Development of machine learning models for reliable prediction of the punching shear strength of FRP-reinforced concrete slabs without shear reinforcements

This study aims to examine the punching shear strength of FRP-reinforced concrete slabs, which is a complex behavior affected by several mechanisms and many variables. In this study, assessment of selected strength models available in the literature using a simplified reliability analysis method showed the need for more ac-curate and consistent strength models. Thus, two machine learning (ML) models were developed and proposed. Both models accurately predict the strength compared to the available models, which provides an alternative method to the available ones. In addition, the effect of the main variables on the strength using the proposed models was compared to that using the available models, which provided an insight on the impact and inter-relationship of effective variables on such a complex problem. Such insight can be helpful in future development of design codes.

Automated summary

This study aims to examine the punching shear strength of FRP-reinforced concrete slabs and proposes two machine learning models that accurately predict the strength. It also provides insight on the impact and inter-relationship of effective variables, which can be helpful in future development of design codes.