Weld Quality Monitoring Using Integrated Sensors and Defect Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing welding quality monitoring systems are inadequate for industrial use due to noise interference and inability to detect volumetric defects, and they suffer from uncertainty due to the complexity of the welding process and variability in work programs.
Innovation Solution
A system and method that uses sensor elements on welding machinery to detect features like voltage, current, and oscillation frequency, and employs machine learning algorithms, specifically random forests, to classify weld defects in real-time, overcoming noise interference and detecting both surface and volumetric defects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If machine learning algorithms use sound signals from external microphones to classify weld defects, then weld quality monitoring is enabled, but the monitoring accuracy deteriorates due to noise interference from the non-protected environment
Solution Approach 1:
The patent introduces an intermediary mechanism (acoustic emission sensor integrated into the welding torch) that mediates between the welding process and the monitoring system. This intermediary captures acoustic signals directly at the source of welding emissions, filtering out environmental noise before the signal enters the monitoring system, thus resolving the contradiction between enabling monitoring and maintaining measurement precision.
Solution Approach 2:
The patent replaces the external acoustic monitoring system with an integrated sensor system that combines acoustic emission sensing with the welding torch mechanism. This substitution eliminates the need for external microphones that are vulnerable to environmental noise, thereby maintaining measurement precision while enabling reliable weld quality monitoring.
2Reliability
If traditional inspection methods are used to detect weld defects, then defect detection is possible, but the detection timing deteriorates to downstream only after welding completion
Solution Approach 1:
The patent implements preliminary action by performing real-time acoustic emission monitoring during the welding process itself, rather than waiting for post-weld inspection. The system continuously analyzes acoustic signals generated during welding to detect defects as they form, enabling immediate identification and correction before the welding process completes, thus eliminating the time loss associated with downstream inspection.
Solution Approach 2:
The patent establishes a feedback loop where acoustic emission data collected during welding is immediately analyzed and used to provide real-time information about weld quality. This feedback mechanism allows the system to detect defects during the welding process and potentially adjust parameters or alert operators, rather than waiting for post-weld inspection, thereby resolving the timing contradiction.
3Measurement precision
If multiple features are monitored during welding operations to improve classification accuracy, then weld defect identification precision is improved, but the system complexity increases
Solution Approach 1:
The patent applies universality by designing a multi-functional monitoring system where a single integrated platform collects and processes multiple types of data (acoustic emissions, electrical parameters, mechanical measurements) simultaneously. This universal system eliminates the need for separate dedicated systems for each measurement type, thereby improving classification precision through multi-feature analysis while avoiding the complexity increase that would result from multiple independent systems.
Solution Approach 2:
The patent merges multiple monitoring functions into a single integrated system that combines acoustic emission sensing, electrical parameter measurement, and mechanical data collection into one unified platform. By combining these functions rather than using separate systems, the patent achieves improved weld defect classification precision through comprehensive multi-feature analysis while maintaining manageable system complexity through integration.
Data Source
AI summary
A method and system for monitoring and identifying the quality of a welding performed by a welding machinery on a metallic component, which includes establishing a plurality of classes that describe possible weld defects on the metallic component; identifying a plurality of features that are relative to the functioning of the welding machinery; acquiring the values assumed by the plurality of features during the execution of a current welding on a metallic component; and processing the values assumed by the plurality of features during the execution of the current welding to establish the probabilities of the future outcome of the welding of belonging to each of the classes of the plurality of classes.