Regresar

An evaluation of the current approaches and recommendations for more rational approaches for assessing the seismic torsional stability of buildings

Abstract:

Building codes for seismic resistant design of building structures generally include requirements for imposing an “accidental” torsion response. Methods for incorporating such effects into the analysis are ad-hoc at best and have not changed significantly over the past 50 years. This is despite the tremendous advanced in analytical and computational capabilities over the same period. The current building code methodology for including accidental torsion depends on the analytical method. For linear static approaches, lateral forces are applied at an eccentricity that depends on the sensitivity of the system to torsional effects. Where modal response spectrum analysis (MRSA) is used, the torsion may be included by imposing a mass offset or adding the accidental torsion as in linear static analysis. Where linear or nonlinear time history analysis is utilized, the mass offset approach has been used almost exclusively. In a recent report published by FEMA (FEMA P-2012), it was found that the dynamic mass offset procedure could produce grossly unsafe designs relative to the static load approach for torsionally sensitive systems. For this reason, it was recommended that the mass-offset approach not be used in MRSA or in linear time history analysis (LTHA) for systems with extreme torsional irregularities. A recent supplement to ASCE 7-16 included the restriction on MRSA but not on time history analysis. In this paper, the mass-offset approach for including accidental torsion in the dynamic analysis is evaluated in detail and compared with the static analysis for different building configurations. It was found that for the analyzed buildings, MRSA pbkp_redicts well the demands obtained using LTHA, and hence, the prohibition of the mass offset approach for accommodation accidental torsion should be extended to response history analysis.