Friction Stir Weld Surface Inspection for Real-Time Thinning Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current friction stir welding technologies rely on manual detection methods, which are inefficient and unable to adapt to high-speed welding processes, leading to limitations in production efficiency and accuracy in detecting welding defects such as thinning and flash.
Innovation Solution
A method involving laser scanning to acquire continuous surface depth images of the welding seam, segmenting the images into reference and thinned regions, and calculating the thinning amount by determining the latest reference height, enabling real-time detection of thinning, grooves, and flashes with high automation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual detection methods are used for welding quality inspection, then detection accuracy can be maintained through human judgment, but production efficiency is greatly limited and cannot adapt to high-speed welding processes
Solution Approach 1:
The patent replaces manual mechanical detection with an automated optical detection system using a line laser scanner to acquire surface depth images of the welding seam. The system automatically processes images to detect defects such as grooves, flashes, and excessive thinning, eliminating the need for manual inspection and enabling adaptation to high-speed welding processes (0.4-2 m/min).
Solution Approach 2:
The detection system performs self-service by automatically acquiring welding seam images, processing the data, identifying defects, and providing feedback without requiring manual intervention. The system autonomously completes the entire detection workflow, from image capture to quality assessment, significantly improving production efficiency.
2Speed
If laser scanning is used to acquire continuous surface depth images for real-time detection, then detection speed and automation are improved, but the complexity of the detection system increases
Solution Approach 1:
The patent segments the welding seam detection into distinct regions: front reference region, thinned region, and rear reference region. This segmentation allows the system to process images in manageable sections, comparing each region against reference standards to identify defects. The segmented approach simplifies the overall detection process while maintaining high speed and accuracy.
3Loss of time
If the welding part is detected while slightly deformed due to stressing and heating, then real-time detection during welding is achieved, but measurement precision may be affected by deformation
Solution Approach 1:
The patent changes the reference parameter dynamically by updating the reference height in real-time based on the front and rear reference regions. This adaptive reference adjustment compensates for workpiece deformation during welding, allowing accurate thinning amount measurement even when the part is slightly deformed due to stressing and heating.
4Measurement precision
If the reference height is updated in real-time for deformed parts, then detection accuracy is maintained, but the computational complexity increases
Solution Approach 1:
The patent divides the welding seam into segmented regions (front reference region, thinned region, rear reference region) and updates the reference height by comparing specific segments. This segmented approach reduces computational complexity compared to processing the entire welding seam at once, while still maintaining accurate real-time reference height updates for deformed parts.
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 allows for accurate and rapid detection of welding quality, improving production efficiency and adapting to high-speed welding processes, reducing manual intervention and enhancing the detection of complex curved surfaces and deformed parts.
Implementation Method 1
acquiring a continuous surface depth image of a welding seam through laser scanning
Data Source
AI summary
A method for detecting a surface welding quality of friction stir welding includes: acquiring a continuous surface depth image of a welding seam; intercepting a surface depth image segment of the welding seam with a proper step size, and dividing the intercepted surface depth image segments of the welding seam into a front reference region, a thinned region and a rear reference region; judging whether flatness of the front reference region is less than a threshold, and if yes, taking a height of the front reference region at this point as a latest reference height; if no, judging whether the flatness of the rear reference region is less than the threshold, and if yes, taking a height of the rear reference region at this point as the latest reference height; if no, taking the reference height of previous depth image segment as the latest reference height; and calculating a difference.