EUROSTEEL 2011, August 31 - September 2, 2011, Budapest, Hungary EFFECT OF COMPOSITE ACTION ON COLLAPSE CAPACITY OF STEEL MOMENT FRAMES UNDER EARTHQUAKE LOADING Dimitrios G. Lignos a , Laura Eads b and Helmut Krawinkler c a McGill University, Dept. of Civil Engineering and Applied Mechanics, Montreal, Canada b Stanford University, Dept. of Civil and Environmental Engineering, California, United States c Stanford University, Dept. of Civil and Environmental Engineering, California, United States INTRODUCTION Collapse assessment of steel structures subjected to extreme earthquake loading necessitates the use of advanced numerical models that are able to simulate cyclic deterioration in strength and stiffness of steel beams and columns that make up the structural system. A large number of such models has been developed and refined through the years [1, 2, 3, 4] to assess the effect of component deterioration on collapse potential of steel frame structures. The focus of these studies is on the collapse mode associated with sidesway instability, in which a story or a number of stories displaces sufficiently so that P-Delta effects accelerated by component deterioration fully offset the first order story shear resistance of a structural system. However, component deterioration is typically concerned with analytical modeling of the bare steel moment resisting frame only, ignoring the effect of composite action on lateral stiffness and strength of the structural system. The focus of this paper is to quantify the effect of composite action on the collapse capacity of steel frame structures under cyclic loading. This assessment is based on experimental data of steel beam- to-column subassemblies that became available through a recently developed database for deterioration modeling and through full scale collapse tests of a 4-story steel structure [3, 4, 5]. In particular the emphasis is on the effect of composite action on (1) bending strength, (2) pre and post-capping plastic rotation and rate of cyclic deterioration of steel beams under cyclic loading. This assessment in based on a phenomenological model, which is able to simulate strength and stiffness deterioration under cyclic loading and has been modified to incorporate slab effects on moment-rotation characteristics of composite steel beams. 1 DETERIORATION MODELING OF COMPOSITE STEEL FRAME STRUCTURES 1.1 Deterioration Model Cyclic deterioration of bare steel beams can be simulated with the modified Ibarra-Krawinkler (I-K) [2,3] deterioration model. This model is defined by a reference backbone curve that is shown in Fig. 1. This curve is defined by an elastic stiffness K e , yield moment M y , capping to yield moment ratio M c /M y , pre-capping rotation θ p , post-capping rotation θ pc , residual strength M r = κM y and an ultimate rotation θ u , at which the component strength drops to zero. This model can deteriorate cyclically in strength and stiffness based on an energy rule that was developed by Rahnama and Krawinkler [6]. The main assumption is that a steel component has an inherent reference energy dissipation capacity E t , which is defined as a multiple θ p and M y , (E t = λ θ p M y , and Λ=λθ p ). In this relationship, Λ is a reference cumulative rotation capacity. The hysteretic response of the modified I-K deterioration model has been calibrated with about 350 component tests contained in a recently developed steel component database that includes information for different types of steel beam-to- column connections. An example of a calibration of the simulated response of the modified I-K model versus the moment rotation diagram of a bare steel beam is shown in Fig. 1. Through multivariate regression analysis different relationships have been proposed [3,4] in order to model the deterioration parameters of steel beams and columns for reliable collapse assessment of bare steel moment resisting frames. These relationships have been adopted by PEER/ATC [7] for modelling and acceptance criteria of steel beam-to-column connections.