Reliability analysis of the punching shear prediction models in reinforced concrete slabs without beams according to ABNT NBR 6118 (2014), EUROCODE 2 (2010) and ACI 318M (2014)

Abstract

The search for new constructive systems in concrete, different from those arranged in columns, beams and slabs, known as conventional system, has been increasingly studied in the world. One of the most well-known alternative systems is the so-called flat slabs. Flat slabs are defined by ABNT NBR 6118 (2014) as slabs supported directly on the columns, without the use of capitals. Although it may have some advantages compared to the conventional system, due to the absence of beams, the main structural problem is the punching shear phenomenon. This, which consists of shear stresses concentrated directly on the slab-column connection, can cause the slab to pierce due to the direct action of the column. His study is of great importance, because his way of rupture is fragile, occurring abruptly, without warning, and should always be avoided. In this context, this work aims to verify the reliability of the theoretical models of the design codes NBR ABNT 6118 (2014), EUROCODE 2 (2010) and ACI 318M (2014), for the internal, edge and corner columns in the design of slabs in reinforced concrete to punching shear. In order to perform this analysis, spreadsheets will be developed for the three normative design codes, in order to automate the calculations of sizing and verifications of the connections. Then, parametric analyzes for the three models of calculation will provide the most influential parameters in the behavior of the slab-column connection to punching shear. In addition to this, an extensive bibliographical review of previous experimental analyzes will be carried out in order to evaluate the calculation models imposed by the three design codes and to compare experimentally and theoretically. Thus, it wants define the variable model error, by the ratio between the experimental response and the theoretical response for each design code. Finally, the study will evaluate the reliability of these slab-column connections for each type of column, considering the design code of the three standards studied without and with the model error variable. With this, it is believed to show, in terms of probabilities of failure, the quality of prediction models of behavior to the punching shear of each norm.