Contribuição à análise da estabilidade global de estruturas em concreto pré-moldado de múltiplos pavimentos

Abstract

MARIN, M. C. Contribution in the analysis of global stability of multi-storey precast concrete framed structures. 2009. 213f. Dissertation (M.Sc) Escola de Engenharia de Sao Carlos, Universidade de Sao Paulo, Sao Carlos, 2009. In this research it was evaluated the main design parameters in the analysis of global stability of multi-storey precast concrete framed structures considering the physical nonlinearity (PNL), geometric nonlinearity (GNL) and semi-rigid connections behavior. The structural systems considered are those consisting of frames with beam-to-column semi-rigid connection and columns inset in the foundation. The beam-to-column connections are defined by two dowels and cast-in-place concrete cap with longitudinal reinforced bars. The PNL is evaluated according the M x N x 1/r diagrams, where are considered the effect of normal force, the reinforcing steel, prestressing steel and creep. Functions are proposed and reducing stiffness coefficients are defined, with were compared with codes that include the approximated PNL. The GNL is evaluated according to ANSYS software by no approximate analisys and it’s approximate according to the coefficient z and (P-) method. This research presents an analytical model characterization of stiffness and strength of beam-to-column connections to the negative and positive bending moment. In the numerical analysis is made a full case study of a typical multi-storey precast concrete framed structure with the aid of the ANSYS software, evaluating different forms of consideration of the PNL and the GNL. It analysed the bending moment distribution according the combinations of actions used and the behavioral models of connections, for some variations at geometry and loading. The reduction coefficients of flexural rigidity obtained according the M x N x 1/r diagrams differ from normative indications for simplified account of PNL. In GNL simplified analysis, the coefficient z showed better results in predicting the second-order effects with respect to those obtained by the coefficient 0,95. z.