DSP-Based Seam Tracking System for Pipeline TIG Welding

Literature Overview and Research Context

The study by Sun Xinhui from Offshore Oil Engineering Co., Ltd., published in China Shipbuilding in 2013, presents the development and implementation of a seam tracking system for pipeline TIG welding based on Digital Signal Processing (DSP) technology. Pipeline welding, particularly in the offshore oil and gas industry, presents unique challenges related to weld quality, consistency, and productivity. The offshore environment introduces additional constraints such as limited access, difficult positioning, and the need for high-integrity welds that must withstand extreme pressures and corrosive environments. Seam tracking, which ensures that the welding torch follows the joint line accurately throughout the welding process, is a critical technology for achieving these quality objectives, especially in automated or semi-automated welding operations.

Core Technical Content and System Architecture

The proposed system employs a DSP-based signal processing architecture to detect and track the weld seam in real time. The fundamental principle involves sensing the deviation between the actual torch position and the desired seam centerline, then generating corrective signals to adjust the torch position accordingly. The DSP processor handles the acquisition, filtering, and processing of sensor signals at high speed, enabling rapid response to seam deviations and maintaining weld quality throughout the joint.

Sensor Configuration and Signal Processing

The system utilizes an arc sensing method, where the welding arc itself serves as the sensing element. By modulating the arc current or measuring the arc voltage at specific frequencies, the system can determine the lateral position of the torch relative to the joint centerline. When the torch is centered on the joint, the arc voltage waveform exhibits a characteristic pattern; deviations from the centerline cause measurable changes in the voltage waveform, which the DSP processor interprets as a position error. The system then generates a proportional corrective signal to move the torch back toward the seam centerline.

The DSP-based approach offers significant advantages over analog signal processing systems. The digital filtering capabilities allow for precise noise rejection, the ability to implement complex control algorithms, and the flexibility to adjust system parameters without hardware modifications. The processing speed of modern DSP processors, typically operating at clock frequencies of 100–200 MHz, enables control loop response times of less than 1 ms, which is sufficient for most pipeline welding applications where travel speeds range from 50 to 200 mm/min.

Control Algorithm and Real-Time Performance

The control algorithm implemented in the DSP processor combines proportional-derivative (PD) control with adaptive gain scheduling. The PD controller provides fast response to seam deviations while minimizing overshoot, which is critical for maintaining weld geometry. The adaptive gain scheduling adjusts the controller gains based on the welding current and travel speed, ensuring stable operation across the full range of welding conditions. For example, at higher travel speeds, the controller gains are increased to compensate for the reduced dwell time at each point along the seam, while at lower travel speeds, the gains are reduced to prevent oscillation.

The system also incorporates a dead zone in the control algorithm, typically set at ±0.3 mm, to prevent unnecessary corrections when the torch is already within an acceptable tolerance of the seam centerline. This dead zone reduces wear on the servo mechanism and prevents the introduction of high-frequency oscillations into the welding arc, which could degrade weld quality.

Engineering Implementation and Performance Evaluation

The system was implemented and tested on pipeline welding operations at Offshore Oil Engineering Co., Ltd. The test pipelines consisted of carbon steel pipes with diameters ranging from 219 mm to 610 mm and wall thicknesses of 8–16 mm. The welding process employed was a hybrid approach: GTAW (TIG) for the root pass, followed by GMAW (MAG) for the fill and cap passes. The seam tracking system was applied specifically to the root pass, where the joint geometry is most sensitive to torch positioning errors.

The performance evaluation demonstrated a significant improvement in weld quality compared to manual or conventional automated welding without seam tracking. The root pass defect rate decreased by approximately 60%, with a notable reduction in lack of fusion, undercut, and excessive reinforcement. The welding productivity increased by approximately 15% due to the elimination of torch repositioning and the ability to maintain consistent travel speed without frequent interruptions for visual inspection.

Defect Reduction Analysis

The following table summarizes the defect reduction achieved with the seam tracking system:

Study Insights and Implications for Engineering Practice

This research demonstrates the practical value of DSP-based seam tracking in pipeline welding applications, particularly in the offshore industry where weld quality is paramount. The system's ability to maintain consistent torch positioning throughout the welding process addresses one of the most persistent challenges in automated welding: joint geometry variation due to fit-up tolerances, pipe misalignment, and thermal distortion.

In the context of bimetal pressure vessel fabrication, where clad plate welds and overlay welds are common, seam tracking technology has direct applicability. The overlay welding of corrosion-resistant alloys on carbon steel substrates requires precise torch positioning to ensure uniform overlay thickness and consistent dilution. The DSP-based approach described in this study could be adapted for overlay welding applications with appropriate modifications to the sensing method and control parameters.

The study also highlights the importance of system integration and operator training. The effectiveness of the seam tracking system depends not only on the hardware and software but also on the proper setup and calibration of the system for each specific joint configuration. Operator training is essential to ensure that the system is correctly configured for different pipe diameters, joint preparations, and welding positions.

Overall, this literature represents a significant contribution to the field of automated welding technology, with clear practical benefits for pipeline manufacturing and related industries. The DSP-based approach offers a cost-effective and reliable solution for seam tracking that can be readily adapted to other welding applications requiring precise torch positioning.