CLADDING TECHNOLOGY SHANXI CO., LTDNumerical Simulation of Dynamic Thermal Process During Double Sided Asymmetrical TIG Backing Welding of Large Thick Plates
Literature Overview
Published in China Welding in 2011, this paper by Liu Dianbao, Zhao Huihui, Li Fuquan, Zhang Guangjun, and Wu Lin from the National Key Laboratory of Science and Technology on Precision Heat Processing of Metals at Harbin Institute of Technology presents a computational approach to understanding the complex thermal dynamics of double-sided asymmetrical TIG welding on large thick plates. The research addresses a practical challenge in heavy equipment manufacturing where thick-section components require backside weld reinforcement without access to both sides simultaneously.
Core Technical Content
Double-sided asymmetrical TIG backing welding involves welding from one side while providing thermal backing from the opposite side, creating an asymmetric thermal field that governs residual stress distribution, distortion, and weld quality. The numerical simulation employs a three-dimensional finite element model with coupled thermal-mechanical analysis to capture the dynamic evolution of temperature fields during multi-pass welding.
Simulation Parameters and Model Configuration
| Model Parameter | Value | Justification |
|---|---|---|
| Mesh element size | 2–4 mm in weld zone, 10–20 mm in far field | Balances accuracy and computational cost |
| Thermal source model | Double-ellipsoidal (Goldak) | Captures asymmetry between leading and trailing edges |
| Temperature-dependent properties | Full range 20–2000 °C | Essential for accurate phase transformation prediction |
| Boundary conditions | Convective cooling with variable h(T) | Accounts for air convection and radiation losses |
| Time step | 0.1–0.5 s adaptive | Resolves rapid thermal transients during arc movement |
Key Thermal Findings
The simulation reveals several critical insights:
- Temperature gradient asymmetry: The front-side temperature gradient reaches 15–25 K/mm, while the back-side gradient is reduced to 5–12 K/mm due to thermal mass effects. This asymmetry directly influences solidification microstructure differences between the cap and root welds.
- Peak temperature distribution: The maximum temperature in the weld pool reaches 1850–2100 °C, with the thermal influence zone (above 200 °C) extending 30–50 mm from the weld centerline for plates exceeding 100 mm thickness.
- Cooling rate variation: The cooling rate from 800 °C to 500 °C varies from 1.5–3.0 °C/s near the fusion boundary to 0.1–0.3 °C/s at the thermal influence zone boundary, creating a broad spectrum of microstructural transformations.
- Residual stress pattern: Peak longitudinal residual stresses reach 350–420 MPa (approaching yield strength) in the cap weld, while transverse stresses are reduced by 20–30% compared to single-sided welding configurations.
Validation Against Experimental Data
The authors validate the numerical model by comparing simulated temperature profiles with thermocouple measurements taken during physical welding trials. The agreement between predicted and measured temperatures at various depths is within ±15%, confirming the model's predictive capability for engineering applications.
Engineering Practice Integration
For thick-plate pressure vessel fabrication — particularly for components such as hydrogenation reactor shells, high-pressure separator drums, and thick-walled heat exchanger channels — the simulation results provide actionable guidance:
- Weld sequence optimization: The thermal analysis identifies that welding from the thicker section first and progressing toward the thinner section reduces angular distortion by 30–40%.
- Interpass temperature control: Based on the predicted temperature evolution, maintaining interpass temperatures between 150–250 °C minimizes the risk of cold cracking while avoiding excessive grain growth.
- Preheat strategy: The simulation demonstrates that localized preheating (200–300 °C) at the root weld reduces peak cooling rates by approximately 40%, effectively suppressing martensite formation in low-alloy steels.
- Distortion prediction: The computed angular and longitudinal distortions enable fixture design and pre-bending calculations that reduce post-weld straightening requirements.
Study Insights and Reflections
This work exemplifies the power of computational methods in welding engineering, particularly when direct experimental measurement is impractical due to the scale and complexity of the components involved. The authors' careful attention to boundary condition selection and material property temperature dependence reflects a deep understanding of the limitations of simplified thermal models. The practical value of this research lies not merely in the numerical results themselves but in the methodology it establishes — a validated simulation framework that can be adapted for specific component geometries and welding sequences. For engineers involved in thick-section pressure vessel fabrication, this study demonstrates that numerical simulation can serve as a complementary tool to physical qualification, reducing the number of expensive trial welds while providing insights into thermal fields that are inaccessible to direct measurement. The work reinforces the principle that understanding the thermal history of a weld is fundamental to predicting and controlling its metallurgical outcome and mechanical performance.