Uncertainty Quantification for the Sensitivity Analysis and Prediction Error of 1D Radiation-Hydrodynamics Simulations of Cylindrical Implosions
When
Cylindrical targets are used to study instability growth in regimes relevant to inertial confinement fusion implosions, as they retain the effects of convergence while allowing for direct diagnostic access. Design parameters for experiments on the National Ignition Facility are informed by one-dimensional (1D) and two-dimensional (2D) radiation-hydrodynamics simulations, and these cylindrical implosions are most sensitive to parameters such as target radius, fill composition, and laser pulse shape. Even in 1D, the parameter space responsible for dictating the behavior of various implosion performance metrics is vast, and it can be difficult to ascertain which model parameters, or combinations thereof, have the most influence on these metrics. Additionally, uncertainty introduced via the addition of ad hoc parameters which are used as a proxy for complex underlying physics can result in mismatches between model results and experimental data. This talk applies model calibration techniques based on fidelity mapping to 1D simulations in order to examine the sensitivity of our model to such parameters, as well as identify regions of space where the error between model predictions and experimental results is relatively low. William Gammel, Joshua P Sauppe, Samy Missoum, Bharath Pidaparthi
Place: Math, 402 and Zoom: https://arizona.zoom.us/j/82075792519 Password: 150721