Experimental And Analytical Study Of Internal Beam To Column Connections Subjected To Reversed Cyclic Loading.

Abstract

In a reinforced concrete building subjected to earthquake type loading the beam to column connections constitute one of the critical regions and they must be designed and detailed to dissipate large amounts of energy without a significant loss of stiffness or strength. In the experimental part of this study, six full size interior beam-column subassemblages were tested under quasi-static loadingwhich was intended to simulate earthquake input. All specimenswere designed following the accepted design philosophy of strongcolumn-weak beam approach. The three variables selected forthis investigation included the percentage of transverse hoopreinforcement in the joint ((rho)(,t) = 0.75% to 1.15%), the joint shearstress level (10SQRT.(f(,c)'(' )to(' )15SQRT.(f(,c)') and the presence of transversebeams and slab on half of the specimens. On the basis of test results it was concluded that the joint shear stress level was critical for the satisfactory performance of beam to column connections without transverse beams and slab. For connections with transverse beams and slab, a well confined joint core was essential for the effective participation of the transverse beams in resisting joint shear. For the evaluation and design of joints, a joint performance index is proposed which integrates the effect of pertinent variables influencing the joint behavior. For detailing, an odd number of layers of hoops in the joint is recommended with a nimimum of three layers. In the analytical part of this study, a hysteretic model for beam-column subassemblages is developed from the hysteretic behavior of specimens observed during testing. The proposed model accounted for pinching of hysteresis loops, stiffness degradation, reduced unloading stiffness and fixed end rotations due to the slippage of bars through the joint. A simple analytical model for computing the maximum story displacements for regular frames is proposed. This model idealized a building frame as an elastic column with rotational restraints at the floor levels. Three building frames of five, seven and ten stories were analyzed using the SIMPLE program developed specifically for this model. The maximum story level displacements were found to be in good agreement with those obtained from more complex models.