[4A1] Selective matrix capture with non-linear array density for laser-induced phased array

Peter Lukacs, Theodosia Stratoudaki and Anthony Gachagan
Department of Electronic and Electrical Engineering, University of Strathclyde, UK 

Laser ultrasound is a non-contact, couplant free technique with a small footprint, thus has the capabilities to overcome some of the challenges conventional transducers current face in the industry. These include inspection at elevated temperature, places of restricted access or on components with complex geometries. Laser Induced Phased Arrays (LIPA) synthesise the ultrasonic phased array in post processing, in order to overcome the low Signal-to-Noise Ratio (SNR) and hardware constraints, common in conventional laser ultrasonics.[1] Using the Full Matrix Capture for data acquisition and the Total Focusing Method for post processing, LIPA can generate images with good SNR and resolution. However, capturing the full matrix for LIPA requires very long scanning times which compromises potential uptake of the technique from the industry.

Conventional, transducer based phased arrays are limited to their predefined design with a fixed element pitch and aperture size. In contrast, LIPA are reconfigurable and variable element spacing can easily be achieved [2]. In this work we present a data capturing technique that places data acquisition points with a varying element spacing to optimised locations based on laser ultrasound propagation theory. By doing so, the number of signals to capture can be significantly reduced.

Using Selective Matrix Capture (SMC) [3] it was possible to reduce element number by utilising the characteristics of laser generated and detected ultrasound, without compromising defect detectability. However, our previous results showed reduced resolution of the image due to the fixed aperture requirement. In this study, results are presented in simulation, for a homogenous solid medium. LIPA are optimised by applying a varying element spacing and aperture size and improved resolution is shown without compromising the SNR.