Comparison of results of soil-structure interaction analyses based on time and frequency domain approaches with emphasis on treatment of damping

Abstract

Comparison of results of soil-structure interaction analyses of the reactor building of a nuclear power plant using different analytical approaches and solution procedures is presented. The emphasis of the comparison was on the treatment of damping in these different approaches and procedures. An axisymmetric model of the reactor building was employed. The analyses were performed for the aircraft impact loadings. Two different locations were used for these loadings. The following four different sets of analyses were performed. 1. (1) Time-domain analysis using frequency-independent soil springs in conjunction with modal damping cut-off. 2. (2) Frequency-domain analysis using frequency-independent soil springs in conjunction with a complex modulus approach. 3. (3) Frequency-domain analysis using frequency-dependent soil-impedance coefficients in conjunction with a complex modulus approach. 4. (4) Frequency-domain analysis using frequency-dependent soil-impedance coefficients in conjunction with Rayleigh damping. The frequency-independent soil springs were computed using the standard approach based on rigid base supported on an elastic layered half-space. The frequency-dependent soil impedance coefficients were computed in the form of a soil substructure matrix which included the uncoupled as well as the coupled terms. The computations were based on the use of a “flexible” base mat supported on a layered half-space. Unit dynamic loads, for each frequency, were applied to the layered half-space corresponding to each degree of freedom and the displacements were xomputed corresponding to all degrees of freedom. The compliance matrix so computed was inverted to obtain the impedance matrix for each frequency. The computations were repeated for all frequencies of interest for the aircraft impact loading. Floor response spectra were developed and compared at various floor elevations of the reactor building using the above four different sets of analyses. Conclusions were developed as a result of these comparisons.