AMOUR AissaENCA/MENASRI Noureddine2025-01-072025-01-072025-01-07DOC/GM/2024https://dspace.univ-msila.dz/handle/123456789/45449Particle-laden flows are crucial for specific industrial applications but can have undesirable consequences, especially in industries like cement manufacturing that heavily rely on centrifugal fans. Operating under demanding conditions with a notable load of erodent solid particles poses challenges for these fans. Despite their widespread use, erosion in various fan types remains insufficiently explored. This thesis focuses on investigating the erosion process of an induced cement mill fan (FN-280). In pursuit of this objective, Unsteady three-dimensional numerical simulations using the Eulerian- Lagrangian approach were conducted to predict the fan’s aerodynamic characteristics and particle dispersion. Results are compared with experimental data and on-site inspections to assess the simulation’s effectiveness in predicting the fan’s performance and erosion patterns. Additionally, a thesis section examines how altering the number of blades and the blade’s outlet angle affects overall performance and erosion patterns. The numerical fan’s performance curve closely aligns with experimental data, validating the model’s reliability. Findings underscore the model’s ability to accurately replicate actual erosion patterns, emphasizing the substantial influence of particle size, flow conditions, and particle load on patterns and erosion rates. It is shown that increasing the blade outlet angle significantly intensifies erosion on the blade’s pressure side, with the primary factor driving higher erosion rates being the interplay between particle impact angle and frequency. Adjusting the blade count has minimal impact on erosion patterns on the blade pressure side compared to changing the blade outlet angle, but a noticeable trend emerges with fewer blades, resulting in more pronounced erosion rates. This is suggested to be attributed to the higher local impact frequency on blades featuring fewer blades.enThesis