Guided-wave tomographic imaging of defects in pipe using a probabilistic reconstruction algorithm

There has been much interest recently in monitoring the structural integrity of predetermined critical zones in pipeline using leave-in-place sensors. Ultrasonic guided waves have significant advantages over vibration and impedance methods for structural health monitoring. With guided waves it is possible to locate and size defect regions. The work presented here introduces a Reconstruction Algorithm for the Probabilistic Inspection of Damage (RAPID), which is used to construct tomographic images of multiple defects in a 16-diameter schedule 30 steel pipe. Imaging is completed using an array of 16 low-cost transducers. It is shown that defect location and severity are accurately predicted by employing the RAPID algorithm with multi-frequency data sets. A sparse array study is completed in which images are reconstructed using only eight of the 16 available transducers. It is found that damage can still be detected using the sparse array, though defect sizing and location accuracy suffer. It is shown that multiple defects can still be resolved using 13 of the original 16 transducers chosen at random.