A Heterogeneous thermal conductivity CFD simulation models how heat moves through special materials. In many materials, the ability to conduct heat is not the same everywhere. This is called non-uniform Thermal conductivity. This is common in advanced parts like composites and electronics. To study this, we use powerful computer tools like ANSYS Fluent. A normal simulation cannot handle this problem. We need to write a special program, called a UDF (User-Defined Function), to tell Fluent how the thermal conductivity changes at every point. To store this special information for each part of our model, we use UDM (User-Defined Memory). This method of using a Heterogeneous thermal conductivity Using UDF & UDM Fluent simulation is essential for creating accurate models for Advanced Thermal Management.

Figure 1: A schematic showing that thermal conductivity (K) in a Heterogeneous thermal conductivity CFD analysis can change based on location or material properties.

Simulation Process: Fluent Setup, Defining Heterogeneous Conductivity with UDF & UDM

To perform this Fluent UDF for Heat Transfer study, we first created a simple 2D rectangular geometry. We then made a high-quality structured grid using quadrilateral cells. A structured grid is very good for this problem because it makes it easy for our custom program to find the location of each cell. In the ANSYS Fluent software, we used the steady-state solver and turned on the Energy equation to calculate heat transfer.

The most important part of this simulation was creating a custom User-Defined Function (UDF) in the C programming language. This UDF was written to define a new, spatially varying material. The UDF works by checking the x and y location of the center of every single cell in the mesh. Based on that location, the UDF finds the correct thermal conductivity value from a special 220×60 data matrix that was written directly into the code. This gives every cell its own unique thermal property. We also used User-Defined Memory (UDM) as a storage space to save the x and y coordinates of each cell for checking our work. Finally, to make heat move, we added a heat source of 10,000 W/m³ in one area and a heat sink of -10,000 W/m³ in another area.

Figure 2: The high-quality structured grid with quadrilateral cells used for the CFD with User Defined Functions analysis.

Post-processing: CFD Analysis, Visualizing Heat Flow in a Non-Uniform Material

The temperature contour in Figure 3 is a direct and powerful visualization of our custom material’s behavior. From an engineering perspective, the results are exactly what we expect from the setup. The simulation correctly identifies a clear hot spot and a cold spot. The left side, where the 10,000 W/m³ heat source was applied, shows the highest temperature, reaching about 335 K (the red area). On the right side, the -10,000 W/m³ heat sink creates the coldest spot, with the temperature dropping to approximately 56 K (the dark blue area). This confirms that the basic heat transfer physics is working correctly.

However, the real engineering story is in the shape of the temperature lines. The contours are not smooth, round, or symmetrical. They are distorted and irregular. This is the absolute proof that the thermal conductivity is heterogeneous, or non-uniform. Heat is not flowing evenly in all directions. Instead, it is forced to follow complex paths based on the conductivity values we defined in our UDF’s data matrix. Some areas conduct heat easily, while others act as insulators, blocking the heat flow. The strange, finger-like shape of the cold blue region spreading from the right is a perfect example of this. The heat is finding the “easy” paths to escape and avoiding the “hard” paths. The most important achievement of this simulation is its ability to visually prove that the UDF and UDM method can successfully create a material with complex, spatially varying properties, demonstrated by the uniquely distorted temperature contours that could not be produced with any standard material model.

Figure 3 The final temperature contour from the Heat Transfer Simulation Fluent, showing the direct result of applying a UDF and UDM to model non-uniform heat flow.