MODELING LASER–PLASMA INTERACTION PHYSICS UNDER DIRECT-DRIVE INERTIAL CONFINEMENT FUSION CONDITIONS Introduction

Introduction Laser–plasma interaction (LPI) processes taking place in indirectand direct-drive targets differ significantly in several ways. Plasma electron densities ne in hohlraum targets are typically a few percent of the critical density n m e c e = ω π 0 2 2 4 , so that the main instability mechanisms are stimulated Raman and stimulated Brillouin scattering (SRS and SBS, respectively), which typically have very large predicted linear gains due to the long scale lengths of near-uniform plasma.1 The theoretical challenge here is to understand the nonlinear saturation mechanisms that are responsible for the small, observed reflectivities. In direct-drive targets the plasma is inhomogeneous, with the linear gain for parametric instabilities often limited by the inhomogeneity of the plasma, rather than by damping of the unstable waves. In direct-drive targets, all electron densities up to critical (nc ~ 8 × 1021 cm−3 for 0.351-μm light) can be accessed by the laser. As a consequence of the dispersion relations of the participating waves, SBS can take place anywhere in the underdense region ne < nc, and SRS can take place anywhere below the quarter-critical surface ne ≤ nc/4. At the quarter-critical surface SRS is in competition with two-plasmon decay (TPD), a particularly dangerous instability because of its low threshold and its ability to produce hot electrons that preheat the target. Complicated physics is expected at the critical surface itself, including but not limited to resonance absorption, profile modification, instability, and surface rippling.2 Interactions in the underdense plasma corona are further complicated by the crossing of multiple beams. These beams can interact parametrically via common decay waves, excited simultaneously by several beams, or via electromagnetic seeding involving specular or parametric reflections at or near the critical-density surface. The need to take into account such complications means that simple theoretical models are of rather limited use. One must adopt multidimensional simulation tools that are able to model the necessary physical processes on a large scale in order to have a hope of interpreting current experimental data and making predictions for future experiments. While modeling LPI in indirect-drive-relevant plasmas has received a great deal of attention, and several semipredictive simulation codes have been developed,3,4–6 the same cannot be said for direct drive. Recently pF3D, a three-dimensional, parallel LPI interaction code developed by LLNL, has been modified at LLE for use in direct-drive conditions. The significant advantage of pF3D3 over the code HARMONHY4 is its efficient parallelization using message passing, which has been exploited using Hydra, a 64-processor SGI Origin. This article describes recent developments in this regard, and in addition to some background on pF3D and similar codes, shows some of the first results that have modeled long-scalelength OMEGA multibeam experiments.