Review of assumptions in simplified multi-component and codified seismic response evaluation procedures

Abstract

Representation of an earthquake by orthogonal normal and principal components, accuracies of several rules to combine collinear and non-collinear responses of MDOF structures, and estimation of the relative importance of the effects of individual orthogonal components are addressed in this paper. Statistically there are no differences between normal and principal components in estimating axial load for the elastic case; the principal components, however, may produce about 11% larger axial load for the inelastic case. There are no major differences in estimating base shear in all cases. To combine the two horizontal components, the 30% and SRSS combination rules underestimate the collinear axial load by about 10%. For the case of three components, the underestimation by the 30% and SRSS rules are about 5% and 10%, respectively. When responses due to axial loads and bending moments are combined, the underestimation could be between 10 and 15% for both combination rules. The assumption that the vertical response is 2/3 of that of the horizontal component appears not to be very conservative. The assumption that the response produced by the minor horizontal component is 85% of that of the major horizontal component appears to be very conservative.