Effects of strain gradients in the onset of global buckling in slender walls due to earthquake loading

Abstract

Global buckling of slender walls, reported only in a few laboratory tests before 2010, became a critical issue in design of reinforced concrete buildings after it was observed following the 2010 Mw 8.8 Chile earthquake and the 2011 Mw 6.3 New Zealand earthquake. Researchers have proposed theoretical buckling models based on prismatic columns subjected to uniform tension/compression cycles, where the key parameters are slenderness ratio, number of curtains of reinforcement, and maximum tensile strain before buckling during load reversal. These models have shown sufficient accuracy in comparison with laboratory tests on columns under such loading conditions. However, buckling in walls is more complex because of variation of strains through the wall depth and variation of moment along the wall height. Nonlinear finite elements are used to evaluate the effects of these more complex loadings on buckling of wall boundary elements. Analyses showed that the maximum tensile strain (averaged over the wall out-of-plane unsupported height) required to buckle the wall during load reversal does not depend on the moment variation along the wall height. Moreover, for typical wall lengths, the wall boundary behaves like an isolated column subjected to axial force cycles, with minimal apparent bracing provided by the wall web. This allows to analyze a broad range of practical cases for buckling susceptibility using simplified approaches based on buckling models of axially loaded columns.