Bulletin of the American Physical Society
60th Annual Meeting of the APS Division of Plasma Physics
Volume 63, Number 11
Monday–Friday, November 5–9, 2018; Portland, Oregon
Session JO4: Analytical and Computational Techniques in Inertial Confinement Fusion |
Hide Abstracts |
Chair: Chuang Ren, University of Rochester Room: OCC B110-112 |
Tuesday, November 6, 2018 2:00PM - 2:12PM |
JO4.00001: Deep learning for physics: teaching neural networks to do ICF Brian K. Spears, Jayson Dean Lucius Peterson, John E Field, Kelli Humbird, Timo Bremer, Jayararaman Thiagarajan, Rushil Anirudh Scientists are embracing machine learning models for analysis of physics data. These models are characterized by an extreme ability to adapt to detailed structures in data. However, there are relatively few ways to force deep learning models to respect known physical relationships. We describe here our efforts to develop deep neural networks for ICF that incorporate physical constraints and rules by design. We first train these new models on simulation data to capture the theory implemented in advanced simulation codes. During the training, we enforce loss functions and constraints that force predicted output to satisfy physical principles. Later, we improve, or elevate, the trained models by incorporating experimental data. The training and elevation process both improves our predictive accuracy and provides a quantitative measure of uncertainty in such predictions. We will present work using inertial confinement fusion research and experiments at the National Ignition Facility as a testbed for development. We will describe advances in machine learning architectures and methods necessary to handle the challenges of ICF science, including rich, multimodal data (images, scalars, time series) and strong nonlinearities. |
Tuesday, November 6, 2018 2:12PM - 2:24PM |
JO4.00002: High-order Conservative, Eulerian, Multi-dimensional Hydrodynamic Simulations of Interpenetrating Plasmas Richard Berger, D. Ghosh, T. D Chapman, W. Arrighi, J W Banks, A M Dimits, J A Hittinger, I Joseph, C. Kavouklis A three-dimensional multi-species, multi-flow set of hydrodynamic equations has been developed. Plasmas moving at high relative velocity evolve on a shared Cartesian mesh nearly independently, coupled weakly by the shared electron pressure, ion-ion drag, electron-ion friction, and temperature equilibration. The electron inertia is neglected in the electron momentum evolution which eliminates the electrostatic potential and the need to solve Poisson's equation. This well-justified approximation for nonrelativistic flows increases the ease of parallelization and eliminates the electron plasma frequency time scale which allows a much larger time step. An arbitrary number of ion species with their own flows and ion temperatures are allowed and simulated in the examples discussed in this talk.
|
Tuesday, November 6, 2018 2:24PM - 2:36PM |
JO4.00003: Kinetic simulations of the thermomagnetic instability in inertial fusion conditions Mark Sherlock In laser-fusion conditions, the heat transport is well known to be reduced relative to theoretical predictions. Theoretical work over the last 4 decades has increased our understanding of the role of non-local effects in explaining the reduced heat flow, but a multiplier-free simulation capability remains elusive. A number of magnetic-field generating instabilities have long been suspected of playing a role in thermal transport in laboratory and astrophysical contexts, including the thermomagnetic instabilities [1]. We discuss the role of various transport instabilities in laser-plasmas and present the first 2D Vlaso-Fokker-Planck simulations of the thermomagnetic instability, highlighting the growth and saturation of the magnetic field, its effect on the heat flow in non-local conditions, and the formation of filamentary structures which disrupt symmetry. [1] D.A.Tidman & R.A.Shanny, Field-generating thermal instability in laser-heated plasmas, Phys. Fluids 17, 1207 (1974). |
Tuesday, November 6, 2018 2:36PM - 2:48PM |
JO4.00004: Hybrid kinetic modeling techniques for inertial confinement fusion targets Nichelle Bennett, Roger A Vesey, Dale Welch, Kyle J Peterson Experiments in direct magnetically-driven inertial confinement fusion are being conducted at Sandia National Laboratories' Z Facility using imploding liner targets. Typically target design is performed using Magnetohydrodynamic (MHD) codes. Kinetic particle-in-cell (PIC) models may evaluate post-shot performance and elucidate the dynamics of Z-pinch plasmas and neutron production. PIC models non-thermal or non-Maxwellian particle distributions, finite mean-free-paths, and charge distributions. A hybrid PIC scheme in the Chicago code allows particles to be described with quasi-neutral, multi-fluid, or fully kinetic equations of motion. These treatments are applied simultaneously to model solid-density liners and a compressed gas fuel. Preliminary results are presented for a generic imploding liner using Z-machine conditions. |
Tuesday, November 6, 2018 2:48PM - 3:00PM |
