Three-Dimensional Dynamic Simulation of Temperature Field in Tungsten Inert Gas Welding and Infrared Thermometry Verification

Overview of the Study

This paper by Lei Yucheng and colleagues from Jiangsu University addresses a fundamental challenge in welding engineering: accurately describing the rapid heating and cooling temperature field during welding. The authors established a three-dimensional dynamic finite element model for the moving arc temperature field, using ANSYS with APDL programming to simulate the welding process of aluminum alloy via GTAW. The double-elliptical heat source model was employed, incorporating temperature-dependent thermophysical properties, radiation, convection, and phase transformation effects. Experimental validation was performed using the SAT-HY6800 infrared thermometer with a filter to exclude arc radiation interference. The results showed good agreement between simulated and measured temperatures, confirming the validity of the approach.

Core Technical Methodology

The heart of this work lies in the construction of a dynamic finite element model that tracks the moving arc. Unlike steady-state approaches, this method captures the transient thermal history of the weld zone, which is critical for predicting phase transformations and post-weld mechanical properties. The double-elliptical heat source model, originally proposed by Goldak, is well-suited for GTAW because it accounts for the asymmetry of heat distribution between the front and rear of the arc.

The APDL program automates the process of mesh updating as the arc moves along the weld path, which is a computationally intensive but necessary step for three-dimensional dynamic simulation. The inclusion of phase transformation latent heat is particularly important for aluminum alloys, where the relatively low melting point and significant specific heat variation with temperature demand careful numerical treatment.

Interpretation of Key Technical Points

The decision to use a double-elliptical heat source rather than a simple Gaussian or point heat source reflects a sophisticated understanding of the GTAW process. In GTAW, the arc column concentrates heat in a narrow region, and the distribution is not symmetric about the arc axis due to the direction of travel. The front half of the arc deposits more energy into the material ahead of the arc, while the rear half has a longer residence time. The Goldak model captures this by using different elliptical geometries for the front and rear regions.

The consideration of temperature-dependent thermophysical properties is not merely an academic refinement but a practical necessity. Aluminum alloys exhibit significant changes in thermal conductivity, specific heat, and density over the temperature range from room temperature to the melting point and beyond. Using constant properties would introduce errors that could propagate into inaccurate predictions of solidification behavior, residual stress, and distortion.

The experimental validation approach using infrared thermometry with an arc filter is noteworthy. Infrared thermometry is the most practical non-contact method for measuring weld surface temperatures, but the intense optical radiation from the welding arc can saturate or distort the sensor reading. The use of a filter to block the arc radiation while allowing thermal radiation from the workpiece to pass is a practical engineering solution that many practitioners overlook.

Engineering Practice Implications

From the perspective of pipe welding and pipe fitting fabrication, the accurate prediction of temperature fields has direct consequences for weld quality and component integrity. In the context of API 5L line pipe welding or ASME B31.3 process piping, the thermal cycle experienced by the heat-affected zone determines the microstructural evolution, which in turn governs the susceptibility to cracking, the residual stress distribution, and the long-term service performance.

For aluminum alloy pipe welding, which is common in cryogenic applications and aerospace structures, the thermal simulation approach described in this paper can be directly applied to optimize welding parameters, predict distortion, and select appropriate post-weld heat treatment conditions. The validated model can serve as a baseline for developing welding procedure specifications (WPS) that minimize defects such as hot cracking and porosity.

The methodology also has implications for welding quality control. By comparing simulated thermal cycles with those predicted from standard procedures, engineers can identify potential problem areas before actual welding takes place. This proactive approach aligns with the preventive philosophy of modern quality management systems and can reduce costly rework in production environments.

Study Insights and Reflections

Reading this paper, I am struck by the elegance of combining computational modeling with experimental validation in a way that is both rigorous and practical. The authors did not simply present simulation results but took the critical step of verifying them against measured data, which is the hathe writing systemark of responsible engineering research. The use of an APDL program to automate the dynamic simulation process demonstrates foresight, as it makes the model reproducible and adaptable to different welding conditions.

One area that could benefit from further exploration is the extension of this approach to multi-layer multi-pass welding, which is common in thick-walled pipe fabrication. The thermal history of subsequent passes is influenced by the residual temperature from previous passes, and a fully dynamic model can capture this coupling. Additionally, the incorporation of mechanical boundary conditions to predict residual stress and distortion would make the model even more valuable for engineering applications.

The paper also raises questions about the accuracy of the double-elliptical heat source model for different welding positions and conditions. In pipe welding, where the joint geometry is curved and the weld position changes around the circumference, the heat source distribution may deviate from the planar assumptions inherent in the Goldak model. Future work could investigate modifications to the heat source model for cylindrical geometries.

In conclusion, this study provides a solid foundation for thermal simulation of GTAW processes and demonstrates the value of integrating numerical methods with experimental measurements. For practitioners in pipe and fitting welding, the approach offers a pathway to more rational and efficient welding procedure development, ultimately contributing to improved weld quality and reduced production costs.