Numerical Analysis of Molten Pool Temperature and Flow Fields Considering Free Surface Deformation in Active TIG Welding

Literature Overview

This study by Huang Yong and colleagues from Lanzhou University of Technology and Tangshan Kaiyuan Automatic Welding Equipment Co. Ltd. presents a rigorous three-dimensional transient numerical model for active TIG (ATIG) welding of stainless steel, funded by the National Natural Science Foundation of China (Grant No. 51265029). The work, published in the Journal of Lanzhou University of Technology in 2015, addresses a critical gap in welding numerical simulation: the treatment of the molten pool free surface. Traditional models often assume a flat or fixed free surface, which oversimplifies the complex fluid dynamics occurring at the weld pool boundary.

Core Technical Approach

The authors employ the Volume of Fluid (VOF) method to track the free surface deformation of the molten pool, coupled with the enthalpy-porosity technique to handle the liquid-solid phase transformation. Three primary driving forces are incorporated into the governing equations: electromagnetic force (Lorentz force), surface tension force (Marangoni effect), and arc pressure. The model is built on a three-dimensional non-steady-state framework, which captures the transient nature of the welding process as the arc traverses the workpiece.

Key Findings and Technical Interpretation

The comparative analysis between models with and without free surface deformation reveals several critical insights:

1. Maximum temperature elevation: When the free surface deformation is accounted for, the maximum temperature in the molten pool increases. This occurs because the deformation of the pool surface alters the heat dissipation pathway. A deformed surface (typically concave or convex depending on the balance of forces) reduces the effective heat transfer area to the atmosphere, trapping more thermal energy within the pool.
2. Penetration depth reduction: The molten pool becomes shallower when free surface deformation is considered. The weakened convection phenomenon within the pool means that heat is less effectively transported to the deeper regions of the weld. This is counterintuitive from a practical standpoint, as engineers often rely on convection to achieve deep penetration.
3. Weld width increase: The weld becomes wider with the inclusion of free surface effects. The altered flow patterns redistribute molten metal laterally rather than vertically, promoting a broader but shallower weld profile.
4. Aspect ratio decrease: The combined effect of reduced penetration and increased width leads to a decreased depth-to-width ratio, which has direct implications for weld quality assessment and joint design.

Engineering Implications

From a practical engineering perspective, this study underscores the importance of free surface modeling in welding process optimization. The finding that convection is weakened when surface deformation is considered suggests that traditional flat-surface models may overpredict the convective heat transfer within the pool. This has several consequences:

• Process parameter selection: Welders and process engineers relying on flat-surface simulations may select arc currents or travel speeds that yield different results than predicted, particularly regarding penetration depth.
• Weld defect prediction: The altered flow patterns affect the entrainment of impurities and the formation of porosity. A shallower, wider pool may be more susceptible to lack of fusion at the root but less prone to undercut.
• ATIG-specific considerations: Active TIG welding introduces additional heat input through the introduction of gas or flux into the arc, which further modifies the surface tension gradient and electromagnetic force distribution. The free surface deformation becomes even more pronounced under these conditions.

Methodological Critique

The VOF method, while powerful for tracking sharp interfaces, introduces numerical diffusion that can smear the free surface over several grid cells. This is a known limitation of the method and may affect the accuracy of surface tension force calculations near the triple point (where solid, liquid, and gas meet). The authors do not explicitly discuss grid sensitivity or time step convergence, which would be critical for validating the numerical results.

The enthalpy-porosity approach for phase change is well-established but introduces a fictitious solid fraction in the mushy zone that affects the effective viscosity and thermal conductivity. The choice of the mushy zone constant (typically between 10^6 and 10^9) significantly influences the predicted flow patterns, and sensitivity to this parameter should be reported.

Study Insights and Reflections

This work represents a meaningful step toward more realistic welding pool simulations. The key takeaway is that the free surface is not merely a boundary condition but an active participant in the thermal-fluid coupling within the weld pool. For engineers involved in welding process development, particularly for applications requiring precise control of weld geometry such as pipe butt welding or overlay repair, incorporating free surface effects into numerical models can lead to more accurate predictions and better-informed process parameter selection. The finding that the deformed-surface model yields results closer to actual weld shapes validates the investment in more complex modeling approaches, even when computational costs are higher.