A random matrix approach to detect defects in a strongly scattering polycrystal: how the memory effect can help overcome multiple scattering

摘要

We report on ultrasonic imaging in a random heterogeneous medium. The goal isto detect flaws embedded deeply into a polycrystalline material. A 64-elementarray of piezoelectric transmitters/receivers at a central frequency of 5 MHzis used to capture the Green's matrix in a backscattering configuration.Because of multiple scattering, conventional imaging completely fails to detectthe deepest flaws. We utilize a random matrix approach, taking advantage of thedeterministic coherence of the backscattered wave-field which is characteristicof single scattering and related to the memory effect. This allows us toseparate single and multiple scattering contributions. As a consequence, weshow that flaws are detected beyond the conventional limit, as if multiplescattering had been overcome.