Research on snow distribution of long-span flat roof considering interference effect

This study employs Computational Fluid Dynamics (CFD) to perform numerical simulations of snow distribution on two categories of long-span flat roofs considering interference effect. Initially, through a comparative analysis of simulation results with field measurements documented in prior literature, the feasibility of the modified Eulerian-Eulerian method was validated. Building upon the foundation of model validation, an exploration was undertaken into the snow distribution on the superimposed-shaped long-span flat roof, considering varying wind velocities and directions. The finding revealed a notable reduction in snow accumulation on most roof areas, while an increased accumulation phenomenon was observed at the roof surface overlap position (maximum snow distribution coefficient reached 1.49). Subsequently, an investigation was conducted into snow distribution on the complex-shaped long-span flat roof, considering the presence or absence of surrounding architectural interference. The result indicated a significant impact of surrounding structures on the distribution of roof snow accumulation, with the maximum snow distribution coefficient recorded at 1.48 (compared to 1.38 without surrounding architectures). Moreover, a comparison was performed between CFD numerical simulation results and prevailing snow load standards. It is evident that ISO 4355:2013 offers accurate predictions with a maximum error of 2%. The predictions of other standards are conservative or unsafe. The results in this paper can provide a reference for predicting snow loads on long-span flat roofs. Graphical Abstract