Prediction of expansion behavior of self-stressing concrete by artificial neural networks and fuzzy inference systems

• Effects of mix design parameters on free expansion strain of self-stressing concrete.
• Regression model were established to calculate the 28-day expansion strain of SSC.
• Free expansion strain values were predicted with the RM, ANN and FIS models using the experimental data.
• FIS model showed a better prediction performance compared with RM and ANN.

Self-stressing concrete (SSC) can be used to inhibit the growth of cracks, and meanwhile create considerable partial pre-stresses. The mix design method of SSC, nowadays, completely depends on experience and experiments, which hinders the application of SSC. In this research, regression model (RM), artificial neural network (ANN) and fuzzy inference system model (FIS) for predicting the free expansion strain of SSC under wet curing conditions have been developed. To construct these models, 730 experimental data were gathered. The data used in the ANN and FIS models are arranged in a format of four input parameters that cover the water/cement ratio, cement abundance coefficient, cross-section area of specimens and curing time, and output parameter, which is the free expansion strain of SSC. Results calculated by RM, ANN and FIS models that were applied were compared. The results show that ANN and FIS models have a strong potential to predict the free expansion strain of self-stressing concrete.