Fusion Science and Technology / Volume 51 / Number 4 / May 2007 / Pages 514-518
The first ignition experiments on the Laser Mégajoule facility will use an indirect drive scheme. Our A1040 point design target is a graded doped plastic capsule filled by permeation within a gold cylinder. The deuterium-tritium ice layer may be formed either by classical slow cooling at 1.5 K below triple point, or by rapid cooling at 2.3 K below triple point. To complete the specifications, we first studied the robustness to all technological defects with the current CEA capabilities for these two options of ice formation. The technological imperfections taken into account are regrouped into 1D errors, which keep the implosion spherical, and 3D errors, which induce a deformation of the shell. The 3D robustness is expressed in terms of deformation at peak velocity and compared to the deformation threshold obtained with 2D simulations. The 1D robustness is given by the probability of exceeding 50% of nominal yield. We have taken into account 22 1D parameters and the fusion energy is approximated by a neural network based on 2000 simulations. Although the studies are not finished yet, the first results show that the A1040 design with rapid cooling has sufficient margins with respect to technological defects.