Multiscale Modeling of Powder Coating with a Moving Spray Gun

Abstract

Powder coating is a finishing process where electrostatically charged particles are sprayed onto metal surfaces, forming a uniform and durable layer after thermal curing. This work addresses the modeling and optimization of the process through high-fidelity numerical simulation, adopting the multiscale P-DNS (Pseudo-Direct Numerical Simulation) method to capture the turbulent flow of charged particles interacting with an electrostatic field. Particles are simulated in representative volume elements (RVEs), and their dynamic response is homogenized. By non-dimensionalizing governing parameters, including electric and gravitational forces, previously trained reduced models can be reused in coarsescale simulations. A boundary condition is also introduced to quantify coating layer thickness. The tool is validated against literature results and applied to the coating of complex geometries, analyzing applied voltages and trajectories of a moving spray gun. The reliable results demonstrate the methodology’s potential as a fast and efficient predictive tool for powder coating technologies.