Numerical Simulation of Intermittent Alternating Longitudinal Magnetic Fields in TIG Welding
Introduction to the Literature and Its Relevance
This 1999 paper from Xi'an Jiaotong University tackles a problem that has been on my mind since I first ran a TIG cell in the early 1990s: how do you predict the magnetic field distribution when you wrap a coil around a tungsten electrode and drive it with a pulsed current? The authors, Luo Jian and colleagues, developed a singularity-free single-integral mathematical model to calculate the intermittent alternating longitudinal magnetic field generated by a coaxial axisymmetric hollow cylindrical coil under bidirectional pulsed rectangular excitation current. The work was funded by the National Natural Science Foundation, which tells you this was considered fundamental research with real industrial promise.
The practical question behind the mathematics is straightforward: if you want to use an external magnetic field to stir the weld pool, control the arc, or improve penetration, you need to know what field you are actually generating. The field is not uniform, it is not constant, and it interacts with the arc in ways that are difficult to measure directly. This paper gives you a tool to calculate it.
Core Technical Content and Key Findings
The paper presents a mathematical model based on the Biot-Savart law, but with a crucial improvement: it avoids the singularity problem that occurs when the field point lies on the coil itself. The single-integral formulation is computationally efficient, which matters when you are doing parametric studies or iterative design of magnetic field-assisted welding systems.
The authors discuss three important aspects of the field distribution:
- Field distribution along the axis and radial direction - The longitudinal component Bz is maximum at the coil center and decreases with distance. The radial component Br is zero on the axis but becomes significant off-axis.
- Field uniformity - For a given coil geometry, there is a region where the field is relatively uniform. This is critical because you want the magnetic field to act consistently on the arc and the weld pool.
- Ratio of longitudinal to transverse components - The paper quantifies Bz/Br as a function of position. This ratio is important because the longitudinal component is what you want for arc constriction and pool stirring, while the transverse component can cause arc deflection and asymmetry.
The key parameters that determine the field are:
| Parameter | Symbol | Typical Range | Practical Significance |
|---|---|---|---|
| Coil inner radius | a | 5-20 mm | Determines how close the coil can be to the arc |
| Coil length | L | 10-50 mm | Affects field uniformity along the axis |
| Number of turns | N | 50-500 | Directly proportional to field strength |
| Excitation current | I | 1-10 A | Limited by coil heating and power supply capability |
| Pulse frequency | f | 0.1-10 Hz | Controls the stirring frequency in the pool |
| Duty cycle | D | 20-80% | Controls the intermittent nature of the field |
Engineering Interpretation and Practical Implications
Now let me tell you what this means for someone who actually has to weld stainless steel pipe on a production line.
First, the magnetic field approach is not a laboratory curiosity. I have seen it used in real applications, particularly for austenitic stainless steel where the weld pool is viscous and penetration is limited. The intermittent alternating field creates a periodic Lorentz force in the pool, which promotes convection and can improve penetration depth. This is particularly useful for root passes in pipe welding where you need to achieve full penetration without excessive heat input.
Second, the calculation method matters because you cannot measure the field inside the arc. The arc plasma is electrically conducting, and inserting a Hall probe would disturb the very field you are trying to measure. So you have to rely on calculation. The model in this paper gives you a practical way to design the coil and choose the excitation parameters.
Third, the uniformity issue is critical in practice. If the field is not uniform across the weld pool, you get asymmetric stirring, which leads to asymmetric weld beads and potential lack of fusion on one side. This is a real problem when welding pipe in the 5G position, where gravity is already causing the pool to sag.
Lessons from the Shop Floor and Delivery Site
I have seen what happens when engineers ignore the field distribution. On one occasion, we had a magnetic field-assisted welding setup for a duplex stainless steel line pipe project. The coil was positioned incorrectly, and the field was stronger on one side of the arc than the other. The result was a weld with excessive penetration on one side and lack of fusion on the other. The pipe had to be cut out and rewelded, costing us two days of production and a significant amount of money.
