CLADDING TECHNOLOGY SHANXI CO., LTDNumerical Simulation of Transient Molten Pool Three-Dimensional Morphology in TIG Welding Based on FLUENT
Literature Overview
This 2009 study from Nanchang University, supported by the National Natural Science Foundation of China and the Jiangxi Provincial Science and Technology Department, presents a comprehensive three-dimensional numerical simulation of the transient molten pool morphology during TIG welding using the FLUENT computational fluid dynamics (CFD) software. The accurate prediction of molten pool geometry is fundamental to understanding weld formation, predicting weld bead dimensions, and ultimately optimizing welding processes for various applications including cladding and overlay welding.
Core Technical Points
The transient molten pool in TIG welding is a complex multiphysics system involving heat transfer, fluid flow, mass transport, and phase change phenomena. The study employs a coupled heat-mass transfer model that solves the Navier-Stokes equations for fluid flow, the energy equation for heat transfer, and the solidification equation for phase change. The molten pool is modeled as a two-phase system with the liquid phase governed by full fluid dynamics and the solid phase treated as a stationary domain with heat conduction only.
Numerical Model Parameters
| Parameter | Value / Range | Description |
|---|---|---|
| Domain Size | 40×20×20 mm | Computational domain |
| Mesh Size | 0.2×0.2×0.2 mm | Uniform mesh |
| Time Step | 0.001-0.01 s | Adaptive time stepping |
| Welding Current | 100-200 A | Arc power input |
| Travel Speed | 5-15 cm/min | Welding speed |
| Arc Length | 3-5 mm | Electrode gap |
| Shielding Gas | Argon | Inert shielding |
| Heat Source Model | Double-ellipse (Goldak) | Surface heat flux |
Heat Source Modeling
The accurate representation of the arc heat source is critical to the fidelity of molten pool simulations. The study employs the Goldak double-ellipse heat source model, which characterizes the non-uniform heat distribution of the TIG arc as two overlapping elliptical distributions - one for the leading edge (keyhole zone) and one for the trailing edge (weld pool zone). The model is defined by:
q(r,θ) = Q f(r) g(θ)
where Q is total heat input, f(r) is the radial distribution function, and g(θ) is the angular distribution function. The key parameters are the front and rear heat source coefficients (a_f, a_r), the front and rear ellipse coefficients (b_f, b_r), and the front and rear ellipse lengths (c_f, c_r).
The study demonstrates that the double-ellipse model provides significantly better agreement with experimental observations than simpler models such as the Gaussian or uniform disk models, particularly in predicting the asymmetric pool shape that characterizes TIG welding. The front-to-rear heat distribution ratio was found to be approximately 0.6:0.4 for typical TIG parameters, meaning 60% of the heat is concentrated ahead of the arc center.
Molten Pool Morphology Analysis
The simulation results reveal several important features of the transient molten pool morphology. First, the pool exhibits a characteristic teardrop shape with a narrower leading edge and a broader trailing edge, consistent with experimental observations from sectioning studies. Second, the maximum pool depth is located slightly ahead of the arc center, reflecting the combined effects of arc pressure and surface tension-driven flow. Third, the pool width at the surface is approximately 1.5-2.5 times the pool width at the maximum depth, creating a characteristic hourglass cross-section.
| Pool Parameter | Simulation Result | Experimental Comparison | Deviation |
|---|---|---|---|
| Pool Depth (mm) | 2.5-3.5 | 2.8-3.8 | -5% to +8% |
| Pool Width (mm) | 6-8 | 6.5-8.5 | -6% to +5% |
| Pool Length (mm) | 8-12 | 9-13 | -8% to +5% |
| Maximum Depth Location (mm) | 1.5-2.0 ahead of center | 1.8-2.2 ahead of center | -10% to +5% |
Flow Patterns and Solidification Behavior
The internal flow patterns within the molten pool are governed by the competing effects of arc pressure (driving flow downward and backward), surface tension gradients (driving flow from hot to cold regions), and buoyancy forces (driving flow from hot to cold regions due to density differences). The study identifies a primary vortex that drives liquid from the leading edge downward and rearward, with a secondary recirculation zone near the pool surface. The solidification front advances from the pool boundaries inward, with the solidification rate varying significantly across the pool cross-section.
The solidification rate at the pool center is found to be approximately 0.5-2.0 mm/s, while at the pool boundaries it increases to 5-15 mm/s due to the steeper temperature gradients. This variation in solidification rate directly influences the microstructure, with coarser dendritic structures at the pool center and finer structures at the boundaries. The thermal gradient at the solidification front ranges from 10^3 to 10^5 K/m, creating conditions favorable for dendritic solidification.
Engineering Practice Implications
For engineers involved in cladding and overlay welding, the molten pool simulation results have direct practical significance. The pool geometry determines the dilution rate between the overlay material and the base metal, which is critical in bimetal fabrication. A deeper pool results in higher dilution, potentially compromising the corrosion resistance or wear resistance of the overlay layer. The simulation provides a tool for predicting dilution rates before actual welding trials, reducing development time and material costs.
The predicted flow patterns also have implications for defect formation. The strong downward flow at the leading edge can entrain gas from the pool surface, increasing the risk of porosity. The recirculation patterns can cause non-uniform mixing of alloying elements, leading to compositional segregation. Understanding these flow patterns enables engineers to develop countermeasures such as optimized shielding gas coverage, controlled welding speeds, and appropriate filler wire placement.
Study Insights and Reflections
This study demonstrates the power of CFD simulation as a tool for understanding and predicting welding phenomena, though several limitations should be acknowledged. The numerical model assumes idealized boundary conditions and material properties that may not fully capture the complexity of real welding conditions. The arc-heat transfer model, while more sophisticated than earlier approaches, still relies on empirical parameters that require calibration against experimental data. For engineering applications, the simulation results should be validated against actual weld cross-section measurements and used as a guide rather than a definitive prediction tool. The methodology presented here is directly transferable to cladding process development, where understanding the interaction between the overlay material and the base metal through the lens of molten pool dynamics is essential for achieving the desired bimetal properties.