Machine Learning‐Based Failure Prediction in Concrete Slabs and Cubes Under Impact Loading

Abstract

ABSTRACT Experts have been interested in the behavior of concrete under impact loading because of its wide range of applications in construction projects. Due to their quasi‐brittle nature, failure modes related to concrete may occur without any prior warning signs of destruction, and they also expose the supporting element to the spread of damage. Finite element modeling and machine learning techniques are essential for conducting an adequate reliability investigation of the behavior of concrete samples as slabs and cubes under impact load. The research uses gradient boosting, random forest, lasso, linear regression, and support vector regression to create predictive models for the behavior of these two concrete models. The models were created by taking experimental concrete slab tests and 20 cubes into consideration. Design standards‐based statistical comparisons such as coefficient of determination and root mean square error are used to assess the efficacy of the generated models. These results show that with increasing impact load intensity, displacements and failures in the slab increase significantly. Using these models allows engineers to design more resistant and optimal structures against impact loads. This research shows that machine learning models, especially random forest and gradient boosting, can provide accurate predictions of failures and cracks in concrete under impact loads and are useful tools for analyzing concrete behavior under dynamic and complex conditions. The linear regression with a coefficient of determination ( R 2 ) of 0.995 and lasso regression with RMSE of 3.9 have the lowest accuracy, while random forest and gradient boosting models with R 2 of 0.9991 and 0.2, 0.991 and 0.5. respectively, showed higher accuracy in predicting concrete cracks and failures.