Earthquake Damage Evaluation Models for Reinforced Concrete Buildings and Wooden Houses

Based on the recent advance on strong motion seismology researchers try to predict strong ground motions for future earthquakes in order to reduce the damage caused by them. From the investigation of the disasters caused by recent inland earthquakes such as the Hyogo-ken Nanbu earthquake and the Northridge earthquake, the cause of casualties in cities is mainly related to building damage. In this regard we have to evaluate building damage quantitatively for urban disaster mitigation. We try to establish a set of building models that can reproduce the observed damage ratios in Kobe during the Hyogo-ken Nanbu earthquake. Once we establish these models we can evaluate the building damage that future earthquakes will cause as long as we have simulated strong motions for future earthquakes. We can also use them to determine the impact of observed ground motions on building damage. First we made a set of models (Building Code Models) by following the building code enforced in 1981 in Japan. We considered the different characteristics with respect to the differing number of stories and building types. We used four representative models with 3, 6, 9, and 12 stories for reinforced concrete (RC) buildings, and one model with 2 stories for wooden houses, respectively. Then we calculated damage ratios from building damage statistics surveyed in Kobe after the Hyogo-ken Nanbu earthquake for each model. For RC buildings we considered the large modification of the building code in 1981, i.e., we created models for buildings built before and after 1981. We made dynamic response analyses based on simulated strong motions by Matsushima et al. (2000) and calculated the damage ratios. We found that BCMs were too weak to reproduce the observed damage ratios even for buildings built before 1981. So we repeated analyses modifying the strength of BCMs until the calculated damage ratios matched the observed damage ratios. We found that the yield strength estimated by the method described above was much larger than that is considered in the code, especially for low-rise buildings. Using the established models we succeeded in reproducing the belt-shaped area with high damage ratios in Kobe. According to the distribution of maximum inter-story drift these models could simulate the observations that middle-rise (6 or 9 stories) buildings collapsed in one of their lower stories (1st to 3rd story) in Kobe. We apply this set of models to ground motions for recent earthquakes and the hypothesized scenario earthquake for Fukuoka City, Kyushu, Japan to see the structural damage impact of these ground motions.