as a compact source of relativistic electron beams and hard x-rays. Femtosecond hard x-ray pulses have important applications such as probing time-resolved x-ray absorption and diffraction. Relativistic electrons from solid targets have superior properties in beam charge and divergence than those from wakefield acceleration in underdense plasmas, and can find applications in warm dense matter creation, electron radiography, seed of wakefield accelerators and fast ignition researches. In this work, the 30fs laser pulses are focused down to near diffraction-limit spot size to achieve relativistic intensity (a\textunderscore 0\textgreater 1) and ablate into a thick (\textasciitilde mm) glass target. We investigate the scaling laws of this interaction in terms of laser wavelength (0.8\textmu m, 1.3\textmu m and 2\textmu m), laser energy (millijoule to joule level), angle of incidence (grazing, 45\textdegree and normal) and preplasma scale length (0.1$\lambda $ to 5$\lambda )$. Particle-In-Cell simulation (PIC) and particle tracking shows that the incident half and reflected half of the laser pulse form a standing wave to accelerate electrons to relativistic (MeV) energy.   

High repetition rate experimental techniques at the Extreme Light Laboratory at AFRL

As ultra-intense lasers transition to a high repetition rate (1-10 Hz) mode of operation, experimental techniques must also adapt to meet the new challenges of complementary experimental design, targetry, and detectors. At the Extreme Light Laboratory at AFRL, we perform relativistically intense laser-plasma interaction experiments at 1 kHz repetition rate. In this poster we present novel, optically synchronized pump-probe techniques well matched to high repetition rate operation. Self-refreshing, liquid microjet-based targets are capable of generating submicron thick sheets and can be applied at repetition rates exceeding 10 kHz. In order to diagnose the energetic particle spectra we design custom digitized particle spectrometers to provide real time experimental feedback.   

Interplay between the Weibel instability and the Biermann battery in realistic laser-solid interactions

In this work, we investigate the interplay between the Weibel (driven by temperature anisotropies) and Biermann (caused by non-parallel density and temperature gradients) magnetic fields in realistic laser-plasma experiments. With ab initio multi-dimensional particle-in-cell simulations, we model the interaction between a short ($\le $ ps) intense (a$_{\mathrm{0}}\ge $1) laser and a solid-density target. We show that Weibel magnetic fields can be observed alongside the Biermann fields usually dominant in experiments. The expanding hot energetic electron population generated by the laser produces an anisotropy in the velocity distribution. This anisotropy supplies the free energy that drives the Weibel instability, consisting of an intense filamentary field growing on the target surface. This field can dominate over the Biermann battery field, provided that the pre-plasma scale length is much larger than the local electron inertial length.   

\textbf{Untangling resistive and collisionless electron filamentation instabilities in dense plasmas over large spatiotemporal scales}

Plasma micro-instabilities induced by high-energy particle currents play an important role in many space or laboratory plasma environments. Here, we report on measurements that reveal, over large temporal (tens of picoseconds) and spatial (hundreds of microns) scales, the growth of a multiplicity of electromagnetic filaments, following localized laser-generation of MeV electrons in a solid foil. The proton radiography data obtained in both low- and high-resistivity targets show two distinct, superimposed electromagnetic field patterns, which point to different field generation processes, namely of collisionless and resistive character. The collisionless Weibel instability is suggested, by particle-in-cell simulations, to build up in the dilute plasma expanding into the vacuum, independently of the target material, and to lead to observed azimuthally symmetric electromagnetic structures. Additionally, when the target resistivity is high enough, an additional resistive filamentation instability arises through the bulk target, resulting in observed radially elongated filaments.   

Increased electron beam energy using structured targets at the Scarlet laser facility

Electron and ion beams accelerated from ultra-intense laser systems have versatile applications including target heating, time resolved probes of EM field structure, and precise imaging of dense material. We describe the use of structured targets to increase the yield and energy of laser accelerated electrons, including a new experiment on the Scarlet laser utilizing arrays of 5 um diameter tubes that are 100 or 300 um long, composed of a glass substrate with both undoped and Ni-doped versions. These structures enhance electron acceleration by generating a sufficiently dense medium to supply electrons but not inhibit laser propagation, resulting in better coupling of laser energy. The Scarlet laser is a 1 shot/min Ti:Sapphire based system capable of producing 10 J, 30 fs pulses with focal spot diameter of 2.4 um (full width at half maximum). A plasma mirror was used for this experiment to improve the \textless 1 ns contrast, resulting in an on-target energy of 6 J and peak intensity \textgreater 4x10$^{\mathrm{21}}$ W/cm$^{\mathrm{2}}$. We observe an increase in the electron cutoff energy from 20 MeV with 2 um copper to 50 MeV with the 100 um-long undoped glass structure.   

Highly Directional Deuteron Acceleration in Nanowire Arrays Irradiated at Highly Relativistic Energies

It has recently been shown that the irradiation of deuterated polyethylene nanowire arrays with an ultra-high contrast laser pulse of relativistic intensity can accelerate ions to multi-MeV energies and efficiently produce quasi-monoenergetic neutrons from DD fusion reactions1. Here we demonstrate the highly directional nature of ions with energy greater than 10 MeV being accelerated off the front surface of the nanowire arrays, with a narrow cone of emittance with a FWHM down to 7.5 degrees. Results of ion energy distribution measurements and beam directionality will be compared with 3D- relativistic PIC simulations. Preliminary results from a pitcher-catcher experiment to produce neutrons from D-Be and D-Li fusion reactions will be presented. 1. Alden Curtis, et al., ``Micro-Scale Fusion in Dense Relativistic Nanowire Array Plasmas'', Nature. Comm. 9, 1077 (2018) Work supported by AFOSR grant FA9550-17-1-0278   

Angularly Resolved Ion and Electron Spectrometer (ARIES) for high repetition rate, relativistic laser-solid target interactions

When high intensity laser pulses interact with solid density materials, one outcome is the generation of energetic ions, electrons, and photons. Common experimental conditions result in a broad spray in both energy and direction of the accelerated particles from this interaction. To better understand what conditions govern the energy and directionality of these particles, it is necessary to collect data that has high angular and energy resolution. Using the kHz repetition rate Red Dragon laser with the extreme light group at Wright Patterson Air Force Base, we demonstrate the use of a magnetic spectrometer situated on a semi-circular track. The Angularly Resolved Ion and Electron Spectrometer (ARIES), in its current configuration, collects electron and proton energy spectra through a slit that subtends a $\sim 0.5$ degree angle at 100 Hz using linear CCDs covered with plastic scintillators. By moving the spectrometer on the track, we are able to extend our angular collection to nearly 135 degrees. We demonstrate the use of the spectrometer in an experiment with a $\sim 10^{19}$ Wcm$^{-2}$ laser interacting with a $\sim 400$ nm thin liquid sheet target.   

High-intense laser propagation into strongly magnetized dense plasma

Recent progress of method for kilo-tesla class magnetic field generation allows us to perform experiments of high-intense laser plasma interactions (LPI) under strong external magnetic field. Such strong magnetic field affects not only fluid dynamics but also fast electrons and laser propagation. We study high-intense laser propagation into dense plasma under strong magnetic field using Particle-In-Cell (PIC) simulations. According to the linear theory of cold plasma in strong magnetic field, where the cyclotron frequency is greater than the laser frequency, right-handed circularly polarized (RCP) component of electromagnetic wave propagates into dense plasma without cut-off density. Simulation results show that initially the RCP component of injected laser can propagate into dense plasma, but after a while following laser cannot propagate. It is found that apparent ion acoustic wave generates at the area where the laser cannot propagate and it inhibits the propagation of the RCP component. This means that the RCP component of the laser can propagate until ion acoustic wave sufficiently grows.   

Enhanced electron acceleration with wide-spot picosecond pulse relativistic laser

High power lasers with relativistic intensities with wide focal spot and picosecond (ps) pulse lengths are available in recent years. In over-ps relativistic laser-plasma interaction, both energy and number of hot electrons become higher than those predicted by the scaling used in the sub-ps regime. The electron energy spectrum changes from the thermal distribution to the nonthermal one, and the accelerated ion energy becomes much higher than the energy predicted by the isothermal model [1]. One of the key mechanisms of such the super-thermal electron generation is the stochastic heating in laser-foil plasma interaction, where hot electrons recirculate around the plasma and suffer multiple kicks from the laser field during the interaction [2]. The blowout of plasma towards the laser [3] enhances the stochasticity in the laser-plasma interaction. Two-dimensional PIC (PICLS) simulations demonstrate the significant enhancement of the electron heating with a large focal spot laser light over ps interaction, while a laser with small focal spot produces thermal hot electrons with the ponderomotive temperature. [1] N. Iwata et al, Phys. Plasmas {\bf 24}, 073111 (2017). [2] Y. Sentoku et al., Appl. Phys. B {\bf 74}, 207 (2002). [3] N. Iwata et al., Nat. Commun. {\bf 9}, 623 (2018).   

Fast electron heating and thermalization using circular polarization

Simple laboratory generation of warm dense matter in a spatially homogeneous, thermalized state would be a valuable tool for studies of high-energy-density physics. We use the particle-in-cell code SMILEI to investigate the effect of Coulomb collisions and ionization on the interaction of ultra-short laser pulses with solid-density copper plasmas. In such moderately high-$Z$ materials, electron-ion collisions result in significant laser energy absorption through inverse Bremsstrahlung. We consider the case of circular polarization~(CP), which is known to lead to inefficient electron energization through collisionless mechanisms such as $j \times B$ heating. In collisional settings, however, not only is CP comparable with linear polarization~(LP) in terms of electron heating efficiency, but it also produces thermalized electrons on $<100\,\rm fs$ timescales, i.e., much faster than LP, which yields a more pronounced and longer-lived energetic electron tail. Such isochoric heating can potentially be used to study atomic physics models or equations of state under extreme conditions.   

Expansion dynamics of foil plasma irradiated by picosecond relativistic laser

Petawatt lasers with picosecond (ps) pulse durations become available today. In the experiments, target normal sheath acceleration (TNSA) of ions with higher energies than those predicted by the isothermal plasma expansion model [1] has been observed [2]. In over-ps laser-foil interactions, the fast electron temperature increases temporally beyond the ponderomotive scaling, which results the non-isothermal TNSA [3]. Such an electron heating is triggered by the transition to the blowout phase due to the change of the pressure balance at the laser-plasma interface [4]. We here study the expansion dynamics of foil plasmas in the over-ps regime. In PIC simulations, we found that the expanding tenuous plasma forms a “skirt”-like density profile in the blowout phase, where laser heating and expansion cooling are balanced to keep the sheath electric field strength constant. The expansion dynamics and its effect on the electron heating will be discussed. [1] P. Mora, Phys. Rev. Lett. 90, 185002 (2003). [2] A. Yogo et al., Sci. Rep. 7, 42451 (2017); D. Mariscal et al., Phys. Plasmas 26, 043110 (2019). [3] N. Iwata et al, Phys. Plasmas 24, 073111 (2017). [4] N. Iwata et al., Nat. Commun. 9, 623 (2018).   

