Numerical challenging of capturing membrane action in reinforced concrete beams and one-way slabs

The experimental data on reinforced concrete beams subjected to large deformations and provided with proper boundary conditions to develop arching action followed up by tensile membrane action are scarce. Moreover, so many different factors such as boundary conditions, concrete compressive strength and reinforcing ratio significantly contribute in membrane behavior of reinforced concrete beams and capturing the effects of these contributing factors by experimental studies is quite costly. Accordingly, numerical models, particularly finite element (FE) models, can be employed as a suitable alternative to experimental studies, however, modeling reinforced concrete members subjected to large displacements is a challenging task which involves both material nonlinearities (i.e. concrete cracking and crushing and yielding of steel) as well as geometrical nonlinearities. In this paper a comparative study between continuum-based and frame finite element models is presented. A nonlinear force-based frame element developed and implemented in a FORTRAN code as well as ATENA software are employed for modeling and nonlinear analysis of sample beams and one-way slabs taken from the literature. The effect of concrete cracking and crushing in both FE models are taken into account and the 1D frame model takes advantage of a nonlocal integral model to restore the objectivity of results whereas the continuum- based FE model developed in ATENA adopts the crack band approach to resolve the mesh sensitivity associated with concrete softening. The results obtained from both FE models (i.e. continuum-based and 1D frame) show a good correlation with the available experimental results, however, the continuum-based FE model cannot always restore the objectivity of results and it is more time demanding than the frame FE element model.