3D Finite Element Analysis of Temperature Field Distribution in 321 Stainless Steel Fillet Weld TIG Welding

Literature Overview

This research by Tao Jing and Wang Xiangjie, published in 2016 in the journal "Hot Working Technology" (热加工工艺), presents a three-dimensional finite element analysis (FEA) of the temperature field distribution during TIG welding of 321 stainless steel fillet welds. The study was supported by the Hubei Provincial Department of Education Youth Talent Research Project (Q20144404) and the Hubei Polytechnic University Excellent Young and Middle-aged Scientific and Technological Innovation Team Program (JD2014001), indicating a well-funded and institutionally supported research effort. The focus on 321 stainless steel and fillet welds is particularly relevant to pressure vessel fabrication, where fillet welds are extensively used in nozzle-to-shell connections, reinforcement pad attachments, and internal component weldments.

321 Stainless Steel Characteristics and Welding Relevance

321 stainless steel (UNS S32100) is a stabilized austenitic stainless steel containing titanium (0.1-0.4%) as a stabilization element. The titanium addition prevents the formation of chromium carbides during welding and high-temperature service, thereby eliminating the risk of intergranular corrosion. This makes 321 stainless steel an excellent choice for pressure vessel applications involving temperatures in the range of 425-870°C, where sensitization would otherwise be a concern with unstabilized grades such as 304 or 316.

The thermal properties of 321 stainless steel are of particular importance for FEA modeling. The low thermal conductivity at room temperature means that heat is concentrated near the weld zone, leading to high thermal gradients and significant residual stresses. However, the thermal conductivity increases with temperature, which affects the heat dissipation pattern during welding. Accurate modeling of these temperature-dependent properties is essential for reliable FEA results.

Finite Element Model Development

The 3D FEA model developed in this study likely employs a moving heat source to simulate the TIG welding process. The heat source model is a critical component of the welding FEA, as it determines the accuracy of the temperature field prediction. Common heat source models used in welding FEA include:

1. Gaussian double-ellipsoidal model (Goldak model): This model distributes heat over an ellipsoidal volume, with different heat distributions in front of and behind the weld center. It is widely used for TIG welding simulations because it can capture the characteristic deep penetration in front of the arc and the wider heat distribution behind the arc.
2. Single-ellipsoidal model: A simplified version of the Goldak model, used when the heat distribution is more symmetric.
3. Conical heat source model: Used for processes with a concentrated heat input, such as laser welding.

For TIG welding of 321 stainless steel fillet welds, the Goldak double-ellipsoidal model is the most appropriate choice. The model parameters include:

• Heat input (Q): typically 2-6 kJ/mm for TIG welding
• Aspect ratio (a): typically 0.5-1.0
• Frontal ellipse width (b_f): typically 2-4 mm
• Rear ellipse width (b_r): typically 3-6 mm
• Frontal ellipse depth (c_f): typically 2-4 mm
• Rear ellipse depth (c_r): typically 3-6 mm

The FEA model must also account for the complex geometry of the fillet weld, which differs significantly from a butt weld in terms of heat flow patterns. The fillet weld junction creates a three-dimensional heat sink effect, with heat being conducted away from the weld in multiple directions. This makes the temperature field distribution more complex than for a simple butt weld and requires a full 3D model for accurate prediction.

Temperature Field Analysis and Results Interpretation

The key outputs of the FEA analysis include:

Peak temperature distribution: The peak temperature at the weld surface is typically 1500-2000°C, depending on the heat input and travel speed. The peak temperature in the interior of the weld is slightly lower due to the heat sink effect of the base metal.

Thermal cycle characteristics: The thermal cycle at different locations in the weld and HAZ is characterized by:

• Peak temperature (T_max)
• Time above 1300°C (t₁₃₀₀): determines grain growth
• Time above 800°C (t₈₀₀): determines sensitization risk
• Time above 400°C (t₄₀₀): determines tempering of the HAZ
• Cooling rate (800°C to 500°C): affects microstructure and residual stress

Residual stress distribution: The residual stress pattern in a fillet weld is complex, with tensile stresses developing in the weld metal and compressive stresses in the surrounding base metal. The peak residual stress can reach 200-400 MPa, depending on the constraint level and the thermal expansion mismatch.

Engineering Practice and Quality Control Implications

The FEA results have direct implications for welding procedure development and quality control in pressure vessel fabrication:

1. Heat input control: The FEA can be used to predict the heat input required to achieve the desired penetration without excessive grain growth or sensitization. For 321 stainless steel, the heat input should be controlled to minimize the time spent in the sensitization temperature range (425-870°C).
2. Preheat and interpass temperature: The FEA can predict the effect of preheat on the thermal cycle and residual stress distribution. For thick-section fillet welds, preheating to 100-200°C can reduce the thermal gradient and residual stresses, while also reducing the risk of cold cracking.
3. Post-weld heat treatment: The FEA results can inform the design of the PWHT schedule. The predicted residual stress distribution and thermal cycle characteristics can be used to determine the appropriate PWHT temperature and duration.
4. Weld sequence optimization: For complex weldments with multiple fillet welds, the FEA can be used to optimize the weld sequence to minimize distortion and residual stresses. This is particularly important for pressure vessel fabrication, where dimensional accuracy is critical for assembly and inspection.

Study Insights and Reflections

The use of 3D FEA for welding analysis represents a significant advancement over the traditional trial-and-error approach to welding procedure development. By understanding the temperature field distribution and thermal cycle characteristics, engineers can make informed decisions about welding parameters, preheat requirements, and post-weld heat treatment. The FEA approach also allows for the prediction of welding-induced residual stresses, which is critical for the assessment of structural integrity and fatigue life.

However, it is important to recognize the limitations of FEA in welding analysis. The accuracy of the FEA results depends on the quality of the input data, including the material properties, heat source model parameters, and boundary conditions. The material properties of 321 stainless steel are temperature-dependent and must be accurately represented in the model. The heat source model parameters must be calibrated against experimental data to ensure that the predicted temperature field is realistic.

Another important consideration is the computational cost of 3D FEA. A full 3D model of a complex weldment can require significant computational resources, and the analysis time can be several hours to several days. This limits the use of FEA for routine welding procedure development and makes it more suitable for critical applications where the cost of welding defects is high.

The study by Tao Jing and Wang Xiangjie demonstrates the value of FEA in welding research and provides a methodology that can be adapted to other materials and weld configurations. The approach is particularly valuable for pressure vessel fabrication, where the consequences of welding defects are severe and the need for reliable welding procedures is paramount.

Reference Value and Outlook

The FEA methodology presented in this study has broad applicability to welding research and engineering practice. The ability to predict temperature field distributions and thermal cycles provides valuable insights into the welding process that cannot be obtained from experimental measurements alone. As computational resources become more available and material property databases become more complete, FEA is expected to play an increasingly important role in welding procedure development and quality assurance.