Towards controlled laser acceleration of electrons in laser-plasma coupling regimes relevant to fast ignition

Direct laser acceleration (DLA) underpins many laser-plasma coupling processes when an intense laser pulse channels through plasma and encloses under its envelope a portion of comoving relativistic electrons. It may play an important role in generating hot electron beams for fast ignition. However, DLA, in general, requires matching between the laser and beam conditions and is therefore vulnerable to the evolving nature of laser-plasma interactions. It often leads to divergent electron beams, significantly reducing the beam coupling efficiency in fast ignition. Here we propose to systematically investigate the underlying physics of DLA in order to control the process to generate desired beam qualities. To this end, we build a simplified model that can describe the evolution of the electron phase-space dynamics. The differences between linear and circular laser polarizations are identified. The onset criteria of DLA and the weighting of DLA over plasma acceleration are analyzed. Key properties of the resulting electron beams when DLA dominates are presented. Potential ways to manipulate the DLA beams towards desired qualities are explored. These results are augmented by particle-in-cell simulations that address the non-ideal initial conditions as well as the effects of evolving laser-plasma dynamics. The implications for near-term experiments under design are also discussed.   

Physics of Electron Beam Generation and Dynamics from a Single Diamond Field Emitter

Many applications such as compact accelerators and electron microscopy demand high brightness electron beams with small beam size and low emittance. Electric-field-assisted diamond emitters manufactured from semiconductor processes are strong candidates for cathodes in such sources. We use the LSP particle-in-cell code to simulate the diamond emitter and obtain the beam size and divergence. ?To account for charge transport/tunneling, a semiclassical Monte Carlo emission method is developed and applied to a model to explain the measured emission characteristics. The beam divergence observed in the simulations is further corroborated with electron trajectories in an empirical field model. The results are compared with experimental observations. An effective mass based model accounting for the conduction band quantization in a high aspect ratio semiconductor nanotip is also developed for electron transport and emission.   

High-repetition neutron generation from ultrashort laser pulse irradiation of electrohydrodynamically dispensed deuterated microdroplets

We report initial findings of laser-driven fusion neutron yield from the interaction of regeneratively amplified several-mJ, 35 fs laser pulses at 1/2 kHz with spatio-temporally resolved microdroplets from a novel electrohydrodynamic jet nozzle. Femtoliter-scale deuterated droplet targets are produced via pulsed high-voltage electrostatic extraction from a 50$\mu m$ I.D./120$\mu m$ O.D. stainless steel capillary. High intensity laser pulses (of order 10$^{19}$ W/cm$^{2}$) are focused under vacuum and collide with the microdroplets to create energetic deuterons via the Target Normal Sheath Acceleration (TNSA) mechanism. 2.45 MeV neutron pulses are generated via the \textbf{d\emph{(d,n)}$^{3}$He} fusion half-reaction. Neutron flux is measured via zero gamma sensitivity calibrated bubble detectors while neutron spectrum is quantified with plastic scintillators in a pulse-shape discrimination neutron time-of-flight (ToF) setup. To our knowledge, this experiment is the first to demonstrate micron-scale monodisperse droplet generation in vacuum utilizing pulsed electrohydrodynamic jetting.   

Simulations of laser-driven ion acceleration using the quasi-cylindrical PIC code CALDER-CIRC and application to the PETAL facility

Realistic simulation of ion acceleration in the TNSA regime is very challenging. Indeed, the hot electrons' transverse expansion, which strongly affects the dynamics and spatial distribution of the accelerating sheath field, can only be correctly described in a 3D geometry. Despite this limitation, 2D PIC simulations often appear to be the only reasonable option due to the excessive computational cost of 3D simulations. As an alternative, we investigate here the benefit of using the quasi-cylindrical PIC code CALDER-CIRC to describe TNSA over experimentally relevant scales. This code enables reduced 3D simulations at a computational cost close to that of 2D Cartesian simulations. To illustrate both its potential and limitations, we will compare simulations of a typical TNSA setup carried out using CALDER-CIRC and the 2D and 3D Cartesian versions of CALDER. Moreover, we will report on a CALDER-CIRC simulation of TNSA under conditions relevant to the PW PETAL laser ($\sim$ 450 J energy, 600 fs pulse duration, 50 $\mu$m focal spot). The effect of the laser prepulse on the relativistic laser interaction and the acceleration processes will be analyzed.   

Characterizing the spatial resolution of scintillators for imaging applictions of laser-driven proton beams

Laser driven proton beams are widely used in visualizing the electromagnetic fields in high-energy-density physics experiments. However, typical detectors for proton imaging, i.e.\ radiochromic film (RCF) and plastic-track (CR39) detectors, are single-use and unable to meet the needs of higher repetition-rate facilities. Scintillators are a viable substitute their reusability and prompt, easy data acquisition by imaging the emitted optical signal onto a CCD camera are both advantageous features for a rep-rated experiment. We perform experiments using the Tcubed laser system at the University of Michigan to diagnose the intrinsic spatial resolution of the scintillators based on resolution grids imprints on the proton beam.The signal-to-noise from the laser-driven experiment, where there is significant relativistic electron and x-ray flux, is compared with Cyclotron based data [1].A configuration where the magnified imprint of a mesh in the proton beam is used to demonstrate that scintillators are capable of comparable overall spatial resolution to RCF for applications in proton beam diagnosis and radiography applications. [1] M.J.-E. Manuel, \textit{et al.}, Nucl.\ Inst. Meth.\ Phys.\ Res.\ A, \textbf{913}, 103 (2019).   

\textbf{A new device for high repetition rate plasma mirrors for petawatt class lasers using liquid crystal films}.

The full use of high repetition rate (HRR, \textgreater 1 Hz) PW class lasers is currently hindered by limitations in target insertion, diagnostics, real time data analysis and more. In particular, HRR plasma mirrors (PMs) are necessary for some experiments either for pulse contrast enhancement or beam redirection. We have previously shown the ability to form quality PMs using films made of the liquid crystal 8CB, including varying thicknesses films (10 nm to \textgreater 1 $\mu $m) for use as targets for ion acceleration (Poole, \textit{et al.}, Applied Physics Letters \textbf{109}, 151109 (2016)), plasma mirrors for pulse contrast enhancement (Poole \textit{et al.} Scientific Reports \textbf{6}, 32041 (2016)) at \textasciitilde 1 shot/minute repetition rates, and beam redirection for laser accelerated electrons at the BELLA Center (LBL). Here we describe a new device that can deliver \textasciitilde 20 nm thick 8CB LC films for use as plasma mirrors at repetition rates of \textasciitilde 1 Hz with good film flatness and improved pointing stability and robustness at a very low cost per film. We also discuss the use of other liquid crystals, including discotic liquid crystals.   

Soft X-ray Measurements of High Intensity Kilohertz Laser-Solid Interactions

Relativistic laser-solid interactions are a pathway to producing bright, attosecond duration, x-rays. The radiation produced is correlated to the driving laser and the specific plasma conditions. We present measurements of this radiation emission from a high-resolution spectrometer. We demonstrate how features in the emission correspond to plasma characteristics such as electron temperature.   

Bremsstrahlung efficiency in ultrahigh intensity laser-solid interactions with thin foil targets

Non-linear Compton scatter (NCS) and bremsstrahlung radiation are responsible for the generation of X-rays in laser-solid interactions. While it is expected that NCS will efficiently produce X-rays for next generation lasers, the weaker NCS signals from laser-solid interactions in existing facilities have not yet been observed due to the bremsstrahlung background. This work derives a quantitative description of the bremsstrahlung efficiency in interactions between ultrahigh intensity short pulse lasers, and thin foil targets. This model was tested against simulations from the particle-in-cell code EPOCH, which has been extended to include bremsstrahlung modelling capabilities. The relative efficiencies of bremsstrahlung and NCS have been compared to seek NCS signatures which may be observed experimentally.   

Numerical Study of Quantum Effects on Non-Linear Plasma Waves~

Quantum plasmas have recently attracted attention due to the miniaturization of electronic devices, the study of warm dense matter and of white dwarf stars. As the quantum analog to Vlasov-Poisson, Wigner-Poisson provides a kinetic model for a quantum plasma but is computationally challenging to solve. We use a new numerical method which combines the techniques from the Vlasov and Wigner literature to study waves in a quantum plasma. We employ Strang splitting to divide Wigner-Poisson into an advection equation and a solve involving the pseudodifferential operator. A forward, Semi-Lagrangian scheme based on the Convected Scheme handles the advection piece which allows the bypassing of the restrictive CFL condition. A Fourier transform handles the pseudodifferential operator. High order is achieved by using WENO to calculate small corrections to the Convected Scheme. We use traditional problems such as Landau damping and the two-stream instability to validate our code with known results. In the context of nonlinear stationary states, KEEN waves attracted attention as an answer to how BGK-like states could be formed from a Maxwellian plasma. We simulate KEEN waves under the Wigner-Poisson model to examine their existence under quantum effects. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under contract DE-AC52-07NA27344. Lawrence Livermore National Security, LLC   

Asymptotic quasioptical theory of mode-converting wave beams

A general quasioptical theory of short-wavelength RF beams is proposed (arXiv:1901.00268). We use the Weyl calculus to develop a rigorous asymptotic approximation of the governing dispersion operator $\hat{D}$ and its projection on modes of interest. A parabolic differential equation (quasioptical equation) is derived for the wave envelope in curved coordinates constructed around some reference ray which propagates close to the beam center. The theory does not assume any specific $\hat{D}$ and, unlike other quasioptical models, can capture mode conversion. Single-mode and mode-converting beams are described on the same footing. Based on this theory, a new quasioptical code PARADE (PAraxial RAy DEscription) has been developed and tested (arXiv:1903.01357, arXiv:1903.01364), as reported in a separate poster by K. Yanagihara {\it et~al.}.   

Quasioptical modeling of the power deposition by mode-converting wave beams in fusion plasmas

Electron-cyclotron heating and current drive with mm waves require modeling of the power deposition with high precision. Multi-dimensional full-wave simulations are prohibitively expensive at these wavelengths, and ray- and beam-tracing techniques are not sufficiently accurate in many practical applications. We report quasioptical modeling of the power deposition by mm-wave beams using a new code PARADE (PAraxial RAy DEscription), which was recently presented in [arXiv:1901.00268, arXiv:1903.01357, arXiv:1903.01364; to appear in Phys. Plasmas]. The beam transverse structure is calculated from first principles with the inhomogeneity of the absorption coefficient taken into account. The simulations also account for beam refraction, diffraction, and mode conversion, such as the X--O mode conversion at the fusion plasma edge caused by the magnetic shear.   

Fast metaplectic algorithm for RF simulations

The WKB approximation for modeling waves in inhomogeneous plasma is known to break down at cutoffs (caustics). However, it can be reinstated if the wave propagation is considered in the ray phase space rather than in the configuration space as usual. In particular, fold caustics can be treated by continually rotating the phase space such that the projection of the dispersion surface onto the coordinate plane is never singular, thereby restoring the geometrical-optics approximation. Here, we discuss a new fast algorithm for performing such phase-space rotations, or more generally, linear canonical transformations of the phase space. We also describe how this algorithm can feature in a reduced code for modeling RF waves which keeps the wave amplitude finite at cutoffs.   

A Hamiltonian formulation for the perturbed Vlasov-Maxwell equations

The Hamiltonian formulation for the perturbed Vlasov-Maxwell equations is expressed in terms of the perturbation derivative $\partial{\cal F}/\partial\epsilon \equiv [{\cal F}, {\cal S}]$ of an arbitrary functional ${\cal F}[f,{\bf E},{\bf B}]$ of the Vlasov-Maxwell fields $(f,{\bf E},{\bf B})$, which are assumed to depend continuously on the (dimensionless) perturbation parameter $\epsilon$. Here, $[\;,\;]$ denotes the standard Vlasov-Maxwell functional bracket, and the perturbation {\it action} functional ${\cal S}$ is said to generate perturbations of the Vlasov-Maxwell fields. The new Hamiltonian perturbation formulation highlights the crucial roles played by polarization and magnetization in Vlasov-Maxwell perturbation theory.   

