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

Literature Overview

This 2017 publication by Li Linmin, Li Baokuan, Liu Lichao, and Cao Xia from Northeastern University and Hohai University presents a comprehensive numerical simulation of TIG welding that integrates fluid flow dynamics, heat transfer analysis, and interface tracking using dynamic mesh methods. Funded by the National Natural Science Foundation of China (Grant 51574068), this research advances the computational modeling of welding processes to a level that enables detailed prediction of weld pool behavior, solidification patterns, and residual stress development.

Core Technical Points

TIG welding involves complex coupled phenomena that are difficult to observe experimentally due to the small scale, high temperatures, and rapid dynamics of the weld pool. Numerical simulation provides a powerful tool for understanding these phenomena and optimizing process parameters. The key challenges in TIG welding simulation include:

1. Electromagnetic effects: Arc plasma generates electromagnetic forces that drive weld pool convection
2. Multi-phase flow: Gas-liquid interface between the molten weld pool and shielding gas
3. Solid-liquid phase change: Stefan problem of moving solidification front
4. Moving heat source: Gaussian or double-ellipsoidal heat source moving at welding speed
5. Thermo-mechanical coupling: Temperature-dependent material properties and plastic deformation

Dynamic Mesh Methodology

The dynamic mesh (remeshing) approach is essential for accurately tracking the gas-liquid interface and solidification front during TIG welding. Unlike fixed-mesh methods that approximate interfaces using volume-of-fluid (VOF) or level-set methods, dynamic mesh methods physically move mesh nodes to follow the interface, providing sharp interface resolution without numerical diffusion.

Key Simulation Results

Weld Pool Geometry

The simulation reveals that TIG weld pool geometry is strongly influenced by the relative contributions of electromagnetic stirring and buoyancy-driven natural convection:

Temperature Distribution

The simulation shows that the maximum temperature in the weld pool reaches approximately 2200-2500°C for typical TIG parameters (150-200 A), with the thermal gradient decreasing rapidly with distance from the arc center. The solidification front temperature (liquidus) for carbon steel is approximately 1450-1500°C, and the simulation accurately captures the shape and position of this front.

Comparison with Experimental Data

The simulation results show excellent agreement with experimental measurements, validating the dynamic mesh approach for TIG welding analysis. The small deviations (typically 4-6%) are attributed to simplifications in the arc heat source model and material property assumptions.

Engineering Practice Integration

Process Optimization Using Simulation

Numerical simulation enables engineers to optimize TIG welding parameters before conducting expensive experimental trials:

1. Parameter selection: Determine optimal current, voltage, and travel speed for desired weld geometry
2. Distortion prediction: Estimate thermal distortion and residual stress for pre-weld fixture design
3. Microstructure prediction: Model cooling rates to predict grain size and phase distribution
4. Defect prevention: Identify conditions that promote porosity, cracking, or lack of fusion
5. Procedure development: Accelerate WPS qualification by narrowing the parameter window for testing

Application to Pressure Vessel Welding

For pressure vessel fabrication governed by ASME VIII or GB/T 150, numerical simulation provides:

Study Insights and Reflections

The most significant advancement represented by this research is the integration of dynamic mesh methods with full TIG welding physics. Previous simulations often used simplified fixed-mesh approaches that could not accurately capture interface dynamics, leading to poor predictions of weld pool geometry and solidification behavior. The dynamic mesh approach, while computationally more expensive, provides the accuracy necessary for reliable engineering predictions.

However, I note several limitations that engineers should consider when applying simulation results to practical welding:

1. Material property uncertainty: Temperature-dependent properties (thermal conductivity, specific heat, viscosity) vary significantly between material batches and grades
2. Arc model simplification: The Gaussian heat source model does not capture arc plasma dynamics, which affect electromagnetic force distribution
3. Scale limitations: Simulation of full-scale pressure vessel welds requires significant computational resources and may need model reduction
4. Validation requirements: Simulation results must always be validated against experimental data before being used for design decisions

For engineers in the cladding and bimetal industry, numerical simulation has particular value in optimizing overlay welding parameters where the interaction between base metal and overlay material creates complex thermal and metallurgical gradients. The ability to predict dilution, HAZ width, and residual stress distribution through simulation can significantly reduce the number of experimental trials needed to qualify overlay procedures.

This research demonstrates that computational welding science has matured to a level where simulation results can reliably support engineering decisions, provided that appropriate validation and uncertainty quantification are performed. The dynamic mesh methodology represents a significant step forward in accurately modeling the complex fluid dynamics and phase change phenomena inherent in arc welding processes, and its continued development will further bridge the gap between computational models and practical welding applications.