Industrial processes involving thermal plasma such as cutting, welding, laser machining with ultra-short laser pulses (nonequilibrium conditions), high temperature melting using electrical discharge or ion-beams, etc., generate non-repeatable fast transient events which can reveal valuable information about the processes. In such industrial environments containing high temperature and radiation, it is often difficult to install conventional lens-based imaging windows and components to observe such events. In this study, we compare imaging requirements and performances with invasive and non-invasive modes when a fast transient event is occluded by a metal window consisting of numerous holes punched through it. Simulation studies were carried out for metal windows with different types of patterns, reconstructed for both invasive and non-invasive modes and compared. Sparks were generated by rapid electrical discharge behind a metal window consisting of thousands of punched through-holes and the time sequence was recorded using a high-speed camera. The time sequence was reconstructed with and without the spatio-spectral point spread functions and compared. Commented MATLAB codes are provided for both invasive and non-invasive modes of reconstruction.
Invasive and Non-Invasive Observation of Occluded Fast Transient Events: Computational Tools
