Optimum Prediction of the Transfer Length of Strands Based on Artificial Neural Networks

As the effectiveness of prestressing is crucially linked to the transfer length (TL) of the strands, this study used Artificial Neural Network (ANN) technique for optimal prediction of TL based on more than 458 data points collected from various literature works. The ANN technique allowed for investigating the effect of various key parameters classified into major categories including: strand characteristics, concrete properties, geometric details, and manufacturing method. The MATLAB software was utilized to build, train, and test the ANN using 19 input variables and one targeted output. The proposed ANN showed high prediction capability with a low mean square error. The sensitivity analysis of the TL gave a good indication regarding the significance of the parameters influencing the TL determination. Mathematical expression was developed considering the most significant parameters according to the ANN results and sensitivity analysis.