Research And Its Application Of Earthquake Loss Based On Detailed Earthquake Damage Data

Earthquake losses are largely caused by the damage to engineering structures.The study of earthquake losses on structures is of great significance for earthquake loss prediction and rapid assessment of post-earthquake losses,and can also provide persuasive evidence for the improvement of existing seismic design methods.Compared with the analytical method,the results obtained by the empirical method based on actual earthquake damage data are more reliable,but destructive earthquakes do not occur frequently,and for various reasons,there are very few detailed earthquake damage data of buildings for seismic damage studies,which leaves a lack of in-depth research on earthquake losses.Based on the detailed damage records of more than 1.5 million buildings in 27 historical earthquakes in New Zealand,and the detailed damage data of nearly 600,000 houses collected after the 2011 Great East Japan Earthquake,this paper firstly analyzes the characteristic relationship between ground motion and structural loss by using actual data.Then,a model is established by using the Open SEES platform for dynamic analysis.The similarities and differences of the results obtained by the analytical method and the empirical method are extracted,which are used to improve the existing seismic design standards.The following conclusions were reached.(1)Based on the analysis of the historical detailed earthquake damage information of nearly one million houses in New Zealand and Japan,we found the existence of two boundary conditions of seismic loss distribution for the same ground motion.The current seismic loss distribution model cannot consider the boundary conditions,so an innovative mixed probability distribution model of seismic loss that can consider both the distribution and its boundary is proposed to describe the uncertainty characteristics of structural loss.The parameters of the proposed distribution model were determined using New Zealand earthquake damage data through parameter estimation and regression analysis,and the applicability of the proposed distribution model was verified by comparing earthquake loss of specific types of buildings in New Zealand and Japan.Finally,the ground motions that induce identical degrees of structural damage are studied,and it is found that there is a large amount of uncertainty associated with relationship between ground motion and structural earthquake loss.(2)Using the Open SEES platform,one-story and two-story wood frame building models were built by simplification method for incremental dynamic analysis,and the seismic loss results obtained were compared with the actual situation,it is found that the structural vulnerability obtained by the analytical method is in good agreement with the actual results,and the uncertainty characteristics of the seismic loss obtained by the two methods have strong consistency within the boundary.However,the accuracy of the analytical methods needs to be improved to consider the boundary conditions,and the main reason is that the results obtained by the analytical method are less discrete compared to the actual seismic loss.It shows that more uncertainties need to be considered in the subsequent seismic loss research based on analytical method,especially the uncertainties of the structural limit states and performance quantitative indicators.(3)Based on the actual seismic performance of buildings with seismic-design and the results of the analytical method,the existing seismic design code are discussed.It is concluded that the existing seismic design standards have the following deficiencies: only a few points of the infinite ground motion intensity that the building may encounter in the future are considered;the uncertainty of structural response is not considered;it is difficult to control the overall performance of the structure using only a single deformation indicator;and the structural performance is described in qualitative terms.Based on the above,this paper proposes a seismic design method that considers full-probability and uncertainty,with an example to exemplify the method’s specific application.