Numerical Simulation of Residual Stress in Circumferential Weld Joints of T91 Grade Pipeline Steel Using SYSWELD

Literature Overview

The paper by Miao Zeyu, Kong Fanyu, and Liu Hui (2016), published in Heat Processing Technology (Vol. 45, No. 17, pp. 156-160), presents a finite element numerical simulation of welding residual stresses in circumferential weld joints of T91 grade pipeline steel using the SYSWELD software. The study was supported by the National Natural Science Foundation of China (Grant No. 51104134) and the Zhejiang Provincial Natural Science Foundation (Grant No. LY14E040001). The authors are affiliated with Zhejiang Wanli University of Science and Technology, specifically the Key Laboratory of Furniture Inspection Technology.

Core Technical Content

T91 is a 9Cr-1Mo-V-Nb-Ti martensitic steel widely used in high-temperature applications such as power plant boiler tubes, supercritical steam pipelines, and nuclear reactor components. The welding of T91 steel presents significant challenges due to its high hardenability, susceptibility to hydrogen-induced cracking, and the formation of brittle microstructures in the heat-affected zone (HAZ). The circumferential weld joints in T91 pipelines are critical because they are subjected to complex stress states during both manufacturing and service, and residual stresses from welding can significantly influence the fatigue life, stress corrosion cracking (SCC) susceptibility, and overall structural integrity.

Simulation Methodology

The authors calibrated the heat source model for the circumferential weld based on experimental results, which is a critical step in ensuring the accuracy of the numerical simulation. The SYSWELD software was then used to simulate the entire welding process, including the thermal cycle, phase transformation, and mechanical response. The simulation tracked the evolution of temperature, stress, and strain throughout the welding sequence, providing a detailed picture of the residual stress distribution in the weld joint.

Key Findings on Residual Stress Distribution

The simulation results revealed that the maximum residual stress is located at a distance of 5 to 10 mm from the weld centerline, specifically in the region just outside the HAZ. The minimum residual stress is found in the base metal region away from the weld. This distribution pattern is consistent with the general understanding of welding residual stress, where the peak tensile residual stress typically occurs at the boundary between the weld metal and the HAZ, where the thermal gradient is most severe.

A particularly interesting finding is that during multi-pass welding, as the number of welding layers increases, the peak residual stress location migrates outward to the outer HAZ. This phenomenon occurs because each subsequent welding pass heats the previously solidified weld metal and HAZ, partially relieving the residual stresses accumulated in earlier passes. The cumulative effect of multiple passes leads to a redistribution of the residual stress field, with the peak stress shifting to the most recently affected region.

Interpretation of Technical Points

Residual Stress Mechanism in T91 Welding

The residual stress distribution in T91 weld joints is governed by the interplay of thermal expansion, plastic deformation, and phase transformation. During welding, the intense heat input causes rapid thermal expansion of the material near the weld, which is constrained by the cooler surrounding base metal. This constraint leads to compressive plastic deformation in the heated region. Upon cooling, the heated region contracts, but the plastic deformation from the heating phase is irreversible, resulting in residual tensile stresses in the weld and HAZ regions and residual compressive stresses in the surrounding base metal.

For T91 steel, the phase transformation from austenite to martensite during cooling introduces additional volumetric expansion, which can further modify the residual stress distribution. The martensitic transformation in T91 is characterized by a relatively low martensite start temperature (Ms) and a broad transformation range, which means that the transformation-induced plasticity (TRIP) effect can be significant. The SYSWELD simulation should ideally account for these transformation effects, although the paper focuses primarily on the thermal-mechanical aspects.

Implications for Weld Integrity and Fatigue Life

The location of the maximum residual stress at 5 to 10 mm from the weld centerline has direct implications for the fatigue life of the weld joint. This region corresponds to the coarse-grained HAZ (CGHAZ) or the fine-grained HAZ (FGHAZ), depending on the specific thermal cycle. In T91 steel, this region is susceptible to intergranular cracking, precipitate coarsening, and microstructural degradation. The presence of high residual tensile stresses in this region can accelerate crack initiation and propagation, reducing the fatigue life of the weld joint.

From an engineering perspective, post-weld heat treatment (PWHT) is commonly employed to relieve residual stresses in T91 welds. The simulation results can inform the PWHT design by identifying the regions of highest residual stress concentration, which are the most critical for stress relief. The migration of the peak residual stress location during multi-pass welding also suggests that the PWHT temperature and duration should be optimized to ensure uniform stress relief throughout the weld joint.

Engineering Practice Implications

For engineers involved in the fabrication and integrity assessment of T91 pipelines, this study provides several practical insights. First, the numerical simulation approach using SYSWELD offers a cost-effective alternative to experimental residual stress measurement, which is typically expensive and time-consuming. Second, the simulation results can be used to predict the effectiveness of different welding sequences and parameters on the residual stress distribution, enabling optimization of the welding procedure specification (WPS). Third, the identification of the peak residual stress location provides a target for non-destructive testing (NDT) and post-weld inspection activities.

In terms of quality control, the residual stress distribution should be considered when interpreting the results of NDT methods such as ultrasonic testing (UT) and radiographic testing (RT). High residual tensile stresses can affect the sensitivity of UT by altering the acoustic impedance of the material, and they can influence the interpretation of RT by affecting the contrast of weld features. Engineers should be aware of these interactions when developing inspection procedures for T91 welds.

Key Questions and Reflections

A question that arises from this study is how well the SYSWELD simulation captures the transformation-induced plasticity effects that are particularly important in T91 steel. The accuracy of the residual stress prediction depends heavily on the material model used in the simulation, and the inclusion of phase transformation effects can significantly improve the prediction accuracy. Future work should validate the simulation results against experimental residual stress measurements, such as those obtained using the hole-drilling method or neutron diffraction, to assess the fidelity of the model.

Another reflection concerns the scalability of the simulation approach. The study focuses on a specific T91 pipe configuration and welding procedure, and the results may not be directly transferable to other geometries or welding parameters. Engineers should exercise caution when applying the findings to different pipe sizes, wall thicknesses, or welding processes, and should consider the need for additional simulations or experimental validation for specific applications.

Study Insights and Implications

The work by Miao, Kong, and Liu demonstrates the value of numerical simulation in understanding and predicting welding residual stresses in T91 pipeline welds. The identification of the peak residual stress location at 5 to 10 mm from the weld centerline and the observation of stress migration during multi-pass welding provide valuable insights for optimizing welding procedures and post-weld treatment. For engineers, the key takeaway is that residual stress management is a critical aspect of T91 weld integrity, and numerical simulation tools such as SYSWELD can serve as powerful aids in the design, fabrication, and assessment of T91 pipeline weld joints. The integration of simulation results with experimental validation and NDT data will provide a comprehensive framework for ensuring the long-term reliability of T91 pipeline systems.