Three-wave interactions in magnetized warm-fluid plasmas

Three-wave coupling coefficients in magnetized warm-fluid plasmas is computed by solving the fluid-Maxwell's equations to second order using multiscale perturbative expansions. A convenient general formula is obtained, whereby numerical values of the coupling coefficient can be determined for any three resonantly interacting waves propagating at arbitrary angles. To illustrate how the general formula can be applied, coupling coefficient governing laser scattering is evaluated as one example. In conditions relevant to magnetized inertial confinement fusion, Raman and Brillouin instabilities are replaced by scattering from magnetized plasma waves when lasers propagate at oblique angles. As another example, coupling coefficient between two Alfven waves via a sound wave is evaluated. In conditions relevant to solar corona, the decay of a parallel Alfven wave only slightly prefers exact backward geometry.   

Solitary zonal structures in subcritical drift waves: a minimum model

Solitary zonal structures have recently been identified in gyrokinetic simulations of subcritical plasmas with background shear flows, while the physical mechanism that underpins these structures is not well understood. These structures share similarities with the drift-wave--zonal-flow (DW--ZF) solitons known as solutions to the modified Hasegawa--Mima equation (mHME). However, the solitons cannot be sustained when background shear flows are introduced to the mHME. We show that, by further including a primary instability in the mHME to model subcriticality, solitary zonal structures can readily be restored. Accordingly, these structures can also be retained in models that subsume the mHME and yield primary instabilities, such as the modified Hasegawa--Wakatani equation. Remarkably, all these structures satisfy the same ``equation of state'' of the DW--ZF solitons, which is a simple algebraic equation that connects the DW envelope and the ZF velocity.   

Topological Waves in Plasmas and Analogs to Topological Insulators

A topological understanding of matter is not only deepening our knowledge of physics, but also leading to novel practical devices and applications. Topological insulators have led to a surge in interest in the topological protection and robustness to backscatter afforded to particular unidirectional edge modes. While these concepts were originally developed in photonic systems and electronic structures, recent advances have demonstrated that continuum fluid systems can support topological waves. We identify a specific candidate for a topological plasma wave. To be concrete, the wave is an electromagnetic RF surface wave propagating at the boundary between magnetized plasma and vacuum. The magnetic field breaks time-reversal symmetry. Using the cold-plasma model, we show that a plasma can be characterized by a nontrivial Chern number, although regularization of the Hamiltonian is required for the integral of the Berry curvature to yield an integer Chern number. Moreover, we perform detailed theoretical calculations and demonstrate that this wave could exist at plasma parameters achievable in existing laboratory devices. Experiments to confirm the existence of this wave would open a new frontier in the exploration of the physics of topological waves in plasmas.   

Excitation and breaking of relativistic longitudinal electron-ion modes in a cold plasma

The excitation and breaking of relativistically intense electron-ion modes in a cold plasma is studied using 1D-fluid simulation techniques. To excite the mode, we have used a relativistic rigid homogeneous electron beam propagating inside a plasma with a velocity close to the speed of light. It is observed that the wake wave excited by the electron beam is identical to the corresponding Khachatryan mode [{\it Phys. Rev. E, 58, 6(1998)}], a relativistic electron-ion mode in a cold plasma. It is also seen that the numerical profile of the excited electron-ion mode gradually modifies with time and eventually breaks after several plasma periods exhibiting explosive behavior in the density profile. We have found that the numerical wave breaking limit of these modes lies much below than their corresponding analytical limit. The discrepancy between the numerical and analytical limit has been understood in terms of phase-mixing process of the mode. The phase mixing time (or wave breaking time) obtained from the simulations has also been scaled as a function of beam parameters and found to follow the existing analytical scaling.   

Very basic, applied, and in between approaches for the nonlinear kinetic description of electron plasma waves and stimulated raman scattering.

In order to address large scale, and complex experiments, such as those involved in laser fusion, one needs to solve Vlasov-Maxwell equations. This requires the introduction of new theoretical results, since a purely numerical approach would be completely out of reach. As a first step, we show how to solve the nonlinear Vlasov equation when the charge density is induced by a sinusoidal electron plasma wave, and when wave-particle interaction is in the trapping regime. Our result is valid whatever the rate of variation of the wave amplitude. It goes beyond an envelope description. Moreover, for slowly-varying waves, we derive coupled envelope equations for stimulated Raman scattering in an inhomogeneous and non-stationary plasma. We also show how to solve our equations in a three-dimensional geometry by using a recently introduced Monte-Carlo ray-tracing method. Finally, we simplify our equations so as to come with a model that is being implemented in our hydrodynamical code at CEA, and will be used to design experiments at the Laser MegaJoule facility.   

Ion-Kinetic Calculations of the Collisional Kelvin-Helmholtz Instability

Using a recently implemented hybrid (kinetic ion/fluid electron) algorithm in the LANL particle-in-cell code VPIC [1], we study Kelvin-Helmholtz (KH) unstable sheared flow layers including full ion kinetics. The ion-ion collisions are treated with the Takizuka-Abe [2] particle-pairing algorithm, which includes the velocity dependence of the screened Coulomb collisions. In agreement with fluid predictions [3], the KH instability is suppressed by ion viscosity for Reynold’s numbers below ~100, corresponding to Knudsen numbers greater than ~0.01 for subsonic flows. In addition to resolving the physical collisional viscosity, the particle treatment allows us to track plasma particle mixing over time. [1] K. J. Bowers, et al., Physics of Plasmas, 15, 055703 (2008). [2] T. Takizuka & H. Abe, Journal of computational physics 25, 205-219 (1977). [3] E. Roediger et al., Monthly Notices of the Royal Astronomical Society, 436, 1721-1740 (2013).   

3D Numerical Simulation of Kink-Induced Rayleigh-Taylor Instability for Fast Magnetic Reconnection

Kink-Induced Rayleigh-Taylor (KI-RT) instability has been observed and proposed as a MHD mechanism to instigate fast magnetic reconnection in the Caltech jet MHD experiment [1]. Experimental observations and preliminary analysis showed that KI-RT occurs more readily for heavy ions and for sufficiently large acceleration [2] and that the disrupted current causes an inductive electric field that accelerates initially 2 eV electrons to sufficiently high energy to produce 6 keV X-rays [3]. However, how plasma parameters govern KI-RT dynamics, the exact requirements for KI-RT to occur, and the process of how KI-RT leads to a magnetic reconnection have not yet been resolved. This work addresses these questions by means of a 3D numerical MHD simulation that uses Chodura resistivity for simulating the localized kinetic enhancement of resistivity when J/ne becomes large. The numerical simulation is consistent with a simplified model for acceleration from kink leading to RT and fast reconnection. The results are compared to experimental and space/solar plasmas. [1] Moser, A. L. and Bellan, P. M., Nature 482, 379 (2012), [2] Zhai, X. and Bellan, P. M., Phys. Plasmas 23, 032121 (2016), [3] Marshall, R. S. and Bellan, P. M., Phys. Plasmas 26, 042102 (2019).   

Linear and nonlinear benchmarks of CLT code and M3D-C1 code

MHD instabilities are a common phenomenon in tokamaks and have been investigated by many 3D toroidal MHD codes for many years. The validation of the codes in the linear regime is accomplished by reproducing analytic scaling laws, such as $\gamma \sim S^{-3/5}$for the tearing mode and $\gamma \sim S^{-1/3}$ for the resistive kink mode. Here $\gamma $is the linear growth rate and $S$is the Lundquist number. Nonlinear benchmarking between different codes presents a new challenge. M3D-C1 is an implicit, three-dimensional high-order finite element code for the solution of the time-dependent nonlinear two-fluid magnetohydrodynamic equations in cylindrical or toroidal geometry. CLT is an explicit three-dimensional finite-difference nonlinear magnetohydrodynamics code for toroidal geometry. They are both used to investigate MHD instabilities in tokamaks. In this work, we present quantitive benchmarks of CLT and M3D-C1 for several instabilities, including linear and nonlinear tearing modes, an ideal internal kink mode, and disruptive instabilities.   

Verification of a PIC-Fluid Hybrid Code With the Two-stream Plasma Instability Problem

While Particle in Cell (PIC) codes accurately model plasmas with a range of densities, high-density plasmas propose a computational challenge for PIC codes.  If the particles in high-density plasma can be assumed to have a Maxwellian distribution, multi-fluid codes offer a significant computational advantage in these regimes.  While these two types of plasma codes are well suited for modeling problems in their respective regimes, issues can arise when trying to model a plasma application that involves high density plasmas alongside particles with non-Maxwellian distributions. These such problems serve as a motivation for a hybrid approach that combines the two codes. This poster presents an implementation of a PIC-Fluid hybrid code, and application of such code to the two-stream plasma instability problem. The results are compared with analytic theory results taken from a linearized version of the governing equations, as well as results from the PIC and multi-fluid parts of the code. Through the comparison of these results, this poster will show the computational and numerical implications of using a Hybrid code over a PIC or multi-fluid code for certain problems.   

Merging and directional drift of electron spots during the nonlinear evolution of the current filamentation instability.

The transport and energy deposition of relativistic electron beams in the radially non-uniform plasmas are investigated with two-dimensional electromagnetic particle-in-cell simulations. For the beam with the radius much larger than plasma skin depth, the current filamentation instability excited by the relativistic electron beam can be clearly observed, which breaks the electron beam into a large number of filaments and leads to the formation of strong magnetic field consequently. As the beam-plasma system evolves self-consistently, asymmetric transverse magnetic field, which is associated with plasma density gradient, contributes to the directional drift of electron focal spots and thus can enhance the merge effectively. The effects of different plasma distributions on transport and energy deposition of the eletron beam are compared. Furthermore, an axial electric field is generated in the middle of two filaments and causes the corresponding energy step change when they merge under the action of the magnetic field. The energetic plasma jet, vertical to the merging direction, is shown to result from the defocusing effect of the magnetic field.   

Study of Streaming Plasma Instabilities in the Extreme Relativistic Regime

Relativistic streaming plasma instabilities are important in a wide range of high-energy astrophysical environments, from blazar jets to gamma-ray bursts. While the linear phase of these instabilities is well studied, important aspects of their saturation and nonlinear evolution are not yet clear. Laboratory experiments using highly relativistic lepton beams (e.g. from FACET-II or laser wakefield beams) could soon shed light on the nonlinear physics of these instabilities and on their ability to produce bright gamma-ray emission. We will present a detailed numerical study, using particle-in-cell simulations, and theoretical analysis of the evolution of streaming plasma instabilities in the extreme relativistic regime (where the Lorentz factor of the lepton beam is > 1000). We will discuss how the background plasma response (including the ions) affects the growth and nonlinear evolution of the instabilities, the slow down of the relativistic beam, and the corresponding high-energy radiation emission.   

Particle-In-Cell Simulations of Stimulated Raman and Compton Scattering for $k \lambda_d$ from 0.2 to 2

We use the particle-in-cell code OSIRIS to study how backward stimulated Raman scattering (SRS) and stimulated Compton scattering, and forward SRS compete and how they are affected by external magnetic fields. The parameter space covered includes regimes of near threshold, weakly coupled, and strongly coupled growth. We have previously shown how small magnetic fields can significantly modify the evolution of backward stimulated Raman scattering (SRS) in the kinetic regime due to the enhanced dissipation of nonlinear electron plasma waves propagating perpendicular to magnetic fields $[1]$. The transition from Raman scattering to Compton scattering has been explored in $[2]$, and the competition between forward and backward Raman scattering has been explored in $[3]$. Driven by the collaboration between UCLA, UCSD and LLNL on the Titan LPI project, we continue this work, examining a range of experimentally possible parameters that could cover a wide range of $k\lambda_d$ growth regimes.   

