Vulnerability functions for wooden buildings using a probability distribution model of seismic performance

Vulnerability functions for wooden buildings are proposed, which employ a probability distribution model of seismic performance. An acceleration response spectrum is used as input to the functions. The probability distribution model is formulated based on two indices of seismic diagnosis data: stifthess and resistance force, and it is clarified that the bivariate normal distribution is a reliable model. Regarding estimation accuracy, the proposed functions are compared with experience-based vulnerability functions in past literature for three earthquake disasters. The proposed functions indicate fairly good accuracy when proper parameters of stiffness increase and resistance force increase due to non-structural walls are used.