TIG Welding Flow Heat Transfer and Interface Tracking Dynamic Mesh Numerical Simulation

Literature Overview

This paper, published by Li Linmin and colleagues from Northeastern University in the Journal of Northeastern University (Natural Science) in 2017, presents a magnetohydrodynamic (MHD) coupled numerical model for TIG welding that incorporates dynamic mesh technology to track the arc-weld pool interface. Funded by the National Natural Science Foundation of China (Grant 51574068), the work addresses a long-standing challenge in computational welding mechanics: the accurate prediction of weld pool geometry and the interaction between arc plasma and molten metal. The authors first validated the model by computing free-burning arc characteristics, then systematically verified four driving forces within the weld pool, and finally applied the fully coupled model to simulate 304 stainless steel TIG welding.

Core Technical Content

The mathematical framework couples four major physical phenomena: electromagnetic field, fluid flow, heat transfer, and phase change (melting and solidification). The arc plasma is modeled using a self-consistent MHD approach that solves for velocity, temperature, pressure, current density, and magnetic field simultaneously. The key innovation lies in the dynamic mesh method (DMM), which tracks the arc-weld pool interface based on a pressure dynamic equilibrium criterion rather than a fixed geometric assumption.

The four driving forces verified in the weld pool are presented below:

Interface Tracking Methodology

The dynamic mesh approach represents a significant departure from conventional weld pool models that assume a fixed paraboloid or ellipsoidal geometry. The pressure-based dynamic equilibrium criterion determines the arc-weld pool boundary at each computational time step. As the arc plasma impinges on the weld pool surface, it creates a localized depression at the center. The model captures the characteristic topography where the pool center dips downward due to electromagnetic and plasma drag forces, while the pool edges rise upward due to Marangoni-driven surface flow and thermal buoyancy.

The numerical results demonstrate that the predicted pool geometry closely matches experimental observations for 304 stainless steel TIG welding. The central depression depth and the rim elevation are quantitatively consistent with high-speed imaging data reported in the literature. This validates the physical fidelity of the coupled MHD-DMM approach.

Process Parameters and Validation

The free-burning arc computation yielded accurate arc zone velocity, temperature, and pressure distributions. The arc root current density, arc pressure distribution, and arc temperature profile all agreed well with established analytical solutions and experimental measurements. This foundational validation is critical because errors in the arc model propagate directly into the weld pool predictions.

Engineering Implications

For engineering practice, this numerical model provides several valuable insights. First, it enables the prediction of weld pool geometry under varying process parameters without requiring expensive experimental campaigns. Second, the explicit coupling of arc plasma and weld pool physics allows engineers to understand how changes in arc characteristics (such as arc length or electrode configuration) affect weld pool shape and, consequently, weld bead geometry. Third, the model can be extended to predict defects such as lack of fusion, undercut, and excessive penetration by analyzing the pool geometry under different welding conditions.

The dynamic mesh method is particularly useful for thin-plate welding applications where the weld pool geometry changes rapidly with process parameters. In pipe welding scenarios, where the geometry is three-dimensional and the pool shape varies along the weld length, this approach could be extended to predict weld quality along circumferential and longitudinal welds.

Key Insights and Reflections

The most significant contribution of this work is the rigorous coupling of arc plasma physics with weld pool fluid dynamics through a physically motivated interface tracking method. Traditional models often treat the arc-weld pool boundary as a fixed geometric surface or use simplified heat source models (such as double-ellipsoidal or Gaussian distributions) that do not capture the true momentum and thermal interaction. The pressure-based dynamic equilibrium criterion provides a physically grounded alternative that does not require empirical calibration of the boundary condition.

However, the model still has limitations that practitioners should be aware of. The two-dimensional axisymmetric assumption restricts applicability to stationary arc conditions and does not account for welding travel speed effects on pool asymmetry. The treatment of surface tension as a function of temperature, while physically sound, requires accurate knowledge of surface contamination levels, which can vary significantly in industrial environments. Additionally, the computational cost of the fully coupled MHD-DMM model is substantially higher than simplified heat source models, which may limit its use for real-time process monitoring.

Reference Value and Outlook

This work establishes a rigorous computational framework that bridges the gap between arc plasma physics and weld pool metallurgy. For pipe and fitting manufacturers, the model can be adapted to predict weld pool behavior in different joint configurations, including tube-to-tubesheet joints, pipe-to-flange joints, and circumferential butt welds. Future extensions should incorporate three-dimensional geometry, travel speed effects, and multi-pass welding scenarios. The integration of this model with metallurgical transformation models would further enhance its predictive capability for microstructure and residual stress prediction.