Overview of the Basic Plasma Science Facility

The Basic Plasma Science Facility (BaPSF) at UCLA is a US national user facility for studies of fundamental processes in magnetized plasmas. The centerpiece of the facility is the Large Plasma Device (LAPD), a 20m long, magnetized linear plasma device\footnote{W. Gekelman, et al., Rev. of Sci. Inst. {\bf 87}, 025105 (2016)}. This LAPD has been utilized to study a number of fundamental processes, including: collisionless shocks\footnote{A.S. Bondarenko, et al., Nat. Physics {\bfseries 13}, 573 (2017)}, dispersion and damping of kinetic and inertial Alfv\'{e}n waves\footnote{C.A. Kletzing, et al., Phys. Rev. Lett. {\bfseries 104}, 095001 (2010).}, turbulence and transport\footnote{D.A. Schaffner, et al., Phys. Rev. Lett. {\bfseries 109}, 135002 (2012). } interactions of energetic ions and electrons with plasma waves\footnote{B. Van Compernolle, et al., Phys. Rev. Lett. {\bfseries 114}, 245002 (2015).} and RF sheaths produced by an ICRF antenna\footnote{M. Martin, et al., Phys. Rev. Lett. {\bfseries 119}, 205002 (2017)}. An overview of the facility and recent upgrades and recent research using the facility will be provided. In addition to a discussion of how prospective users can apply for experimental time.   

Modifications produced on a large magnetized plasma column by a floating end-plate that is partially emissive: experiment and theory

This poster reports on an experiment performed in the Large Plasma Device (LAPD) at the University of California, Los Angeles (UCLA) in which an electrically floating structure is placed near the end of a 20-meter magnetized plasma column. The structure consists of a flat carbon plate that acts as a mask for a smaller, ring-shaped LaB6 emissive surface whose temperature can be externally controlled. This configuration has been previously used to study electron heat transport and pressure-driven avalanches by biasing the LaB6 ring-cathode with respect to a distant anode in a cold afterglow plasma. In contrast, the present study is performed during the active portion of the steady-state discharge in which the nominal plasma parameters are determined by injection of an electron beam from a BaO cathode at the opposite end. Even without an applied bias on the LaB6 cathode, the self-consistent potential and current profiles are modified near the end plate as the LaB6 temperature is increased, resulting in density increases on the field lines in contact with the emissive surface. In the absence of enhanced ionization, at the largest cathode temperatures the ambient density can be doubled. A theoretical model is presented that provides a quantitative explanation for the observations.   

Driven Thermal Waves in Magnetized Plasmas: Diagnostic Uses, Wave-Field Modeling, and Cross-Field Structure

Results are presented from basic heat transport experiments performed in the LAPD at UCLA using a magnetized electron temperature filament. A CeB$_6$ cathode injects low energy electrons along a magnetic field into the center of a pre-existing plasma forming a hot electron filament embedded in a cold plasma. Previous experiments observed spontaneous thermal waves corresponding to a quarter-wave resonance. In new experiments, perturbations are added to the cathode bias to create an oscillating heat source. Probe measurements allow for the determination of the amplitude and parallel phase velocity of the resulting thermal oscillations over a range of driver frequencies. The results demonstrate the presence of a thermal resonance and are used to verify the parallel thermal wave dispersion relation based on classical transport theory. This technique provides a measure of the density normalized thermal conductivity, independent of the electron temperature. A heat equation in the form of a reaction-diffusion equation has been derived and the solution closely matches the observed thermal resonance. More recently, the cross-field structure of the waves has been investigated in both classical and anomalous transport conditions; a preliminary analysis of these experiments is presented.   

An experiment to investigate parametric interaction and mixing of microwaves in plasma

Building on earlier research, investigating the mechansim for Auroral Kilometric Radiation [1,2], a “linear plasma” experiment is being developed to investigate the coupling of multiple microwave beams in magnetised plasma. The magnetic flux desnity is expected to reach 0.09T. The plasma will be generated by a helicon system using a 'flat sprial' type of antenna. This will produce a large (~0.5m diameter, 2-3m long), dense, cool plasma, potentially with a high ionisation fraction. Microwave beams from fixed frequency magnetrons and wideband TWT ampligiers will be used in multi-signal interaction experiments. The paper will present progress on this system. [1] Ronald K., Speirs D.C., McConville S.L., Phelps A.D.R., Robertson C.W., Whyte C.G., He W., Gillespie K.M., Cross A.W., Bingham R., 2008, Phys. Plasmas, 15, art.056503 [2] Speirs D.C., Bingham R., Cairns R.A., Vorgul I., Kellett B.J., Phelps A.D.R., Ronald K., 2014, Phys. Rev. Lett., 113, art 155002   

Development of a versatile permanent magnet based compact helicon plasma source

A compact helicon plasma source, which is advancement over the conventional inductively coupled plasma sources for producing high-density plasmas, is developed at IPR, India. Plasma and wave studies are carried out in Argon and hydrogen plasmas. The experimental setup uses a permanent magnet instead of the electromagnets. The plasma is produced in the narrow source tube by applying RF Power at $\omega \quad =$ 13.56 MHz. It expands in a diverging field into the expansion chamber. In Argon plasma, a drastic increase in plasma density is achieved in the expansion chamber with the use of full line cusps. Higher radial modes are observed in the complicated field configuration. Using hydrogen gas, with the right antenna helicon plasma is created. In the expansion chamber the plasma conditions are conducive to the high yield of H- without the use of Caesium. The diverging magnetic field also enables resonance cone absorption when $\omega $ approaches $\omega _{ce} $.   

Argon Ion Temperatures in a 10kW Helicon Source

There is a need to conduct experiments in plasmas with nearly equal ion and electron temperatures to suppress temperature-driven effects such as ion acoustic instabilities. Typical laboratory helicon plasmas operate below 1 kW, where the ion temperature is much less than electron temperature. Previous ion temperature measurements in the Controlled Shear Decorrelation Experiment (CSDX) and the Resonant Antenna Ion Device (RAID) exist only up to a maximum of 4 kW forward power. Projections based on measurements in RAID suggest that nearly equal ion and electron temperatures will occur at helicon source powers of 10 kW. Laser induced fluorescence (LIF) measurements of the ion velocity distribution functions and the ion temperatures in CSDX are presented for forward powers up to 10 kW. These high powers are achieved with a novel water-cooled dielectric chamber through which the rf is coupled to the plasma. Perpendicular ion temperatures up to \textasciitilde 2.5 eV are observed at 10 kW, assuming Maxwellian fits to the ion velocity distributions. These results are in agreement with the RAID scaling at 4 kW. Based on these observations, we extrapolate trends perpendicular ion temperatures at even larger RF powers in helicon sources.   

Rayleigh Taylor instability in the solar corona loop experiment

We are observing a Magneto Rayleigh Taylor instability (MRTI) in a laboratory experiment that simulates solar corona loops. Unlike a previous experiment in our lab where MRTI instigated by the effective gravity of a kink instability was observed $^{\mathrm{[1]}}$, here the acceleration from the hoop force acting on the loop provides the effective gravity. Detailed measurements indicate a scaling where the observed axial wavelength $\lambda $ increases when a larger bias magnetic field is used. This scaling is possibly consistent with the theoretical MRTI growth rate $\gamma ^{2}=gk-\frac{\left( {{\rm {\bf k}}\cdot {\rm {\bf B}}_{{\rm {\bf 0}}} } \right)^{2}}{\mu_{0} \rho }$, because this theoretical growth rate implies that if ${\rm {\bf k}}$is parallel to ${\rm {\bf B}}_{{\rm {\bf 0}}} $ (i.e., undular mode), the fastest growing mode has $\lambda =\frac{2\pi }{k}=\frac{4\pi {\rm {\bf B}}_{{\rm {\bf 0}}}^{2}}{\mu_{0} \rho g}$. We are also exploring other features such as appearance of a kink after the MRTI, different experiment parameters and whether a fast magnetic reconnection happens during this process. [1] A. L. Moser and P. M. Bellan, Nature 482, 379 (2012).   

Observation and simulation of wave particle interaction in a Traveling Wave Tube upgrade

Beside telecommunications [Minenna et al., Eur. Phys. J. H 44, 1 (2019)], Traveling Wave Tubes (TWT) are useful to mimic plasma-like wave-particle interaction [Tsunoda et al., Phys. Rev. Lett. 58, 1112 (1987), Doveil et al., Cel. Mech. Dyn. Astr. 102, 255 (2008)]. We use a TWT with a 4 m long helix (diameter 3.4 cm, pitch 1 mm) slow wave structure. At one end, a triode produces an electron beam radially confined by a constant axial magnetic field. Movable probes, capacitively coupled to the helix, launch and monitor waves generated at a few tens of MHz with arbitrary waveform. At the helix other end, a trochoidal analyzer reconstructs the beam energy distribution function. The observed dispersion relation agrees very well with a sheath model also used to estimate the TWT impedance measured by the Kompfner dip method. For sufficiently large beam intensity, growth and saturation of a launched wave is observed. A new symplectic code DIMOHA [Minenna et al., Europhys. Lett. 122, 44002 (2018)] based on an N-body Hamiltonian approach for particles and waves, which has shown big success (computation time divided by 100) for commercial TWTs, is applied to our unconventional TWT. This work also leads to revisit the Abraham-Minkowski dilemma about light momentum.   

Experimental observation of drift wave turbulence in an inhomogeneous cusp magnetic field of MPD

This paper presents a detailed study on the controlled experimental observation of drift wave instabilities in an inhomogeneous Six pole cusp magnetic field generated by an in-house developed Multi-pole line cusp magnetic field device (MPD) \textbf{[}Patel \textit{et al}. Rev. Sci. Instrum., \textbf{44}, 726 (2018)]. The device is composed of six axially symmetric cusps and non-cusp (in between two consecutive magnets) regions. The observed instability has been investigated in one of these non-cusp regions by controlling the radial plasma density gradient with changing pole magnetic field which is a unique feature of this device. It has been observed that the frequency of the instability changes explicitly with the density gradient. Moreover the scale length of plasma parameters, frequency spectrum, cross-correlation function, and fluctuation level of plasma densities has been measured in order to identify the instability. The cross field drift velocity due to fluctuation in plasma parameters have been measured from the wave number- frequency $S (k_{z}, \omega )$ spectrum and verified with the theoretical values obtained from density scale length formula. .   

Characteristics of Ion Acoustic Wave in Multi Cusp Plasma Device.

In Multi-cusp Plasma Device (MPD) six electromagnets on the circumference of the device, have been used for the multi-cusp magnetic field profile production. This geometry has the centre of radius of curvature outside the confined plasma that provides magneto-hydrodynamic stability to the plasma. Such plasmas are very quiescent with density fluctuations $\le 0.1\% $. In the quiescent background, any external perturbation to plasma either in ion or electron regime can be examined better. Hence in this paper the experimental study of potential perturbation propagated as ion-acoustic wave (IAW) and characteristics of IAW interaction in MPD will be presented.   

