Numerical Simulation and Residual Stress Prediction for TIG Welding of Aluminum Alloy Plates

Literature Overview

The paper by Jin Cheng, Niu Jitai, He Shiyu, and Chen Yong, published in Materials for Mechanical Engineering (Vol. 31, No. 3, 2007), presents a finite element analysis of the TIG welding process for aluminum alloy plates using ABAQUS software. Funded by the National Natural Science Foundation of China (Grant No. 90205035), the study employs a moving elliptical Gaussian distribution surface heat source to simulate the temperature field and stress field during welding, and predicts the residual stress distribution after cooling. The numerical results are validated against experimental measurements, demonstrating good agreement.

Core Technical Approach

Heat Source Modeling

The moving elliptical Gaussian heat source is a widely accepted model for TIG welding heat input. The heat flux distribution is defined by the following expression:

q(x, y) = (6√3 Q) / (π ab √π) × exp(-3(x²/a² + y²/b²))

where Q is the total heat input, a is the major axis of the ellipse (typically in the direction of travel), and b is the minor axis. The elliptical shape accounts for the asymmetric heat distribution observed in TIG welding, where more heat is deposited ahead of the arc due to the tilting of the tungsten electrode and the direction of travel.

Thermal-Mechanical Coupled Analysis

The simulation employs a coupled thermal-mechanical analysis approach:

1. Thermal analysis: The moving heat source is applied to the surface of the model, and the temperature field is computed using the heat conduction equation with appropriate boundary conditions. Convective and radiative heat losses from the surface are modeled using Newton's cooling law and Stefan-Boltzmann radiation, respectively.
2. Mechanical analysis: The temperature field from the thermal analysis is used as input to compute the stress field. The material is modeled with elastic-plastic behavior, including temperature-dependent yield strength, thermal expansion, and plastic strain hardening. The welding process is simulated by sequentially activating elements as the heat source moves across the plate.

Residual Stress Prediction

After the welding simulation is complete and the model has cooled to room temperature, the residual stress distribution is extracted. The residual stresses result from the complex interplay of thermal expansion and contraction, plastic deformation during heating and cooling, and the constraints imposed by the surrounding cooler material.

Technical Analysis

Temperature Field Characteristics

The TIG welding temperature field is highly non-uniform, with peak temperatures exceeding 2000°C at the arc-weld pool interface and rapidly decreasing with distance from the weld centerline. The temperature gradient creates a complex stress state that drives plastic deformation. For aluminum alloys, the yield strength decreases significantly with temperature, reaching near-zero values above approximately 400°C. This temperature-dependent softening is a critical factor in the development of residual stresses.

Residual Stress Distribution

The predicted residual stress distribution typically exhibits the following characteristics:

• Longitudinal residual stresses: Tensile stresses develop in the weld zone and near-weld region, while compressive stresses develop in the far field to maintain equilibrium. The peak longitudinal tensile stress can approach the yield strength of the material at room temperature.
• Transverse residual stresses: Generally lower in magnitude than longitudinal stresses, with a similar tensile-compressive distribution pattern.
• Through-thickness residual stresses: Compressive stresses develop on the top surface of the weld due to the restraint from the cooler base material, while tensile stresses develop at the bottom surface.
• Stress relaxation: The residual stress distribution evolves with time as the material cools; the final state represents the equilibrium configuration after all thermal and mechanical effects have been accounted for.

Validation Against Experimental Data

The study reports good agreement between the predicted and measured residual stresses. This validation is essential for establishing confidence in the numerical model and its predictive capability. Common experimental methods for residual stress measurement include:

• Hole-drilling strain gauge method (ASTM E837)
• Neutron diffraction
• X-ray diffraction
• Ultrasonic measurement

The hole-drilling method is the most widely used due to its relative simplicity and non-destructive nature (when performed on coupons or test pieces).

Engineering Practice Integration

Application to Aluminum Alloy Welded Structures

Aluminum alloy welded structures are widely used in aerospace, automotive, and marine applications where weight reduction is critical. Residual stresses in these structures can have significant effects on:

• Fatigue life: Tensile residual stresses at the weld toe significantly reduce fatigue life by adding to the applied stress range.
• Corrosion resistance: Residual stresses can promote stress corrosion cracking in susceptible aluminum alloys.
• Dimensional stability: Residual stresses can cause distortion during machining or during service under thermal cycling.
• Hydrostatic test performance: Residual stresses can affect the pressure-bearing capacity of welded pressure vessels.

Mitigation Strategies

Based on the residual stress predictions, the following mitigation strategies can be implemented:

1. Post-weld stress relief: Thermal stress relief at 400–450°C can reduce residual stresses by 50–80%. However, this may cause additional distortion and must be carefully controlled.
2. Peening: Shot peening or ultrasonic impact treatment can introduce beneficial compressive residual stresses at the weld toe, improving fatigue performance.
3. Welding sequence optimization: For multi-pass or multi-weld joints, the welding sequence can be optimized to minimize residual stresses. Back-step welding, zigzag welding, and symmetric welding patterns are common strategies.
4. Fixture design: Rigid fixtures can constrain welding distortion but may increase residual stresses. A balance between distortion control and residual stress management must be achieved.
5. Welding parameter optimization: Lower heat input reduces the extent of plastic deformation and consequently reduces residual stresses. However, this may compromise weld penetration and require more passes.

Finite Element Model Development Guidelines

For engineers developing their own residual stress prediction models, the following guidelines are recommended:

• Use a coupled thermal-mechanical analysis with appropriate temperature-dependent material properties.
• Employ element activation/deactivation to simulate the sequential welding of multiple passes.
• Include convective and radiative boundary conditions with appropriate heat transfer coefficients.
• Use a sufficiently fine mesh in the weld zone to capture the steep temperature and stress gradients.
• Validate the model against experimental data for at least one representative configuration before applying it to new designs.
• Consider the effect of welding sequence, fixture constraints, and preheat temperature on the residual stress distribution.

Study Insights and Implications

This research demonstrates the effectiveness of finite element simulation in predicting residual stresses in TIG welded aluminum alloy plates. The good agreement between numerical and experimental results validates the modeling approach and provides confidence in its application to more complex geometries. For practicing engineers, the key implication is that residual stress prediction can be integrated into the welding process design workflow, enabling proactive mitigation of stress-related issues rather than reactive correction after welding. The methodology can be extended to other welding processes, materials, and geometries, making it a versatile tool for welding process optimization. The study also highlights the importance of accurate heat source modeling and temperature-dependent material properties in achieving reliable predictions, which should be considered essential components of any residual stress simulation effort.