Coded Aperture Tests for Fusion Source Imaging

For Coded Aperture Imaging (CAI) of moderately extended sources, it is shown that mask patterns based on Singer sets offer superior Signal to Noise Ratio (SNR) by comparison with those based on, more commonly used, Hadamard sets. This is because Singer sets inherently have a wide range of available open fractions, allowing the design of the CAI mask to be tailored to the spatial extent of the source being imaged. The basis of CAI is reviewed and relevant SNR formulae are derived. The performance of a Singer set mask pattern with 400 pixels and 57 open square holes (representing ρ=1/7=14.3% open fraction) is studied by both numerical simulation and experiment. Monte Carlo simulation studies confirm the validity of our decoding algorithm and the derived SNR formulae. The experimental setup comprised: a 226Ra source of alpha particles, a laser-machined CAI mask, and CR-39 nuclear track detectors, arranged inside a vacuum enclosure. After chemical etching of the detectors, the area, circularity, mean optical density and positions of all candidate tracks were measured by an automated scanning system. Appropriate criteria were used to select alpha particle tracks, and a correlation algorithm applied to the (x,y) data produced the de-coded image of the source. The results demonstrate that the Singer set CAI masks can achieve the resolution and SNR requirements for our intended application: imaging of the DD fusion source in pulsed plasma focus devices. (author)