Experimental study of Zonal flows in low pressure linear magnetized plasma

Low frequency (0.2-0.3kHz) coherent mode is observed in Inverse Mirror Plasma Experimental Device (IMPED). Measurement of radial ($k_{r} )$, poloidal ($k_{\theta } )$ and axial ($k_{\vert \vert } )$ wavenumber shows that $k_{\theta } $ and $k_{\vert \vert } $ are approximately zero while $k_{r} $ is finite and radial variation in polarity. The potential fluctuations are much stronger than the density fluctuations. Initial analysis shows that this mode is zonal flow in nature. The fluctuations due to zonal flow are strongest at the minimum of the electron temperature gradient scale length($L_{T_{e} } )$. On changing the radial location of minimum of $L_{T_{e} } $, the strongest zonal flow fluctuation is found to follow it. Further, the strength of zonal flow is found to increase on increasing the ratio of hot to cold electron population. The experimental results are presented and discussed.   

A comparison of electron velocity distribution measurements in the SPSC

Velocity-resolved particle measurements are important for many laboratory studies at the forefront of fundamental plasma physics. Energy analyzers have been used for decades to record field-aligned velocity distributions. Unfortunately, energy analyzers typically lack resolution to measure distributions on the timescale of waves in laboratory plasmas, making interactions of waves and particles difficult to study in the lab. An alternate technique that enables faster measurements, termed wave absorption, uses a probe wave whose absorption depends on the phase-space density of resonant particles. Wave absorption has been used to study electron heating in tokamak plasmas and wave-particle interactions between Alfven waves and electrons. Wave absorption measurements have not previously been directly compared with energy analyzer measurements. Using the Space Physics Simulation Chamber (SPSC) at the Naval Research Laboratory, whistler-mode wave absorption measurements of the background electron distribution are compared with energy analyzer measurements in the same plasma. Complicating effects, like the frequency-dependent radiation pattern of whistler-mode waves and the finite gyroradius of electrons, will be considered.   

Plasma instabilities in ExB plasma devices

Inter-streaming instabilities between electrons and ions are studied in a range of EXB plasma devices. Study 1: Nonlinear development of Electron Drift Instability is studied in azimuthal-radial PIC simulations of the annular channel of a Hall thruster. It is shown that in the nonlinear stage, the instability which starts as a short length scale linear instability, undergoes a sequence of transitions into longer wavelengths modes. The transitions in mode wavelengths are accompanied by related transitions of the magnitude of anomalous axial current. Study 2: Collisional PIC-MCC Simulations of the cross-sections of a Penning Discharge and a Cylindrical Magnetron are used for studying instabilities driven by ioniziation of neutrals. Evolution of these modes to a steady state and the possibility of modulating these ionization modes and associated plasma currents by tuning the background neutral pressure and the discharge voltage is investigated. Study 3: An E X B plasma existing in the parametric space where conventional quasi-neutrality just transitions over to non-neutrality, is simulated in the cross-section of a Penning-Malmberg trap. The differential drift between the ion and the electron component drives a spectrum of transient azimuthal-radial modes.   

Plasma and Radio Wave Generation by an Electron Beam in a Laboratory Plasma

Interaction between relativistic electron beams and a magnetized plasma is a fundamental and practical problem relevant to many challenging issues in space physics and astrophysics. For example, compact high-energy electron beam sources may be used on future spacecraft to map magnetic field lines during substorms, provided the beam is not disrupted by wave generation. Similar classes of waves (x-mode, Langmuir, etc.) may also be generated by naturally occurring electron beams, possibly explaining type II/III solar radio emissions. We present preliminary results from a $20$~keV beam setup on the Large Plasma Device (LAPD) at UCLA aimed at addressing wave generation mechanisms, wave properties, and dependence on plasma parameters. Results show strong emission between the plasma and upper hybrid frequencies; the emission peaks on the edge of the beam profile and is detectable as radiation outside the plasma. The parallel phase speed is measured to be consistent with wave production via a Landau resonance process. Observed second harmonic power suggests a role for non-linear processes under certain parameters. Future experiments will extend our parameter space to a $\sim 1$~MeV electron beam. Comparison of experimental results with theory and kinetic simulations is presented.   

MHD stability constraints on divertor heat flux width in DIII-D

The spatial width of heat flux flowing into the divertor is examined in DIII-D in the context of MHD stability limits. At low power the SOL width remains consistent with the ITPA scaling, dependent only on the midplane poloidal field. The midplane separatrix pressure gradient remains below the MHD ballooning stability limit even for very high density with divertor detachment. At high power with increasing density the midplane pressure gradient approaches the MHD stability limit at a midplane density near half of the Greenwald density limit. Further increases in density result in a broadening of the SOL and divertor temperature and density profiles to keep the pressure gradient below the MHD limit. However, the increased turbulence and transport required to maintain midplane separatrix MHD stability does not degrade the edge pedestal pressure and overall performance. The implications of these effects for divertor heat flux and its control in future reactor scale tokamaks will be explored.   

\textbf{Non-axisymmetric heat flux patterns on tokamak divertor tiles}

An updated model for divertor heat flux simulation in 3D plasmas with applied RMP finds peak heat fluxes and layer widths compare well to infrared camera measurements in DIII-D. A heat flux model for perturbed plasmas based on guiding center ion drift in vacuum fields [A. Wingen et al., PoP (2014)] is reintroduced. Divertor footprints are simulated for multiple ion kinetic energies and summed, using a Maxwellian distribution as weight factors to account for their respective contribution. Ion drifts cause the heat flux to shift towards the private flux region. Recently, the model was extended to add E\texttimes B drift effects. It is found that a radial electric field E$_{\mathrm{r}}$ in the near SOL can considerably shift the footprints toroidally, leading to a smear out effect of the incident heat flux, while the E$_{\mathrm{r}}$ inside the separatrix has little impact on footprints. The modeled toroidally averaged heat flux patterns can be fit well to an Eich profile [T. Eich et al., PRL (2011)].   

SOL and Divertor Fluctuations and Transport During Detachment

Turbulence in the DIII-D divertor and main chamber is characterized in attached and detached L and H-mode discharge conditions revealing the impact of distinct physics on each of particle and energy transport. Plasma density is increased in successive repeat discharges until T$_{\mathrm{e}}$ at the divertor plate is \textasciitilde 2-5 eV. As T$_{\mathrm{e}}$ drops at the plate, the heat flux profile width, measured by IRTV, varies little while the particle flux profile, measured with probes as , narrows by a factor of 2 until detachment. Density fluctuations increase 50-100{\%} as density increases towards detachment, but relative fluctuation levels, actually drop by 10X. However, for a given density, near-plate fluctuation levels always increase with divertor T$_{\mathrm{e}}$, suggesting that heat is the free energy source feeding the fluctuations.   

Edge neutral density radial profiles in open and closed divertor conditions using Balmer series spectroscopy on DIII-D

Upstream radial profiles of D$\alpha $656.3 nm and D$\gamma $434.0 nm have been measured across the outboard midplane separatrix, while the strike points are placed in the closed, upper divertor and the open, lower divertor in L-mode and ELMy H-mode plasma. This provides a radial neutral density profile, n(r), of the main fueling species in this region. Analysis of edge n(r) ranging from strongly attached to detached conditions will be presented to compare the effects of divertor geometry on neutral leakage. Initial results using only D$\alpha $ comparing lower and upper single-null configurations show neutral densities to be higher at the open, lower-single-null case far from the separatrix, but lower than the closed upper-single-null configuration near the separatrix with very large uncertainty. By using multiple Balmer emission lines (D$\alpha $, D$\gamma )$ at the same lines-of-sight, a more consistent edge n(r) is expected by cross-checking the profiles from each calculation against one another, thus reducing the possible uncertainties when estimating n(r) from a single Balmer emission line, e.g. from atomic physics effects such as photon reabsorption at high n$_{\mathrm{e}}$. The use of the integrated synthetic diagnostic code, CHERAB, provides an analysis work-flow to facilitate this multi-emission line approach.   

Enhanced tungsten prompt re-deposition during ELMs in the DIII-D divertor

Recent measurements conducted in the DIII-D divertor indicate a substantially higher tungsten prompt re-deposition fraction, $f_{redep} $, occurs in H-mode discharges with edge localized modes (ELMs) relative to L-mode discharges. A value of $f_{redep} $ of \textasciitilde 0.72 was inferred on a 15 mm diameter W-coated DiMES sample exposed to 3 repeat L-mode discharges, whereas an $f_{redep} $ value of \textasciitilde 0.88 was calculated on an identical sample exposed to one H-mode discharge and 2 repeat L-mode shots. This result is interpreted using the recently developed free-streaming plus recycling model (FSRM), extended to include the effect of W prompt re-deposition using analytic geometric approximations. The FSRM predicts that $f_{redep} $ should increase from \textasciitilde 0.4 in the inter-ELM phase to 0.9 during the peak of the intra-ELM phase for strongly attached DIII-D discharges, and from \textasciitilde 0.7 to \textasciitilde 0.95 for a typical detached divertor case. This increase in $f_{redep} $ is due to the strong increase in divertor electron density, and correspondingly shorter W ionization mean free path during the ELM. This work represents the first experimental evidence of enhanced prompt re-deposition during ELMs and is a promising result for ITER where nearly all the W sputtered near the strike-point will be during ELMs. Additional ELM-resolved W re-deposition measurements using spectroscopic interpretation of WI/WII line ratios will also be presented.   

Gross and Net Erosion of Silicon from SiC-coated PFCs in the DIII-D Divertor.

Gross and net erosion rates of silicon from silicon carbide (SiC) coatings were measured in the divertor of DIII-D under well diagnosed reactor-relevant plasma conditions. Amorphous and crystalline SiC coatings on graphite with a thickness of 80 nm and 250 microns, respectively, were exposed near an attached outer strike point of L-mode plasmas using the Divertor Material Evaluation System (DiMES). Plasma density and electron temperature near the center of the coatings were $n_{e\thinspace }=$ 4x10$^{\mathrm{19\thinspace }}$m$^{\mathrm{-3}}$ and $T_{e\thinspace }=$ 23 eV. Gross erosion of Si from all samples was measured spectroscopically using the Si II 636 nm line. It was found to be a factor of 4 higher for the amorphous coatings compared to the crystalline one, possibly because of the higher surface binding energies in the latter. An average net Si erosion rate of 3x10$^{\mathrm{16\thinspace }}$cm$^{\mathrm{-2}}$s$^{\mathrm{-1}}$ was measured with Rutherford backscattering on the amorphous coatings with toroidal extent of 1 mm, in good agreement with ERO-OEDGE modeling. Using this rate and corrections from the modeling, an effective SXB coefficient for the Si II 636 nm line of 52 and a Si sputtering yield of 0.017 Si/D were calculated. Compared to previous DiMES measurements of net C erosion from pure carbon coatings at 3 times lower $n_{e}$ and comparable $T_{e}$, the effective reduction of the C erosion rate from SiC is about an order of magnitude. Recently experiments with SiC were repeated at lower $n_{e}$; data analysis is in progress.   

Assessing Scrape-Off Layer Impurity Content using Isotopic Methane (13CH4) Tracer Experiments and Collector Probes at DIII-D

Recent experiments at DIII-D have enabled the development of an upper single null plasma shape and a non-perturbative methane (CH4) injection flow rate (10 T-L/s) through the upper outer baffle for isotopically tagged 13CH4 tracer experiments. With an elongated plasma, the `crown' is brought in close proximity to the DiMES station where samples for obtaining deposition profiles can be placed. During these experiments, Langmuir probe measurements have been used to assess heat flux in order to prepare for the insertion of impurity collector probes (CPs). In addition, a new design for the large diameter MiMES CPs, refined from those used in the 2016 Metal Rings Campaign, is presented. A wider collection face and a sampling length of approximately 4 m on either side of the insertion location allows for higher fidelity evaluation of the C content in the far scrape-off-layer. Laser ablation mass spectroscopy analysis of DIII-D graphite samples from previous 13CH4 experiments will be presented demonstrating the high sensitivity (\textgreater ppm) of distinguishing 13C deposition. Data collected during these experiments is being used to compare with impurity transport modelling to enhance the predictive capabilities through a multi-reservoir impurity particle balance and codes such as OEDGE.   

