Noise removal using adaptive filtering for ultrasonic guided wave testing of pipelines

Abstract

One of the most recent technologies for NDT of pipelines is Ultrasonic Guided Waves (UGW), using low-frequency ultrasound signals to enable long-range inspection. In inspection, the practice is to excite a pure single axisymmetric wave mode in their nondispersive region. Due to mode conversion and impure testing condition, various other wave modes exist in the received waveforms. To find the defect location, one needs to be able to identify the axisymmetric wave mode from this coherent noise in the signal. Due to the variability of coherent noise, there is a need for a procedure to update adaptively with regard to the local characteristic of the received signals. Adaptive filtering is a powerful tool in removing time-varying additive noise from the coherent signals of interests in the observations and is able to achieve higher Signal-to-Noise Ratio (SNR). In this paper, a strategy is provided to use adaptive leaky filters and is validated by using laboratory trials on an experimental pipe which resulted in significant increment of SNR.