Abstract
The creation of transverse openings in beams triggers the shear performance. The dual impact of height and length on the overall shear performance and strain variations in reinforcements of deep concrete beams with and without fibres was assessed to investigate the effect of opening in the beam. This effect of opening was explored and modelled using finite element software Abaqus and predicted using an artificial neural network (ANN) model. The data set for ANN was 56 deep concrete beams, while for the finite element model (FEM), 12 deep concrete beams were used. The effect of input parameters in the ANN model was assessed through sensitivity analysis. Results show that with an increase in opening depth, the strain in top steel reinforcement shifted to tensile strain, resulting in premature beam failure. In addition, experimental and FEM shear resistance had a mean absolute error (MAE) of 4.1, 5.0, and 20.6% for deep beams without fibres, with fibres and fibre mesh, respectively. Compared to available analytical models, the ANN model reasonably predicts the shear resistance with an R2of 0.84 and a mean square error (MSE) of 0.01. The use of the ANN and FEM models is recommended as they save time, and the prediction does not involve degradation of the environment, hence demonstrating sustainable construction practices. Doi: 10.28991/CEJ-2024-010-08-02 Full Text: PDF
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call