In February 2014, owing to heavy snowfall in the Kanto-Koshin region, several buildings were severely damaged. A major cause for the damage was the load on building surfaces caused by heavy snow followed by a surcharge load due to the ensuing rainfall. Countries such as the United States and Canada have established standards on snow loads, which also take into account the surcharge load due to rainfall. In contrast, Japan has not adopted such standards, and hence it is imperative to establish a method for calculating loads added by rainfall to snow loads. Therefore, in this study, experiments were conducted using artificial rainfall on roofs covered with snow to evaluate and propose a calculation method for loads added by rainfall. First, outdoor experiments as well as indoor experiments in a low-temperature experimental facility were conducted. Model roofs with different span lengths and gradients were built and loaded with natural and artificial snow. These model roofs were then treated with rainfall at constant intensity by using an artificial rainfall simulator. The increase in load was measured and several observations were made from the experiments. A smaller roof gradient and larger roof span resulted in a greater peak value of the load added by rainfall. For roofs of the same shape, the peak value of the load added by rainfall increased as the initial snow depth increased. Next, the relationship between surcharge loads due to rainfall and the span length and gradient of the roofs was quantified. The load added by the rainfall was treated as the approximate square root of the snow depth, and a coefficient a, which depends on the span length and gradient of the roof, was defined. However, valuable data were obtained from the experiments for only five types of roof geometries. To estimate the values of the coefficients a for other roof geometries, regression analysis between the coefficient a, roof gradient, and span length was conducted. Comparing the surcharge loads, which were calculated using the estimated coefficients a, with the loads calculated using the O'Rourke equation, a relatively good correspondence between both the results was confirmed. Snow loads that take rainfall into account were obtained for the given regions with corresponding values of the design snow depth. This was achieved by including the load due to rainfall, which is correlated to the snow depth, computed from the proposed formula and the estimated coefficients. In addition, methods that include coefficients reflecting the regional climatic conditions were proposed. One such method is based on the fact that winter rainfall is not always observed during the period when the deepest snow cover occurs. Extreme values of the snow depth were obtained from the meteorological data to derive the ratio of snow loads on rainy days against the maximum winter snow loads; the ratio was treated as the coefficient k1. Another method considers the regional meteorological cases wherein the surcharge load due to rainfall does not reach the peak value. The ratio of the surcharge load value due to such rainfalls against the surcharge load calculated by the proposed formula was treated as the coefficient k2. However, because the influence of coefficient k2 is relatively small, it can be considered as 1 in practice. Moreover, because the impact of loads added by rainfall is relatively small in heavy snowfall regions, the regions where it is necessary to consider the loads due to rainfall may be limited to general regions where the design snow depth is 1 m or less.
各種検索
サポート
ヘルプ