Zhuojin Pipe Fitting Co., LtdFinite Element Simulation of Residual Stress Field in A6061 Aluminum Alloy Pulsed MIG Welded T-Joints
Literature Overview
The paper by He Qi, Li Shichun, Gu Jinliang, Xiao Gang, and Huang Hao, published in Materials in Mechanical Engineering (2023, Vol. 47, No. 9, pp. 70–75), presents a finite element simulation study of the temperature field and residual stress field in A6061 aluminum alloy T-joints welded using pulsed MIG welding. Supported by multiple funding sources including the National Natural Science Foundation of China (52075159) and the Hunan Provincial Natural Science Foundation (2022JJ30019), this research introduces two key methodological innovations: a B-spline-based weld bead model and a composite heat source model consisting of a Gaussian surface source and a conical volume source. The study validates the simulation model against experimental results and demonstrates significant improvements in prediction accuracy compared to conventional simplified models.
Core Technical Points
B-Spline Weld Bead Model
The traditional approach to modeling weld beads in finite element analysis uses simplified geometric shapes such as rectangles, ellipsoids, or prisms. These simplified models fail to capture the true three-dimensional geometry of the weld bead, which has a complex nonlinear surface profile with varying reinforcement height, penetration depth, and width. The B-spline approach proposed in this paper addresses this limitation by using non-uniform rational B-splines (NURBS) to fit the actual weld bead surface. B-splines are particularly well-suited for this purpose because they can represent smooth, complex curves with a small number of control points, and they provide local control over the shape of the curve.
| Modeling Approach | Geometric Accuracy | Computational Complexity | Prediction Accuracy |
|---|---|---|---|
| Rectangular weld bead | Low | Low | Poor |
| Ellipsoidal weld bead | Moderate | Moderate | Fair |
| Double-ellipsoid heat source | Moderate | Moderate | Fair |
| B-spline weld bead model | High | Higher | Excellent |
The B-spline model allows the simulation to capture the true shape of the weld bead, including the asymmetric profile that results from the welding process. In T-joint welding, the weld bead geometry is inherently asymmetric due to the geometry of the joint, and this asymmetry has a significant influence on the distribution of residual stresses. By accurately representing the weld bead geometry, the B-spline model enables more realistic predictions of stress concentrations and distortion patterns.
Composite Heat Source Model
The second key innovation is the composite heat source model that combines a Gaussian surface heat source with a conical volume heat source. This approach is based on the understanding that pulsed MIG welding delivers energy through two distinct mechanisms: the arc, which primarily heats the surface of the workpiece, and the molten metal transfer, which deposits energy into the volume of the weld pool. The Gaussian surface source represents the arc heat input, which is concentrated at the point of arc attachment on the workpiece surface. The conical volume source represents the energy deposited by the transferred metal droplets, which extends into the weld pool volume.
The heat source model is parameterized based on the base and peak current characteristics of pulsed MIG welding. During the base current phase, the heat input is relatively low and primarily comes from the arc, which is well-represented by the Gaussian surface source. During the peak current phase, the heat input increases significantly, and the energy from the transferred metal droplets becomes more important, which is captured by the conical volume source. The amplitude and distribution of the two sources are adjusted to reflect the temporal variation in heat input during the pulse cycle.
| Heat Source Component | Type | Represents | Key Parameters |
|---|---|---|---|
| Gaussian surface source | Surface | Arc heat input | Peak intensity, radius, position |
| Conical volume source | Volume | Metal transfer energy | Cone angle, length, energy density |
Simulation Results and Validation
The simulation results demonstrate excellent agreement with experimental measurements for the weld pool geometry. The relative error between simulated and experimental penetration depth and weld width is no greater than 1.4%, which is remarkably accurate for finite element welding simulations. This level of accuracy validates the combined use of the B-spline weld bead model and the composite heat source model for predicting weld pool geometry.
