Solving Inverse Problems of Ultrasound Tomography in a Nondestructive Testing on a Supercomputer

This paper is concerned with the use of a supercomputer to solve tomographic inverse problems of ultrasonic nondestructive testing in the framework of a scalar wave model. The problem of recovering the velocity of a longitudinal wave in a solid is formulated as a coefficient inverse problem, which in this formulation is nonlinear. First, the algorithms were tested on real data obtained in experiments on a test bench for ultrasound tomography examinations. Ultrasound in the 2-8 MHz band was used for sounding. The experiment employed a rotating transducer system. A rotating transducer system substantially increases the number of emitters and detectors in a tomographic scheme and makes it possible to neutralize the image artifacts. An important result of this study is an experimental confirmation of the adequacy of the underlying mathematical model. The proposed scalable numerical algorithms can be efficiently parallelized on CPU-supercomputers. The computations were performed on 384 computing CPU cores of the "Lomonosov-2" supercomputer at Lomonosov Moscow State University.