A stochastic model for predicting wind-induced damage to wooden houses in snowy cold region has been developed, focusing on roofs. The model consists of ‘wind loading model’ and ‘wind resistance model’. These models consider the regional features of the structures and the aged deterioration of the roofing. The evaluation methods of the wind resistant performance and the wind load acting on the roofing are based on the building code and previous studies. Furthermore, pulling-up load test was carried out to evaluate the wind-resistant performance of roof frame joints. It is found that the materials and construction of roofs commonly used in snowy cold regions tend to cause much wind damage to roofing and roof structures. The results of a probabilistic analysis of the damage indicate that damage to roofs can be reduced significantly by appropriate maintenance.
The effects of support condition and dimension of glass plate on the wind load resistance are investigated by performing full-scale breakage tests and numerical simulation that can predict glass failure pressures. Full-scale breakage tests were carried out using window specimens whose edges were supported as used in practice, called “elastic support condition”, in order to investigate the application of the existing numerical simulation to such a support condition. The results indicate that the numerical simulation can capture the load resistance obtained from the full-scale breakage tests by adjusting the parameters related to the material properties and testing environment, especially relative humidity. By using the numerical simulation, the examination of the effects of support condition and dimension glass plate, and the comparisons with the Japanese and USA design codes are performed. The results show that the support condition of glass plates affect slightly on the wind load resistance. On the other hand, the load resistance depends on size and aspect ratio more significantly than expected in the Japanese code. Therefore, it is found that the Japanese code seems to overestimate the wind resistance of glass plates because it does not reflect the effects of size and aspect ratio of glass plate, which are considered in the USA code.