Penetration modeling of ultra‐high performance concrete using multiscale meshfree methods

Abstract

Summary Terminal ballistics of concrete is of extreme importance to the military and civil communities. Over the past few decades, ultra‐high performance concrete (UHPC) has been developed for various applications in the design of protective structures because UHPC has an enhanced ballistic resistance over conventional strength concrete. Developing predictive numerical models of UHPC subjected to penetration is critical in understanding the material's enhanced performance. This study employs the advanced fundamental concrete (AFC) model, and it will run inside the reproducing kernel particle method (RKPM)‐based code known as the nonlinear meshfree analysis program (NMAP). NMAP is advantageous for modeling impact and penetration problems that exhibit extreme deformation and material fragmentation. A comprehensive experimental study was conducted to characterize the UHPC. The investigation consisted of fracture toughness testing, the utilization of nondestructive microcomputed tomography analysis, and lastly projectile penetration shots on the UHPC targets. To improve the accuracy of the model, a new scaled damage evolution law (SDEL) is employed within the microcrack informed damage model. During the homogenized macroscopic calculation, the corresponding microscopic cell needs to be dimensionally equivalent to the mesh dimension when the partial differential equation becomes ill posed and strain softening ensues. To ensure arbitrary mesh geometry for which the homogenized stress‐strain curves are derived, a size scaling law is incorporated into the homogenized tensile damage evolution law. This ensures energy‐bridging equivalence of the microscopic cell to the homogenized medium irrespective of arbitrary mesh geometry. Results of numerical investigations will be compared with results of penetration experiments.

Assessment of the ballistic impact response of Cor-Tuf UHPC concrete using the HJC constitutive model

“…The HJC model’s prediction of the damaged area on the front and rear faces of the concrete target is compared to the results of the experimental data available in (Sparks et al, 2019), which are presented for a projectile impact with an initial velocity of 817 m/s, which is shown in Figure 8. The cratering sizes are determined by measuring the radius between the center of the impact site to the edge of the cratering region in eight radial directions, each 45° apart for both the experiments and simulations.…”

“…The HJC model is evaluated for high strain rate loadings through a validation procedure by comparing the parameters for the Cor-Tuf concrete to existing experimental data of ballistic impacts to Cor-Tuf concrete targets. Ballistic impact experiments of the Cor-Tuf concrete were taken from the literature (Sparks et al, 2019) which included a range of impact velocities. In this study, the Cor-Tuf concrete was shaped into cylindrical specimens with a diameter of 406.4 mm and an average target thickness of 31.75 mm.…”

“…Additionally, this model can be readily fit using compressive experiments available in the current literature and is available in commercially available software such as Abaqus (Simulia). The model's response for this concrete is evaluated by comparing finite element simulations and experimental data of ballistic impacts (Sparks et al, 2019) with impact velocities ranging from 617 m/s up to 1112 m/s.…”

“…Figure9. Results of the ballistic impact finite element simulations residual velocities using the HJC concrete model with the calibrated parameters for the Cor-Tuf concrete compared against the residual velocities of the experimental data presented in(Sparks et al, 2019).…”

Simulation of debris flow on an instrumented test slope using an updated Lagrangian continuum particle method

“…The past decades have seen significant development in computational methods for problems in fracture mechanics, damage modeling, strain localization, and unsaturated soil mechanics. For this type of problems, the finite element method remains the preferred computational platform [35,43,52,53,62,73,75], even as a number of continuum particle methods have also emerged as viable alternatives in recent years [35,[58][59][60]62]. However, when it comes to debris flow modeling, where the motion is so chaotic that element connectivity is difficult to impose, the finite element method may not be an appropriate platform to use, since it suffers from severe mesh distortion that impacts on its accuracy and overall performance.…”