Simulating Electric Motor Cooling with Ansys FreeFlow: Oil Spray and Water Jacket Methods

Hello everyone, this is Batuhan from Ozen Engineering. In this video, I will be demonstrating how to set up and run electric motor cooling simulation in Ansys FreeFlow. FreeFlow is Ansys' smooth particle hydrodynamics, or SPH-based solver.

Unlike traditional CFD, which relies on complex meshing, SPH uses particles to represent the fluid. This makes it especially powerful for simulating free surface flows, splashing, spray nozzles, and oil jet cooling, which are applications that are very important for modern electric motors.

Now let's go to the interface of Ansys FreeFlow. Here is the geometry of the motor we will be using. The housing contains an embedded water jacket, and the endcaps are drilled with nozzles that spray oil directly onto the end windings.

The rotor shaft, bearings, and magnets are also included to simulate the rotation effects. The cooling circuit has one inlet at the entrance of the water jacket and three outlets: one at the jacket end and two at the bottom of the housing for oil drainage.

To start the workflow setup, we will boot the simulation by following the workflow tree on the left side of FreeFlow. First, we will enable the SPH density monitor, mass flow rate, and density monitor. Then we will change the gravity's direction to the negative x-axis.

We will change the default fluid properties from water to engine oil. For the rotor rotation, we assign 1000 rpm clockwise; for now, you can increase it later based on your requirements, to the rotor core, shaft, magnets, and bearings.

This allows us to study how the rotating components interact with the oil spray. Now let's move into SPH settings. We set the kernel size to 1 m; a smaller kernel will give more detail but also increase the computational time.

The artificial sound speed is set to 100 m per second, which is roughly 10 times the maximum expected velocity, ensuring numerical stability. The time step is 1.1 seconds, and we will simulate this simulation for 10 seconds. Now let's assign the boundary conditions.

The inlet has a surface area of about 188 mm^ 2. We assume a flow of 10 liters of oil per minute, which corresponds to an inlet velocity of 1.54 m/s. With this design, FreeFlow generates around 664,000 SPH particles. And the full simulation takes about 4 hours to complete.

Once the simulation is complete, we can switch to post-processing. Here, we are using an oiler reconstruction to visualize the oil as a continuous flow. We can see the oil filling the jacket, flowing through the channels, and spraying from the nozzles directly onto the end windings.

At the nozzle exit, the velocity peaks at around 1 to 1 m/s, creating a strong spray effect on the windings. In the rising channels, velocity drops to around 1 m/s.

Within the first second of the simulation, the nozzle starts spraying as the channels fill, pressure builds, strengthening the spray and improving the heat transfer at the windings.

One of the most powerful features of FreeFlow is that it can be coupled with other Ansys solvers for multi-physical analysis. With a one-way thermal coupling, FreeFlow can export heat transfer coefficients directly into Ansys Mechanical for extended thermal analysis.

In a two-way fluid-structure interaction setup, FreeFlow sends the SPH-calculated forces to Mechanical, which computes the deformation and then fits the displacements back into FreeFlow. This allows us to capture the interaction between the fluid and the structure in real-time.

Thank you for watching, and stay tuned for more demonstrations on advanced electric motor simulation methods. Please contact us at https://ozeninc.com/contact for more information.