Numerical Simulation of Molten Pool Morphology Evolution in TIG Welding

Literature Overview and Context

This 2009 study by Wu Jianhua from Shenyang Vocational and Technical College presents a numerical simulation of the molten pool morphology evolution during TIG (Tungsten Inert Gas) welding. The work addresses the fundamental challenge of predicting and controlling the weld pool shape, which directly determines weld penetration, bead geometry, residual stresses, and mechanical properties. Numerical simulation of the welding process has become an indispensable tool for process optimization, defect prediction, and quality assurance, particularly for complex geometries where experimental investigation is impractical or prohibitively expensive. The study employs computational fluid dynamics (CFD) and heat transfer modeling to simulate the coupled phenomena of arc heat input, fluid flow in the molten pool, and solidification.

Core Technical Findings

Governing Equations and Model Assumptions

The simulation is based on the following governing equations:

• Mass conservation: ∂ρ/∂t + ∇·(ρu) = 0
• Momentum conservation: ρ(∂u/∂t + u·∇u) = -∇p + ∇·τ + ρg + F_body
• Energy conservation: ρCp(∂T/∂t + u·∇T) = ∇·(k∇T) + Q_arc + Q_source
• Solidification model: The enthalpy-temperature relationship with latent heat accounted for through the mushy zone model

Key assumptions include:

Key Modeling Parameters

Simulation Results and Validation

Molten Pool Shape Prediction

The simulation predicts the following molten pool characteristics for typical TIG welding parameters (200 A, 15 V, 300 mm/min):

• Pool depth: 2.5–3.5 mm for 6 mm thick steel plate
• Pool width: 8–12 mm at the surface
• Pool aspect ratio (width/depth): 3:1 to 4:1
• Maximum temperature: 2000–2200°C at the arc root
• Cooling rate at fusion boundary: 50–150°C/s

The simulation captures the characteristic teardrop shape of the molten pool, with the deepest penetration occurring slightly ahead of the arc center due to the direction of travel. The surface flow pattern shows a clockwise circulation (when viewed from the side) driven by the thermocapillary force, with the surface flowing away from the arc center and returning through the bulk of the pool.

Validation Against Experimental Data

The simulation results were validated against experimental measurements of weld bead geometry and cooling rates:

The deviations are within acceptable limits for engineering prediction, with the simulation slightly overestimating penetration due to the simplified arc heat source model.

Engineering Applications

Process Optimization

Numerical simulation enables the following engineering applications:

1. Parameter optimization: Identifying the optimal combination of current, travel speed, and torch angle for maximum penetration with minimum distortion.
2. Defect prediction: Predicting the likelihood of porosity, lack of fusion, and hot cracking based on the predicted cooling rate and solidification path.
3. Residual stress prediction: Estimating the residual stress distribution in the weld and HAZ, which is critical for fatigue life assessment.
4. Distortion prediction: Predicting the angular and longitudinal distortion during multi-pass welding, enabling fixture design and welding sequence optimization.
5. Microstructure prediction: Coupling the thermal simulation with a solidification model to predict grain size, phase fraction, and hardness distribution.

Limitations and Challenges

Despite its utility, numerical simulation of welding processes has inherent limitations:

• Arc modeling: The arc heat source is typically modeled as a Gaussian or double-ellipsoidal distribution, which does not capture the complex physics of the arc itself.
• Material properties: Temperature-dependent material properties are essential but difficult to obtain for all relevant temperatures, particularly in the liquid and mushy zones.
• Phase transformation: The transformation of austenite to martensite, bainite, or ferrite in the HAZ is difficult to model accurately without detailed metallurgical data.
• Computational cost: 3D simulations of multi-pass welding are computationally expensive, often requiring days of computation time for a single weld sequence.
• Boundary conditions: The convective and radiative heat loss from the weld surface is difficult to model accurately and can significantly affect the predicted cooling rate.

Study Insights and Reflections

The numerical simulation of TIG welding molten pool morphology represents a powerful tool for understanding and controlling the welding process. The ability to predict weld pool shape, cooling rates, and solidification behavior without physical experimentation has revolutionized welding process development and optimization. However, engineers must be aware of the limitations of simulation and use it as a complement to, not a replacement for, experimental validation. The simulation results provide insight into the underlying physics of the welding process, enabling engineers to make informed decisions about process parameters and quality control. The study's approach of coupling heat transfer, fluid flow, and solidification modeling is the standard methodology for welding simulation, and the validation against experimental data demonstrates the credibility of the approach. For engineers involved in cladding and bimetal fabrication, where the weld pool shape directly affects dilution, bond strength, and microstructural integrity, numerical simulation is an invaluable tool for process development and quality assurance. The key insight is that simulation provides a deeper understanding of the process, enabling engineers to move from trial-and-error to rational design of welding procedures.