Porcelain tile is a building material characterized for its low water absorption (less than 0.5%) and other useful aspects, such as high mechanical and abrasion resistances, in addition to its aesthetical applications. High stable and homogeneous material handling is required in order to attain the final product specifications for each plate. After preparation of raw materials through milling and spray drying, the atomized powder is composed of granules with different sizes and moisture contents. The powder is stored in silos until the temperature is reduced and residual moisture is more evenly distributed. A practical concern of granular material handling is segregation. During silo loading, particles of different sizes may occupy preferential regions, decreasing mass transfer rates and generating non-uniform moisture distribution. In case of funnel flow during silo discharge, segregation effects increase. Thus, powder discharged from silos have variable properties over time, resulting in operational issues during later processes and standardization problems.

From a mathematical modeling perspective, flow of particles of various sizes while addressing the dynamics of moisture transfer and segregation is a major challenge. The aim of this research is to develop a coupled approach of Discrete Element Method (DEM) and Finite Element Method (FEM) to properly simulate the phenomena that occur during the storage of spray-dried powder in silos used for the production of porcelain tiles. Simulations will also provide a powerful platform to design distributors to minimize segregation and inserts to improve discharge flow.

DEM can be used to model wet granular flow during silo filling, with particular emphasis on estimating segregation (as shown in Figure 1) and the evolution of distinct flow regimes during silo discharge. Additionally, FEM can be applied to solve mass transfer among particles and moisture distribution inside the silo (Figure 2). Since granular flow is affected by the moisture content of the powder and cohesion parameters are significant in contact models, multiphysics coupled DEM-FEM is required to solve this problem.