Comparison of Site Periods Derived from Different Evaluation Methods

Abstract

Abstract As a part of our microzonation research activities for the city of Ottawa, the fundamental site period, T 0 , was investigated based on different methods, including (1) the horizontal-to-vertical spectral ratio (HVSR) using microtremor ambient-noise measurements; (2) equivalent single-layer (ESL) modeling, as noted in the current National Building Code of Canada (2005); (3) earthquake weak motion observations; (4) multilayer soil modeling; and (5) finite element modeling for linear and nonlinear soil. The differences between these methods are discussed. We have discovered that T 0 based on the HVSR method systematically deviated from T 0 based on the equivalent single-layer modeling method. The variance was more than 30% for periods longer than 2 s, corresponding to impedance boundary depths of more than 75 m in the study area. The effect of the shear-wave velocity gradient on T 0 was investigated by applying multilayer soil modeling, which confirms that the actual velocity-depth gradient shifts the evaluated T 0 to shorter periods compared to equivalent single-layer modeling. The effects of soil nonlinearity on T 0 were examined using a finite element method analysis. The dependency of T 0 on the level of shaking is well defined at higher levels of shaking when the peak ground acceleration (PGA) exceeds 80 Gal because the soft soil behaves nonlinearly. It has been concluded that, for the case of nonlinear soil response, damping plays an important role in reducing the fundamental frequency. These findings will be used in our future research activities directed at providing fundamental period maps for the city of Ottawa, including fundamental period maps for the strong motion design earthquake.