Plasticity based material model for concrete subjected to dynamic loadings

Abstract

A new plasticity based material model has been proposed in the present work for concrete subjected to dynamic loads. The model has then been used for blast analysis of Reinforced Concrete slabs. The model has three parts; first the equation of state is described by a pressure versus volumetric strain relationship, second a method for defining damage and in the last part of the model we define a modified strength surface. Tensile and compressive strain rates are accounted separately. Three invariant failure surface is proposed by accounting for the effect of third invariant in the form of lode angle. Tensile and compressive damage are treated separately by considering strain rate effects and damage functions for tensile and compressive behavior. In order to account for the effect of shear dilation, a fully associative flow rule is considered. A user defined material model UMAT code has been developed for the proposed model and implemented in an FEM framework in LS-DYNA. The performance of the proposed model is then compared with in-built LS-DYNA concrete material models. The following inbuilt models namely : K&C Concrete Model (KCCM), Continuous Surface Cap Model (CSCM) and Winfrith Concrete Model (WCM) are considered for comparison. An unconfined and confined compressive and tensile test are done numerically and the advantages of proposed model over in built material model KCCM, WCM and CSCM has been brought out. A parametric study has been performed with the proposed model by varying the thickness of the slab, reinforcement ratio, scaled distance, and concrete strength. It seen that the slab thickness and scaled distance play a significant role in influencing the blast analysis results. The value of maximum deflection decreased by increasing the slab thickness and by increasing the scaled distance. But in case of increasing the strength of concrete, the slab shows a very small reduction in maximum deflection. From both the comparative and parametric studies the proposed model was found to be more efficient for blast analysis.