The residual stress predictions, however, show larger discrepancies between simulation and experiment. At a distance of 10 mm from the weld center, the relative error in longitudinal residual stress is 28.0%, and at 30 mm, the error is 20.6%. While these errors are significant, the paper notes that the accuracy is at least 15.7% better than what would be achieved using simplified weld bead and double-ellipsoid heat source models. This improvement is meaningful, as it demonstrates that the more sophisticated modeling approach yields substantially better predictions of residual stress distributions.
| Validation Metric | Simulated Value | Experimental Value | Relative Error |
|---|---|---|---|
| Penetration depth | Within 1.4% of experiment | Measured by macrograph | ≤ 1.4% |
| Weld width | Within 1.4% of experiment | Measured by macrograph | ≤ 1.4% |
| Peak temperature at key points | Within 1.4% of experiment | Thermocouple measurement | ≤ 1.4% |
| Longitudinal residual stress at 10 mm | Simulated | Measured by XRD | 28.0% |
| Longitudinal residual stress at 30 mm | Simulated | Measured by XRD | 20.6% |
Engineering Practice Integration
Residual stress prediction is critical in engineering practice because residual stresses can significantly affect the fatigue life, stress corrosion cracking resistance, and dimensional stability of welded structures. In aluminum alloy structures, which are increasingly used in transportation and aerospace applications for weight reduction, residual stresses are particularly concerning because aluminum alloys are susceptible to stress corrosion cracking in certain environments.
For A6061 aluminum alloy T-joints, which are common structural configurations in vehicle frames, aircraft structures, and industrial equipment, the residual stress distribution directly influences the joint's resistance to fatigue failure. The T-joint geometry creates stress concentrations at the intersection of the two plates, and the residual stresses from welding can either exacerbate or mitigate these stress concentrations depending on their magnitude and direction. Tensile residual stresses at the weld toe are particularly detrimental to fatigue life, and their accurate prediction is essential for life assessment and design optimization.
The simulation results have direct implications for post-weld treatment strategies. If the simulation can accurately predict the residual stress distribution, engineers can design targeted stress relief treatments such as thermal stress relief, vibration stress relief, or shot peening to specifically address the most critical stress concentrations. The B-spline and composite heat source model provides the accuracy needed to make these engineering decisions with confidence.
Study Insights and Reflections
The residual stress prediction errors of 20–28% are not negligible, and they highlight the inherent challenges of welding residual stress simulation. The accuracy of residual stress predictions depends not only on the heat source model and weld bead geometry but also on the constitutive model used to describe the material behavior, the boundary conditions, and the cooling rate assumptions. Aluminum alloys exhibit complex thermomechanical behavior, including significant plastic deformation during welding, precipitation hardening and softening effects, and thermal cracking susceptibility, all of which influence the final residual stress state.
The improvement of at least 15.7% in accuracy over simplified models is a meaningful result, but it also suggests that there is still significant room for improvement. Future work could explore the use of more sophisticated constitutive models that account for the precipitation behavior of A6061 aluminum alloy, the effects of welding sequence on residual stress accumulation, and the interaction between residual stresses and residual distortions. The integration of data analysis techniques with finite element simulations could also provide a path to further accuracy improvements, although this is beyond the scope of the current paper.
The B-spline approach to weld bead modeling is particularly valuable because it can be easily adapted to different weld geometries and joint configurations. Unlike simplified geometric models that require manual parameterization for each new weld shape, B-spline models can be generated directly from experimental measurements of the actual weld bead geometry. This makes the approach highly practical for production welding applications where weld bead geometry can vary due to process parameter changes, joint fit-up variations, and welding position effects.
Reference Value and Outlook
This paper makes a significant contribution to the field of welding simulation by demonstrating that the combination of B-spline weld bead modeling and composite heat source modeling can substantially improve the accuracy of temperature field and residual stress predictions in aluminum alloy welded joints. The validation results, while not perfect, provide a solid foundation for engineering applications where residual stress prediction is critical for design, life assessment, and post-weld treatment planning. For engineers working on aluminum alloy structures in transportation, aerospace, and industrial applications, this research provides a practical and reliable simulation methodology that can be adapted to specific welding processes and joint configurations. The continued development of welding simulation technology, including the integration of more sophisticated material models and the use of advanced computational methods, will further improve prediction accuracy and expand the range of engineering problems that can be addressed through simulation.