Bayesian Data Assimilation of Argon Gas Puff X-ray Source Data Collected on Sandia National Laboratories' Z-machine

Argon gas puff implosions regularly produce > 300 kJ of ~ 3 keV photons on Sandia National Laboratories' Z-machine, making them one of the brightest laboratory x-ray sources in this energy range. Recent measurements have shown that only ~ 16 of the ~ 28 MA available on Z couple to the Ar gas puff. To better understand this high current loss phenomenon, Z gas puff plasmas are observed using several spectrometers, calorimeters, and photoconducting diamonds. Each of these diagnostics is analyzed separately and then combined to arrive at a final spectrum and x-ray yield, an important source performance metric. Unfortunately, this data analysis approach usually results in higher-than-desired x-ray yield uncertainties (~ 30 %), which provides insufficient precision to resolve the high current loss problem. In this poster, we present a high-precision x-ray yield analysis using a Bayesian data assimilation approach for Ar gas puffs on Z. Advantages of this new approach include simultaneous and self-consistent analysis of all experimental data, more rigorous treatment of experimental uncertainties, an automated approach to excluding faulty diagnostics, and inherent diagnostic value-of-information testing capabilities. We present results using the Bayesian data assimilation approach to reproduce emitted spectra with an analytic model, as well as 1D, and 3D simulations. We also show initial application of our technique to Ar gas puff experimental data. Finally, we discuss the potential of the Bayesian assimilation to reduce experimental x-ray yield uncertainties to < 10 % , which should enable us to progress in our understanding of poor current coupling in Ar gas puffs on Z.

This work was supported by the Laboratory Directed Research and Development program at SNL. SNL is managed and operated by NTESS under DOE NNSA contract DE-NS0003525.