Research On Near-fault Velocity Pulse Period

The period of near-fault velocity pulse has become an important researchtopic in the field of earthquake engineering because of its influence on near-fault groundmotion, as well as its impact on the long-period structure. Currently, research on thevelocity pulse period is mainly depend on the statistical analysis on near-fault records,regression analysis are mainly performed to get the relationship among the velocity pulseperiod, the moment magnitude,the fault distance and other source parameters. In this paper,thanks to the two rich databases from US and Japan, we are able to pick the favorablequantitative classification method on velocity pulse period from several calculationmethods, get the long-period signal from massive near-fault records, and then divided therecords into groups according to different site conditions, different fault types and differentPGV threshold levels. Empirical relationships of several period-magnitude in differentsituations are then given, which may do help to further study. The main researches in thispaper are as follows:1. The main formation mechanism, the main incentives and other aspects of groundmotion characteristics of near-fault velocity pulses are introduced in detail, while twocommon genres of velocity pulses are given.2. Two earthquake record databases, PEER from US and K-NET from Japan, areselected in the paper, as well as a brief introduction to them. Meanwhile, the baselinecorrection, which involved in seismic data processing, were introduced in several parts: thereasons, methods and verification. Also, the data from K-NET are renovated by mean ofbaseline correction.3. Several methods that help to get velocity pulse period are proposed, as thezero-crossing method and the pseudo-velocity response spectral method are introducedbriefly, the quantitative classification method based on wavelet analysis is introduced indetail, supported by several examples in different methods to get the comparison.4. Quantitative criteria are built to determine the selection of records with velocitypulse characteristics from a large number of near-fault earthquake records. All Records aredivided into groups according to different site conditions, different fault types and differentPGV threshold levels. Regression coefficients of several period-magnitude relationships indifferent situations are then given.5. The description of PGV (peak ground velocity) is given as its significant use in forecasting for certain engineering practice. Also, the relationship among PGV, momentmagnitude and fault distance are focused. Finally, empirical relationships among PGV,moment magnitude and fault distance is given after referring to other researchers’attenuation relationships.At last, the research contents are summed up and the problems need to be furtherstudied are put forward.