Critical evaluation of torsional provision in IS-1893: 2002

Abstract

For a building to be symmetric it must have, at each floor level, coincident centers of mass and stiffness that lie on common vertical axis. In practice, this condition is rarely encountered and most buildings are unsymmetrical to varying degrees, due to asymmetry in plan, elevation, distribution of vertical members or mass distribution on the floors. Although considerable research on response of asymmetric structures under seismic excitation has been reported in the literature, the performance of asymmetric structures designed in accordance with the Indian code has not been studied adequately. Frequent occurrences of devastating earthquakes in India clearly call for the need of evaluation of Indian buildings for seismic safety.
The earthquake resistant code in India, IS: 1893 (Part1), has been revised in 2002 to include provisions for asymmetric buildings. An attempt has been made in the present study to investigate the gap in the seismic design of asymmetric RC structures in the Indian context. Three reinforced concrete moment-resisting frame buildings with different types of asymmetry are designed based on prevailing Indian codes as test examples. Nonlinear static (pushover) and dynamic (time-history) analyses are performed on these structures and a comparison is made of displacements, inter-storey drift ratio, ductility and hinge pattern of the frames to show the changes in their behaviour due to torsion which is recognized as a principal cause of severe damage in eccentric multi-storey buildings during earthquakes. Nonlinear dynamic analysis (time history analysis) is done for ten recorded ground motion and five generated ground motion consistent to IS-1893: 2002 (Part1) response spectrum.
Results obtained from this study show that the plan asymmetry in the building makes it non-ductile even after design with code provision. The maximum base shear demands for the three building variant are almost same. This is because the fundamental periods of all the three building are almost identical. It is found that there is a considerable amount of variation in the maximum roof displacement responses of the three building variants subjected to generated earthquake ground motion. The maximum roof displacement responses for symmetric building variants are found to be lesser compared to the two asymmetric buildings for all the cases studied here. However the average maximum roof displacement responses for two asymmetric buildings are found to be approximately same.
Base moment demand obtained from nonlinear dynamic analyses is found to be almost same for both of the two asymmetric buildings for all the cases studied here. Also, in most of the cases base shear – roof displacement and base moment – roof rotation hysteresis curves are found to be similar for both the asymmetric buildings with small translational/rotation shift.
It is found that the yielding for both the asymmetric building occurs at the same time step of the dynamic analyses after following the same elastic eccentricity even though ASYM2 (designed with code provision) has a greater strength compared to ASYM1 (designed without code provision).
Considering that all the building structures will undergo inelastic deformation under an expected earthquake it is meaningless to relate the design criterion to the elastic centre of rigidity. Design criterion given in IS 1893:2002 (Part-1) with regard to plan asymmetry seems to be not very efficient. Code criterion for plan asymmetry recommends increasing the strength distribution in the building but it does not look for changing the stiffness distribution of the building. Change in the stiffness distribution to reduce eccentricity can be a useful for such buildings.