Please use this identifier to cite or link to this item:
Title: Improved defect detection using adaptive leaky NLMS filter in guided-wave testing of pipelines
Authors: Nandi, AK
Nakhli Mahal, H
Yang, K
Keywords: adaptive filtering;leaky normalized mean square;ultrasonic guided waves;pipeline inspection;SNR enhancement;signal processing
Issue Date: 15-Jan-2019
Publisher: MDPI
Citation: Applied Sciences
Abstract: Ultrasonic guided wave (UGW) testing of pipelines allows long range assessments of pipe integrity from a single point of inspection. This technology uses a number of arrays of transducers, linearly placed apart from each other to generate a single axisymmetric wave mode. The general propagation routine of the device results in a single time domain signal, which is then used by the inspectors to detect the axisymmetric wave for any defect location. Nonetheless, due to inherited characteristics of the UGW and non-ideal testing conditions, non-axisymmetric (flexural) waves will be transmitted and received in the tests. This adds to the complexity of results’ interpretation. In this paper, we implement an adaptive leaky normalized least mean square (NLMS) filter for reducing the effect of non-axisymmetric waves and enhancement of axisymmetric waves. In this approach, no modification in the device hardware is required. This method is validated using the synthesized signal generated by a finite element model (FEM) and real test data gathered from laboratory trials. In laboratory trials, six different sizes of defects with cross-sectional area (CSA) material loss of 8% to 3% (steps of 1%) were tested. To find the optimum frequency, several excitation frequencies in the region of 30–50 kHz (steps of 2 kHz) were used. Furthermore, two sets of parameters were used for the adaptive filter wherein the first set of tests the optimum parameters were set to the FEM test case and, in the second set of tests, the data from the pipe with 4% CSA defect was used. The results demonstrated the capability of this algorithm for enhancing a defect’s signal-to-noise ratio (SNR).
ISSN: 2076-3417
Appears in Collections:Dept of Electronic and Computer Engineering Research Papers

Files in This Item:
File Description SizeFormat 
Fulltext.pdf9.54 MBAdobe PDFView/Open

Items in BURA are protected by copyright, with all rights reserved, unless otherwise indicated.