Study On Nonlinear Statistical Characteristics Of Surface/downhole Response Spectrum Ratio

Ground motion field effect is an important factor to be considered for seismic fortification of engineering structures.The nonlinear dynamic behavior of soil under strong motion and its influence on the nonlinear ground motion field effect are one of the difficulties.In recent years,with the accumulation of observation records of vertical drilling platform array,the surface/downhole spectral ratio method using downhole platform as a reference station is still widely applied to study the nonlinear effect of ground motion in the field.Since the establishment of the KiK-net strong motion observation network after the Hanshin earthquake in 1995,a large amount of strong motion observation data has been accumulated,which provides the basic data for the empirical study of nonlinear dynamic characteristics of soil based on strong motion observation.In this paper,based on 141881 sets of surface/downhole acceleration records from 136 vertical borehole array of KiK-net strong motion observation network in Japan,the nonlinear variation of surface/downhole response spectrum ratio platform value and the main influencing factors of the ground motion intensity are studied.The main work of this paper is as follows:1.If the ground motion intensity is weak,the dispersion of the platform value of the surface/downhole response spectrum ratio will be large,and the sliding window average method can effectively enhance the linearity of the data.The slope of the regression line is defined as the ratio of surface/underground response spectrum platform value of the nonlinear attenuation index,its value between 0.24 to 0.08 and the median is 0.168.When the downhole acceleration peak value is 500 cm/s2,the platform value of the surface/downhole response spectrum ratio is about 35%to 78%of that under the action of weak vibration(peak downhole acceleration 0.316-1cm/s2),its median is 54%.2.Based on the selected 136 arrays in the ground motion intensity grouping of the average length of sliding window to determine the minimum record related to the intensity of earthquake ground motion stations,sample size,think up in sliding window width takes 30 set of strong motion records,at least 1 set of downhole>100 cm/s2 peak acceleration of strong motion records can ensure the statistical conclusion has higher reliability.3.The shear wave velocity profile of KiK-net array was collected and sorted,and 12 field condition characterization parameters were calculated:The equivalent shear wave velocity(VSE)defined by Chinese code,the average shear wave velocity(VS30)defined by American code,the shear wave profile gradient(B30)of 30m,the overburden thickness(Z0.5?Z0.8?Z1.0)of 0.5,0.8,1.0 km/s,the average shear wave velocity(Vsav)of the overburden,and the field Ground preeminent frequency(fpred),site fundamental frequency(fs),site fundamental period(Ts),peak value of average Fourier spectrum amplification coefficient(As)and peak value of response spectrum amplification coefficient(Asr)under weak earthquake.4.In the study of 12 site characteristics characterization parameters,the weaker the intensity of the ground motion under the action of surface/underground Fourier spectrum ratio of peak(As)and the ratio of the surface/subsurface response spectrum peak(Asr)are negatively correlated with the the nonlinear attenuation index of the platform value of the surface/downhole response spectrum ratio(determination coefficient R2 were 0.25 and 0.28,respectively);The 30 m mean shear wave velocity(VS30)and the site predominant frequency(fpred)are positively correlated with the nonlinear attenuation index of the surface/downhole response spectrum ratio(determination coefficient R2 were 0.08 and 0.18,respectively).5.A multiple regression model for nonlinear attenuation exponent of surface/downhole response spectrum ratio platform value considering four field parameters,i.e.,30 meters on average shear(Vs30)wave velocity,basic frequency(fpred),peak surface/downhole Fourier spectrum ratio under weak vibration(As)and surface/underground response spectrum peak(Asr),has good correlation coefficient with the nonlinear attenuation index(determination coefficient R2=0.68);The coefficient of determination between the multiple regression model considering Vs30,fpred,Asr and the nonlinear attenuation index is only 0.04(R2=0.64)lower than that of the multiple regression model considering four parameters.Considering that only three site parameters are used,the three-parameter multiple regression model is recommended in this paper to predict the nonlinear attenuation index.6.Through the statistics of the distribution of the nonlinear attenuation index under different types of sites,We want to give the recommended values of the nonlinear attenuation index for various types of sites.According to the statistics,the mean values of nonlinear attenuation indices of the four types of sites are-0.150,-0.169,-0.182 and-0.218,respectively,showing a decreasing trend with the increase of sites types.However,due to the large discreteness of the results,the reliability of determining the nonlinear attenuation index according to the site category is not enough,and it is worthy of further research in this future.