UV Spectroscopy Advancements for Measurements of Tungsten Erosion {\&} Re-deposition

A new high-resolution and high-throughput spectrometer has been constructed to resolve the most promising W line radiation in the DIII-D divertor between 200 and 400 nm arising from erosion during and between ELMs. Recently completed spectral surveys in the Compact Toroidal Hybrid experiment have identified over 30 low charge state W emission lines in the UV region, which can be combined with atomic predictions to determine the erosion and re-deposition of plasma facing W surfaces. The importance of metastable level populations on the W spectrum requires that multiple W emission lines be monitored simultaneously to accurately characterize erosion rates. The new UV spectrometer has a maximum resolving power of 1.6 {\AA} at 250 nm with better than 1 kHz temporal resolution. The designs for fiber-coupled collection optics and instrument shielding required for installation on the DIII-D tokamak will be presented along with expected signal levels for DIII-D plasma conditions.   

Spectroscopic measurements of neutral tungsten for gross erosion measurements

Tungsten erosion at the plasma boundary is diagnosed spectroscopically with a new high- resolution UV spectrometer in combination with improved predictions for atomic coefficients representing the ionizations per photon (S/XB). New collisional radiative modeling suggests that neutral tungsten emission in low density linear plasma experiments such as the PR-2 facility as well as fusion relevant experiments are dominated by the non-steady state W metastable populations. The time evolution of neutral W metastables is tracked using the recently released collisional radiative solver ColRadPy. A scheme for measuring non-steady state metastable populations will be presented using W I spectral lines around 260 nm. A new high- resolution UV optimized spectrometer has been tested on the Compact Toroidal Hybrid (CTH) experiment in preparation for installation on DIII-D. High- resolution tungsten spectra from CTH will be presented and compared to modeled spectra using W I R-matrix excitation data and exchange classical impact parameter (ECIP) ionization data.   

An update on new tungsten ionization and excitation data for use in erosion, redeposition, and transport studies

Spectroscopic techniques to measure erosion, redeposition, and transport for tungsten plasma facing components require accurate atomic data, with the near neutral ion states being the most critical. In support of tungsten experiments on DIII-D, the available atomic data has been undergoing improvements over the last few years using large-scale quantal calculations. Recent modeling suggests that it is important to include the role of non-steady-state metastables in the interpretation of many diagnostic measurements, which led to a new R-matrix with PseudoStates (RMPS) calculation for neutral W metastable ionization. This calculation also included new neutral W ground state ionization results. The impact of the new data on W I S/XBs and effective ionization rate coefficients is presented. New non-perturbative results for W$^{\mathrm{+}}$ ionization and excitation were also performed, for applications in redeposition studies. The role of double ionization is also explored.   

Progress in Coherence Imaging Techniques on DIII-D

Coherence Imaging Spectroscopy CIS systems are used on DIII-D for measurement of 2D distributions of impurity ion flow velocities, Imaging Motional Start Effect-(IMSE), and a newly developed impurity ion temperature diagnostic [1]. High-quality calibrated flow images previously required careful temperature control; rapid inter-shot system calibration with a laser and wave meter has reduced this requirement. The tunable laser has facilitated CIS characterization and calibration, and results from interferometers with three different optical delays (1000-flow, 1300, 2900-ion temperature waves) will be presented. More than one emission line in the CIS bandpass requires extra analysis; an example with two emission lines is shown. Calibrated tomographically inverted CIII flow images of the DIII-D divertor plasma will be presented. In the case of IMSE, the fidelity of the measurement depends on a switchable polarization retarder being a true half-wave. We have worked to improve the liquid crystal variable retarder and initial results will be shown. [1] C.M. Samuell, et al., these proceedings   

Progress Towards a Coherence Imaging Ion Temperature Measurement in the DIII-D Divertor

To address the role of ion temperature in setting the total pressure balance, Doppler Coherence Imaging Spectroscopy (CIS) on DIII-D is being extended to make impurity ion temperature measurements. A key challenge in divertor physics is understanding the transport of particles and energy throughout the scrape-off-layer which is largely driven by strong pressure gradients for which the ion temperature contribution is relatively unknown compared to the routinely measured electron temperature. Here we present progress in using CIS to measure 2D impurity ion temperature profiles in the DIII-D divertor in addition to the routine measurement of flow velocities. While CIS has been used extensively for flow velocity measurements on DIII-D, the technique is being extended using a mixture of small and large delay birefringent crystals in synchronized interferometers to measure impurity ion temperature. A number of diagnostic configurations are compared using a custom-built laser calibration technique for concurrent velocity and temperature calibration. Progress will be presented towards absolutely calibrated measurements of the evolution of the ion temperature as detachment progresses and the pressure profile along field lines is modified.   

The Wide Emission Spectral (WiSE) Diagnostic on DIII-D

The Wide Spectral Emission (WiSE) diagnostic is a set of 10 absolute intensity calibrated, moderate spectral and temporal resolution spectrometers co-viewing vertically through the plasma being implemented on the DIII-D fusion device for study of neutral, ions, and molecules. Working together with existing extreme ultraviolet (EUV) and vacuum ultraviolet (VUV) diagnostics, this system provides a spectral `footprint' of a tokamak plasma from 185 nm up through 5000 nm, all along a coincident line-of-sight, spanning the deep ultraviolet (DUV), ultraviolet (UV), visible (VIS), near infrared (NIR), short-wavelength infrared (SWIR) and medium wavelength infrared (MWIR) bands. Light from the plasma passes through a UV-grade sapphire viewport, then is collected with a fused silica-sapphire triplet lens and is transmitted from the machine to up to 10 separate instruments using a multi-pronged fiber bundle. Each spectrometer is capable of 0.5-4.5 kHz operation and is paired with a dedicated compact PC for operation and data acquisition. Details of design choices for the WiSE diagnostic will be presented, with implications for study of plasma parameters, impurity content, line-ratios, radiated power, and transients, along with beneficial implications for boundary code validation in DIII-D.   

Expansion of the Thomson Scattering Diagnostic on DIII-D

The Thomson scattering diagnostic on DIII-D consists of 3 laser beam lines, 10 lasers, 3 collection lenses, and 54 spatial locations. Recently the system has been expanded to include the choice of 8 spatial points in the divertor to be relocated to the upper divertor or redistributed as supplemental points to the core systems. In addition, the lower divertor system can be moved to a smaller radius using in-vessel mirrors and optics to measure high triangularity shaped plasmas. Currently, 16 additional polychromators are being built to permanently provide upper divertor and additional core measurements. In-vessel lenses and fiber optics are being installed so that 3 locations inside the SAS divertor can be measured. Low T$_{\mathrm{e}}$ (\textless 5 eV) capability is being added to all the core polychromators in order to make measurements during disruption mitigation and run-away electron experiments. Future expansion includes adding more lasers to the core horizontal and divertor systems, building a scanning 2D divertor system, adding a 20 kHz burst mode laser to study transients, and adding a wide angle lens to the core horizontal system so that midplane measurements can be extended to the plasma edge for better comparison with other diagnostics. Details of the designs and results are presented.   

Isotope Effect on H-mode Pedestal Characteristics in DIII-D Hydrogen and Deuterium Discharges

H-mode pedestal characteristics in `ITER baseline' deuterium discharges on DIII-D with D neutral beam heating were compared to H discharges with H neutral beams. The discharges had ITER cross-sectional shape and aspect ratio, q$_{\mathrm{95}}=$3.2, and $\beta_{\mathrm{N}}=$1.8. The electron density at the top of the H-mode pedestal, n$_{\mathrm{e}}^{\mathrm{PED}}$, was matched using main chamber gas puff fueling. The H cases required more gas and more heating power to match n$_{\mathrm{e}}^{\mathrm{PED}}$, and $\beta_{\mathrm{N}}$, giving ELM frequencies 4 to 8 times higher, contributing to lower carbon impurity concentration, 6n$_{\mathrm{C}}$/n$_{\mathrm{e}}=$0.1-0.25 compared to 0.6 for D. T$_{\mathrm{i}}$/T$_{\mathrm{e}}=$1for both H and D. The H discharges had outwardly shifted n$_{\mathrm{e}}$ profiles and narrower T$_{\mathrm{e}}$ pedestals giving lower T$_{\mathrm{e}}^{\mathrm{PED}}$, and T$_{\mathrm{i}}$. The total pedestal pressure varied in H discharges with the ELM frequency but was generally lower than in D. At the lowest H ELM frequency H and D pressures were comparable partially due to the lower impurity dilution. Both H and D discharges had pressure gradients consistent with peeling-ballooning stability, and had widths consistent with the EPED width scaling, although the higher ELM frequency H discharges had pedestal pressures significantly below the EPED prediction. Results from recent experiments covering a wider range of conditions in H and D will be presented.   

Analysis of Turbulence Dynamics in the H-mode edge using Phase Contrast Imaging on DIII-D

Analysis of Phase Contrast Imaging (PCI) measurements shows that the dynamics of electron density turbulence near the edge of high-performance plasmas are dominated by strong correlations between turbulent modes and spatial variation in turbulence parameters. These observations are not consistent with expectations for fully-developed turbulence (FDT), which describes a regime with weakly-correlated turbulent modes described by a few parameters (correlation lengths, mean wavenumbers in $r$, $\theta$) that change slowly in space, although previous work [Rost et al, Phys\. Plasmas \textbf{21} (2014) 062306] showed that a simple FDT model with velocity shear could explain important characteristics of the PCI measurements in QH-mode. Newer work, including an upgraded FDT model, more rigorous fitting criteria, and observations during ramps of poloidal rotation in QH-mode, provides additional constraints on turbulence location and radial structure, showing radially-limited turbulent modes in the E$_r$ well driven by strong coupling to turbulent modes in the pedestal, violating assumptions of the local FDT model. Analysis of the data must therefore go beyond simplified parameters like correlation length and gyrokinetic modeling is needed to interpret short scale length turbulent regimes.   

ECE Imaging Cross Correlation for Coherent Fluctuation Measurement

The DIII-D ECE imaging (ECE-I) system is applied to measure radial mode structures and growth rates of coherent fluctuations whereby the noise is suppressed using correlation techniques between the 160 channels of the ECE-I system. These channels are arranged in a rectangle around the low-field side mid-plane. Because the spacing between neighboring channels along the flux co-ordinate is well outside the spatial coherence length for ECE radiation, radial correlation techniques were used to suppress the thermal noise and enhance the coherent modes. The ECE-I system can give the radial location of these modes even if they are highly localized near the edge. The radial resolution of our technique is limited by the natural line width of the ECE resonance, optical thickness effects at the plasma edge, and finite receiver bandwidths. For typical modes in the optically thick region of the plasma edge, we have achieved a radial resolution of 15 mm, which is supported by synthetic ECE-I simulation. In a proof-of-principle demonstration, we have applied our technique on sawtooth pre- and post-cursors and on Alfven eigenmodes to determine the spatial location and growth rates for those modes. The same technique is also applied to mode activity that is localized near the edge.   

