A Jet in crossflow CFD simulation using VOF TO DPM method in ANSYS Fluent  is a critical tool for engineers designing the combustors inside modern jet engines. This process, called JICF, is how fuel is injected. A high-speed jet of liquid fuel is shot out of an injector directly into a stream of very fast-moving air, called the crossflow. The violent collision between the liquid and the air tears the fuel apart into millions of tiny droplets in a process called atomization. These droplets then mix with the air and burn to produce thrust. The success of the entire engine depends on this process being perfect. A Jet in crossflow fluent simulation is essential because the physics is incredibly complex, involving a liquid that starts as a solid stream but shatters into a dispersed spray cloud in microseconds.

A VOF TO DPM CFD simulation is special because it uses two different models that work together. First, the Volume of Fluid (VOF) model is used near the injector. The VOF model is an expert at tracking the surface of the continuous liquid jet as it bends and starts to break up due to the intense aerodynamic forces from the crossflow. As the jet breaks into large pieces, the simulation automatically converts these pieces into particles for the second model: the Discrete Phase Model (DPM). A VOF TO DPM Fluent simulation is very efficient because the DPM can track millions of individual droplets without needing an impossibly fine mesh. This allows a Jet in crossflow ANSYS simulation to capture the entire event: the initial jet bending, the primary breakup into large fragments, and the final spray cloud of tiny droplets. This gives engineers the data they need on jet penetration, spray pattern, and droplet size to design more efficient and lower-emission engines.

• Reference [1]: Leparoux, Julien, et al. “Primary atomization simulation applied to a jet in crossflow aeronautical injector with dynamic mesh adaptation.” 14th Triennal International Conference on Liquid Atomization and Spray Systems. Vol. 137. 2018.

Figure 1: The computational domain for the Jet in Crossflow CFD simulation, showing the boundary conditions, including the crossflow inlet and the fuel injector location. [1]

Simulation process: Fluent VOF-to-DPM Setup, Modeling Jet Breakup with Adaptive Meshing and SBES

The simulation process for this Jet in crossflow cfd study began with creating a structured mesh of hexahedral cells using ICEM CFD. This high-quality grid provides a strong foundation for accuracy. A key innovation in this ANSYS Fluent setup was the use of dynamic mesh adaptation. This powerful feature means that as the simulation runs, the software automatically makes the mesh cells smaller in the exact location of the liquid jet’s surface. As the jet bends and breaks apart, this zone of fine mesh moves with it. This technique is critical for a VOF TO DPM CFD simulation because it provides maximum accuracy for tracking the breakup without the massive computational cost of having a fine mesh everywhere.

Inside the Fluent solver, the simulation was set up as a transient (time-dependent) case, as atomization is an extremely fast, unsteady event. The k-ω SST turbulence model was chosen, enhanced with the Stress-Blended Eddy Simulation (SBES) submodel. This advanced hybrid turbulence model is perfect for JICF because it uses efficient RANS calculations in the simple parts of the flow and automatically switches to a more detailed, LES-like simulation in the chaotic jet breakup region, correctly capturing the large turbulent vortices that rip the fuel apart. The multiphase physics was handled by the VOF to DPM model. The VOF model was used to track the continuous liquid jet as it left the injector. The simulation was configured so that when a piece of VOF liquid broke off and became smaller than a specific size, it was automatically converted into a DPM particle. The DPM model was set up with two-way coupling, which is very important because it means the air moves the droplets, and the droplets also push back on the air, which is a critical piece of the physics. Finally, unsteady particle tracking was used, ensuring the droplets reacted realistically to the fast-changing turbulence, leading to an accurate prediction of the final spray pattern.

Figure 2:  close-up view of the structured mesh after dynamic mesh adaptation has been applied. The contours show that the grid cells have been automatically refined (made smaller) around the VOF liquid interface to ensure a highly accurate capture of the jet breakup.

