Iterative processing of ultrasonic measurements to characterize flaws in critical optical components

Real-time nondestructive evaluation is crucial for the safety and maintenance of critical optics in high energy, laser physics experiments. Fluence levels in short pulse, high-energy lasers can produce pits and cracks in the surfaces of the laser's optical components. These flaws in the optical glass can adversely affect the production of the laser light, or even result in a catastrophic failure of the optical component itself. Consequently, the detection, localization, and characterization of these flaws is critical. This paper describes the novel application of several signal and image-processing techniques that detect, localize, and characterize flaws in optical components. These techniques are embedded into an optic scanning system to automatically identify and report on the condition of the vacuum windows used in high fluence laser systems. These techniques exploit measurements made from two orthogonal acoustic arrays mounted on adjacent edges of the optic. After preprocessing the raw channel measurement data from two orthogonal, narrow beamwidth, transducer arrays, a two-dimensional (2-D) power image is created. A physics-based 2-D matched filter is then developed for detecting and localization. An iterative solution to sequentially search the resulting image to extract and characterize the flaws is discussed.