Coherent boundary density fluctuations in DIII-D advanced tokamak hybrid plasmas

Turbulent transport into the scrape-off layer (SOL) results in enhanced plasma-wall interactions which could impact the performance of advanced tokamak fusion plasmas. Coherent density fluctuations have been observed by Doppler backscattering systems in the scrape-off layer (SOL) region of advanced tokamak (AT) hybrid plasmas in the DIII-D tokamak. These coherent fluctuations occur between ELMs, and are associated with a step-like density profile at the separatrix, as well as high pedestal temperature, i.e. $T_e^{\rm ped}>1000$ eV. The coherent fluctuations have a mean frequency of $f_0\sim 2000$ kHz and wavenumbers of $k_\perp \sim 4-5$ rad/cm. As these SOL density modes burst, fluctuations of pedestal density and divertor ion saturation currents increase correspondingly. This suggests that these coherent SOL fluctuations play a role in the boundary transport and power deposition in high-$\beta$ AT-class plasmas, thus strongly motivating further work.   

Characterization of pedestal modes through fast vertical oscillations at DIII-D

Controlled vertical oscillations (jogs) of the plasma can be used to trigger ELMs by perturbing the edge current. In a dedicated experiment at DIII-D, oscillations were applied as tools to probe the structure and character of inter-ELM modes. Induced current at the plasma edge is shown to couple with instabilities resident in the edge transport barrier in between ELMs. These instabilities appear towards the end of the ELM cycle after clamping of the pressure gradient and are similar to fluctuations reported in a variety of experimental studies. High-resolution fluctuation diagnostics indicate that these modes are localized near the separatrix. Oscillation events are shown to modify the frequency of the inter-ELM fluctuations depending on the direction and magnitude of the induced current, an effect attributed to changes in the plasma rotation and a shift of the instability location. The radial electric field well is also influenced by jogging events, responding to changes in the current and q-profile near the plasma edge. These data provide evidence that the observed pedestal microinstabilities are affected by current density and play a role in inter-ELM profile evolution.   

Testing Ion-orbit Loss Models in MHD

The dynamics of the tokamak H-mode edge pedestal are known to depend strongly on the flow and its associated shear. The flow profile is critical to determining accessibility to operation in regimes free from edge-localized modes (ELMs), such as those with resonant magnetic-field perturbations or quiescent H-mode. While MHD simulations of these ELM-free regimes is now routine, the physics that determines the pedestal flow structure is outside the scope of the MHD model. Without a fully coupled momentum-transport model, these simulations are limited to being interpretive in nature. Long term, a transport model must incorporate the dominant transport physics in the edge: neoclassical stresses which include ion-orbit loss, high-k turbulent fluxes if needed, neutral fueling, and impurity physics. We present an analysis of DIII-D shot 164988 that incorporates the forces from ion-orbit loss from a drift-kinetic calulation [2] and ion poloidal-flow damping with impurity corrections and fluid neutrals within the NIMROD code [3]. The time-independent forces as well as the time-dependent evolution is computed and compared to the values measured by charge-exchange recombination spectroscopy.   

Impact of Divertor Closure on Nitrogen Seeded Super-H mode Plasmas in DIII-D

Integration of a high pressure pedestal and high performance core with a radiative divertor is assessed in DIII-D Super H-mode (SH) plasmas. SH is a promising regime for future devices due to the high pedestal pressures able to be obtained via increased shaping and density. The peeling limited pedestal in the SH regime allows for high densities in the scrape of layer and pedestal foot, without degradation of the pedestal height. Previous DIII-D experiments have shown significant temperature reduction at the divertor plate and near-detachment conditions with nitrogen seeded SH-modes in an open divertor. In this poster, similar plasmas are compared with a flipped shape in the upper, more closed, divertor configuration in DIII-D. Advances in feedback on nitrogen radiation in the upper divertor using bolometry allow for control over the level of radiative power, and by proxy, the heat flux to the divertor plate. The impact on global confinement and correlation with edge profiles and stability, along with divertor radiated power, temperatures, densities, and heat flux are analyzed and compared between divertor configurations. The EPED model is used to compare experimental conditions with theoretical predictions of access to the super H-mode channel.   

Effect of ECRH on pedestal structure, turbulence and ELM dynamics in DIII-D

Pedestal structure and instabilities, in different operating scenarios, is a key area of tokamak research as pedestal pressure is a factor for core plasma performance. In DIII-D, ECRH is applied (at $\rho =$0.2) in increments in NBI shots to study the effects of $T_{e}$/$T_{i\thinspace }$\textgreater 1 and density pump-out on the pedestal. For NBI, there are rapid ELMS of varied amplitude, while, for ECRH, the ELM frequency is well-defined, lower than in the NBI shots, and each type-I ELM is followed by one or two very small ELMs. Fast-magnetics show a group of modes at 13\textasciitilde 116 kHz consisting of 3 distinct modes in ECRH shots only. Phase-locked analysis shows that 2 modes at 13\textasciitilde 70 kHz do not grow for the first 10 ms of the inter-ELM period ( $\Delta t=$0\textasciitilde 10 ms, w.r.t. the ELM crash) and then grow to saturation for $\Delta t=$10\textasciitilde 25 ms. Note that $\nabla n_{e}$, $\nabla T_{e}$ and $\nabla p_{e}$ saturate after $\Delta t=$10 ms. These results may indicate that, either $\nabla p_{e}$ needs to reach a certain threshold to trigger the growth of these modes and/or the growth of the modes clamps $\nabla p_{e}$. An alternative possibility is that the changes in $\nabla E_{r}$ (and hence \textbf{E}x\textbf{B}) may affect $\nabla p_{e}$ and/or the modes. Note that changes in rotation have been observed at the pedestal for $\Delta t=$0\textasciitilde 25 ms. Analysis of pedestal recovery, transport and stability with ECRH will be reported.   

Character and Behavior of Pedestal Density Turbulence in Inter-ELM periods of Type-I ELMing DIII-D Discharges

Changes in character and temporal behavior of electron and ion scale density turbulence going from the foot to the top of the pedestal in type-I inter-ELM periods are reported for H-mode discharges of DIII-D tokamak. ITG-scale (k$_{\mathrm{\theta }}\rho_{\mathrm{s}}$\textasciitilde 0.3) \~{n} near pedestal foot increases just after ELM crash and gradually decreases as inter-ELM period progresses. Amplitude of this ITG scale \~{n} correlates well with the divertor D$\alpha $ light intensity indicating its role in edge particle transport. TEM scale (k$_{\mathrm{\theta }}\rho _{\mathrm{s}}$\textasciitilde 0.7-1.2) \~{n} measured in the steep gradient region of the pedestal shows a critical gradient behavior i.e. increases once a critical pedestal pressure gradient ($\nabla $P$_{\mathrm{e,ped}})$ is reached and its saturation correlates with that of $\nabla $P$_{\mathrm{e,ped}}$. TEM-scale (k$_{\mathrm{\theta }}\rho _{\mathrm{s}}$\textasciitilde 1-2) \~{n} measured around pedestal top has stronger amplitude compared to that in the steep gradient region, consistent with higher ExB shear in the steep gradient region suppressing more of this TEM-scale \~{n}. Adding ECH at $\rho $\textasciitilde 0.5, pedestal density gradient decreases, temperature gradient increases, and ELM frequency increases. Steep gradient localized TEM-scale \~{n} are suppressed and pedestal top TEM-scale \~{n} are unchanged after adding ECH indicating a possible role of \~{n} (in steep gradient region) in determining the inter-ELM period. Understanding these fluctuation behaviors will help gaining predictive capability of H-mode pedestal evolution.   

Plasma Response in Single Null and Double Null Plasmas and Implications for ELM Suppression

The 3D plasma response to applied non-axisymmetric magnetic fields decreases as the plasma shape transitions from single null to double null divertor configuration. Measurements from the DIII-D tokamak show the response measured on the high-field side (HFS) drops while remaining relatively constant on the low-field side (LFS) in both the external magnetics and internal SXR. The drop in the HFS response is shown over a range of $n_{e,ped}$, $\beta_{n}$, and $q_{95}$ for both $n=2$ and $n=3$ perturbations indicating a robust effect. Linearized time-independent extended MHD modeling similarly shows a reduction in HFS response in double null configurations. Conceptually, the additional null adds radial shear to externally driven field-aligned modes on the LFS and therefore may inhibit the coupling to the HFS. This may provide a reason as to why RMP ELM suppression in double-null configurations has been elusive—coupling to the HFS plasma response has been previously correlated with ELM suppression.   

Controlling magnetic footprints wetted area using resonant magnetic perturbations

The radial extension of the heat load distribution at the divertor plates due to 3D magnetic fields of a tokamak is determined by the resonant component of the non-axisymmetric field perturbations. Whether they are intrinsic, like error fields, or they are applied through 3D coils, the non-axisymmetric fields produce complex 3D edge magnetic topologies that alter the properties of the heat and particle flux distributions on the target plates. A study of the impact of applied 3D fields on the footprints wetted area is done for the DIII-D tokamak for several equilibria using the MHD code M3D-C1 coupled with the field line tracing code TRIP3D. To highlight the impact of the resonant component of the magnetic perturbation (MP) versus the non-resonant one, the poloidal spectrum of the MP is modified by varying the relative phase of the 2 rows of 3D coils used to produce n<3 perturbation. This shows that the largest footprint is achieved when the relative phase of the 2 rows is close to zero, that corresponds to the maximum resonant coupling with the plasma. A comparison of the predictions with experimental data from particle flux and infrared images is also shown. Work supported by US DOE under DE-FC02-04ER54698, DE-FG02-07ER54917 and DE-FG02-05ER54809.   

D$_{\mathrm{2}}$ pellet ELM triggering toward ITER-relevant conditions in DIII-D

While the operational availability of RMP ELM suppression and/or ELM-free scenarios are being explored, ITER will use hydrogenic pellet ELM triggering to increase ELM frequency and mitigate transient heat fluxes to material surfaces in the divertor. In order to understand the physics of this triggering and extrapolate the mechanism and heat fluxes to ITER, experiments have been performed using both pacing-sized (1.3 mm) and fuelling-sized (1.8 mm) pellets in DIII-D discharges for the first time with low collisionality pedestals ($\upsilon $* \textless 0.7) in order to push towards the more peeling-limited pedestal conditions expected in ITER. Changes to target heat and particle flux patterns have been documented, including a shift in the peak heat flux from the inner to the outer strike point when ELMs were triggered by LFS pellets compared to ELMs that occurred due to natural pedestal evolution.   

Utilizing M3D-C1 to understand triggering of ELMs in pellet pacing experiments in DIII-D ITER-like plasmas

M3D-C1, a code for solving the linear and non-linear extended-MHD equations in toroidal geometry, is currently being used to model pellet ELM triggering in DIII-D ITER-like plasmas. ELM pacing via injection of hydrogenic pellets can trigger small ELMs at a rate exceeding the natural ELM frequency and has been shown to be a successful method to mitigate effects of large ELMs. Understanding of the physical mechanisms of ELM triggering and improved modeling are required for confident extrapolation to ITER and beyond. A feature of M3D-C1 is that an unstructured triangular mesh provides sufficient resolution to capture the sharp gradients present in the pellet deposition layer. Additionally, the code provides high toroidal resolution that is important for investigating the ballooning mode physics of ELM triggering by pellets. Recent M3D-C1 modeling efforts have focused on 3D nonlinear- time-dependent simulations incorporating a pellet ablation model. Simulated pellets are injected at a speed of 150 m/s with varying pellet size and ablation location to examine the affect of ELM triggering in DIII-D ITER-like plasmas. Initial 3D results show toroidally localized perturbations in the pressure and current profiles due to the presence of the pellet.   

