Detailed high-resolution three-dimensional simulations of OMEGA separated reactants inertial confinement fusion experiments

Abstract

We present results from the comparison of high-resolution three-dimensional (3D) simulations with data from the implosions of inertial confinement fusion capsules with separated reactants performed on the OMEGA laser facility. Each capsule, referred to as a “CD Mixcap,” is filled with tritium and has a polystyrene (CH) shell with a deuterated polystyrene (CD) layer whose burial depth is varied. In these implosions, fusion reactions between deuterium and tritium ions can occur only in the presence of atomic mix between the gas fill and shell material. The simulations feature accurate models for all known experimental asymmetries and do not employ any adjustable parameters to improve agreement with experimental data. Simulations are performed with the RAGE radiation-hydrodynamics code using an Implicit Large Eddy Simulation (ILES) strategy for the hydrodynamics. We obtain good agreement with the experimental data, including the DT/TT neutron yield ratios used to diagnose mix, for all burial depths of the deuterated shell layer. Additionally, simulations demonstrate good agreement with converged simulations employing explicit models for plasma diffusion and viscosity, suggesting that the implicit sub-grid model used in ILES is sufficient to model these processes in these experiments. In our simulations, mixing is driven by short-wavelength asymmetries and longer-wavelength features are responsible for developing flows that transport mixed material towards the center of the hot spot. Mix material transported by this process is responsible for most of the mix (DT) yield even for the capsule with a CD layer adjacent to the tritium fuel. Consistent with our previous results, mix does not play a significant role in TT neutron yield degradation; instead, this is dominated by the displacement of fuel from the center of the implosion due to the development of turbulent instabilities seeded by long-wavelength asymmetries. Through these processes, the long-wavelength asymmetries degrade TT yield more than the DT yield and thus bring DT/TT neutron yield ratios into agreement with experiment. Finally, we present a detailed comparison of the flows in 2D and 3D simulations.

Impact of imposed mode 2 laser drive asymmetry on inertial confinement fusion implosions

“…Under these conditions, on-target illumination nonuniformity better than 2% rms 33 is expected without any external perturbations applied. However, previous work has shown that a number of external asymmetry seeds, including unintended drive asymmetry, 11 unintended capsule misalignment (offset), 9,34 and the capsule stalk mount, [35][36][37] can be expected to modify this picture. The anticipated effect of these factors is illustrated in Fig.…”

“…Several different simulation tools were brought to bear on these experiments to interpret the results. These include 1D LILAC, 51 1D and 2D xRAGE, 36,52,53 3D ASTER, 9,10 and 3D Chimera, 54 with the latter initialized with 1D HYADES. 55 Results from each are briefly discussed below.…”

“…Unlike Chimera and ASTER, the Eulerian radiation hydrodynamics code xRAGE 52,53 does have the capability to simulate the impact of the capsule stalk mount. 36…”

“…These simulations were performed with a maximum grid resolution of 0.25 lm and in a similar fashion to those detailed in Ref. 36. Two types of 2D xRAGE simulations were run: (i) with the glue spot but excluding the imposed drive asymmetry and (ii) with the as-shot drive but excluding the glue spot (the 2D geometry prevents simultaneous simulations of both asymmetry seeds).…”

“…62 With the interplay hypothesis in mind, it is easy to imagine how interference of flows from these various sources could give rise to large flows but small observable T ion asymmetries. OMEGA implosions, on the other hand, are expected to be impacted by a mode 1 asymmetry in the laser drive, 11 mode 1 due to the stalk, 35,36 and mode 1 due to capsule offsets. 9 When the various asymmetry seeds align, we would expect large T ion variations, while when they do not, smaller variations would be expected.…”