SHEAR FAILURE OF REINFORCED CONCRETE BEAM

Finite element modelling of an experimental test conducted on reinforced concrete beam to evaluate its shear strength. An experimental investigation has been designed to scrutinize the shear strength of a continuous reinforced concrete beam. The beam has 150 x 400 mm rectangular cross-section and two 1.90 m spans. The beam was designed to fail in shear in the vicinity of the middle support in order to ensure that failure occurs at the region of combined shear and normal stress. A full 3D solid model is constructed to obtain detailed information of field outputs. It should be mentioned that this problem is governed by tensile cracking of the concrete due to shear forces, which require to construct the model and input suitable material properties for the concrete to ensure that the model undergoes successive cracking and dowel action of steel reinforcement occurs. 20-noded tetrahedral elements are used from the beginning to soften the solution. The reason beyond the choice of these elements is that the problem is crack sensitive, so increasing the nodes per element will help predicting the correct solution accurately without increasing number of elements too much. For the steel reinforcement whether flexural or shear, 3-noded truss elements are used. Vertical supports are assigned to areas each of them is 100 mm length and of width equals to the beam width (150 mm) to avoid the premature failure at support regions. Load is applied at two areas, too, to have a total 1 kN load acting downwards. The interaction between steel cage and the concrete is defined as embedded region constraint. This type of constraint allows you to embed a region of the model within a "host" region of the model or within the whole model, tying the displacements of each embedded node to the displacements of the surrounding nodes. Solution of the FE equations is performed using the arc length convergence algorithm method (Static/Riks method). Experimental first crack load was 306.5 kN and the first cracking load from the finite element model is 319 kN which results in 4.1 % difference. Experimental peak load was 544 kN and the peak load from the finite element model is 531.05 which results in 2.4 % difference. In terms of first cracking load and peak load, the finite element model shows a great agreement with the experimental results.  Reference for experimental data: Akroush, N, Almahallawi, T, Seif, M and Sayed-Ahmed, E. "CFRP Shear Strengthening of Reinforced Concrete Beams in Zones of Combined Shear and Normal Stresses". Composite Structures Journal, Submitted July 2016, Under-review.