ABSTRACT
The floor response spectra (FRS) are of primary importance for seismic qualification of equipment, components, and secondary systems at complex structures [1]. It is a common practice to use a deterministic method for generation of FRS. However, this method is methodologically not rational and expensive, since it utilizes time histories for analyses. It cannot take into account the randomness of the seismic movement of the soil and its effect on the FRS. The deterministic method usually needs generation of seismic time histories compatible with the design response spectrum (RS), a form that usually is used for definition of seismic movement of soil for designing of nuclear power plants and other complex structures. The probabilistic method uses analysis of random vibrations for defining the probabilistic FRS from influence of random seismic movement of the soil. This method can directly generate FRS, for different levels of confidence, from the prescribed design RS without conducting time-history analysis. Although probabilistic methods enable direct generation of FRS, they still do not have wider application. Systematic evaluations are needed to justify their reliability and preciseness for practical application. The probabilistic methods for generation of FRS are grouped in two approaches: implicit and explicit. In the implicit approach, i.e. power spectral density function (PSDF)–RS from random stationary process is used implicitly to determine the FRS as a direct function of the prescribed design RS [2]. The explicit method utilizes the relationship PSDF–RS explicitly to generate FRS: (i) calculation of PSDF of seismic movements of the soil from the prescribed RS using the relationship PSDF–RS; (ii) calculation of a transfer function; (iii) calculation of RS in PSDF type; and (iv) calculation of FRS using inverse PSDF–RS relationship.
