Three-Dimensional Finite Element Simulation of Welding Temperature Field in Aluminum Alloy Thin Plates Based on ANSYS
Literature Overview and Core Findings
This paper by Chen Yuxi, Zhu Jinhong, Shi Hongxin, and Ding Gaojian from Henan University of Science and Technology, published in Hot Working Technology (2009, Vol. 38, No. 9), presents a three-dimensional finite element simulation of the welding temperature field in aluminum alloy thin plates using ANSYS. The study employs a double ellipsoid heat source distribution model and develops a finite element numerical analysis model for the welding process under a moving arc.
The key technical achievement is the use of ANSYS APDL (Parametric Design Language) to effectively simulate the moving arc heating process and the transient temperature field evolution during welding. The parameterised programming approach allows for efficient simulation of the dynamic welding process. The simulated temperature field distribution provides a foundation for subsequent analysis of welding stress and strain.
Technical Interpretation and Simulation Methodology
Double Ellipsoid Heat Source Model
The double ellipsoid heat source model, originally proposed by Goldak, is a widely used approach for simulating arc welding processes. It represents the heat input from the welding arc as two semi-ellipsoids, one in front of and one behind the arc centre, with different heat flux distributions.
The double ellipsoid model is particularly suitable for TIG welding because it accounts for the asymmetric heat distribution caused by the arc movement. The front semi-ellipsoid has a steeper heat flux gradient, while the rear semi-ellipsoid has a more gradual distribution, reflecting the heat transfer characteristics of the moving arc.
The heat flux distribution in the double ellipsoid model is given by:
| Parameter | Description | Typical Value for TIG |
|---|---|---|
| qf | Heat flux in front ellipsoid | Higher peak, shorter length |
| qr | Heat flux in rear ellipsoid | Lower peak, longer length |
| af | Front ellipsoid length | 2-4 mm |
| ar | Rear ellipsoid length | 6-10 mm |
| b | Ellipsoid width | 3-5 mm |
| c | Ellipsoid depth | 2-4 mm |
APDL Parameterised Programming
The use of APDL for parameterised programming is a significant advantage for welding simulation. APDL allows the user to define variables, create geometry, apply loads, and solve the model using a scripted approach. This enables:
- Parametric studies: Varying welding parameters (current, voltage, travel speed) without recreating the model.
- Automated mesh generation: Creating meshes with different element sizes for convergence studies.
- Moving heat source implementation: Defining the heat source position as a function of time and moving it across the workpiece.
Temperature Field Simulation Challenges
Simulating the welding temperature field in aluminum alloy thin plates presents several challenges:
- High thermal conductivity: Aluminum has a thermal conductivity of approximately 200 W/m·K, which is about 10 times that of steel. This means heat dissipates rapidly, requiring careful time stepping to capture the temperature distribution.
- Low melting point: Aluminum melts at approximately 660°C, which is much lower than steel. This affects the heat input required and the size of the molten pool.
- Thin plate geometry: Thin plates (typically 2-6 mm) have a high surface area-to-volume ratio, leading to rapid heat loss by convection and radiation.
- Temperature-dependent material properties: The thermal conductivity, specific heat, and density of aluminum change significantly with temperature, requiring the use of temperature-dependent material properties in the simulation.
Engineering Practice and Delivery Considerations
Application in Thin-Walled Structure Fabrication
The simulation of welding temperature fields in aluminum alloy thin plates is directly relevant to the fabrication of thin-walled structures such as:
- Aircraft fuselage panels: Where distortion control is critical for aerodynamic performance.
- Ship superstructures: Where lightweight construction is important for stability.
- Heat exchanger components: Where weld quality affects thermal performance.
- Automotive body panels: Where dimensional accuracy is essential for assembly.
Distortion and Residual Stress Prediction
The temperature field simulation provides the foundation for predicting welding distortion and residual stresses. The thermal gradients during welding generate differential expansion and contraction, leading to:
- Transverse shrinkage: Causing angular distortion in butt joints.
- Longitudinal shrinkage: Causing bowing in long welds.
- Rotational distortion: Causing twisting in asymmetric joints.
For thin plates, distortion is a particular concern because the section modulus is low, and even small residual stresses can cause significant deformation.
Validation with Experimental Measurements
Numerical simulation is only as good as its validation. The temperature field simulation must be validated against experimental measurements using:
- Thermocouples: Placed at various locations to record temperature-time histories.
- Infrared thermography: For non-contact measurement of surface temperature distributions.
- Metallographic examination: To verify the predicted HAZ width and weld pool size.
Reflections and Lessons Learned
This paper demonstrates the value of numerical simulation in welding engineering, but I would offer several practical observations based on my experience:
- Simulation is a tool, not a substitute for experience. The accuracy of the simulation depends on the correctness of the input parameters, including material properties, heat source parameters, and boundary conditions. Garbage in, garbage out.
- The double ellipsoid heat source parameters must be calibrated. The values of af, ar, b, and c are not universal constants; they must be calibrated against experimental measurements for each specific welding process and parameter set.
- Thin plate welding requires careful control. The high thermal conductivity of aluminum means that heat dissipates rapidly, and the welding parameters must be adjusted accordingly. The simulation can help optimise parameters, but the final validation must be done on actual test pieces.
- Distortion control is a production reality. In my experience, the most common problem with thin aluminum plate welding is distortion, not weld quality. The simulation can predict distortion, but the real solution is often a combination of proper joint design, welding sequence, and fixturing.
- The transition from simulation to production requires a validation protocol. The simulated temperature field must be verified against experimental measurements before the model is used for parameter optimisation or distortion prediction.
In conclusion, this paper provides a solid methodology for simulating the welding temperature field in aluminum alloy thin plates using ANSYS and the double ellipsoid heat source model. The parameterised APDL approach is efficient and flexible, enabling parametric studies and optimisation. However, the practical value of the simulation depends on proper calibration and validation against experimental data. The key lesson is that numerical simulation is a powerful tool for understanding the welding process, but it must be used in conjunction with experimental verification and practical experience to achieve reliable results in production.
Zhuojin Pipe Fitting Co., Ltd