Optical-Electrical Penetration Signal Extraction in Steel and Aluminum TIG Welding

Literature Overview

This paper by Zhao Chongyi, Yin Shuyan, Geng Zheng, and Cao Jiming, published in Metal Science and Engineering (1991, Vol. 10, No. 3), originates from Harbin Institute of Technology. The research focuses on the development of a real-time weld penetration monitoring system using PbS (lead sulfide) infrared detectors for TIG welding of both steel and aluminum. The study addresses the challenge of distinguishing high-temperature molten zone radiation from lower-temperature near-weld zone radiation to achieve reliable penetration detection.

Core Technical Content

Penetration Monitoring Principle

Weld penetration monitoring is critical for ensuring full weld joint fusion and preventing defects such as lack of fusion and incomplete penetration. The optical-electrical method detects infrared radiation emitted from the weld zone, particularly from the back side of the workpiece where the penetration front reaches.

Detection System Configuration

Component	Specification
Detector	PbS (Lead Sulfide) infrared detector
Operating wavelength range	1.5–2.5 μm (primary band)
Operating temperature	77 K (liquid nitrogen cooled) or room temperature with filtering
Response time	<10 ms
Sensitivity	0.1–1.0 mV/W
Filter	Cut-off filter (long-pass) to isolate specific wavelength bands
Signal processing	Amplification, threshold detection, comparison circuit

Signal Characteristics in Steel vs. Aluminum TIG Welding

Parameter	Steel TIG Welding	Aluminum TIG Welding
Peak emission wavelength	1.0–2.0 μm	0.8–1.5 μm
Emission intensity at penetration	Moderate	High (due to higher reflectivity)
Background radiation	Lower	Higher (due to higher thermal conductivity)
Signal-to-noise ratio	Better	Lower (requires filtering)
Penetration detection threshold	Lower current	Higher current

Penetration Resolution Technique

The key innovation described in this paper is the use of a low-cost cut-off filter (long-pass filter) to achieve spectral discrimination between the high-temperature molten zone and the lower-temperature near-weld zone. The filter transmits radiation from the molten pool (T > 1500°C) while blocking radiation from the HAZ (T < 1000°C), enabling reliable penetration detection.

Filter Type	Cut-off Wavelength	Transmitted Band	Application
Glass long-pass	1.2 μm	1.2–2.5 μm	Steel TIG welding
Glass long-pass	0.9 μm	0.9–1.8 μm	Aluminum TIG welding
Multilayer dielectric	1.0 μm	1.0–2.0 μm	High-precision applications

Signal Processing and Threshold Determination

The penetration signal is processed through the following stages:

Raw signal acquisition: Infrared radiation detected by PbS sensor
Amplification: Signal amplified to measurable voltage level
Filtering: Cut-off filter removes unwanted spectral bands
Threshold comparison: Signal compared against pre-determined penetration threshold
Output: Binary (penetration/no-penetration) or analog (penetration degree) signal

The penetration threshold is determined through calibration experiments, where the detector response is measured at known penetration depths. For steel, the threshold typically corresponds to a back-side temperature of 1200–1400°C, while for aluminum, the threshold is set at 800–1000°C due to the lower melting point and higher thermal conductivity.

Engineering Applications

Real-Time Process Control

The penetration monitoring system can be integrated into a closed-loop control system that:

Adjusts welding current to maintain constant penetration
Detects and compensates for joint fit-up variations
Identifies and flags incomplete penetration before the weld is completed
Reduces the need for post-weld destructive testing

Quality Assurance Benefits

Benefit	Description
Reduced rework	Early detection of penetration issues
Lower inspection costs	Fewer RT/UT examinations required
Improved consistency	Uniform penetration throughout weld length
Operator assistance	Real-time feedback for manual welding
Process documentation	Continuous penetration record for quality traceability

Applicable Standards

AWS D1.1 (Structural Welding Code - Steel)
AWS D10.1 (Structural Welding Code - Aluminum)
ASME BPV Section V (Nondestructive Examination)
ISO 9712 (Qualification and Certification of NDT Personnel)
GB/T 3323 (Radiographic Testing of Welds)

Study Insights and Limitations

The optical-electrical penetration monitoring approach represents an early but significant contribution to real-time welding process control. The use of PbS detectors with spectral filtering provides a cost-effective solution for penetration detection in both steel and aluminum TIG welding.

However, several limitations should be noted:

Detector cooling requirement: PbS detectors typically require liquid nitrogen cooling for optimal performance, which adds complexity to field deployment.
Alignment sensitivity: The detector must be precisely aligned with the weld zone, which is challenging in production environments with varying joint geometries.
Environmental interference: Ambient thermal radiation, reflections from surrounding equipment, and shielding gas effects can introduce noise.
Material-specific calibration: Each material combination requires separate calibration, limiting universal applicability.

For modern applications, these limitations have been largely addressed through the development of uncooled microbolometer detectors, fiber-optic sensors, and multi-sensor fusion systems. Nevertheless, the fundamental principle of spectral discrimination for penetration detection remains valid and continues to influence contemporary weld monitoring technologies.

The research demonstrates that real-time penetration monitoring is technically feasible using relatively simple optical components, providing a foundation for the development of more sophisticated process control systems that can improve weld quality, reduce rework, and enhance manufacturing efficiency in steel pipe and fitting production.