Statistical modeling of the fusion yield and areal density in DT-layered implosions on OMEGA

In recent years, the application of statistical modeling has enabled development of accurate predictive models of fusion yield in direct-drive inertial confinement fusion experiments on OMEGA. Through the choice of separable basis functions and a careful selection of parameters guided by physical considerations, the statistical model has been used to quantify the effect of all the major sources of fusion yield degradation on OMEGA. The strongest dependencies include the age of the deuterium tritium fuel from filling to shot time, the asymmetry of inferred ion temperatures as a proxy of the L=1 mode, the ratio of the laser beam radius to target radius as a proxy of the illumination nonuniformity of overlapping beams, and parameters related to the hydrodynamic stability such as the shell adiabat, in-flight aspect ratio and convergence ratio. Here, we present statistical models of the areal density, ion temperature and fusion yield trained on a large database of OMEGA implosions. It is shown that a reasonably accurate predictive capability can be developed enabling the design of DT-layered implosions with incrementally higher performance.