C ompared to the traditional liquid coating technology, powder coating eliminates the problem associated with solvent emission, reduces the extra cost of solvent and allows easy recycle of over sprayed powder. It is a more environmentally friendly and also more economical process. Traditionally, powder coating has been mainly applied to parts where the surface fi nish quality is not critical, so that relatively larger (35 to 60µm) powders can be used. This is because smaller powders tend to agglomerate and lead to fl ow problems. In the past decade, however, there have been strong efforts to apply powder coating to new markets where powder coating has not been typically used-such as automotive clear top coatings, can and coil coatings, and coatings on high-end products (Misev and van der Linde, 1998). To ensure the quality of coating fi nish, fi ner powders must be used to avoid obvious orange feel on the coating surface. Finer particles, if applied properly, can produce much more uniform coatings that results in superb coating fi nish that cannot be achieved by regular-sized (35 to 60 µm) coating powder (Zhu and Zhang, 2004).As charged particles are conveyed out from a powder spray gun by air, particle trajectories are governed by aerodynamic This paper reports on a research project that studies the effect of particle size on the coating effi ciency and coating uniformity in a powder coating system using the computational fl uid dynamics as a modelling tool. The numerical simulations are conducted for different particle sizes with different distances between the spray gun and the coating part and different positions of the powder spray gun pattern adjuster sleeve (PAS). This study can provide detailed information on air fl ow pattern and particle trajectories inside the powder coating booth, and the coating fi lm thickness on the coated part as well as the particle transfer effi ciency (PTE). In numerical simulations, the air fl ow fi eld is obtained by solving three-dimensional Navier-Stokes equations with standard κ-ε turbulence model and non-equilibrium wall function. The second phase, the coating powder, consists of spherical particles that are dispersed in the continuous phase, the air. In addition to solving transport equations for the air, the trajectories of the particles are calculated by solving the particle motion equations using the Lagrangian method. It is assumed that particle-particle interaction can be neglected. The electrostatic fi eld is modelled by solving the Laplace equation.On a étudié l'effet de la taille des particules sur l'effi cacité et l'uniformité de l'enrobage dans un système d'enduction de poudre à l'aide de la simulation numérique des écoulements. Les simulations numériques sont effectuées pour différentes tailles de particules et différentes distances entre le dispositif d'atomisation et la région d'enduction, ainsi que différentes positions du dispositif de contrôle du mode d'atomisation (PAS). Cette étude fournit des informations précises sur le mode d'écoulement de l'air et l...