JO4.00005: Thermal transport modeling of ICF Hohlraums and laser-irradiated spheres Mehul V Patel, Kevin H Ma, Christopher W Mauche, Michael M Marinak, Gary D Kerbel, Jonathan P Brodrick, Christopher P Ridgers Radiation hydrodynamics simulations of ICF hohlraums commonly use a Spitzer-Harm thermal diffusion model that relies on an ad-hoc flux-limiter and fails to include pre-heat that would be driven by steep temperature gradients present in laser-heated targets. In this study, we examine the effects of improved electron heat transport modeling by comparing flux-limited diffusion models with an improved implementation[1] of the SNB[2] model in HYDRA. The impact of recent physics and algorithmic improvements to the nonlocal thermal transport models is quantified by performing simulations of ICF hohlraums (2D NIF Au-hohlraum post-shot models). We find that differences in modeled temperatures may be significant enough to cause variations in drive symmetry. The overall radiation drive, however, appears less sensitive to thermal conduction modeling than other physics models (e.g. NLTE kinetics in Au wall). In order to focus on the thermal transport, we also study laser illuminated spherical targets composed of lower-Z materials (e.g. Be, Al) in which non-LTE kinetics uncertainties are smaller and therefore less impactful on observables. [1] Brodrick et. al, Phys. Plasmas 24, 092309 (2017) |
Tuesday, November 6, 2018 3:00PM - 3:12PM |
JO4.00006: Fokker Planck and Krook models for nonlocal energy transport in laser produced plasmas*, Wallace Manheimer Several Krook models have appeared in the literature attempting to describe nonlocal electron energy transport, but there seems to be little consensus. We find that all models so far have had errors, which we hope to correct. One of the main ones involves the Coulomb Logarithm, which is the log of the ratio of maximum to minimum impact parameters in a collision. Using local parameters, instead of the actual nonlocal energies actually involved in the collision gives in some cases a value too small by a factor of between 3 and 5. We have derived approximate analytic solutions for nonlocal transport, which can be used to check and guide more detailed code results. We find that in general, the fuel preheat is significant. However we have also developed a Fokker Planck model and come up with approximate steady state solutions for it. We find that this latter model gives much less preheat. We explain this in terms of the differences between the two collision processes. *@ Consultant to NRL through RSI Corporation, wallymanheimer@yahoo.com # Consultant to NRL through Syntek Technologies |
Tuesday, November 6, 2018 3:12PM - 3:24PM |
JO4.00007: Kinetic Structure of Multi-Ion Collisional Plasma Shocks and Its Implications for ICF Implosions* Brett D Keenan, Andrei N. Simakov, William T. Taitano, Luis Chacon, Steven Anderson, William S Daughton Shock-driven ion-species stratification is a critically important effect for ICF-capsule implosions, resulting in the shock-front enrichment with the lighter ion species and the ICF-capsule-center enrichment with the heavier species upon the shock reflection. This imprint persists through the high-collisionality stagnation phase of a gas-filled OMEGA capsule implosion [1], influencing its yield. Yet, despite the apparent significance of shock structure, basic features of multi-ion plasma shocks remain poorly understood, inviting contradictory claims in the literature [cf., Ref. 2 and Ref. 3]. Using the new, multi-ion Vlasov-Fokker-Planck code iFP [4] developed at LANL, as well as direct comparisons to multi-ion hydrodynamic simulations and semi-analytical predictions, we disentangle the complicated structure of multi-ion plasma shocks, resolving along the way several controversies in the literature. 1. W. T. Taitano et al., Phys. Plasmas 25, 056310 (2018). |
Tuesday, November 6, 2018 3:24PM - 3:36PM |
JO4.00008: Hohlraum Simulations with Tabulated Non-LTE Data Howard Scott, Judy Harte Non-LTE atomic kinetics is necessary in ICF hohlraum simulations for adequately modeling high-Z walls and dopants. Detailed treatments of these materials are possible, but at a computational cost far beyond that amenable to use in radiation-hydrodynamic codes, which are restricted to using highly averaged atomic models. An alternative to inline non-LTE calculations is the use of tabulated non-LTE data, assuming a thermal electron distribution and steady-state kinetics. Non-LTE tables have been successfully used in situations with negligible radiation, but generalizing the tables to incorporate the entire range of anticipated radiation spectra seems prohibitive. The Linear Response Matrix (LRM) provides an economical method for tabulating and using non-LTE material data including the effects of radiation. We discuss the formulation of this approach and demonstrate its use in radiation-hydrodynamics hohlraum simulations. The good results obtained from these simulations provide confidence that this method may be used to simultaneously speed up hohlraum simulations and incorporate data from detailed non-LTE models. |
