Ultrasonic Flaw Detection Using Deep Learning for Composite Materials
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing ultrasonic flaw-detection methods for composite materials in wind turbine blades lack objectivity and are prone to human error due to inspector variability, making it difficult to consistently detect defects like debonding, delamination, and cracks.
Innovation Solution
An ultrasonic flaw-detection system utilizing a deep learning-based approach that includes signal preprocessing, defect candidate group selection using VAE or ResNet, and defect determination using YOLO or Faster R-CNN algorithms to generate B-Scan and C-Scan images for accurate defect identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual ultrasonic flaw detection is used, then detection flexibility is maintained, but detection objectivity and consistency deteriorate due to inspector variability
Solution Approach 1:
The patent creates a digital copy of the inspection process through automated image processing and deep learning algorithms. The system captures ultrasonic images, processes them through predefined algorithms, and uses trained neural networks to detect defects, replacing the variable human inspection process with a consistent automated system that produces objective results.
Solution Approach 2:
The patent replaces the mechanical human inspection process with an automated computational system. The deep learning model (such as ResNet or YOLO) processes ultrasonic images through neural networks, substituting the inspector's subjective judgment with algorithmic analysis that maintains consistent detection criteria and eliminates human variability.
2Reliability
If deep learning algorithms are applied for automatic defect detection, then detection objectivity and continuity are improved, but system complexity increases
Solution Approach 1:
The patent divides the defect detection system into distinct functional modules: image acquisition module, image processing module, deep learning detection module, and result output module. Each module performs a specific function, making the complex system manageable and easier to implement. The segmentation allows for independent optimization of each component while maintaining overall system functionality.
Solution Approach 2:
The patent employs preliminary actions through image preprocessing steps (noise reduction, contrast enhancement, segmentation) before the deep learning analysis. These preparatory processing steps simplify the input data for the neural network, reducing the computational complexity required for defect detection and improving algorithm efficiency.
3Measurement precision
If comprehensive defect detection is performed, then detection accuracy is improved, but inspection time increases
Solution Approach 1:
The patent applies partial action by focusing the deep learning analysis on specific regions of interest identified through preliminary image processing. Instead of analyzing every pixel of the entire ultrasonic image, the system first identifies potential defect areas through edge detection and contrast enhancement, then applies the computationally intensive deep learning algorithms only to these regions, reducing overall processing time while maintaining high detection accuracy.
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
The system provides objective and reliable defect detection by minimizing human error, enhancing the continuity and accuracy of flaw detection results in composite materials.
Implementation Method 1
an ultrasonic wave transducer configured to transmit an ultrasonic wave to a detection target and collect an ultrasonic echo signal reflected from the detection target
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The embodiments of the present disclosure relate to an ultrasonic flaw-detection system and an ultrasonic flaw-detection method. The ultrasonic flaw-detection system may include: an ultrasonic flaw-detection device configured to transmit an ultrasonic wave to a detection target, collect an ultrasonic echo wave reflected from the detection target, and then generate a signal data; a signal data preprocessor configured to preprocesses the signal data; a defect candidate group selection unit configured to select a defect candidate group based on the preprocessed signal data and generate defect candidate signal data based on the selection; an image data generator configured to generate image data based on the defect candidate signal data included in the defect candidate group; and a defect determination unit configured to determine whether there is a defect in the defect candidate group based on the image data.