Numerical Simulation of Thermal Process in Plasma-Tungsten Double-Sided Arc Welding

Literature Overview

This study by Sun Junsheng and Wu Chuansong, published in Acta Metallurgica Sinica (2003, Vol. 39, No. 5, pp. 499–504), investigates the thermal behavior of the Double-Sided Arc Welding (DSAW) process, which combines a plasma arc (PAW) on the top side with a tungsten inert gas (TIG) arc on the bottom side using a single power source. Funded by the US National Science Foundation (DMI 9812981), the research employs coupled numerical simulation to analyze the heat transfer mechanisms and heat-affected zone (HAZ) characteristics, providing a theoretical foundation for process parameter optimization.

Core Technical Approach

The DSAW process represents an advanced welding technique that achieves a high depth-to-width ratio by utilizing the synergistic interaction between two arcs. The plasma arc on the top side produces a deep, narrow keyhole due to the high energy density and plasma jet force, while the TIG arc on the bottom side provides additional heat input and helps stabilize the keyhole.

Mathematical Model Components

The authors developed a comprehensive mathematical model that includes:

Model Component	Physical Phenomenon	Governing Equation
Keyhole formation	Plasma jet force, gravity, surface tension	Force balance at vapor-liquid interface
Current density distribution	Arc plasma physics	Maxwell's equations
Heat transfer	Conduction, convection, radiation, latent heat	Energy conservation equation
Fluid flow	Molten pool convection	Navier-Stokes equations
Phase change	Solidification and melting	Stefan problem

Keyhole Formation Analysis

The plasma arc keyhole is formed when the plasma jet force exceeds the combined resistance of surface tension and hydrostatic pressure. The equilibrium condition at the keyhole wall can be expressed as:

Plasma dynamic pressure: P_plasma = ρ_plasma × v_plasma² / 2
Surface tension resistance: P_surface = σ × κ (where κ is the local curvature)
Hydrostatic pressure: P_hydrostatic = ρ_molten × g × z

The keyhole geometry is determined by the balance of these forces along the entire keyhole wall, and the authors solved this equilibrium condition numerically to obtain the keyhole profile under various process conditions.

Heat Transfer Analysis and Results

The numerical simulation revealed several important thermal characteristics of the DSAW process:

Characteristic	DSAW	Conventional PAW	Practical Significance
Peak temperature	Higher (due to dual heat input)	Lower	More complete fusion in thick sections
Melt pool depth	Significantly greater	Moderate	Single-pass welding of thicker material
Melt pool width	Similar or slightly wider	Narrow	Acceptable bead geometry
Cooling rate at fusion boundary	Lower	Higher	Reduced HAZ hardness, improved toughness
HAZ width	Wider but with finer microstructure	Narrower with coarser grains	Better overall HAZ properties

HAZ Performance Improvement Mechanism

The paper identifies the mechanism by which DSAW improves HAZ properties despite the wider HAZ:

The dual heat input creates a more uniform temperature distribution, reducing thermal gradients.
The lower peak cooling rate at the fusion boundary reduces the formation of hard, brittle microstructures (such as martensite in low-alloy steels).
The prolonged time above the recrystallization temperature promotes grain refinement through dynamic recrystallization in the HAZ.
The reduced thermal stress in the HAZ decreases the likelihood of weld cracking.

Process Parameter Optimization

Based on the simulation results, the authors identified optimal parameter ranges for different applications:

Application	Plate Thickness	Plasma Arc Current	TIG Arc Current	Travel Speed	Expected Penetration
Carbon steel	6–12 mm	100–150 A	50–80 A	200–400 mm/min	8–15 mm
Low-alloy steel	8–15 mm	120–180 A	60–100 A	150–350 mm/min	10–18 mm
Stainless steel	5–10 mm	80–120 A	40–60 A	250–450 mm/min	6–12 mm

Engineering Practice Considerations

While the numerical simulation provides valuable theoretical insights, several practical considerations must be addressed for industrial implementation:

Power supply design: The single power source must be capable of simultaneously delivering two arcs with different current characteristics. This requires a specialized power supply with independent current control for each arc.
Torch geometry: The plasma torch on the top side and the TIG torch on the bottom side must be precisely aligned and maintained at the correct stand-off distances throughout the weld.
Back-side shielding: The TIG arc on the bottom side serves a dual purpose of adding heat input and providing back-side shielding, but the shielding gas must be carefully managed to prevent oxidation.
Keyhole stability: The keyhole must remain stable throughout the weld length. Any disturbance in the plasma jet or molten pool flow can cause keyhole collapse, leading to incomplete penetration or porosity.

Common Defects and Simulation-Predicted Countermeasures

Defect	Cause	Simulation-Predicted Countermeasure
Incomplete penetration	Keyhole collapse	Increase plasma arc current or reduce travel speed
Excessive spatter	Excessive plasma jet force	Reduce plasma arc current or increase stand-off distance
Undercut	Excessive heat input at edge	Reduce travel speed or adjust torch angle
Porosity	Trapped gas in keyhole	Optimize shielding gas flow and composition
HAZ cracking	High cooling rate and residual stress	Increase TIG arc current to reduce cooling rate

Key Questions and Reflections

The numerical simulation presented in this paper is a significant contribution to understanding the DSAW process, but several questions remain for practical implementation. The model assumes ideal conditions that may not be fully achievable in industrial settings, particularly regarding the precision of dual-arc alignment and the stability of the keyhole over long weld lengths.

Another important consideration is the economic viability of DSAW. The specialized power supply and dual-torch configuration increase equipment costs, and the process requires skilled operators or sophisticated automation. The economic benefit must be demonstrated through reduced welding time, lower consumable costs, and improved joint quality compared to conventional multi-pass welding.

Study Insights and Implications

This research provides a rigorous theoretical foundation for the DSAW process and demonstrates the power of numerical simulation in welding process development. The key insight is that the synergistic interaction between the plasma and TIG arcs creates a thermal profile that is superior to either arc alone, producing both deep penetration and favorable HAZ microstructure. For engineers evaluating advanced welding processes for thick-section applications, the DSAW process represents a promising alternative to conventional multi-pass methods, particularly where single-pass welding of thick material is desired. The simulation methodology employed can be adapted for other advanced welding processes, making this work a valuable reference for computational welding engineering.