Temperature Field Numerical Simulation of YAG-MIG Hybrid Welding of Thick Aluminum Plates

Literature Overview

This paper by Jiang Youqing, Gu Lei, and Liu Jianhua from Huazhong University of Science and Technology, published in The International Journal of Welding (2006, Vol. 27, No. 6), presents a finite element temperature field simulation of YAG laser-MIG arc hybrid welding of 12 mm thick aluminum alloy plates. The authors employed a moving rotating Gaussian volume heat source model in ANSYS and validated the simulation against experimental weld results. The work extends the validation to 5 mm plate thickness as well, demonstrating the model's applicability across thickness ranges.

Core Technical Approach

The study addresses a well-recognized challenge in welding simulation: accurately modeling the heat source in hybrid welding processes. Unlike single-process welding, where the heat source is relatively simple (e.g., a point source for laser or a double-ellipsoid for arc), hybrid welding involves the superposition of a highly concentrated laser heat source and a broader arc heat source.

The authors adopted a rotating Gaussian volume heat source function, which accounts for:

The rotating nature of the model is significant: it represents the dynamic interaction between the laser beam and the arc, including the arc's tendency to deflect due to the laser-induced plasma plume and the resulting asymmetry in heat distribution.

Technical Analysis of the Heat Source Model

The rotating Gaussian volume heat source is defined by the superposition of two Gaussian distributions—one for the laser component (narrow, high intensity) and one for the arc component (broader, lower intensity)—with a rotation term that models the arc deflection. The general form can be expressed as:

q(r, θ) = (3√3 / (π^(3/2) · R_L · R_A)) · I_L · exp(-3r²/(R_L²)) + (3√3 / (π^(3/2) · R_A²)) · I_A · exp(-3r²/(R_A²))

where R_L and R_A are the characteristic radii of the laser and arc heat sources, and I_L and I_A are their respective energy fractions.

The inclusion of convection and radiation boundary conditions is essential for aluminum welding, given the high thermal conductivity and rapid surface oxidation of aluminum alloys. Aluminum's thermal conductivity (approximately 200–230 W/(m·K) for 6xxx-series alloys) means that heat dissipates rapidly from the weld zone, creating steep thermal gradients that influence weld geometry, residual stress, and microstructure.

Validation and Engineering Relevance

The authors validated the model by comparing simulated weld shapes and dimensions with actual welds produced under identical process conditions. The results showed good agreement for both 12 mm and 5 mm plate thicknesses. This dual-validation is important because the heat source model must account for different thermal diffusion regimes:

• 12 mm plates: Greater heat accumulation, deeper weld penetration, wider thermal affected zone.
• 5 mm plates: Rapid heat dissipation, shallower weld, more sensitive to process parameter changes.

For engineering practice, this validation means the model can be used to predict weld geometry for intermediate thicknesses (e.g., 6–10 mm) without requiring full-scale experiments, reducing development time and cost.

Integration with Engineering Practice

In my decades of experience with hybrid welding processes, the ability to predict weld geometry through simulation is invaluable for several reasons:

1. Process development: Before committing to production trials, engineers can simulate a range of parameters (laser power, arc current, travel speed, stand-off distance) to identify promising process windows.
2. Thickness range extension: A process qualified for 12 mm plates can be extrapolated to thinner sections with reasonable confidence, provided the heat source model is validated.
3. Defect prediction: By analyzing the thermal gradient and cooling rate distributions, engineers can predict regions susceptible to hot cracking, porosity, or incomplete fusion.

The specific application to thick aluminum plates is particularly relevant for aerospace structural components, shipbuilding, and heavy machinery, where aluminum alloy plates of 10–20 mm thickness are common.

Key Questions and Reflections

One limitation of this study, as I see it, is that the model focuses on the temperature field only. In modern welding simulation, coupled thermal-mechanical analysis is increasingly important, especially for aluminum alloys where residual stress and distortion are critical concerns. The temperature field data from this model could serve as input for subsequent mechanical analysis, but the authors do not extend the study to this level.

Another consideration is the accuracy of the heat source model under varying process conditions. The rotating Gaussian model was validated for specific parameter combinations, but its behavior under extreme conditions (e.g., very high laser power or very low travel speed) may deviate from the assumed functional form. In practice, I have observed that at very high power densities, the laser keyhole dynamics become complex, and simple Gaussian models may not capture the full thermal distribution.

The study's publication in 2006 also means that computational resources and software capabilities have advanced significantly since then. Modern implicit/explicit finite element methods with adaptive mesh refinement could achieve higher accuracy and faster convergence, making real-time simulation during welding feasible.

Study Insights and Implications

This paper represents an important step in the maturation of hybrid welding simulation. The validation across two plate thicknesses demonstrates the robustness of the rotating Gaussian heat source model and provides engineers with a reliable tool for process development.

The key insight for practitioners is that numerical simulation, when properly validated, can significantly reduce the cost and time of welding process development. However, simulation should complement, not replace, experimental validation. The model provides a first-principles-based prediction, but real welding involves numerous unmodeled phenomena (spatter, arc instability, material surface condition) that can cause deviations from simulated results.

For engineers working on thick aluminum alloy hybrid welding, I recommend using this type of simulation as a starting point for process parameter optimization, followed by targeted experimental verification of the most promising parameter combinations. The thermal field data should also be used to guide post-weld heat treatment specifications, as the peak temperature and cooling rate distributions directly determine the microstructural evolution and mechanical properties of the weld joint.