Development and application of L-shell spectroscopic modeling for plasma diagnostics

L-shell spectra are a rich source of information about the plasmas from which they are emitted. This work details the development, validation, and application of flexible and reliable L-shell spectroscopic models for Fe, Kr, and Mo. Spectroscopic diagnostics for electron temperature, electron density, opacity, and the presence and characteristics of non-thermal or suprathermal (hot) electrons are presented through a broad investigation of line and continuum emission. The reliability and accuracy of the models are tested by investigating how data sources and level structure affect diagnostic predictions and by comparing modeled data to measurements from well-characterized sources, including Fe spectra from an electron beam ion trap, Kr spectra from an ion beam experiment, and Mo spectra from a laser-produced plasma. The models thus developed and tested are applied as diagnostics to L-shell spectra from a variety of plasma sources: Analysis of spectra from Kr clusters irradiated by femtosecond-scale laser pulses indicates the presence of hot electrons whose fractions are correlated with the initial cluster size; spectra from clusters irradiated with a nanosecond-scale laser are shown to be affected by opacity. Spectra from Mo X pinch plasmas of various wire diameters generated at the ∼1 MA Nevada Terawatt Facility are diagnosed to have fractions of hot electrons that correlate with the measured current; time-gated spectra from the same source indicate that the plasma conditions of different x-ray bursts from a single X pinch can vary widely. The dynamic evolution of Mo X pinch plasmas from Comell University's ∼400 kA XP pulser is investigated through an analysis of streak camera spectra with picosecond-scale time resolution. Finally, an analysis of L-shell spectra with spatial and temporal resolution indicates significant axial variations in Mo wire array plasmas generated on the ∼20 MA Z machine at Sandia National Laboratories. A genetic algorithm is described and used to investigate issues of optimization and uniqueness in diagnostic fitting of modeled spectra to measured data.