Experimental study of pedestrian flow through right-angled corridor: uni- and bidirectional scenarios

Abstract

Corners can be commonly observed in most building facilities. However, pedestrians’ turning behavior at the corners, especially in collective movements, is rarely studied and not fully understood. To investigate the effects of such configuration on pedestrian flow, both uni- and bidirectional experiments were conducted in a right-angled corridor. From the fundamental diagram, it is found that pedestrians in our experiments are less sensitive to high-density situations and the velocity at high densities tends to be larger than observed values in former studies. Besides, in our experiments, no noticeable difference is observed between the fundamental diagrams in uni- and bidirectional scenarios for densities below 2 ped m−2. According to the density profile, pedestrians in unidirectional turning movements tend to seek the shortest path, whereas their followed path is more influenced by the detour behavior against encounters when it comes to bidirectional scenarios. Besides, due to the collision avoidance behavior and lane formation phenomenon in bidirectional scenarios, the highest density does not emerge at the innermost side of the corner, but is slightly outward in high-density case. According to the position-velocity relation, generally speaking, pedestrians’ velocity will decrease as they approach the corner. But for the outer most lane in bidirectional scenarios, pedestrians’ velocity may increase as there is more space available. As for angular velocity, we observe that generally in bidirectional streams, pedestrians from the inner lane are more likely to possess a higher value than those in outer lane. But some exceptional cases exist, for example, when a certain inner lane is severely constrained or pedestrians from outer lane try to avoid physical contact with those in inner lane. The results obtained can serve as a basis for model validation and offer more insights into the dynamics of pedestrians in geometries with corners.