Fully Digital Control Strategy for Dual-Pulse MIG/MAG Welding

Literature Overview

The paper by Sha Deshang and Liao Xiaozhong, published in the Journal of Beijing Institute of Technology (2009, Vol. 29, No. 7, pp. 605-607), presents a fully digital control strategy for dual-pulse MIG/MAG welding. Funded by the National Natural Science Foundation of China (Grant No. 50807005) and the China Postdoctoral Science Foundation (Grant No. 20060390506), this work introduces a three-loop closed-loop control model that achieves equal arc length control and process stability through variable-speed wire feeding.

Core Technical Content

Dual-pulse MIG/MAG welding is an advanced arc welding process that combines the benefits of pulsed transfer (fine droplet transfer, low spatter) with the higher deposition rate of spray transfer. The dual-pulse waveform consists of a high-current pulse that detaches the droplet and a lower-current pulse that preheats the next droplet, resulting in a more stable and controllable transfer process. However, the control complexity of dual-pulse welding is significantly higher than conventional pulsed welding, as multiple current and voltage parameters must be coordinated within each droplet transfer cycle.

The proposed control model features three nested loops:

1. Arc length stabilization loop: Maintains constant arc length by adjusting the wire feed speed based on arc voltage feedback.
2. Voltage instantaneous value feedback loop: Corrects for transient voltage fluctuations within the droplet transfer cycle.
3. Current instantaneous value feedback loop: Adjusts current parameters to compensate for changes in wire stick-out length and other process disturbances.

The entire control system is implemented using a Digital Signal Processor (DSP), enabling fully digital signal processing and control without analog circuitry.

Control Architecture and Technical Parameters

The variable-speed wire feeding is a critical innovation. In conventional pulsed MIG welding, the wire feed speed is constant, and arc length is controlled by adjusting the pulse current amplitude. In the proposed system, the wire feed speed itself becomes a control variable, allowing the system to respond to arc length changes within a single droplet transfer cycle rather than waiting for the next cycle.

Process Stability and Weld Quality

The experimental results demonstrate that the proposed control strategy achieves:

• Stable welding process with minimal arc length variation
• Aesthetic weld bead formation with uniform reinforcement
• Real-time compensation for stick-out length changes
• Robust performance under process disturbances

From a metallurgical perspective, the stability of the dual-pulse transfer mode is crucial for achieving favorable microstructural characteristics in the weld metal. Unstable droplet transfer leads to irregular solidification patterns, increased porosity risk, and potential hot cracking susceptibility—particularly problematic in high-strength steels and stainless steels used in pipeline applications.

Engineering Practice Considerations

In pipeline welding applications, particularly for large-diameter pipes requiring multi-pass welding, process stability is paramount. The fully digital control approach offers several advantages for industrial deployment:

1. Parameter flexibility: Digital systems allow rapid parameter changes between passes, materials, and joint configurations without hardware modification.
2. Diagnostic capability: DSP-based systems can store and analyze process data for quality traceability, a requirement under ASME B31.3 and API 5L welding qualification standards.
3. Scalability: The control architecture can be extended to multi-wire or hybrid welding configurations with minimal hardware changes.

However, the implementation of fully digital control in a production environment requires careful attention to electromagnetic compatibility (EMC). Arc welding generates intense electromagnetic interference that can corrupt digital signals. Shielding, filtering, and proper grounding strategies must be incorporated into the system design. In my experience with production welding systems, the difference between a laboratory demonstration and a reliable production system often lies in the quality of the signal conditioning and noise rejection.

Key Questions and Reflections

The paper raises an important question about the optimal division of control authority between the three loops. If all three loops simultaneously adjust the wire feed speed, potential conflicts could arise, leading to oscillatory behavior. A hierarchical structure with clearly defined bandwidth separation between loops would be more robust. The arc length loop should operate at the slowest bandwidth (one transfer cycle or slower), while the instantaneous voltage and current feedback loops should operate at much higher frequencies with limited authority to prevent interaction.

Another consideration is the computational burden on the DSP. Real-time image processing, signal conditioning, and multi-loop control must all execute within strict timing constraints. The paper does not detail the specific DSP architecture or the computational budget allocation, which would be essential information for engineers attempting to replicate or extend this work.

For pipe welding applications, the dual-pulse control strategy could be particularly valuable in root pass welding of large-diameter pipes, where precise heat input control is critical to avoid burn-through while maintaining adequate penetration. The ability to compensate for stick-out length changes in real time would be especially beneficial in positional welding, where the effective stick-out changes as the torch angle varies with the weld position around the pipe circumference.