ADVANCED METHOD OF SEISMIC VULNERABILITY FUNCTIONS OF WOODEN BUILDINGS IN CONSIDERATION OF REGIONAL CHARACTERISTIC AND CONSTRUCTION YEAR

－Based on the seismic index distribution of wooden houses in case of Hokkaido－

Abstract

 We must take measures to mitigate damages from large-scale earthquakes effectively and efficiently. It is important to determine the amount of damage reduction as a numerical target, select effective countermeasures through analysis of the damage factors, and intensively promote them. For estimating building damage in the effort to reduce loss and destruction, a damage index function was used; accordingly, a building was considered as a unit. For vulnerability function, area was a unit. To set disaster reduction targets according to region, a vulnerability function that strongly takes account of regionality and construction year is necessary. In this paper, the proposed methodology is the use of a vulnerability function incorporating regionality and construction year.
 To construct the vulnerability function, an estimating method using past earthquake damage data was used. This method reflects the characteristics of the area where earthquake damage occurred. We constructed the regional damage rate function of wooden houses by considering the results of a regional seismic diagnosis based on the damage index function. With this method, it becomes possible to evaluate the earthquake resistance of buildings in the area by the probability density distribution of the seismic performance index. If there is a distribution of seismic performance indices for each area, it is possible to construct a highly accurate regional damage rate function without damage data from the region.
 The following is shown as a method of constructing the regional damage rate function:
 a) We estimated the new damage index function with diagnostic results obtained from utilizing the 2004 version of the seismic evaluation method for wooden buildings.
 b) For construction year without diagnostic results, we proposed a method to estimate the seismic index distribution from the new damage index function and the damage rate of destructive earthquakes occurring close to the construction year.
 c) If the area's seismic diagnostic results were small, we proposed estimating the seismic index distribution from Bayesian inference with the diagnostic results of Mokutaikyo as the prior distribution and the results in the area as the likelihood function.
 d) From the above results, we estimated the seismic index distribution of wooden houses in Hokkaido. In houses built before 1950, the average of the scores of Hokkaido and Mokutaikyo (nationwide average) was almost the same. In houses built after 1960, the average for Hokkaido was 1.4 to 1.7 times that of Mokutaikyo. The standard deviation in the case of Hokkaido was larger than for Mokutaikyo, and diversity of seismic performance by various building construction methods was found.
 e) From the new damage index function and the regional seismic performance index distribution of wooden houses, the regional damage rate function in Hokkaido was estimated.