With increased globalization and the exponential growth of human population, underground transit systems have become a necessity in metropolitan areas around the world. In addition, there is an increasing awareness of human comfort and safety inside these complex structures, and the need for accurate numerical modelling and analysis has become a priority as part of the design process. This modelling is generally carried out both network-scale and station-scale. Network-scale phenomena include the piston effect of trains, longitudinal ventilation of tunnel fires, and the long-term thermal absorption of the tunnel lining and surrounding soil. One-dimensional network flow solvers are well-suited to network-scale modelling. Station-scale flow features include smoke migration through stations as well as flow separations and their corresponding pressure losses. These complex flows necessitate more sophisticated three-dimensional flow solvers. As the two scales are highly interdependent, achieving good consistency between the 1D and 3D models is essential especially since the boundary conditions for the 3D models are most often supplied by the 1D runs. However, the complicated 3D designs rarely translate well into accurate 1D models. Thus, the simplification and assumptions made while creating the models introduce inconsistencies. In this article, a novel coupling approach is proposed to mitigate these inconsistencies by optimizing the local pressure loss coefficients within the 1D models. This approach is validated and tested on representative station models. Results promise better flow distribution and prediction using the 1D model. Furthermore, this coupling method would not just be limited to 3D computational data but could also be used with field measurements.
展开▼