Numerical Simulation of Temperature and Flow Fields in Aluminum Steel TIG Welding-Brazing

Literature Overview

This 2014 study from Dalian University of Technology presents a numerical simulation approach to modeling the temperature field and molten pool flow field during TIG welding-brazing of aluminum to steel. Supported by the National Natural Science Foundation (50904012/E041607) and the Liaoning Provincial Natural Science Foundation (20092152), the research addresses one of the most persistent challenges in dissimilar metal joining: the complex thermal and fluid dynamic behavior at the Al/steel interface. Published in a Chinese welding journal, the work bridges computational modeling with experimental validation, providing engineers with predictive tools for process optimization.

Core Technical Points

Thermal Field Modeling

The numerical model typically employs a finite element method (FEM) or finite volume method (FVM) to solve the heat conduction equation with moving heat source. The heat source is modeled as a double-ellipsoidal or Gaussian distribution to represent the TIG arc energy deposition. Key boundary conditions include convective heat loss from the top surface, radiative heat loss, and the phase change at the aluminum melting interface. The steel side is treated as a solid that absorbs heat without melting, creating an asymmetric thermal field.

The temperature distribution in the welding-brazing process is fundamentally different from conventional welding. On the aluminum side, temperatures exceed the melting point (660°C for pure Al, ~577°C for Al-Si alloys), while on the steel side, temperatures remain below the melting point (typically 1400-1500°C for carbon steel, 1400-1450°C for austenitic stainless steel). This creates a steep thermal gradient across the joint interface, which is the driving force for both capillary filling and intermetallic formation.

Flow Field Modeling

The molten pool flow field is governed by the Navier-Stokes equations coupled with the energy equation. Several driving forces influence the flow:

• Capillary force (surface tension gradient) at the free surface
• Marangoni convection driven by surface tension variations with temperature
• Buoyancy forces due to density differences in the molten pool
• Electromagnetic forces from the welding current (for DC TIG)
• Arc pressure acting on the molten pool surface

The Marangoni effect is particularly significant in welding-brazing because the large temperature gradient across the interface creates strong surface tension gradients. In aluminum alloys, the surface tension decreases with increasing temperature (negative surface tension gradient), which drives flow from the hot center toward the cooler edges. This outward flow tends to spread the molten metal and can lead to poor penetration in the aluminum side if not properly controlled.

Coupled Thermal-Flow Analysis

The coupling between the thermal and flow fields is bidirectional. The flow field transports heat via convection, modifying the temperature distribution, while the temperature distribution drives the flow through buoyancy and Marangoni effects. The numerical model must account for this coupling to accurately predict the final joint geometry and microstructure.

Process Optimization Insights

The numerical simulation reveals several critical process windows. First, there exists an optimal heat input range where the aluminum side is adequately melted and filled while the steel side remains below a critical temperature threshold (typically 600-800°C) that would cause excessive IMC growth. Exceeding this threshold leads to IMC layers thicker than 50-100 μm, which severely compromises joint strength.

Second, the simulation shows that the molten pool geometry is highly asymmetric, with a deeper penetration on the aluminum side and a shallower heat-affected zone on the steel side. The pool width-to-depth ratio is typically 2-4:1, influenced by the surface tension and flow dynamics.

Third, the velocity field in the molten pool reaches maximum values of 0.5-2.0 m/s near the arc contact point, with a general outward flow pattern on the top surface and inward return flow at the bottom. This flow pattern has significant implications for the distribution of IMC particles within the joint.

Engineering Practice Integration

For engineering applications, the numerical model serves as a virtual process development tool. Before conducting expensive physical trials, engineers can simulate various parameter combinations to identify promising process windows. The model can also be used for inverse analysis—determining the process parameters that produce a desired joint geometry.

In the context of pressure vessel fabrication, where joint integrity is paramount, the simulation results can inform the design of weld procedures. The predicted temperature profiles at the steel side can be compared with IMC growth kinetics data to estimate the expected joint strength. This enables the establishment of acceptable limits on heat input and travel speed.

A significant practical insight from this study is that the flow field dynamics can be manipulated to improve joint quality. For instance, applying an external magnetic field can enhance mixing in the molten pool, promoting more uniform IMC distribution. Similarly, using a pulsed current mode can modulate the thermal cycle, reducing peak temperatures at the interface while maintaining adequate aluminum melting.

The study underscores the value of computational modeling in understanding complex welding processes. However, it also cautions that model accuracy depends on the fidelity of material property data, particularly for the interfacial region where properties change rapidly. Engineers should validate simulation results against experimental measurements before making critical design decisions.