High resolution tomography of pipeline using multi-helical Lamb wave based on compressed sensing

Corrosion and erosion threaten the material integrity of pipeline structure. The ultrasonic guided wave can facilitate the detection and visualizing of defect. However, the high resolution tomography of pipeline considering the wave diffraction is not established well. This work proposes the compressed sensing scheme for limited projections in cross-hole tomographic reconstruction of pipeline. The Fourier diffraction theorem is constructed to describe the arrangement of double-ring array of transducers around pipeline. Based on the idea of compressed sensing, the image reconstruction is converted into an optimization problem. The continuous iterations are applied to obtain high resolution images from limited projections. In order to verify the effectiveness of the proposed tomographic model, variable-depth defect is built in finite element model and the obtained ultrasonic projections are used for tomography. The estimation error of the thinnest wall thickness and the reconstruction error of the image are defined for quantitative evaluation of the imaging results. In addition, noise is introduced to observe the image quality under the conditions of 20 and 10 dB. The actual complex defect is also studied to carry out experimental verification. The results show that the pipeline thickness distribution achieves high accuracy. The corresponding comparison demonstrates the advantage of using multi-helical Lamb waves for imaging.