Modeling of bracing members and seismic behavior of concentrically braced steel structures.

Abstract

It is now commonly recognized that bracing systems can provide necessary stiffness to resist moderate earthquakes and similar dynamic excitation, and a source of energy dissipation through post-buckling hysteresis behavior of braces. Bracing systems are classified into two broad categories: concentric and eccentric. In the case of eccentric scheme, the prevention of brace buckling is required in order to force the inelastic deformation into a short segment of the girders which acts as a shear hinge. In concentric bracing systems the braces undergo cyclic deformations beyond buckling and yielding under a severe dynamic excitation. In such a situation, understanding the cyclic inelastic behavior of bracing members becomes essential for evaluation of dynamic response of concentric braced structures. Modeling the hysteresis loops of bracing members under a dynamic loading is essential for inelastic analysis of braced multistory steel frames and similar steel structures. The accuracy of the analytical model affects significantly the computed response of such frames. Several analytical models have been developed by previous researchers to represent the cyclic behavior of steel braces. One of the purposes of this study is to formulate a refined and practical hysteresis model for the bracing members on the basis of experimental results of previous researchers. Development of a better understanding of the behavior of concentric (chevron) braced structures subjected to severe earthquakes is an important purpose of this study. An evaluation of the consequences of using lateral design forces in accordance with the 1988 Uniform Building Code, in particular the $R\sb{w}$ factors (lateral force reduction factor to account for ductility) for concentrically braced structures, is carried out. Recommendations for more rational values of $R\sb{w}$ for both concentrically braced moment frames and concentrically braced non-moment resisting frames are developed. Non-linear dynamic response of six-story concentrically braced structures designed according to different design procedures and subjected to four past earthquake records are determined. Included in the study are concentrically braced structures with and without moment resisting frames. The analysis of the computed response shows that column buckling and brace failure at small ductilities present a major problem with these structures. Bracing failure early in the time history response results in excessive story drifts. Alternative methods of design are investigated which can improve the response of these types of structures. Using ductile braces with smaller lateral design forces (compared with those used in current practice) for both moment and non-moment resisting structures showed very good behavior. Also, using moment connections at the exterior girder-to-column connections in frames without braces improved the response of non-moment resisting structures significantly. The results of this research can have direct significance in improving the methods of analysis and design of concentrically braced structures.