Mesh size effect on finite-element modeling of blast-loaded reinforced concrete slab

Abstract

The dramatic increase in computation efficiency over the last three decades has made the finite-element simulation approach the dominant tool for investigating blast phenomena and their effects on building structures. However, the simulation results depend greatly on the mesh size selected for the numerical model, which may generate an inaccuracy in predicting the phenomena correctly. Thus, mesh size for finite-element simulations, appears to be an epistemic uncertainty that is inherent to the simulation process. This paper investigates the reliability of finite-element simulations of a reinforced concrete slab subjected to a certain level of explosion, in terms of mesh size sensitivity. Full-scale free air blast experiments are performed to monitor the blast-structure interaction phenomenon and help validating the numerical simulations. Fifteen finite-element simulations are evaluated with coarse to fine range mesh sizes, and the effects of mesh resolution on the numerical simulations are analyzed. An optimum mesh size is determined for a given set of loading conditions, and established by the amount of internal energy that gives rise to the expected damage on the slab.