CLADDING TECHNOLOGY SHANXI CO., LTDExtraction and Analysis of Penetration Signals During High-Speed TIG Welding
Literature Overview and Historical Context
This 1990 publication by Wang Qilong, Zhang Jiuhai, and Yang Chunli from Harbin Institute of Technology represents pioneering work in weld penetration monitoring during high-speed TIG welding. Published in the Welding Journal, this research was conducted during a period when welding process monitoring was transitioning from qualitative visual inspection to quantitative signal-based approaches. The work addressed a critical practical challenge: how to ensure adequate penetration during high-speed welding operations where visual monitoring becomes impractical.
Core Technical Content on Penetration Signal Characteristics
High-speed TIG welding, typically defined as travel speeds exceeding 1.0 m/min for plate thicknesses of 3-6 mm, presents unique challenges for maintaining consistent penetration. At these speeds, the weld pool becomes elongated and narrow, and the window for maintaining adequate penetration narrows significantly. The research identifies several signal characteristics that correlate with penetration quality:
| Signal Parameter | Normal Penetration | Insufficient Penetration | Excessive Penetration |
|---|---|---|---|
| Arc voltage mean value | Stable, within 5-10% variation | Gradual decrease | Gradual increase |
| Arc voltage standard deviation | Low, <5% of mean | Increasing trend | Increasing trend |
| Arc current fluctuations | Minimal, <2% variation | Reduced fluctuations | Increased fluctuations |
| Arc sound intensity | Stable baseline | Decreasing intensity | Increasing intensity |
| Optical radiation intensity | Stable | Decreasing | Increasing with instability |
The key finding of this research is that the arc voltage signal, when properly processed, provides reliable real-time information about weld penetration. Specifically, the mean arc voltage correlates positively with penetration depth, while the voltage fluctuations indicate penetration instability.
Signal Processing Methodology
The research develops a systematic approach to penetration signal extraction and analysis:
- Signal acquisition: Arc voltage and current signals are acquired at sampling rates of 10-50 kHz using high-resolution data acquisition systems.
- Noise removal: Low-pass filtering removes high-frequency noise while preserving the penetration-relevant signal components in the 0-100 Hz range.
- Feature extraction: Statistical parameters including mean value, standard deviation, peak-to-peak amplitude, and spectral components are calculated over sliding time windows.
- Correlation analysis: The extracted features are correlated with actual penetration measurements obtained through destructive testing of sample welds.
- Threshold determination: Critical thresholds are established to distinguish between adequate, insufficient, and excessive penetration conditions.
Engineering Practice Applications
The practical applications of this research extend across multiple welding operations:
- Butt welding: Monitoring penetration during high-speed butt welding of thin plates ensures consistent weld quality without requiring operator visual judgment.
- Overlay welding: In cladding applications, penetration depth directly affects dilution ratio, making real-time penetration monitoring essential for controlling overlay composition.
- Pipe welding: For orbital TIG welding of pipes, where visual access is limited, signal-based penetration monitoring provides critical quality assurance.
- Automated welding: Integration of penetration signals into automated welding systems enables closed-loop control of travel speed and current to maintain optimal penetration.
For bimetal pressure vessel fabrication, the ability to monitor penetration in real time is particularly valuable when welding dissimilar metal joints where dilution must be carefully controlled. The signal processing approach described in this research provides a non-invasive, real-time method for ensuring that the penetration depth remains within the required range throughout the welding operation.
Study Insights and Long-Term Impact
This 1990 research, while conducted with the computational resources available at that time, established fundamental principles that remain relevant to modern welding process monitoring. The signal characteristics identified—particularly the correlation between arc voltage and penetration depth—have been validated and refined through subsequent research over the past three decades.
The work also highlights an important methodological principle: effective process monitoring requires not only signal acquisition but also systematic correlation with actual weld quality through destructive testing. This calibration step, while time-consuming, is essential for establishing reliable monitoring criteria. The research methodology employed here—combining signal processing with metallurgical analysis—provides a template for developing monitoring systems for other welding processes and materials.
The enduring value of this work lies in its practical orientation. Unlike purely theoretical studies, this research directly addresses a critical manufacturing challenge and provides actionable solutions that can be implemented with relatively modest equipment investment. This practical focus has contributed to the widespread adoption of signal-based welding monitoring in industrial applications worldwide.