Abstract

Profiled steel sheet-mixed aggregate recycled concrete composite slabs are widely studied and applied in engineering applications by virtue of their good durability and recyclability. As an alternative to concrete composite slabs, concrete hollow composite slabs that can reduce their self-weight are also attracting people’s attention. However, the internal holes will increase the risk of collapse and shear failure of the concrete composite slabs. In order to reduce the risk and ensure the flexural performance of the concrete hollow composite slabs, this paper designs the recycled concrete hollow composite slabs with PVC pipes. A full-scale experiment is carried out on five open-mouth profiled steel sheet-mixed aggregate recycled concrete hollow composite slab specimens, to explore the influence of whether the bottom of the plate is reinforced, the thickness of the concrete surface layer and the hollow size on the working performance of the composite slabs. In order to predict the mechanical performance of the proposed hollow composite slabs, a prediction model based on artificial neural network (ANN) consisting of several sub-ANN modules is established. Moreover, according to the priority of the factors affecting the mechanical performance of composite slabs, a decision tree is designed to select appropriate testing sub-ANN module and thus improve the testing accuracy. Predicted results and experimental results are given to validate the feasibility and effectiveness of the developed ANN-based prediction framework for the mechanical performance prediction of composite slabs.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call