DPM to VOF, Color Spray Gun CFD Simulation Tutorial

Description

A spray gun is a painting tool that uses compressed air from a nozzle to atomize a liquid into a controlled pattern. The spray nozzle operates by impinging high-velocity turbulent air on the surface of filaments or films of liquid, causing them to collapse into droplets with a wide range of sizes.

Spray guns are used to paint any surface. Liquid droplets leaving the nozzle with high velocity can sit over the surface and create a liquid layer, forming the paint layer.

The geometry of the present project was created using ANSYS Design modeler and is meshed in ANSYS Meshing. The mesh type is unstructured, and the initial mesh number equals 560,134.

Color Spray Gun with Wall Film Methodology

In this project, the formation of a paint layer over an arbitrary wall is simulated. Different modules, including the VOF multiphase model along with the DPM, are enabled to model such a process.

When enabling the two models of VOF multiphase and dpm simultaneously, two transition mechanisms of DPM-to-VOF and VOF-to-DPM become available, which the user can use to switch between discrete phase elements to the eulerian VOF model or vice versa.

This project uses the DPM-to-VOF model to model the transition of liquid droplets into a paint layer over a wall boundary. The particles are injected from the spray gun’s nozzle in a conical shape with a velocity of 2 m/s.

Furthermore, the liquid wall film boundary condition is also enabled over the painted wall so that the particles are first immobilized on the painted wall.

After some time, when the volume fraction of liquid droplets is high enough to form the paint layer, the discrete particles are converted into the VOF layer, forming the final paint layer on the wall.

Moreover, the automatic mesh adaptation is a mandatory option that needs to be enabled. Of course, small mesh cells are not needed in the stage where the trajectories of particles are calculated.

Once the particles are converted into VOF, the mesh cells must be made as small as possible to capture the paint layer interface with the surrounding matter (i.e., the air). Otherwise, the simulation results are subjected to error and inaccuracies.

On top of the mentioned activated models, the Realizable k-epsilon model is used to solve the turbulent fluid equations. The present study is performed in transient format and 3D.

Conclusion

The results show that the paint layer is formed on the wall. The particles accumulating on the painted wall were converted to a VOF layer, forming separate layers of paint. Due to the flow pattern formed inside the domain due to the particle injection jet, the formed VOF layers on the painted wall tend to move radially away from the wall’s center point.

This motion triggers the automatic mesh adaption, and as depicted in the final mesh image, mesh cells are made distinctively smaller than the mesh cells in some specific locations.

Moreover, even in the final time step, some wall film particles are still on the painted wall.  The value of lagrangian wall film particles under the dpm-to-of session box must be altered to remove the existing wall film particle when they touch the VOF layer to prevent the existence of both wall film particles and the VOF layer simultaneously.

Also, viewing the velocity vectors inside the computational domain makes it interesting to observe the flow pattern created by the injection jet flow.