Post-processing: Engineering Investigation of a Jet in Crossflow

The simulation results provide a complete, microsecond-by-microsecond record of the jet atomization event. By examining the contours as forensic evidence, we can reconstruct the journey of the fuel and deliver a clear engineering verdict on the injector’s performance. The velocity magnitude contour in Figure 3 shows the first stage of the fuel’s journey. The liquid jet shoots upwards from the injector at a very high speed, indicating velocities of 60-75 m/s. It immediately runs into the 50 m/s crossflow of air. The evidence of the interaction is the classic curved path of the jet. The jet starts off vertical, where its own momentum is dominant, but it is immediately bent over by the powerful sideways force of the crossflow. This path is determined by the momentum flux ratio—the competition between the jet’s upward push and the crossflow’s sideways push. From this contour, we can make a critical engineering measurement: the jet penetration depth. We can see that the solid VOF liquid core penetrates about 15-20 injector diameters into the crossflow before it is completely broken apart into a spray. This is a key performance metric for a designer, as it tells them exactly how far the fuel will travel into the combustor.

Figure 3: Velocity magnitude contour from jet in crossflow CFD simulation in Ansys Fluent showing liquid fuel jet (orange-red, 60-77 m/s

The real magic of this simulation is the ability to see the breakup process. The solid orange structure in Figure 3 is the liquid being tracked by the VOF model. The scattered blue dots are the droplets being tracked by the DPM model. The point where the solid orange structure ends and the cloud of dots begins is the region of primary atomization. This is where the simulation performs the successful VOF to DPM conversion. The particle diameter contour in Figure 4 gives us a forensic look at this moment. The large, curved structure, which tells us that the very first droplets created by the breakup have diameters of 49-70 microns. This is physically correct; the initial breakup, driven by aerodynamic shear (known as Kelvin-Helmholtz instability), rips off relatively large fragments and ligaments of fuel, not fine mist.

After the primary breakup, the spray cloud develops, and the droplet diameter contour reveals its internal structure. This contour provides a complete spatial map of the droplet size distribution, which is one of the most important goals of the simulation. The evidence is clear:

• Largest Droplets (70-90 µm): These are found concentrated near the jet’s original path. Because they are heavy, they have high inertia and are not easily pushed aside by the turbulent air; they tend to follow the original trajectory of the liquid jet.
• Smallest Droplets (8-30 µm): These are found scattered widely around the edges of the spray. Because they are very light, they have low inertia and are easily thrown around by the chaotic turbulent eddies in the crossflow. This is evidence of turbulent dispersion. Some of these small droplets may also be created by secondary breakup, where a larger droplet becomes unstable and shatters into smaller ones.

Figure 4: Particle diameter

The particle velocity contour in Figure 4 tells the final part of the story. The colors show that droplets freshly created from the breakup are traveling very fast, in the range of 79-111 m/s. Interestingly, the simulation shows some droplets reaching up to 142 m/s. This is not an error; it is a real physical effect where the surface tension energy stored in the stretched liquid is released explosively when a ligament snaps, launching the new droplets at very high speed. As the droplets travel further into the spray cloud, they slow down due to air drag. We can see droplets colored cyan (49-79 m/s) and finally blue (17-49 m/s). This shows that the smallest droplets (which have a high surface-area-to-mass ratio) are slowed down much more effectively by air resistance than the large droplets.

The most important achievement of this simulation is the successful and physically accurate capture of the entire jet in crossflow atomization process, from the continuous liquid jet to the final dispersed spray.

For a gas turbine combustor designer, this is not just a set of contours; it is invaluable engineering intelligence:

1. It Validates the Injector Placement: The prediction of a 15-20 diameter penetration depth tells the designer exactly where to place the injector. If it is too close to the opposite wall, the liquid fuel will hit the hot metal (a dangerous condition called wall wetting). If it is too far, the fuel will not reach the hot core of the combustor for efficient burning.
2. It Predicts the Fuel-Air Mixture: The simulation provides a complete map of where the fuel droplets of different sizes are going. This allows the designer to shape the combustor and the airflow to ensure that the fuel and air mix perfectly, leading to higher combustion efficiency and lower pollutant emissions (like NOx).
3. It Enables Virtual Prototyping: With this validated model, the designer can now test dozens of new injector designs on the computer. They can change the injection angle, the pressure, or the orifice shape and see the results in a few days. This allows for rapid innovation and saves millions of dollars and months of time compared to building and testing physical prototypes.