Data-Driven Analysis of Ultrasonic Inspection Data of Pressure Tubes

Panagiotis Zacharis,Graeme West,Gordon Dobie,Timothy Lardner,Anthony Gachagan

DOI: https://doi.org/10.1080/00295450.2017.1421803

IF: 1.667

2018-03-01

Nuclear Technology

Abstract:Pressure tubes are critical components of CANDU reactors and other pressurized heavy water–type reactors because they contain the nuclear fuel and the coolant. Manufacturing flaws as well as defects developed during in-service operation can lead to coolant leakage and can potentially damage the reactor. The current inspection process of these flaws is based on manually analyzing ultrasonic data received from multiple probes during planned, statutory outages. Recent advances in ultrasonic inspection tools enable the provision of high-resolution data of significantly large volumes. This highlights the need for an efficient autonomous signal analysis process. Typically, automation of ultrasonic inspection data analysis is approached by knowledge-based or supervised data-driven methods. This work proposes an unsupervised data-driven framework that requires no explicit rules or individually labeled signals. The framework follows a two-stage clustering procedure that utilizes the Density-Based Spatial Clustering of Applications with Noise density-based clustering algorithm and aims to provide decision support for the assessment of potential defects in a robust and consistent way. Nevertheless, verified defect dimensions are essential in order to assess the results and train the framework for unseen defects. Initial results of the implementation are presented and discussed, with the method showing promise as a means of assessing ultrasonic inspection data.

nuclear science & technology

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to efficiently identify and evaluate potential defects in the ultrasonic testing data of CANDU reactor pressure tubes. The current testing process relies on manual analysis of a large amount of ultrasonic data, which is not only time - consuming but also error - prone. With the progress of ultrasonic testing technology, higher - resolution data can be obtained, but this also increases the complexity of data analysis. Therefore, an automated signal analysis method is required to improve the testing efficiency and accuracy. Specifically, the paper proposes an unsupervised data - driven framework that uses the Density - Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm to group ultrasonic testing signals, thereby distinguishing between defective areas and normal areas. This method does not require pre - labeled signals or explicit rules and can adapt to the surface conditions of pressure tubes under different conditions and defects with irregular shapes. Through a two - stage clustering process, the sample size is first reduced, healthy signals are excluded, and then further clustering is carried out based on location attributes, and finally the defective areas are determined. The preliminary results of this method show potential in detecting and measuring defects, but there are also some limitations, such as being sensitive to the selection of clustering parameters. Future work will focus on developing a training process to automatically select the optimal clustering parameters and exploring methods for fusing multi - probe results to improve the accuracy and reliability of detection.

Data-Driven Analysis of Ultrasonic Inspection Data of Pressure Tubes

Study on Industrial Computed Tomography Detection and Identification of Inclusion Defects in Thick-Wall Polyethylene Pipe Butt Fusion Joint Within Nuclear Power Plant

An Ultrasonic In-line Inspection Data Processing Method Considering Invalid Data Caused by Sensor Failure

Unsupervised machine learning for flaw detection in automated ultrasonic testing of carbon fibre reinforced plastic composites

Pipeline Monitoring using Ultrasonic Sensors

Application of deep learning for structural health monitoring of a composite overwrapped pressure vessel undergoing cyclic loading

Automated Real-Time Eddy Current Array Inspection of Nuclear Assets

Quasi-Distributed Fiber Sensor-Based Approach for Pipeline Health Monitoring: Generating and Analyzing Physics-Based Simulation Datasets for Classification

Processing Ultrasonic Inspection Data from Multiple Scan Patterns for Turbine Rotor Weld Build-Up Evaluations

An adaptive PCA‐based method for more reliable ultrasonic guided waves SHM: Data‐driven modeling and experimental validation in high attenuating medium

Circumferential Ultrasonic Testing for Ultra-high Pressure Vessels and Pipes(2)

Circumferential Ultrasonic Testing for Ultra- High Pressure Vessels and Pipes(1)

3-DUSSS: 3-Dimensional Ultrasonic Self Supervised Segmentation

Ultrasonic Weld Quality Inspection Involving Strength Prediction and Defect Detection in Data-Constrained Training Environments

Enhancing Image Quality in Limited-Angle CT Reconstruction for Reactor Pressure Pipelines Detection

Detection and Classification of Uniform and Concentrated Wall-Thinning Defects Using High-Order Circumferential Guided Waves and Artificial Neural Networks

Using Deep Learning to Automate the Detection of Flaws in Nuclear Fuel Channel UT Scans

A study of machine learning object detection performance for phased array ultrasonic testing of carbon fibre reinforced plastics

Application of Object Detection Algorithms in Non-Destructive Testing of Pressure Equipment: A Review

Ultrasonic array imaging of highly attenuative materials with spatio-temporal singular value decomposition

Physics-informed deep filtering of ultrasonic guided waves for incipient defect inspection of large-scale square tube structures