Estimation method of roof snow load using acceleration measurements: a case study on actual wooden building

Abstract In Japan's snowy regions, the significant accumulation of snow makes it impractical to design structures that can withstand the maximum snow load. To address this, many buildings are designed under the assumption that snow will be removed from the roof, allowing for a reduction in the maximum snow load. This approach is commonly used for one- and two-story wooden buildings. However, snow removal is a hazardous task, leading to casualties almost every year. Therefore, determining the optimal timing for snow removal is critical to prevent structural collapse while minimizing the frequency of removal. Typically, snow depth serves as the criterion for this decision. However, estimating roof snow load based solely on snow depth is challenging due to significant variations in snow density caused by consolidation and rainfall. This study proposes a method to estimate roof snow load by measuring the acceleration response of the structure. The snow load is estimated by analyzing fluctuations in the natural frequency before and after snowfall. The practicality of this method was validated through measurements of acceleration response, snow depth, and snow density on an actual two-story wooden building over three years. The estimated snow load, derived from changes in natural frequency, was compared with measurements obtained using a snow sampler and continuous photographic monitoring of the building. The snow load estimated from the acceleration response showed good agreement with the sampler measurements, with a maximum error of approximately 30%. And it was found that snow removal was often carried out when the actual snow load was between 30 and 70% of the design snow load. This indicates that snow removal is performed more frequently than necessary, as the estimated snow loads are well below the design safety limits. In summary, the quantitative evaluation of snow load in this study suggests that unnecessary snow removal is prevalent. The method proposed in this paper has the potential to reduce accidents and costs by minimizing unnecessary snow removal from roofs.