Improved intensity measures considering soil inelastic properties via multi-regression analysis

Abstract

At certain depths, the elastic properties of the ground are not affected by seismic waves. However, as they reach the surface, the soil density decreases and so does its elastic limit. This means that the expected ground motion acting at the foundation of a structure cannot be adequately described without considering the inelastic response of the soil near the surface. Therefore, one of the key elements in characterizing the seismic response of civil structures is the site effect. These depend mainly on the parameters of the soil beneath the structure and the features of the ground motion acting at the depth, where non-linear effects are negligible. Therefore, the main objective of this paper is to find an intensity measure that incorporates the information provided by the soil profile under the structure and the ground motion acting at the bedrock level. Due to the random nature of both elements, a probabilistic framework using Monte Carlo simulation has been developed to analyze this problem. For this purpose, random soil profiles have been generated to obtain a representative sample of likely scenarios of the study area. A large database of Colombian ground motion records has been used to model the seismic hazard. Finally, power functions capable of relating the input variables to the dynamic response of a large set of reinforced concrete structures have been derived by considering multi-regression analysis. It has been observed that, in several cases, intensity measures extracted from the displacement spectrum appear in the mathematical arrangements. These functions could be used to improve the efficiency of seismic risk prediction at the urban level.