Pattern Recognition for Pipeline Weld Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing transverse flux technology struggles to accurately detect narrow axial anomalies in pipeline welds, leading to incomplete fusion and hook cracks, which are often misclassified as non-reportable, resulting in potential pipeline failures.
Innovation Solution
A sophisticated pattern recognition process utilizing transverse magnetic flux leakage technology to identify and characterize longitudinal-seam anomalies, including incomplete fusion and hook cracks, by analyzing anomaly signals and adjusting for pipeline steel properties and operating conditions, with a multi-level screening method to confirm and validate potential defects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional metal-loss sizing algorithms are applied to narrow axial anomalies, then the calculation speed is fast, but the detection precision deteriorates resulting in shallower predicted depth than actual depth
Solution Approach 1:
The patent changes the fundamental parameters used for anomaly characterization. Instead of using metal-loss sizing parameters (volume, area, depth), the patent uses flux leakage signal parameters including peak amplitude, half-width, full-width at half-maximum (FWHM), and signal shape characteristics. This parameter transformation enables accurate detection of narrow axial anomalies by matching the physical characteristics of flux leakage signals rather than forcing volumetric interpretations on crack-like defects.
Solution Approach 2:
The patent replaces the mechanical/volumetric interpretation system with a magnetic field-based signal analysis system. Instead of treating anomalies as volumetric metal loss requiring depth calculation from volume measurements, the patent directly analyzes the magnetic flux leakage signal characteristics, which naturally encode the depth and geometry information of narrow axial defects without requiring complex volumetric reconstruction algorithms.
2Reliability
If transverse flux technology focuses on volumetric metal loss anomalies, then the measurement process is simple, but the detection capability deteriorates for narrow axial crack-like anomalies
Solution Approach 1:
The patent applies local quality analysis by examining the specific characteristics of flux leakage signals at different locations and orientations. The analysis focuses on the local signal shape, amplitude distribution, and spatial pattern of flux leakage, which differ between volumetric metal loss and narrow axial cracks. This localized signal characteristic analysis enables differentiation and accurate detection of narrow axial anomalies that would be missed by general volumetric assessment methods.
Solution Approach 2:
The patent transitions from analyzing anomalies in terms of volumetric dimensions (volume, area, depth) to analyzing them in terms of flux leakage signal dimensions (amplitude, width, shape, orientation). This dimensional transformation allows the detection system to capture the distinctive signature of narrow axial cracks, which have different flux leakage characteristics than volumetric defects, thereby improving reliability for detecting previously missed anomaly types.
3Productivity
If conventional algorithms report anomalies above minimum threshold, then the reporting process is straightforward, but many narrow axial anomalies remain non-reported due to low calculated depth
Solution Approach 1:
The patent enables the flux leakage signal to serve itself by using its own inherent characteristics (amplitude, width, shape) as the basis for anomaly characterization and reporting. The signal's natural properties directly provide the information needed for detection and assessment without requiring transformation into volumetric equivalents. This self-service approach allows narrow axial anomalies to be reported based on their actual flux leakage signature rather than being filtered out by volumetric depth calculations.
Solution Approach 2:
The patent introduces flux leakage signal characteristics as an intermediary between the physical anomaly and the reporting decision. Instead of directly reporting based on volumetric depth calculations, the system uses flux signal parameters (amplitude, FWHM, shape factors) as intermediate metrics that accurately represent the severity and type of narrow axial anomalies, enabling them to meet reporting thresholds based on their true characteristics rather than misleading volumetric equivalents.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach effectively detects and characterizes narrow axial crack-like anomalies, differentiating them from trim issues, and provides a detailed evaluation of potential defects, ensuring accurate reporting and preventing pipeline failures.
Implementation Method 1
The flux leakage method is primarily influenced by anomaly air gap, which is a function of anomaly length and depth, steel properties, and hoop stress
Data Source
AI summary
Embodiments of the present invention provide systems, program product and methods to detect crack-like features in pipeline welds using magnetic flux leakage data and pattern recognition. A screening process, for example, does not affect or change how survey data is recorded in survey tools; only how it is analyzed after the survey data is completed. Embodiments of the present invention can be used to screen for very narrow axial anomalies in the pipeline welds, and may also be used to predict the length of such anomalies. Embodiments of the present invention also produce a listing of the anomalies based on their relative signal characteristics.