\textbf{Effects of Magnetic Geometry on Pellet Fueling in DIII-D and CFETR Plasmas}

Pellet injection is a promising option to provide the necessary fueling for a tokamak reactor. In this work, self-consistent simulations are carried out to explore the effects of magnetic geometry on pellet ablation and deposition in lower single-null, double-null, positive and negative triangularity configurations. The OMFIT STEP workflow is applied, which predicts pedestal with EPED, core profiles with TGYRO/ONETWO, and equilibrium with EFIT. The newly developed Pellet Ablation Module (PAM) includes a comprehensive deposition model with magnetic drift effects. Combined ONETWO and PAM simulations show that deeper fueling by high-field side (HFS) injection is boosted in all cases because the pellet ablation rate is slowed by the higher magnetic field. Initial predictions find that vertical HFS pellet injection into negative triangularity plasma could achieve deeper core fueling than that in positive ones, which may provide an attractive fueling solution.   

Absorption and Current Drive With Top Launch ECH on DIII-D

For the first time, the nearly vertical injection of ``top launch'' electron cyclotron waves is being tested on the DIII-D tokamak to verify the prediction of a high off-axis current drive efficiency. The top launch configuration takes advantage of the near-vertical geometry to combine a large Doppler shift with a long interaction zone to damp the waves on energetic electrons. Initial top launch experiments on DIII-D using a single gyrotron at 117.5 GHz and B$\approx $1.7 T (i.e., 2$^{\mathrm{nd}}$ harmonic absorption) will measure the power deposition profiles for both X-mode and O-mode by modulating the ECH power as part of a wave polarization verification. The top launch current drive efficiency also will be measured in low density, high electron temperature L-mode plasmas. The greatest current drive efficiency is expected when the wave power is damped mostly by tail electrons, which is controlled by shifting the vacuum resonance closer to or further away from the electron cyclotron waves. On DIII-D this is accomplished by varying B, and measurements of the absorbed power and current drive profiles will be shown.   

Gyrotron’s Collector Power Distribution Analysis

Gyrotrons are typically used in tokamaks to launch EC waves in the 90-170 GHz frequency range. At DIII-D, the individual gyrotrons can deliver up to 1 MW of RF power for up to 5 seconds. In order to launch these waves, a cathode generates a cylindrically symmetrical electron beam that is then accelerated by a potential difference in the order of 80-100 kV. These electrons, guided by the field from a superconducting magnet, generate an rf beam while passing through a microwave cavity, losing about $30\%$ of their kinetic energy to the generated rf, and impact on the water-cooled collector surface at ground potential. Ideally, the electron distribution over the collector wall is uniform, as the beam is swept by an externally applied magnetic field. But in practice there are regions on the collector where the current density is not uniform, turning these areas into hot spots which can be diagnosed by an array of RTDs on the outer surface of the collector. The presence of these hot spots can decrease the collector fatigue lifetime at the hottest locations, making the collector more prone to failure, and reducing the lifetime of the gyrotrons. Using COMSOL Multiphysics, a commercial finite element software, we have identified that the non-uniformity on the cathode emitted electron beam, the misalignment of the magnets, and the asymmetry of the electric field can contribute to this non-uniform power deposition.   

Overview of High Field Side Lower Hybrid Current Drive for Off-axis Current Drive Experiment in DIII-D

A key enabling technology for economical, steady state tokamak is developing efficient, robust off axis current drive compatible with the harsh reactor environment. DIII-D provides an opportunity to investigate high field side launch lower hybrid waves (HFS LHCD) for off axis current drive. The HFS launch position was selected to balance the effects of toroidicity and poloidal field up/down shift to improve wave penetration and allow single pass absorption in region $\rho$~0.6-0.8 with peak current density approaching 0.4 MA/m$^2$. Simulations indicate wave penetration will be strongly influenced by local q value. In AT discharges, the wave penetration is dominated by the poloidal upshift resulting in good wave penetration. Several technical challenges needed to be addressed to implement HFS LHCD coupler in DIII-D. The center post tile thickness needed to be increased to accommodate the coupler. The waveguide routing required a minimal number of flanges and flanges that were both vacuum tight and have good RF contacts. The latest simulations, design and system status will be presented.   

Additive Manufacturing, Laser and Electron Beam Welding of GRCop-84 for a High Field Side LHCD Launcher for DIII-D

Recent advances in selective laser melting (SLM) 3D printing technology allow additive manufacture of RF launchers from a new material, Glenn Research Copper 84 GRCop-84, a Cr$_{\mathrm{2}}$Nb (8 at. {\%} Cr and 4 at. {\%} Nb) dispersion hardened alloy, in configurations otherwise unachievable with conventional machining processes. We present the design and construction techniques of an additively manufactured high field side (HFS) LHCD launcher for installation on DIII-D. We present studies of the metallurgical and mechanical properties of SLM printed GRCop-84 and techniques for brazing, laser and electron beam welding GRCop-84 to itself, oxygen free copper, and titanium-zirconium-molybdenum (TZM) alloy. Welds are analyzed with scanning electron microscopy and focused ion beam milling to verify distribution of Cr$_{\mathrm{2}}$Nb nano-crystals within the copper matrix. Additive manufacturing allows the integration of a novel impedance matching structure into the aperture of the launcher module for a reduction in electric fields in the vacuum section of the launcher and reduction in power reflection from the plasma surface when operating near the cut-off density. Work supported by the USDOE, OFES, using User facility DIII-D, under Award Number DE-FC02-04ER54698.   

Control System for the High-Field-Side Lower Hybrid Current Drive System on DIII-D

A high-field-side LHCD system will be implemented on DIII-D for off-axis current drive. The LHCD system will have a new launcher but will re-use many major parts from the LHCD system of Alcator C-Mod, including the klystrons, carts and the HV power supply. The low-power control system needs to be improved. The main task of the control system is to feedback control the power of the 4.6 GHz klystrons and their relative phases so that the LH wave spectra is maintained for current drive operation. The C-Mod system used 2 PXI modules running MatLab Simulink for real-time control. Its interface with the MDSplus system was very cumbersome. Moreover, some major parts of the software and hardware are no longer upgradable. We plan to replace it with a Linux-server-based controller like the one used for the C-Mod ICRF fast-ferrite-tuning system. This setup will allow a much simpler interface with the MDSplus data system and have better upgradability. The GUI of the LH control will also be upgraded to work smoothly with the new control system.   

Design of a Compact Antenna for High Field Side Reflectometry on DIII-D

The coupling of lower hybrid waves to the plasma for non-inductive current drive in tokamaks has been shown to be dependent on the scrape off layer density profile in front of the lower hybrid launcher. The measurement of this density profile is critical for understanding the coupling mechanism and for code validation. Reflectometry allows for high spatial and temporal resolution measurements of the scrape off layer density profile and is planned to support the upcoming high field side lower hybrid current drive experiments on DIII-D. The reflectometry system will utilize the O-mode cutoff frequency in the range of 6-27 GHz using frequency sources from an existing, unused reflectometer. However, due to limited space on the high field side, a new compact antenna will be required for the reflectometer to fit in the available space. This work outlines the design of a compact, ultrawide bandwidth antenna for the reflectometry measurement. The details of the optimization process using COMSOL and the current antenna design will be presented.   

Predictive Control of the Rotation Profile using Neutral Beam Injection and Variable 3D Magnetic Fields

Model predictive control (MPC) of the rotation profile using a control-oriented momentum balance model which incorporates empirical models of the NBI and 3D field torques has been developed for DIII-D. Tokamak plasma rotation is widely recognized to significantly affect the energy confinement, plasma stability, and access to high performance operating scenarios. In this work, a generalized control capability for aiding rotation-related physics studies is developed. To obtain a control-oriented model, a simplified version of the momentum balance equation is combined with empirical models of the momentum sources. Recent progress in modeling the torque density profile driven by 3D fields as a function of the non-axisymmetric field coil currents has been embedded into the control design (N.C. Logan EPS 2018). MPC is well suited to a variety of control objectives because it can explicitly incorporate various types of constraints. For example, control of the edge rotation to adjust the ELM suppression threshold while fixing the stabilizing $q=2$ rotation. A simulation study is presented to demonstrate the control performance and flexibility of the approach.   

Machine learning plasma profile prediction for model-predictive control at DIII-D

At DIII-D, operators are able to control plasmas through a variety of ``actuators'' during shots: from neutral beams that heat and rotate the plasma, to coils which induce plasma current. Experimental proposals entail physicists specifying a desired plasma ``state'' of interest, which can be described by profiles of density, temperature, pressure, safety factor (q), and rotation. Operators and physicists usually work together to pre-specify a ``path'' of actuator signals through time that will successfully realize the state, using feedback control for realtime adjustments. However, the process of finding a successful actuator path is difficult and entails a lot of trial-and-error. ``Model-predictive control'' could make this process more efficient, saving physicists time and ensuring more successful physics experiments during future DIII-D campaigns. In model-predictive control, a realtime model predicts the way the plasma state will evolve given various settings on actuators, then chooses settings which yield the plasma state closest to the physicist's desired end-state. Realtime physics models are not always accurate for regimes of interest. We therefore developed a machine-learning model which generates a single prediction in under 100 microseconds, using only realtime diagnostics. Comparisons of predicted and measured profile evolution during various realtime experiments at DIII-D are made, and the algorithm used by the model to select the actuator path in realtime is discussed.   

A Neural Network Version of Multi-Mode Model for Control-oriented Fast Simulations in DIII-D.

Multi-Mode Model (MMM) is a physics-oriented model that is used to predict thermal, particle and poloidal/toroidal momentum transport in tokamak plasmas. It includes a model for ion temperature gradient, trapped electron, kinetic ballooning, peeling, collisionless and collision dominated magnetohydrodynamics modes as well as model for electron temperature gradient modes, and a model for drift resistive inertial ballooning modes [1]. While MMM is a relatively accurate model, it is too computationally intense for control design, which demands a model capable of producing similar predictions with a significantly faster run time. Neural networks offer the potential to replicate complicated calculations with a high level of accuracy while simultaneously producing results with a run time orders of magnitude faster than that of MMM. In this work, a database of predictive TRANSP runs for DIII-D was built using MMM. This database was used to train and test a neural network (MMMnet). The trained network is shown to match the results of MMM with high accuracy while producing results in a run time on the order of microseconds. This allows MMMnet to be used for model-based optimization and control applications with the potential of running in real time. [1] T. Rafiq et al., Phys. Plasmas \textbf{20}, 032506 (2013).   

Progress Towards Integrated Profile Control and NTM Suppression in DIII-D

New integrated control algorithms for the regulation of plasma profiles and scalar variables in conjunction with neoclassical tearing mode (NTM) suppression in DIII-D have been successfully tested in simulations. The Modified Rutherford Equation, which is used to estimate the NTM island width, has been integrated into the Control Oriented Transport SIMulator (COTSIM) and utilized for simulation-based testing of control algorithms based on robust feedback linearization and Lyapunov redesign techniques. These algorithms allow for controlling a combination of profiles (current, temperature, and/or pressure), scalar variables (central safety factor, $q_{0}$, edge safety factor, $q_{edge}$, stored energy, $W$, and bulk toroidal rotation, $\Omega_\phi$), and NTM island width. An actuator manager based on real-time optimization has been included to handle these competing control objectives. Previous [1] and on-going experimental efforts have the goal of further testing and ultimately validating the aforementioned integrated control architectures and tools within the DIII-D plasma control system. [1] A. Pajares et al., Integrated current profile, normalized beta and NTM control in DIII-D, Fusion Engineering and Design, 2019.