Numerical Simulation of MIG Welding Molten Pool Formation and Solidification

Literature Overview

This paper by Wu Dongsheng et al., published in Welding (2015, No. 9, pp. 6-11), presents a three-dimensional numerical simulation of the MIG welding molten pool formation and solidification process for A36 low-carbon marine steel. Using the FLOW3D software with a double-ellipsoidal heat source model, the authors simulate the transient temperature field and flow field, analyzing both the molten pool formation after arc ignition and the solidification after arc extinction. This is a foundational study for understanding the physics of MIG welding and has direct implications for weld quality prediction and process optimization.

Simulation Methodology

The simulation approach is rigorous and comprehensive, incorporating multiple physical phenomena that govern weld pool dynamics:

The double-ellipsoidal heat source model is a significant improvement over simpler point or Gaussian models because it captures the asymmetric heat distribution in MIG welding, where the arc is concentrated in front of the weld pool and the heat input is distributed differently in the direction of travel.

Molten Pool Formation Process

The authors describe the molten pool formation in three stages:

Stage 1: Initial Heating and Pool Initiation

Upon arc ignition, the heat input rapidly raises the temperature of the surface layer of the base metal. The surface depresses due to the combined action of arc pressure and thermal expansion. The initial pool is shallow and narrow, with limited fluid motion.

Stage 2: Pool Development and Flow Establishment

As heating continues, the pool deepens and widens. The temperature gradient across the pool surface generates Marangoni convection, driving molten metal outward from the arc center toward the pool edges. Simultaneously, electromagnetic forces drive molten metal inward, creating a complex flow pattern with inward flow near the surface and outward flow at the surface.

Stage 3: Quasi-Steady State

After a transient period, the pool reaches a quasi-steady state where the geometry and flow pattern remain essentially constant. The penetration depth stabilizes, and the temperature distribution under the arc reaches a maximum that varies little with time.

Molten Pool Solidification Process

After arc extinction, the molten pool solidifies through a different physical regime. The authors identify several key features:

1. Flow Pattern Change: After arc extinction, the electromagnetic force and arc pressure disappear. The only remaining driving force is surface tension (Marangoni effect), which drives outward flow. The oscillation amplitude of the flow decreases over time as the pool cools and solidifies.
2. Solidification Front: The solidification front advances from the pool edges inward. The shape of the solidification front is determined by the temperature distribution and the solidification rate, which depends on the cooling rate.
3. Pool Stability: Despite the presence of liquid metal oscillations, the pool remains stable because the large external contact angle (greater than π/2) provides a geometric constraint that prevents pool collapse or overflow.
4. Final Weld Geometry: The simulation-predicted weld dimensions and shape agree well with experimental measurements, validating the model's accuracy.

Engineering Implications

The simulation results have several direct implications for welding practice:

Heat Input Control

The double-ellipsoidal model reveals that heat input is not uniformly distributed. The front of the weld pool receives more intense heating, which explains the asymmetric weld geometry commonly observed in MIG welding. Process engineers should account for this asymmetry when setting travel speed and current parameters.

Cooling Rate and Microstructure

The cooling rate, determined by the thermal gradient at the solidification front, directly influences the weld microstructure. Higher cooling rates promote fine-grained acicular ferrite formation, while lower cooling rates favor coarse polygonal ferrite. The simulation provides a quantitative tool for predicting cooling rates under different process conditions.

Residual Stress Prediction

The flow pattern and solidification sequence predicted by the simulation can be used to estimate residual stress distributions. Inward flow during solidification creates tensile stresses in the center of the weld, while outward flow creates compressive stresses at the surface. These residual stresses are critical for fatigue performance and stress corrosion cracking resistance.

Process Optimization

The simulation can be used to optimize welding parameters for specific applications:

Model Validation and Limitations

The authors validate the model by comparing predicted weld dimensions with experimental measurements, reporting good agreement. However, several limitations should be noted:

1. Thermophysical Property Variations: The simulation assumes constant or simplified temperature-dependent properties. In reality, properties such as viscosity, thermal conductivity, and surface tension vary significantly with temperature and composition.
2. Droplet Transfer Modeling: The dynamic impact of droplets is included but may not capture all droplet transfer modes (globular, spray, pulsing). Different transfer modes produce different flow patterns and weld geometries.
3. Shielding Gas Effects: The simulation does not model the shielding gas flow field, which affects arc stability, heat transfer, and weld pool surface conditions.
4. Solidification Modeling: The simulation focuses on the liquid phase. Solidification modeling (including dendritic growth, solute partitioning, and shrinkage porosity) would provide additional insight into weld quality.
5. Multi-Pass Welding: The simulation is for a single pass. Multi-pass welding involves complex interactions between passes, including reheating of previously deposited metal and changes in residual stress state.

Study Insights and Implications

This paper represents a significant contribution to welding simulation methodology. The comprehensive inclusion of multiple physical phenomena in a single model provides a realistic representation of the MIG welding process. The agreement between simulation and experiment validates the approach and gives confidence in using such models for process optimization.

For engineers involved in pipeline and structural welding, the simulation approach offers a powerful tool for:

• Predicting weld geometry before physical trials, reducing the number of coupon tests required.
• Optimizing process parameters for specific joint configurations.
• Understanding the physical mechanisms behind weld defects.
• Developing new welding processes and filler metals through virtual experimentation.

However, simulation should complement, not replace, physical testing. The complexity of real welding processes, with their numerous interacting variables, means that simulation results must be validated through experimental verification. The most effective approach combines simulation-driven parameter optimization with targeted experimental validation.

The large external contact angle observed in the simulation is particularly interesting. It suggests that the weld pool geometry is governed more by surface tension than by gravity, which has implications for welding position flexibility. In thin-plate welding, where gravity effects are less significant, the surface tension-dominated pool geometry may be more stable than commonly assumed.

Concluding Remarks

This simulation study provides a valuable framework for understanding MIG welding physics and optimizing process parameters. The double-ellipsoidal heat source model, combined with comprehensive flow force modeling, captures the essential physics of weld pool formation and solidification. Engineers should leverage such simulation tools as part of a comprehensive welding procedure development process, combining virtual optimization with physical validation to achieve consistent, high-quality welds in production applications.