Artificial neural network for predicting drying shrinkage of concrete

Concrete is the most used construction material for a century. After casting and setting, concrete shows various dimensional physical and mechanical evolutions, of which drying. It is a physical process, having though mechanical consequences. Drying accompanies the hardening of concrete and leads to significant dimensional changes that appear as cracks. vent or absence of cracking depends not only on the potential contraction but also on the extensibility of concrete, it’ s strength (the cracking due to shrinkage appears only in case the tensile strength of concrete has been over passed by the stresses, induced by shrinkage deformations), and its degree of restraint to the deformation that may lead to cracking. These cracks influence the durability of the concrete works. tudy presents the application of a nonparametric approach called Artificial Neural Network (ANN) in order to predict effectively dimensional variations due to drying shrinkage.Using this approach allows to develop models for predicting shrinkage. These models use a multi layer back propagation. They depend on a very large database of experimental results issued from literature (RILEM Data Bank), and an appropriate choice of architectures and learning processes. These models take into account the different parameters of concrete preservation and making which affect drying shrinkage of concrete as: Relative humidity (RH), Cure Period, volume to surface area ratio (V/S), water to cement ratio (W/C), and fine aggregate to total aggregate ratio or sand to total aggregate ratio (S/Ta). To validate these models, they are compared with parametric models as: B3, ACI 209, CEB and GL2000. In these comparisons, it appears that ANN approach describes correctly the evolution with time of drying shrinkage. metric study is also conducted to quantify the degree of influence of some the different parameters used in the developed neural network model.