Numerical Simulation of Keyhole Formation in PAW-TIG Double-Sided Welding

Literature Overview

This paper by Sun Junsheng and colleagues from Shandong University and the University of Kentucky, published in Acta Metallurgic Sinica in 2003, presents a numerical simulation of the keyhole formation process during double-sided welding with a combined plasma arc and TIG arc (PAW+TIG) process. The research is supported by the US National Science Foundation (DMI 9812981) and the Open Fund of the State Key Laboratory of Modern Welding Production Technology at Harbin Institute of Technology.

The study establishes a mathematical model that considers the plasma flow force, gravity, and surface tension as the primary mechanical factors influencing keyhole formation. The model is coupled with the welding heat transfer control equations and solved using numerical simulation techniques to quantitatively analyze the dynamic process of keyhole formation.

Mathematical Model and Keyhole Formation Stages

The authors identify three distinct stages in the keyhole formation process:

Stage	Description	Key Characteristics
Stage 1	From welding start to breakthrough	Molten pool forms and deepens; no keyhole yet
Stage 2	From breakthrough to keyhole initiation	Molten pool reaches full thickness; keyhole begins to form
Stage 3	From keyhole initiation to stable keyhole morphology	Keyhole stabilizes; steady-state welding achieved

The authors propose that the minimum span of the molten pool can serve as an evaluation indicator for keyhole establishment. This is a practical and measurable parameter that can be used to monitor the welding process in real time.

The keyhole formation is described as a result of the combined action of thermal and mechanical forces. The plasma arc provides the intense heat input required to create the keyhole, while the mechanical forces (plasma flow force, surface tension, gravity) shape the keyhole geometry and determine its stability.

Force Analysis in Keyhole Formation

The force balance in the keyhole region is critical to understanding the welding process. The primary forces acting on the molten pool and keyhole are:

Plasma flow force — The high-velocity plasma jet exerts a downward force on the molten pool surface, driving the keyhole formation. This force is proportional to the plasma current and inversely proportional to the arc length.
Surface tension — Surface tension acts to minimize the surface area of the molten pool and keyhole. At the keyhole walls, surface tension acts to close the keyhole, opposing the plasma flow force.
Gravity — Gravity acts on the molten metal, promoting downward flow. In thin-plate welding, gravity can assist keyhole formation, but in thick-section welding, its effect is relatively minor compared to the plasma flow force.
Electromagnetic force (Lorentz force) — The interaction between the welding current and the self-magnetic field generates a Lorentz force that stirs the molten pool. This force contributes to the keyhole stability.

The numerical simulation reveals that the keyhole formation is a dynamic process where these forces continuously interact. The keyhole is not a static feature but rather a dynamic equilibrium between the forces that create it and the forces that tend to close it.

Implications for Pipeline Welding

For pipeline welding, particularly of thin-walled pipes where double-sided welding is common, the PAW+TIG process offers several advantages:

Single-pass welding — The combined heat input allows for single-pass welding of thinner pipe walls, reducing welding time.
Double-sided penetration — The keyhole process ensures full penetration through the pipe wall, which is essential for structural integrity.
Controlled weld geometry — The keyhole process produces a weld with predictable geometry, which is important for post-weld inspection.

However, the keyhole process is sensitive to several parameters that must be carefully controlled in pipeline welding:

Parameter	Effect of Variation	Control Requirement
Plasma current	Too low: no keyhole; too high: excessive penetration	Must be set within narrow window
Arc length	Too long: reduced penetration; too short: torch damage	Must be maintained consistently
Welding speed	Too fast: incomplete keyhole; too slow: excessive heat input	Must be synchronized with pipe rotation
Fit-up gap	Too large: excessive penetration; too small: incomplete fusion	Must be within specified tolerance

Connection to FMEA and Quality Control

From a quality control perspective, the keyhole welding process requires careful attention to potential failure modes. Using a Failure Mode and Effects Analysis (FMEA) approach:

Failure Mode	Potential Cause	Effect	Detection Method
Incomplete penetration	Insufficient plasma current; excessive welding speed	Lack of fusion at root	RT, UT
Excessive penetration	Excessive plasma current; too fast welding speed	Burn-through; distortion	Visual inspection
Keyhole instability	Arc length variation; fit-up variation	Irregular weld geometry	RT, visual inspection
Porosity	Inadequate shielding; keyhole instability	Reduced weld strength	RT, UT

The FMEA highlights the importance of process parameter control and fit-up quality in achieving consistent keyhole welding results. For pipeline applications, where weld integrity is critical, these factors must be rigorously controlled.

Study Insights and Conclusion

This numerical simulation study provides valuable insight into the fundamental physics of keyhole formation in PAW+TIG double-sided welding. The identification of three distinct stages in the keyhole formation process is particularly useful for process development and monitoring. The proposal to use the minimum molten pool span as an evaluation indicator offers a practical approach to real-time process monitoring. For pipeline engineers, the key takeaway is that the keyhole welding process, while powerful, requires careful parameter control and fit-up quality to achieve consistent results. The numerical simulation approach demonstrated in this study can be extended to other welding processes and materials, providing a valuable tool for process development and optimization. The interdisciplinary approach combining computational modeling, fluid dynamics, and welding engineering exemplifies the depth of understanding required for advanced welding process development.