Analysis of wake characteristics of clustered pedestrians in transportation hub transfer passages using proper orthogonal decomposition (POD)

The walking of clustered pedestrians strongly influences the indoor flow field structure in transportation hub with high passenger flows. This study analyzes the wake characteristics of passengers in clusters walking in long and narrow channels by combining the proper orthogonal decomposition (POD) method with the dynamic mesh technique. The large eddy simulation (LES) is used to simulate the wake distribution of full-scale human motion. The snapshot POD method is used to decompose the instantaneous velocity field to obtain different modes energy and temporal coefficients. POD results show that the first mode contributes approximately 84–87% of the total energy, indicating that a small number of modes can accurately reconstruct the primary flow structures. The influences of the wake-core region and wake-interference region are studied by analysing the wake velocity field and vorticity field in different regions, revealing multi-scale vortex interactions and the evolution of secondary vortices in clustered pedestrian wakes. The proposed approach establishes a reduced-order analysis framework for understanding clustered pedestrian wake dynamics and provides a physical and methodological foundation for future flow field prediction and ventilation optimisation in crowded indoor environments.