Defect Localization Sensing System for TIG Welding Rapid Manufacturing Metal Bodies

Literature Overview

Published in 2009 in Hot Working Technology (热加工工艺) by Luo Yong, Wang Fuming, and Zhang Hua from Jiangxi University of Science and Technology and Nanchang University, this study presents a defect localization sensing system for metal bodies fabricated through TIG welding-based rapid manufacturing. Funded by the National 973 Program (2005CCA04300) and Jiangxi Provincial Natural Science Foundation (0650092), this work addresses the critical challenge of in-process defect detection and localization in additive metal deposition processes, which is directly relevant to quality assurance in modern cladding and overlay welding operations.

System Architecture and Sensing Principles

The defect localization sensing system integrates multiple sensing modalities to detect and locate defects in real-time during the TIG welding deposition process. The system architecture comprises:

Arc voltage sensing: Monitors arc stability and detects porosity through voltage fluctuation analysis
Arc current monitoring: Tracks current variations that indicate lack of fusion or crater defects
Visual sensing (CCD camera): Captures weld pool morphology and bead profile deviations
Acoustic emission (AE) sensors: Detects crack initiation and solidification cracking through high-frequency acoustic signals
Thermal imaging: Maps temperature distribution to identify hot spots and thermal anomalies

Sensing Modality	Detectable Defects	Spatial Resolution	Temporal Resolution
Arc Voltage	Porosity, arc instability	1-2 mm	10 kHz sampling
Arc Current	Lack of fusion, spatter	2-3 mm	10 kHz sampling
CCD Camera	Bead profile deviation, undercut	0.5 mm/pixel	30 fps
Acoustic Emission	Cracking, solidification defects	5-10 mm	1 MHz bandwidth
Thermal Imaging	Thermal anomalies, incomplete fusion	1-2 mm	30 Hz

Signal Processing and Defect Classification

The system employs signal processing algorithms to extract defect-relevant features from raw sensor data:

Arc voltage signal processing: Bandpass filtering (100 Hz-5 kHz) to isolate porosity-related fluctuations from background noise; statistical analysis of voltage variance over sliding windows to identify defect probability
Current signal analysis: Derivative-based detection of sudden current drops indicating lack of fusion; harmonic analysis for arc stability assessment
Acoustic emission feature extraction: Wavelet transform for time-frequency analysis; energy thresholding for crack detection; source localization through time-of-arrival differences between multiple AE sensors
Image processing: Edge detection for bead boundary identification; morphological analysis for bead shape deviation quantification; color analysis for oxidation and contamination detection

Engineering Application to Cladding Quality Control

The principles of this in-process sensing system are directly applicable to quality assurance in modern cladding operations:

Multi-pass overlay monitoring: In multi-pass weld overlay for pressure vessel fabrication, each pass must be monitored for defects before the next pass is applied. The sensing system provides real-time feedback for parameter adjustment between passes
Bond line quality assessment: For bimetallic cladding, the bond line quality is critical. Arc voltage and current monitoring during the first overlay pass can identify incomplete fusion at the base metal/overlay interface
Hot cracking detection: In nickel-based alloy overlay (Inconel 625, Hastelloy C276), solidification cracking is a primary concern. Acoustic emission monitoring provides early warning of crack initiation before it propagates to the surface
Porosity control: Gas porosity in overlay layers degrades corrosion resistance. Arc voltage monitoring combined with visual sensing can identify porosity formation in real-time, allowing immediate parameter correction

Integration with Standards-Based Quality Requirements

For pressure vessel fabrication under NB/T 47002, GB/T 150, and ASME VIII Div.1 requirements, the in-process sensing system complements post-weld NDT (RT, UT, MT, PT) by providing:

Early defect detection enabling immediate corrective action rather than post-weld repair
Statistical process control data for procedure qualification under NB/T 47014
Defect trend analysis for process optimization and capability improvement
Documentation of in-process quality parameters for regulatory inspection

Standard Requirement	Sensing System Contribution	Verification Method
NB/T 47002 weld quality	Real-time defect detection	Post-weld RT/UT verification
NB/T 47014 procedure qualification	Process parameter stability data	Witness coupon testing
ASME VIII Div.1 radiography	Pre-screening of suspect areas	Targeted RT inspection
API 934 overlay thickness	Deposition rate monitoring	Ultrasonic thickness measurement
Bond strength requirements	Interface quality assessment	Peel/shear bond testing

Study Insights and Future Direction

This 2009 research established foundational principles for in-process defect detection in additive metal deposition that remain highly relevant to modern cladding quality assurance. The multi-modal sensing approach—combining electrical, optical, acoustic, and thermal sensing—provides comprehensive defect coverage that no single modality can achieve alone. For cladding engineers responsible for overlay welding on pressure vessels, the key insight is that in-process monitoring should complement rather than replace post-weld NDT. The sensing system provides early warning and process control capability, while post-weld NDT provides the definitive quality verification required by codes and standards. The evolution of these sensing principles into modern digital twin and data analysis-based systems represents the natural progression of this research, enabling predictive quality assurance rather than merely reactive defect detection.