Dispersion of particles and airflow optimization for a semiconductor cleanroom

Abstract

Electronic cleanrooms' airflow and particle behavior significantly impact product quality and energy consumption. The airflow and particle diffusion at different supply air velocities are calculated for a semiconductor cleanroom, well-validated against established data. An RNG k-ε model and discrete phase model (DPM) are employed to simulate airflow and particle transport. Transient distributions, residual counts, and removal times are compared for six air supply velocities ranging from 0.2 m/s to 0.12 m/s. The results indicate that vortices are easily generated in the corners formed by the walls and the ceiling edge. The local vortices result in a longer particle removal time. Under stable conditions, all particles released from the ceiling, walls 1, 2, 4, and the raised floor are removed from the cleanroom across six supply air velocities. Two particles released from wall 3 remain as residuals when the supply air velocity is reduced. The final residual particles are primarily located near equipment 6. When the supply air velocity decreases from 0.22 m/s to 0.16 m/s, the stable time remains at 14 min. However, it increases significantly to 17 min and 23 min at 0.14 m/s and 0.12 m/s, respectively. The addition of two Fan Filter Units (FFUs) in the corners of the ceiling enhances both airflow and particle diffusion, reducing removal times by 2–7 min across different air speeds. These results provide theoretical guidance for selecting supply air velocity and optimized the arrangement of FFUs and equipment.