Zhuojin Pipe Fitting Co., LtdNumerical Simulation of TIG Welding Flow Heat Transfer and Interface Tracking with Dynamic Mesh Method
Literature Overview
This paper by Li Linmin and colleagues from Northeastern University, published in the Journal of Northeastern University (Natural Science Edition) in 2017, presents a comprehensive magnetohydrodynamic (MHD) numerical model for TIG (Tungsten Inert Gas) welding. The work was supported by the National Natural Science Foundation of China (Grant No. 51574068). The authors developed a coupled model integrating fluid flow, heat transfer, solidification-melting, and dynamic mesh interface tracking to simulate the complex physics occurring during TIG welding of 304 stainless steel.
Core Technical Approach
The fundamental innovation lies in the dynamic mesh method (DMM) used to track the arc-liquid pool interface. Traditional numerical models for welding often treat the arc zone and the weld pool as separate domains with simplified boundary conditions. This approach, by contrast, establishes a pressure-based dynamic equilibrium at the interface, allowing the arc plasma to physically deform the liquid pool surface and vice versa.
Driving Forces in the Weld Pool
The authors validated four primary driving forces acting within the molten pool:
| Driving Force | Mechanism | Typical Magnitude Range |
|---|---|---|
| Electromagnetic force | Lorentz force from arc current in self-magnetic field | 10^3 to 10^4 N/m^3 |
| Thermal buoyancy | Density variation due to temperature gradient | 10^1 to 10^2 N/m^3 |
| Plasma drag force | Momentum transfer from arc plasma jet to pool surface | 10^2 to 10^3 N/m^3 |
| Marangoni force | Surface tension gradient driven by temperature variation | 10^2 to 10^4 N/m^3 |
Interface Deformation Results
The simulation captured the characteristic phenomenon where plasma impact creates a central depression (dimple) in the pool surface with raised edges. This dimple geometry is critical for understanding weld penetration depth, as it effectively channels molten metal downward, increasing weld depth-to-width ratio. The model predicts that the dimple depth is primarily governed by the balance between plasma dynamic pressure and surface tension forces.
Engineering Practice Integration
From a practical welding engineering perspective, this numerical approach has several important implications. First, the ability to predict pool geometry allows engineers to optimize welding parameters before physical trials, reducing development costs. Second, understanding the relative contributions of different driving forces enables targeted parameter adjustments: increasing current intensifies electromagnetic and plasma drag forces, while adjusting the arc length modifies plasma jet momentum. Third, the model provides a framework for predicting how alloying elements that affect surface tension (such as sulfur and oxygen in stainless steels) influence pool shape and weld geometry.
Application to Pipe Welding
In the context of pipe and fitting welding, particularly for stainless steel pipes used in petrochemical and power generation applications, this type of simulation can inform the selection of welding parameters for thin-walled pipes where excessive heat input risks distortion and burn-through. The predicted pool geometry helps in designing fixturing and back-gas protection strategies for circumferential welds on pipes with diameters ranging from 100 mm to 600 mm.
Key Insights and Reflections
The most significant contribution of this work is the demonstration that coupled arc-pool modeling with dynamic interface tracking yields substantially more accurate predictions than decoupled approaches. The predicted pool geometry, including the central dimple and raised edges, matches experimental observations from high-speed photography. However, the model still relies on certain simplifications: the arc plasma is modeled using a local thermodynamic equilibrium assumption, and the turbulence within the pool is captured through a simplified model rather than a full-scale LES or DNS approach.
For practitioners, the key takeaway is that the interplay between plasma momentum and surface tension forces governs weld geometry in ways that are difficult to predict empirically alone. This numerical framework provides a rational basis for optimizing TIG welding parameters, especially for dissimilar material joints and thin-walled components where precise control of heat input is essential. The methodology can be extended to analyze welding of pipe fittings such as elbows and tees, where complex geometries make empirical parameter optimization particularly challenging.