Guided wave tomography based on least-squares reverse-time migration

Abstract

A key to successful damage diagnostics and quantification is damage imaging through ultrasonic guided wave tomography. We propose the implementation of least-squares reverse-time migration in a circular array for damage imaging in an aluminum plate. The theory of least-squares reverse-time migration is formulated for guided wave applications along with the summary of an efficient optimization algorithm: the conjugate gradient method. Numerical simulation and laboratory experiments are used to evaluate its performance with a circular array setup. In order to improve the data processing efficiency, the concept of using a limited number of actuators but a relatively large number of sensors is tested. Studies are conducted on three numerical cases, including a rectangular-shaped damage site, a complex-shaped damage site, and six other damage sites varying in size. As an inversion-based method, least-squares reverse-time migration shows significantly improved shape reconstruction with the amplitude quantification capability, compared to conventional reverse-time migration. Our experimental data are generated by piezoelectric wafers as actuators, measured by a scanning laser Doppler vibrometer to form a circular array on an aluminum plate, with a rectangular notch located in the inner region of the array. The damage images using experimental data show consistency in both the simulations using Born scattering and in altered material properties in the damaged region. According to the comparison, least-squares reverse-time migration for guided wave tomography is a promising technology to provide high-resolution large area damage imaging for plate-like structures.