An approach for estimating the capacity of RC beams strengthened in shear with FRP reinforcements using artificial neural networks

An artificial neural network model is developed to predict the shear capacity of reinforced concrete (RC) beams, retrofitted in shear by means of externally bonded wrapped and U-jacketed fiber-reinforced polymer (FRP) in this study. However, unlike the existing design codes the model considers the effect of strengthening configurations dissimilarity. In addition model also considers the effect of shear span-to-depth ratio (a/d) ratio at the ultimate state. It is also aimed to develop an efficient and practical artificial neural network (ANN) model. Therefore, mechanical properties of strengthening material and mechanical and dimensional properties of beams are selected as inputs. ANN model is trained, validated and tested using the literature of 84 RC beams. Then neural network results are compared with those ‘theoretical’ predictions calculated directly from International Federation for Structural Concrete (fib14), the American guideline (ACI 440.2R), the Australian guideline (CIDAR), the Italian National Research Council (CNR-DT 200) and Canadian guideline (CHBDC) for verification. Performed analysis showed that the neural network model is more accurate than the guideline equations with respect to the experimental results and can be applied satisfactorily within the range of parameters covered in this study.

► The effect of a/d ratio is considered for the ANN model.
► This is one of the most extensive literature reviews of tests concluded with wrapped and U-jacketing.
► A parametric study was carried out to decide the inputs of the ANN model.
► ANN provided the best predictions compared to experimental results.
► CIDAR predicted the closer results according to experimental results.