CFD and experimental analysis of snow distribution characteristics on double-gable roofs under multi-snowfall conditions

In cold regions, snow on roofs can persist throughout the entire winter. With the continuous occurrence of multi-snowfalls, the snowpack undergoes complex processes, like snow drift, accumulation, melting, recrystallization, and redistribution, which make snow distribution characteristics differ significantly from single snowfall events and pose threats to structural safety. This paper mainly investigates snow distribution on double-gable roofs under wind-snow combined effects during multi-snowfall events. Firstly, based on the combination of the modified Mixture model and the dynamic mesh, a numerical method for the multi-snowfall study was proposed. Then, multi-snowfall experiments using real snow particles were conducted in a cryogenic wind tunnel to validate the numerical model. Finally, using the validated approach, simulations were conducted on the double-gable roof with a height-to-span ratio of 1/10 under different wind velocities, snow densities, and volume fractions. The results showed that the wind velocity had the greatest impact on snow distribution on the double-gable roof, with lower height-to-span ratios being more susceptible. As more snowfall events occur, the location of the maximum snow depth on the roof moves downstream along the direction of the inlet wind, and the average normalized snow depth decreases. However, when the wind velocity is high, the degree of non-uniformity in the snow distribution on the roof will increase. It should be noted that thermal effects such as melting were not considered in this study, and the snow surface was assumed to harden after each snowfall, thus no longer participating in subsequent snow drift processes.