A Hybrid Arc Model for Pulsed TIG Welding
Literature Overview and Core Findings
This paper, published in the Journal of Welding in 2015, addresses a problem that has puzzled welding engineers for decades: how to mathematically model the dynamic behaviour of a pulsed TIG arc. The authors, from Tianjin University, propose a hybrid arc model that combines the classic Mayr and Cassie models with a newly designed arc blocking coefficient equation. The Mayr model is good for low-current arcs, while the Cassie model is good for high-current arcs. A pulsed TIG arc transitions between low and high current states, so neither model alone is adequate. The authors introduce a blocking coefficient that varies with current and time, and they validate the model against static and dynamic characteristics. The results show that the hybrid model can effectively simulate the pulsed TIG arc.
Technical Interpretation and Process Analysis
Why We Need a Better Arc Model
In the steel pipe and fitting industry, pulsed TIG is widely used for root passes on stainless steel and alloy pipes, especially for orbital welding of small-diameter tubes. The pulsed current helps control heat input, reduce distortion, and improve weld bead shape. But the arc behaviour during the pulse is complex. When the current rises from the base level to the pulse level, the arc expands and its voltage changes. When the current falls, the arc contracts. The dynamic response of the arc is not instantaneous; it has a time constant. The Mayr model assumes that the arc conductance changes exponentially with time, which works well for low-current arcs. The Cassie model assumes that the arc voltage remains constant, which works well for high-current arcs. In a pulsed arc, the current swings between these two regimes, so a single model cannot capture the behaviour. The hybrid model addresses this by using a blocking coefficient that blends the two models.
The Blocking Coefficient Equation
The authors propose an arc blocking coefficient equation that determines the relative weight of the Mayr and Cassie models. The coefficient is a function of the arc current and the rate of change of current. At low current, the coefficient favours the Mayr model; at high current, it favours the Cassie model. The key parameters are the blocking time constant and the current threshold. The authors give a method for determining these parameters from experimental data. This is a practical contribution, because it allows engineers to calibrate the model for their specific welding equipment.
Validation and Practical Use
The authors validate the model by comparing its static characteristics with experimental data. The static characteristic of a TIG arc is the voltage-current curve at steady state. The hybrid model matches the experimental curve well. They also validate the dynamic characteristics at different pulse frequencies. The model correctly predicts the arc voltage waveform during the pulse cycle. This is important for controlling the arc in real time. For example, if you want to maintain a constant arc length, you need to know how the voltage responds to current changes. The model provides this information.
Integration with Engineering Practice
Application to Orbital TIG Welding of Pipes
In our industry, we use pulsed TIG for orbital welding of stainless steel pipes in the food, pharmaceutical, and semiconductor industries. The typical parameters are: base current 40–60 A, pulse current 100–150 A, pulse frequency 2–5 Hz, and travel speed 100–200 mm/min. The arc model helps us understand how the arc behaves during the pulse. For example, at the start of the pulse, the arc voltage rises sharply because the current increases faster than the arc can expand. This can cause a momentary increase in heat input, which may lead to burn-through on thin-wall pipe. The model allows us to predict this and adjust the pulse rise time to avoid it.
Parameter Selection Based on the Model
Using the hybrid model, I can propose a set of guidelines for selecting pulse parameters:
| Parameter | Typical Range | Effect on Arc | Recommended Setting |
|---|---|---|---|
| Base current | 30–60% of pulse current | Maintains arc stability | 50% of pulse current |
| Pulse current | 80–150 A | Provides penetration | 120 A for 3 mm wall |
| Pulse frequency | 1–10 Hz | Controls weld bead ripple | 3 Hz for orbital welding |
| Pulse duty cycle | 30–70% | Controls heat input | 50% for balanced weld |
| Rise time | 5–20 ms | Affects arc expansion | 10 ms for stable arc |
| Fall time | 5–20 ms | Affects arc contraction | 10 ms for stable arc |
The model helps us understand that the rise and fall times are critical. If the rise time is too short, the arc may not expand fast enough, causing a high voltage spike and possible arc instability. If the rise time is too long, the heat input may be insufficient, leading to lack of fusion. The model provides a quantitative way to optimise these parameters.
Common Problems and Solutions
- Arc instability at low current – the Mayr model predicts that the arc conductance decays rapidly at low current, which can cause the arc to extinguish. Countermeasure: increase the base current to at least 30% of the pulse current, or use a higher frequency to keep the arc ionised.
- Arc voltage spikes at pulse start – the Cassie model predicts that the arc voltage remains constant, but the hybrid model shows that the voltage can spike if the rise time is too short. Countermeasure: increase the rise time to 10–15 ms.
- Weld bead ripple – the pulse frequency controls the ripple spacing. If the ripple is too coarse, the bead surface is rough. If it is too fine, the bead may be too smooth but the heat input may be excessive. Countermeasure: use a frequency of 2–5 Hz for most orbital welding applications.
Key Questions and Reflections
This paper is a theoretical contribution, but it has practical value. The question I would ask is: how accurate is the model for real welding arcs, which are affected by shielding gas flow, electrode geometry, and magnetic fields? The authors validated the model against a specific experimental setup. In our shop, we have different torch designs and gas mixtures, so the model parameters would need to be re-calibrated. This is not a limitation, but a practical requirement. The model provides a framework for calibration. Another question is: can the model be extended to pulsed TIG with a background current that is not zero? In many applications, the base current is not zero, so the arc does not extinguish. The hybrid model should handle this, because the Mayr model works at low current and the Cassie model works at high current. The transition is smooth.
Study Insights and Implications
For a welding engineer, this paper is a reminder that the arc is not a simple resistor. It has dynamic behaviour that affects weld quality. The hybrid model gives us a tool to understand and predict this behaviour. In practice, I would use this model to design a pulsed TIG welding procedure for a new application. I would start with a set of parameters from experience, then use the model to simulate the arc voltage and current waveforms. If the simulation shows excessive voltage spikes or unstable arc, I would adjust the rise time or the base current. This approach saves time and reduces the number of trial welds. The paper also highlights the importance of understanding the physical processes in the arc. Many welding problems, such as arc instability, burn-through, and lack of fusion, can be traced back to arc dynamics. A good model helps us diagnose these problems. In the future, I expect to see more advanced models that include the effect of electrode geometry and gas flow. But for now, the hybrid model is a solid foundation. For our industry, the key takeaway is that pulsed TIG welding is not just a matter of setting current and frequency. The rise and fall times, the base and pulse currents, and the frequency all interact. A good model helps us optimise these parameters for consistent weld quality.
Zhuojin Pipe Fitting Co., Ltd