Thermal-Mechanical Coupled Simulation and Parameter Optimization of Narrow-Gap TIG Oscillating Welding

Literature Overview

This study by Gao Hui, Qu Jiajun, and Li Xianhui from Beijing Institute of Petrochemical Technology addresses a persistent challenge in narrow-gap welding: undercut defects on the sidewalls during TIG oscillating single-pass multi-layer welding. The authors employed a sequential thermal-mechanical coupled finite element analysis in Abaqus combined with Response Surface Methodology (RSM) to establish a multi-parameter synergistic control strategy. The work was published in Ordnance Materials Science and Engineering (2026, Vol. 49, No. 2, pp. 147-152) and carries significant practical relevance for thick-wall pipe and fitting fabrication where narrow-gap welding offers substantial productivity advantages.

Core Technical Approach

The research methodology integrates computational simulation with experimental validation through a well-structured workflow. The authors developed a dynamic oscillating heat source model based on a double-ellipsoid heat source formulation, which captures both the thermal input distribution and the lateral movement of the arc during oscillation. This model was implemented within the Abaqus finite element framework using a sequential coupled approach, where the thermal analysis results serve as boundary conditions for the subsequent mechanical analysis. The thermal field solution provides temperature distributions across the weld cross-section, while the mechanical analysis yields residual stress and deformation patterns that directly correlate with undercut formation mechanisms.

The key process parameters investigated include oscillation width, welding speed, and sidewall dwell time. These three variables represent the primary levers available to the welding engineer when controlling sidewall fusion quality in narrow-gap configurations. The selection of these parameters reflects a deep understanding of the physics governing sidewall wetting: oscillation width determines the lateral coverage of the heat input, welding speed governs the energy density deposited per unit length, and sidewall dwell time controls the duration of heat exposure at the critical fusion boundary.

Process Parameter Optimization

The authors employed a Box-Behnken experimental design to systematically screen optimal parameter combinations. This design choice is particularly appropriate for the problem at hand because it efficiently explores the interaction effects among three factors while requiring fewer experimental runs than a full factorial design. The response surface model constructed from the experimental data enables prediction of undercut susceptibility across the entire parameter space.

The optimal parameter combination identified through this optimization is summarized below:

These parameters represent a balanced operating window where the thermal input is sufficient to achieve complete sidewall fusion without causing excessive weld pool depression or burn-through. The sidewall dwell time of 2 seconds is particularly critical, as it ensures that the arc remains at the sidewall long enough for the base metal to reach the melting point and achieve proper wetting before the oscillation cycle moves the arc away.

Experimental Validation and Results

The optimization results were validated through actual single-pass multi-layer welding experiments. The weld cross-sections were examined macroscopically, and hardness profiles were measured across the weld zone. The experimental results confirmed that under the optimal parameters, undercut was effectively eliminated, each weld layer fused well with the base metal, and weld hardness uniformity was significantly improved.

From a practical engineering perspective, the improvement in hardness uniformity is noteworthy. In narrow-gap welding of thick-walled pipes, non-uniform hardness distributions can lead to stress concentrations at layer boundaries, which may initiate fatigue cracks under cyclic loading. The optimized parameters produce a more homogeneous microstructure across the weld cross-section, which translates to better fatigue resistance in service.

Engineering Practice Implications

This research has direct applicability to the fabrication of large-diameter pipes and fittings, particularly those requiring narrow-gap welding for economic efficiency. In the pipeline industry, narrow-gap TIG welding is increasingly adopted for pipe bodies with wall thicknesses exceeding 25 mm, where traditional multi-pass welding becomes prohibitively expensive. The ability to predict and prevent undercut through parameter optimization is essential for ensuring code compliance, as undercut is classified as a discontinuity that can reduce the effective load-bearing cross-section and serve as a crack initiation site.

For pipe fitting manufacturing, where complex geometries such as elbows, tees, and reducers require multi-directional welding, the principles established in this study can be adapted to various weld orientations. The sidewall dwell time concept, in particular, is transferable to any configuration where sidewall fusion quality is critical. Engineers should note that the optimal parameters identified here are specific to the material and joint configuration studied, and adaptation to other materials or geometries requires recalibration of the response surface model.

Key Reflections

The study exemplifies the power of integrating finite element simulation with statistical experimental design for welding process optimization. Rather than relying solely on trial-and-error experimentation, which is time-consuming and costly, the authors used simulation to understand the underlying physics and RSM to efficiently navigate the parameter space. This approach is particularly valuable for narrow-gap welding, where the process window is narrow and the consequences of parameter misselection are severe.

One area for further investigation would be the extension of this methodology to automated narrow-gap welding systems with real-time monitoring. The ability to adjust oscillation parameters in response to measured thermal signals could provide additional robustness against variations in material properties, joint fit-up, and ambient conditions. The current study establishes the theoretical foundation for such adaptive control systems, but the implementation of closed-loop feedback remains an open challenge for future work.