Hiroki YASUI

Study on the Effect of Variation in Damage Functions on the Number of Damaged Houses

Takaaki IKEDA

Various earthquake disaster prevention activities are being conducted in Japan, and earthquake damage estimation conducted by each municipality is one of them. Although there is no default evaluation method, the procedure used in many cases is to divide the target area into a mesh, evaluate the surface earthquake external forces for the assumed earthquake, and determine the number of damages within the mesh by using the damage rate corresponding to the earthquake external forces from the damage function. The surface seismic external force is obtained by multiplying the seismic external force at the base by the amplification factor of the surface soil, and site characteristics are taken into account. Although there are some issues in terms of damage mechanisms in actual earthquakes, such as the fact that the seismic external force within the mesh is constant, it is recognized to be useful to a certain extent because it can be used to assess damage over a wide area. However, it is recognized to be useful to a certain extent because it can be used for wide-area damage assessment.
In this study, we focus on the damage function and amplification characteristics, and examine the effects of each on the assumed results for a specific region. The damage function is an important factor that directly affects the calculation of the number of damages, and the amplification property is an important factor that directly affects the calculation of the earthquake external force on the ground surface. The damage assumption targets detached houses, and the target area is Maebashi City. The assumed earthquakes are those used in the Gunma Prefecture Earthquake Damage Assumption and a hypothetical near-field earthquake.
The damage function currently used is obtained by the cumulative distribution function of the standard normal distribution based on the residential damage rate data according to the earthquake ground motion level in the 1995 Hyogo-ken Nanbu earthquake. Due to the paucity of data, the damage function changes as additional damage data is added. Therefore, we created multiple damage functions by adding new data to the distribution range of the actual damage rate data, and examined the influence of the damage functions on the assumed results by intercomparing the damage assumption results.
The influence of nonlinear characteristics of the surface soil was examined for the amplification rate. Using seismic observation records of the base and the surface at the same time, we calculated the change of the amplification factor with the level of seismic motion and reflected it in the amplification characteristics used in the damage estimation. The results showed that the influence of the damage function was large, increasing by about 34% in the case of total destruction or more, by about 20% in the case of half destruction or more, and by about 6% in the case of partial destruction or more in the case of the most different case. Therefore, it was found that it is important to prepare not only existing damage functions but also multiple damage functions with high potential for mutual evaluation in earthquake damage estimation. The amplification characteristics of the damage functions were found to be practically useful in that the number of damages decreased compared to the existing damage functions for total destruction or more, half destruction or more, and partial destruction or more.
Although earthquake damage estimation should reproduce the damage process of actual earthquakes, it is necessary to evaluate a wide area, and there is a contradictory issue that detailed data and characteristics of each region cannot be fully reflected. Continued efforts to improve the accuracy of individual assessments while ensuring simplicity are necessary.