The amount prediction of concrete fragments after impact using smoothed particle hydrodynamics

Concrete median barriers on highways are typical road safety facility that requires predicting correctly the amount of fragments generated during a vehicle collision. The fragmented pieces from the median barrier can cause secondary accidents to a vehicle coming from the opposite lane. Therefore, predicting the amount of fragments depending on the impact severity is important to prevent any secondary accident. Many researchers have studied to predict the damaged area and strain of concrete. Such predictions of concrete structural behavior following impact loads mostly used FEM. However, FEM has a limitation in predicting the fragmentation amount since it simulates fragmentation through element deletion. As an alternative, Smooth Particle Hydrodynamics (SPH) can be used for predicting the amount of fragments or the motion of fragments since these are not affected by the mesh. In the present study, impact analysis was performed to predict the amount of concrete fragments due to vehicle collision. The obtained results of SPH analysis showed that the amount of fragments can change depending on different velocity-to-mass ratios at a fixed local impact energy. Using the results of the SPH analysis, multiple regression analysis (MRA) was conducted further. The MRA showed a rather low coefficient of determination (R²) compared with the SPH analysis results. Therefore, as a future study, with the expectation of improvement, a method such as ANN (Artificial Neural Network) that can predict the amount of fragments including uncertainty is necessary.