Particle-resolved simulation of particle removal dynamics from fibrous filters in the pulse-jet cleaning process

This paper presents a detailed numerical investigation of particle-removal dynamics in fibrous filters subjected to pulse-jet cleaning. A moving wall boundary model based on the signed distance function was developed to simulate a realistic filter motion and its interaction with the fluid and particles. The computational domain was reconstructed using x-ray computed tomography imaging, which enabled an accurate representation of the microstructural geometry. A coupled computational fluid dynamics–discrete element method framework was employed to analyze particle trajectories, clustering behavior, and cleaning efficiency under varying operating conditions. Three key parameters were systematically varied: the filter movement acceleration, reverse-flow velocity, and initial deposition density. The results showed that higher filter acceleration and increased reverse-flow velocity significantly enhanced particle detachment and dispersion. Lower initial deposition volumes promoted early and complete removal owing to reduced interparticle cohesion. Quantitative metrics, including cleaning efficiency, particle number distribution, nondimensional boundary area (NBA), and average kinetic energy, were used to characterize the evolution of particle dispersion and agglomeration. The NBA metric effectively captured the transitions between the dispersed and aggregated states, whereas kinetic energy analysis revealed the influence of deposition density on energy transfer and particle ejection. This study confirmed that both mechanical impulse and aerodynamic force play critical roles in cleaning performance. These findings contribute to a deeper understanding of the pulse-jet cleaning performance and offer guidance for optimizing filter design and operational parameters.