Effect of beam-column connection-stiffness on the lateral behavior of buildings

Abstract

Modern numerical structural analysis techniques of buildings rely on basic assumptions and idealizations that simplify analysis procedures. However, these assumptions are not always accurate and conservative. Therefore, it is necessary to investigate the accuracy of these assumptions, in order to verify that the analysis performed does not lead to unsafe designs. This project deals with one of the most researched areas in structural engineering analysis, the beam-column connections. All analysis methods and commercial computer software (such as Etabs, STAAD) implement in their analysis the assumption of full rigidity of beam-column connections as a default option. This leads to an equal rotation of beam and column (angle between them is always 90°). And the stresses in members are also calculated by taking into consideration a fully rigid connection. However, it is widely known among designers, as is covered in all design codes, that the stiffness of connections depends largely on the detailing of the joint. In reinforced concrete frames, joint stiffness is largely affected by the anchorage and bond-slip mechanism of reinforcing bars. The main purpose of this research is to determine the effect of flexible connections on the building characteristics. For that purpose, a multi-story multi-span frame was modeled with flexible connections, and main characteristics as well as results were compared with the same frame using rigid connections. The modeling of the frames was done using a direct stiffness software that we wrote on the computer mathematical platform "MATHEMATICA". The flexible connections are modeled as linear rotary springs between columns and beams, and the software includes this spring as a fourth DOF on each joint. An analytic model was also created using Euler beam equations and proper boundary conditions. This model was limited to one span and 2 storeys only, and used rotary spring as connections between beams and columns. The main purpose of this analytic software was to verify the direct stiffness software, and after comparison, both model gave the exact same results. The main software was also compared and verified with ETABS, where a model was created with beams having end releases and assigning rotary spring stiffness at the end of beams. Using the direct stiffness software, many runs were performed for different cases, and plots were generated to illustrate the effect of the rotary spring stiffness on lateral stiffness, shears, and moments. Secondary variables included the number of stories, column stiffness and beam stiffness. From the parametric study, the main conclusion if that flexible connection decrease considerably the lateral stiffness of buildings, and that effect becomes more obvious with the increase in number of stories. On the other hand, the span lengths (which appear in α) has little to no effect.