In this CFD Analysis of Magnetic Force Effect on Nanofluid tutorial, we provide a complete training guide on enhancing heat transfer in compact heat exchangers. Modern cooling systems often use nanofluids—a mixture of a liquid base like water and tiny solid particles—to carry more heat. When these particles are magnetic (like Fe3O4), we can control their movement using an external magnet. This interaction creates a Kelvin Force that physically pushes the fluid, mixing it near the walls to cool surfaces faster. This study focuses on Nanofluid in tube flattening, a design where round tubes are pressed flat to increase surface area. While flattening helps air cooling, it often slows down the internal fluid flow.

By using CFD – Magnetic Force Effect on Nanofluid methods, we can simulate how magnetic fields improve this flow. We use ANSYS Fluent combined with a custom MHD field UDF to model the complex interaction between the magnetic field and the nanoparticles. This simulation helps engineers design radiators that are smaller but more efficient. In this lesson, we will focus specifically on how to interpret the custom physics defined in the software. For more details on custom coding, please explore our UDF tutorials.

• Reference [1]: Safikhani, Hamed, and Abbas Abbassi. “Effects of tube flattening on the fluid dynamic and heat transfer performance of nanofluids.” Advanced Powder Technology3 (2014): 1132-1141.

Figure 1: Mesh Grid of the flattened tube cross-section showing the high-quality structured hexahedral cells generated for the Fluent – Magnetic Force Effect on Nanofluid simulation.

Simulation Process: MHD UDF Physics and Nanofluid Mixture Setup

For this technical CFD – Magnetic Force Effect on Nanofluid simulation, we generated a high-quality structured grid with 780,288 hexahedral cells using ICEM CFD. A fine mesh is critical to capture the subtle velocity changes induced by the magnetic force. We set up the physics in ANSYS Fluent using the Mixture Model. The primary phase is water, and the secondary phase is Fe3O4 (Magnetite) with a 5% volume fraction.

The most critical part of this setup is the MHD field UDF. Standard ANSYS Fluent solvers do not have a built-in Kelvin Force model for non-uniform magnetic fields. Therefore, we had to define this physics manually using a C-code function. Inside this code, we calculated the Magnetization (M) of the nanoparticles using the Langevin function, which changes based on the local temperature and particle diameter. We also defined the Magnetic Field Intensity (H) generated by a wire located at a specific distance. Finally, the code multiplies these values to calculate the momentum source terms (xsource and ysource). This tells the software exactly how much force to apply to the fluid at every single cell, simulating the magnetic pull on the Fe3O4 particles.

Post-processing: CFD Analysis of Magnetic Force Effect on Nanofluid and Field Distribution

In this section, we provide a deep engineering analysis of the results. We interpret the contours and graphs to evaluate the effectiveness of the magnetic mixing. First, we must verify the flow regime using the Velocity Profile in Figure 2. This graph plots the velocity magnitude along the vertical centerline of the tube. The curve forms a perfect parabola, which is the textbook definition of fully developed Laminar Flow. The peak of the parabola confirms the Maximum Velocity is 0.033 m/s. From an engineering perspective, this graph tells us the main problem: the velocity near the walls (where the heat transfer happens) is almost zero. This thick boundary layer acts like insulation, trapping heat. The goal of our Magnetic Force Effect on Nanofluid simulation is to distort this perfect parabola and push fresh fluid toward the walls.

Next, we analyze the Velocity Contour in Figure 3. The flow pattern is elliptical, matching the flattened tube shape. While the red core represents the high-speed center, we look for asymmetries caused by the magnet. The contour shows a slight disturbance near the bottom right, but the overall flow remains dominant. This suggests that while the force is present, the magnetic field might need to be stronger or the magnet placed closer to have a global effect on the flow field. We can see the cause of this behavior in the Magnetic Field Distribution (Figure 5). This contour displays the field intensity () stored in our UDM-1 slot. The majority of the tube is Red (low field), but there is a distinct Blue Spot at the coordinates we defined in the UDF. In this small region, the Peak Magnetic Field is -2.41 × 10⁴. This proves that our UDF is working correctly: it is generating a massive magnetic gradient exactly at the wire location. This hotspot pulls the nanoparticles violently toward the wall in that specific corner.

Figure 2: Graph of Velocity Profile along the vertical centerline, verifying the parabolic flow shape and maximum velocity of 0.033 m/s.

Figure 3: Contour of Velocity Magnitude for the Fe3O4 mixture, showing the elliptical high-speed core and low-speed zones near the walls.

Figure 4: Temperature field distribution showing the thermal gradients and asymmetric cooling caused by the magnetic force.

Figure 5: Contour of Magnetic Field (Hy Component) stored in UDM-1, highlighting the strong field concentration region (-2.41e4) near the external magnet location.

Finally, the Temperature Field in Figure 4 confirms the result. We can see a localized cooling effect near the magnet’s location. The magnetic force has successfully disrupted the thermal boundary layer in that zone. Conclusion for Manufacturers: The simulation proves that a single wire magnet creates a local improvement. To fully optimize this Nanofluid in tube flattening design, engineers should place multiple magnets around the perimeter. This would create mixing forces on all sides (not just one corner), destroying the laminar boundary layer entirely and significantly boosting the radiator’s efficiency.