Dynamic mechanical properties and penetration behavior of reactive nano-inorganic cement-based composites

The dynamic compression, tensile, and penetration properties of a reactive nano-inorganic cement-based composite (RNICC) were tested using a split Hopkinson pressure bar (SHPB) and a light gas gun. The parameters of the Holmquist-Johnson-Cook (HJC) model suitable for describing the RNICC were determined according to experimental results and feed forward neural network, which provide a theoretical framework for the use of these materials in engineering applications. The stress-strain curves, strain rate effect, and dynamic compressive/tensile strength of the RNICC were obtained within the range 50-150 s(-1). The penetration tests for the RNICC target were performed, and the penetration resistant properties were determined according to the penetration depth and radius of the RNICC based on a 270 g projectile within the impact velocity range 100-250 m/s. The comparison on DIF and anti-penetration performance between RNICC and other kinds of UHPCC was conducted. The DIF growth rate of RNICC is larger than that of other types of UHPCCs. The RNICC targets has excellent anti-penetration performance. The HJC model is employed to describe the dynamic mechanical properties of the RNICC. A feed forward neural network model was used to train samples and determine the strength parameters of the RNICC in the HJC model. Based on discussions on the damage of concrete influenced by the principal strain and the failure shear strain, this study analyzed the failure type of the HJC model that caused compression damage in the concrete. The parameters of the HJC constitutive model were verified and could be used to effectively predict the mechanical properties and penetration behavior of the RNICC.