The lesson is simple: if you are going to use magnetic field assistance, you must calculate the field distribution and verify that the coil position and excitation parameters give you a uniform field across the weld pool. The model in this paper is the tool you need.
Another practical point: the intermittent nature of the field. With a pulsed excitation current, you get a periodic field that alternately strengthens and weakens. This creates a periodic stirring effect in the pool, which can be beneficial for grain refinement and defect reduction. But the frequency and duty cycle must be matched to the welding speed and the pool dimensions. If the frequency is too high, the pool does not have time to respond. If it is too low, the stirring is too coarse and can cause turbulence and porosity.
Standards and Quality Considerations
For stainless steel pipe welding, the relevant standards are ASME B31.3 for process piping and API 5L for line pipe. Neither standard explicitly addresses magnetic field-assisted welding, but both require that the welding procedure be qualified and that the weld quality meet specific acceptance criteria.
When you introduce a new technique like magnetic field assistance, you need to:
- Qualify the welding procedure according to ASME Section IX or equivalent.
- Verify that the weld meets the required mechanical properties, including tensile strength, impact toughness, and hardness.
- Perform non-destructive testing, typically RT or UT, to verify that there are no defects.
- Document the procedure parameters, including the magnetic field parameters, so that the welder can reproduce the results.
The magnetic field parameters are not typically specified in the WPS, but they should be. If the field is not applied consistently, the weld quality will vary. I have seen this happen: the welder forgets to turn on the magnetic field, and the weld has shallow penetration. The WPS should specify the coil position, excitation current, frequency, and duty cycle, just as it specifies the welding current, voltage, and travel speed.
Key Questions and Reflections
The paper raises several questions that are worth considering:
- What is the optimal field strength for a given material and thickness? The paper does not answer this directly, but the model provides the basis for determining it. For austenitic stainless steel, I would expect the optimal field strength to be in the range of 10-50 mT, but this depends on the welding parameters and the pool dimensions.
- How does the field interact with the arc plasma? The paper focuses on the field calculation, not the plasma interaction. This is a separate problem that requires magnetohydrodynamic modeling. The interaction is complex because the arc is a conducting fluid, and the magnetic field affects the current density distribution and the Lorentz force.
- What is the effect on weld pool dynamics? The field creates a periodic stirring effect, but the pool dynamics are also influenced by surface tension, gravity, and the arc pressure. The relative importance of these forces depends on the welding position and the pool size.
Study Insights and Implications for Practice
This paper is a good example of how fundamental research can support practical engineering. The mathematical model is elegant, but the real value is in the engineering application. For anyone working on magnetic field-assisted welding, this paper provides a solid foundation for coil design and parameter selection.
My advice to fellow engineers:
- Use the model to design your coil. Do not guess. Calculate the field distribution and verify that it is uniform across the weld pool.
- Start with a conservative field strength. Too much field can cause arc instability and spatter. Too little field has no effect. Start with a low value and increase it until you see the desired effect.
- Monitor the weld quality carefully. The magnetic field can mask other problems, such as poor fit-up or incorrect torch angle. Do not assume that the magnetic field will solve all your problems.
- Document everything. The magnetic field parameters should be included in the WPS and the procedure qualification record. This ensures that the welder can reproduce the results and that the quality is consistent.
Conclusion
This paper from 1999 remains relevant today because the problem it addresses is still not fully solved. Magnetic field-assisted welding is a powerful technique, but it requires careful design and control. The numerical model presented here gives you the tools to do that. In my experience, the key to success is understanding the field distribution and ensuring that it is uniform across the weld pool. If you do that, you can achieve deeper penetration, better grain refinement, and fewer defects. If you do not, you will get asymmetric welds and quality problems that will cost you time and money. This paper is a valuable reference for anyone serious about magnetic field-assisted TIG welding.
Zhuojin Pipe Fitting Co., Ltd