Tuesday, November 6, 2018 3:36PM - 3:48PM |
JO4.00009: Universal scaling laws for pulse propagation in plasma and non-linear media Raoul Trines, Holger Schmitz, Jorge M Vieira, Luis O Silva, Ricardo Fonseca, R.A. Cairns, Robert Bingham Pulse propagation in non-linear media is an important topic plasma physics and non-linear optics. Such pulse propagation is subject to scaling laws that go beyond the non-linear extension of the linear dispersion of the carrier wave. These laws govern the relationships ("ideal lines") between pulse amplitude, spatial width, temporal duration and propagation speed, and the evolution of all these quantities in time. They have "attractor" properties: a pulse that is not on an ideal line initially will reshape itself until it is, and then stay on that line. |
(Author Not Attending)
|
JO4.00010: Three-Dimensional Electric-Field Reconstruction at Fluid Scales: Application to Inline Modeling of Cross-Beam Energy Transfer in the Presence of Caustics Arnaud Colaitis, John Palastro, Russel Follett, Igor Igumenshchev, Valeri N Goncharov The description of laser propagation inline to large-scale hydrodynamic–radiative codes has traditionally relied on ray-optics methods, which have been limited to effects such as refraction and inverse bremsstrahlung heating. Nonlinear laser–plasma interaction models have emerged, relying on various methods for the computation of the laser-intensity distribution in plasma. However, self-consistent computation of these effects requires more advanced propagation models that can provide the local electromagnetic field. Furthermore, many laser configurations produce caustics, where conventional ray optics models break down. We present a novel method based on inverse ray tracing in the complex plane coupled to an adaptively refined propagation mesh and etalon integrals, which allows CPU-efficient reconstruction of the electromagnetic field at large fluid scales and at caustics. Notably, this decouples the hydrodynamic mesh from the propagation mesh. The model is interfaced with the ASTER 3-D hydrodynamic–radiative code. Applications to inline 3-D cross-beam energy transfer calculations are presented. |
Tuesday, November 6, 2018 4:00PM - 4:12PM |
JO4.00011: Monte-Carlo formulation of both time dependent and stationary ray-tracing methods including cross-beam energy transfer, Raman and Brillouin scatterings Arnaud Debayle, Charles Ruyer, Olivier Morice, Paul-Edouard Masson-Laborde, Pascal Loiseau, Didier Benisti A complete theoretical understanding of inertial confinement fusion experiments requires a quantitative description of all laser/plasma phenomena at stake such as Brillouin and Raman scattering, cross-beam energy transfer (CBET), and including the laser-smoothing techniques. This level of description is alas too expensive in regards of the simulation size. |
Tuesday, November 6, 2018 4:12PM - 4:24PM |
JO4.00012: Particle-in-Cell Simulations of Laser-plasma Instabilities Relevant to Shock Ignition Alexander Seaton, Tony Arber In the shock-ignition (SI) approach to direct-drive inertial confinement fusion (ICF) a high-intensity laser pulse is used to drive a converging shock through a pre-compressed direct-drive target to achieve ignition. Studies over the last decade have indicated that laser-plasma instabilities (LPIs) play a key role in determining the effectiveness of this ignitor shock. In particular, the hot electron distribution produced by LPIs is vital; electrons with energy over ~100keV may preheat the target ahead of the shock, while those of lower energy deposit their energy in the dense shell behind it and enhance its strength. |
Tuesday, November 6, 2018 4:24PM - 4:36PM |
JO4.00013: Particle-in-cell simulations of laser plasma instabilities and hot electron generation in shock ignition regime Jun Li, Shu Zhang, Farhat Beg, Eli B Borwick, Chuang Ren, Christine M Krauland, Mingsheng Wei Experiments conducted on the OMEGA EP laser facility with high-intensity, multi-kJ UV laser (1×1016 W/cm2, 1.25 kJ, 1 ns) interacting with a long scalelength keV corona plasma have shown strongly directional hot electrons with moderate temperature (~45 keV), quite favorable for electron assisted shock ignition. To understand the underlying physics, we performed 2-dimensional particle-in-cell (PIC) simulations with a long density range (0.01~0.3nc) using the OSIRIS code to study the laser plasma instabilities (LPI) and resultant hot electron generation in the experiments. The simulation results show that the hot electrons are mainly generated by two-plasmon decay (TPD) near the nc/4 surface with half angle of 30 degrees, temperature of 40 keV and ~3% of the laser energy, which agree with the experiments. In the lower density region (< 0.1nc), stimulated Brillouin scattering (SBS) reflects significant amount of laser energy and suppresses stimulated Raman scattering (SRS). Their competition is susceptible to whether a laser speckle or plane wave is used. More details of the simulation results will be presented in the meeting. |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700