Pipe Surface Flaw Detection Using Wavelet Noise Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional nondestructive testing methods for surface flaws in pipes, such as magnetic leakage flux tests, face challenges in distinguishing flaw signals from noise, especially under unfavorable conditions of pipe wall thickness and flaw depth, leading to ambiguous results and the risk of filtering out relevant signals.
Innovation Solution
The implementation of wavelet transformation and filtering techniques to reduce noise levels in near-real-time data processing of leakage flux signals, allowing for unambiguous identification of flaw-based signals by converting analog signals to digital, buffering, and applying wavelet transformations to filter and modify wavelet coefficients for comparison with reference values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If conventional analog filtering techniques are used to reduce noise, then noise reduction is achieved quickly and simply, but the flaw-based signals are filtered out along with the noise because they appear in a similar frequency range
Solution Approach 1:
The patent transforms the filtering approach from conventional frequency-domain analog filtering to wavelet transformation, which operates in the time-frequency domain. This parameter change allows the system to distinguish between noise and flaw signals by their temporal characteristics rather than just frequency, preventing the loss of valid flaw signals while still achieving effective noise reduction.
Solution Approach 2:
The patent replaces the mechanical analog filtering system with a digital wavelet transformation system. This substitution enables more sophisticated signal processing that can adaptively separate noise from flaw signals based on their different temporal patterns, resolving the contradiction between noise reduction and signal preservation.
2Loss of energy
If difference techniques are used to suppress slowly varying background components, then background noise is reduced, but signals of interest are filtered out along with the background
Solution Approach 1:
The patent changes the processing domain from simple time-domain differencing to wavelet transformation, which provides multi-resolution analysis. This allows the system to suppress slowly varying background components at coarse scales while preserving flaw signals at finer scales, preventing information loss.
3Measurement precision
If wavelet transformation is applied to reduce noise levels, then noise reduction effectiveness is significantly improved, but processing time increases making near-real-time analysis difficult
Solution Approach 1:
The patent segments the continuous signal processing into discrete wavelet decomposition levels, allowing selective processing at different resolution scales. This segmentation enables the system to achieve effective noise reduction at critical scales while skipping less critical computations, thereby reducing overall processing time for near-real-time analysis.
Solution Approach 2:
The patent applies wavelet transformation selectively at the most critical decomposition levels rather than processing all possible scales. This partial action approach maintains the essential noise reduction benefits while significantly reducing computation time to enable near-real-time flaw detection.
4Device complexity
If conventional filtering is used, then the processing is simple and fast, but the separation between flaw-based signal and noise level is too small to arrive at meaningful results
Solution Approach 1:
The patent replaces simple analog filtering with digital wavelet transformation, which provides superior signal-to-noise separation through multi-resolution analysis. Although this increases processing complexity, the dramatic improvement in measurement precision justifies the added complexity for critical flaw detection applications.
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 enables reliable and cost-effective near-real-time measurement and processing of surface flaw data in pipes, allowing for immediate intervention in the production process by effectively separating noise from flaw signals, thereby improving the accuracy and efficiency of nondestructive testing.
Implementation Method 1
transforming the copied data with a wavelet transformation and filtering or modifying, or both, the resulting wavelet coefficients
Implementation Method 2
transmitting the signals to a pre-amplifier
Data Source
AI summary
A method for nondestructive testing of the pipes for detecting surface flaws using magnetic leakage flux is disclosed. With of the method, flaws can be detected and analyzed in near-real-time while the pipe is produced. The data obtained with inductive coils, Hall sensors or GMR sensors are digitized, the digital data are buffered in a first memory, and a subset of the digital data are copied into a second memory. The copied data are transformed with a wavelet transformation and the resulting wavelet coefficients are filtered and/or modified. In an alternative embodiment, the digital data can be continuously supplied to a routine for wavelet transformation, which is performed using cascaded digital signal processing routines. The evaluated variable is compared with a reference value, wherein a determined flaw-based signal can be unambiguously associated with the position of the flaw.