Expansion prediction of alkali aggregate reactivity-affected concrete structures using a hybrid soft computing method

The phenomenon of alkali aggregate reactivity (AAR) in concrete structures corresponds to the reaction between aggregates with some ingredients and alkali hydroxide in concretes. This AAR could potentially lead to concrete deformation, micro-cracks and eventually wide visible cracks. In this study, to predict the expansion of the concrete caused by AAR, a novel hybrid model is proposed based on support vector machine (SVM). In the proposed model, the inputs are the aggregate components and concrete age, while the output is the induced expansion in the concrete. To improve the generalisation capacity of the proposed model, the enhanced particle swarm optimisation algorithm is employed to select optimal SVM parameters. The proposed method is evaluated and compared with other conventional soft computing methods based on the experimental data. Finally, the evaluated results endorse the effectiveness of the proposed hybrid model.