The effect of input ground motion location on design spectrum in a seismic ground response analysis performed for an alluvial site

Abstract

Site response analyses are performed to estimate the change in the characteristics of the input rock motion occurred during the vertical propagation of seismic waves and to obtain the local variation in intensity and spectral characteristics of surface ground motion. The rock outcrop motions which are selected by considering the source mechanism, fault distance and earthquake characteristics that are compatible with the standard design ground motion level. It is generally agreed that time domain simulated or spectral matched ground motions may be applied as input ground motions at seismic bedrock in the dynamic response analysis of soil layers under earthquake excitation and thus shear stresses and deformations, acceleration-time histories and site-specific design spectrum can be obtained for surface layers. In “Turkish Building Earthquake Code (TBEC), 2018” it is being stated that the input ground motion can be exerted at the base of a soil layer lying at a shallower depth in comparison with seismic bedrock depth by using the method of spectral matching the accelerograms to the target spectrum of corresponding soil layer where the input motion is applied. In this study, by evaluating and combining the findings of the geotechnical borings and microtremor array measurements, the shear wave velocity profiles extending to seismic bedrock (VS≥760 m/s) were obtained for the investigation sites. The results of the site response analyses that were conducted by using the two different methods proposed in TBEC (2018) were interpreted. At the first method, in deeper 1D dynamic analyses, the input ground motions were applied at seismic bedrock defined as site class ZB where the shear wave velocity (VS30) is higher than 760 m/s2. Whilst at the second method, the inputs were applied at the base of the shallower 1D soil column having site class either ZC or ZD where VS30 is in the range of 360 m/s2 < VS30 < 760 m/s2 and 180 m/s2 < VS30 < 360 m/s2 respectively. As a result, the acceleration spectra obtained on the surface of deep and shallow models were compared for 31 strong ground motion records, selected in accordance with the design spectra corresponding to seismic code depending on the location of the investigation area and local site classes.