FEM- and ANN-based design of CFRP-strengthened RC walls under close-in explosions

Abstract

Reinforced concrete (RC) structures are vulnerable to explosion loading, especially under close-in detonations. Recently, externally bonded carbon fiber reinforced polymer (CFRP) sheets have been used as strengthening layers to improve the blast resistance of structural components. Due to limited published research studies, there are very few tools or models capable of predicting blast-induced damage on RC walls strengthened with FRP subjected to blast effect. In this paper, an Artificial Neural Network (ANN) based model is developed for damage predictions. The ANN model is trained by a database of numerical simulations, which was previously established and validated by field blast tests. Good agreement among experiments, existing design diagrams, numerical simulations, and ANN predictions show that the model is capable of quantitively predicting local damage of structures with strengthening layers on the rear face. For fast engineering applications, design diagrams are developed as a quick assessment tool to evaluate the extent of damage on CRFP-strengthened RC walls caused by close-in explosions.