Thin Plate Overlay Welding Deformation Prediction Based on Initial Deformation

Literature Overview

This paper by Guo Nan, Ma Xiqiang, Yin Xianqing, and Yu Yongjian, published in China Mechanical Engineering in 2020 (Vol. 31, No. 3, pp. 367–372), addresses the critical problem of how initial deformation in thin plate structures affects the accuracy of welding deformation prediction. The authors propose an optical detection method based on Digital Image Correlation (DIC) technology to capture initial deformation profiles and incorporate them into numerical simulation models for more accurate prediction of overlay welding deformation.

Core Technical Content

The research addresses a practical problem that is often overlooked in welding simulation: the assumption that the base structure is perfectly flat before welding begins. In reality, thin plate components used in pipe fitting fabrication, heat exchanger construction, and pressure vessel manufacturing exhibit residual deformations from prior manufacturing processes such as rolling, cutting, and machining.

Methodology

The research methodology follows a systematic approach:

1. Initial deformation acquisition: Structural light scanning is used to obtain the three-dimensional contour of the initial deformation in the welded structural component.
2. Numerical model construction: Two models are developed—one based on an ideal (perfectly flat) geometry and another incorporating the measured initial deformation.
3. Finite element simulation: Both models undergo welding thermal-mechanical simulation using sequential coupling or fully coupled approaches.
4. Experimental validation: DIC technology is employed to measure both static (final) and dynamic (in-process) deformation fields for comparison with simulation results.

Key Findings

Interpretation of Technical Points

The finding that both models predict saddle-shaped deformation after welding completion is consistent with classical welding deformation theory. However, the critical insight is that initial deformation fundamentally alters the deformation evolution pathway. The initial geometric imperfection acts as a perturbation that influences:

• Heat flow distribution: Local variations in thickness or curvature alter the thermal conductivity path, causing non-uniform heating.
• Stress state initialization: Pre-existing residual stresses from initial deformation interact with welding-induced thermal stresses, potentially leading to stress concentration.
• Constraint conditions: Initial curvature changes the effective boundary conditions of the plate, modifying the constraint ratio and thus the magnitude of longitudinal residual stresses.

The paper's conclusion that initial deformation defects affect the longitudinal residual compressive stress distribution and are the root cause of different welding deformations is particularly significant for engineering practice.

Process and Standards Analysis

For thin plate overlay welding applications in pipe fitting manufacturing, this research has direct implications:

• ASME B31.3 and GB/T 20801 specify allowable flatness tolerances for pipe supports and structural components. The research demonstrates that even within these tolerances, initial deformation significantly affects final dimensional accuracy.
• FEM analysis standards such as ASME VIII Div. 2 and BS 7910 require consideration of residual stresses in fitness-for-service assessments. The non-uniform stress distributions predicted by the initial deformation model are more appropriate inputs for such analyses.
• Welding procedure qualification per AWS D1.1 or ISO 15614 should consider the actual geometry of the test coupon rather than assuming ideal conditions.

Integration with Engineering Practice

In my experience with pipe fitting fabrication, particularly for thin-wall alloy fittings used in cryogenic and high-pressure service, the following observations align with this research:

1. Flatness control before welding: For thin plate components (t < 6 mm) intended for overlay welding, pre-weld flatness inspection and correction should be implemented as a quality gate.
2. Fixture design: Welding fixtures should account for the expected deformation based on actual initial geometry rather than nominal dimensions.
3. Post-weld correction: The prediction of saddle-shaped deformation allows for pre-compensation in fixture design or post-weld mechanical correction.
4. Material selection: For thin plate overlay welding, materials with lower coefficients of thermal expansion (such as austenitic stainless steels) may reduce the sensitivity to initial deformation.

Key Questions and Reflections

This research raises several important questions for further investigation:

• How does the amplitude and wavelength of initial deformation affect the prediction accuracy? Is there a threshold below which initial deformation can be neglected?
• What is the interaction between initial deformation and welding sequence for multi-pass overlay welding?
• Can the DIC-based methodology be adapted for in-situ monitoring of deformation during production welding, enabling real-time correction?
• How does the initial deformation analysis extend to cylindrical geometries typical of pipe and pipe fitting manufacturing?

Study Insights and Implications

This paper represents a significant advancement in welding simulation methodology by recognizing that real-world structures are never ideal. The integration of experimental geometry acquisition with numerical modeling creates a more predictive framework that bridges the gap between simulation and reality. For pipe fitting engineers, the practical implication is clear: dimensional accuracy in thin plate overlay welding requires attention to the initial condition of the base material, not merely optimization of welding parameters. The DIC technology described offers a non-contact, full-field measurement approach that is well-suited for production environments where traditional contact measurement methods are impractical.