A new rate-dependent unidirectional composite model – Application to panels subjected to underwater blast

Abstract

In this study, we developed a finite element fluid–structure interaction model to understand the deformation and failure mechanisms of both monolithic and sandwich composite panels. A new failure criterion that includes strain-rate effects was formulated and implemented to simulate different damage modes in unidirectional glass fiber/matrix composites. The laminate model uses Hashin’s fiber failure criterion and a modified Tsai–Wu matrix failure criterion. The composite moduli are degraded using five damage variables, which are updated in the post-failure regime by means of a linear softening law governed by an energy release criterion. A key feature in the formulation is the distinction between fiber rupture and pull-out by introducing a modified fracture toughness, which varies from a fiber tensile toughness to a matrix tensile toughness as a function of the ratio of longitudinal normal stress to effective shear stress. The delamination between laminas is modeled by a strain-rate sensitive cohesive law. In the case of sandwich panels, core compaction is modeled by a crushable foam plasticity model with volumetric hardening and strain-rate sensitivity. These constitutive descriptions were used to predict deformation histories, fiber/matrix damage patterns, and inter-lamina delamination, for both monolithic and sandwich composite panels subjected to underwater blast. The numerical predictions were compared with experimental observations. We demonstrate that the new rate dependent composite damage model captures the spatial distribution and magnitude of damage significantly more accurately than previously developed models.

Introduction

Benefiting from lower material density and with strength comparable to steel, fiber-reinforced composite materials have been widely used in the design of aircrafts, marine hulls, and automobiles under dynamic loadings (Bakis et al., 2002, Lee et al., 2000, Porfiri and Gupta, 2010, LeBlanc and Shukla, 2011, Arora et al., 2011, Arora et al., 2012, Dear et al., 2005, Dear and Brown, 2003). In our previous work (Wei et al., in press), a three-dimensional finite element model was formulated and implemented in ABAQUS (Abaqus 6.9-ef online documentation, 2009) to simulate the deformation and failure of monolithic and sandwich composite panels subjected to underwater blast eliciting fluid–structure interactions (FSI) (Latourte et al., 2011). In such model, the strain-rate effect was modeled for the inter-lamina interface and the foam core in the sandwich panels using a rate-dependent cohesive law and a crushable foam plasticity model, respectively. The model successfully predicted the deformations of both monolithic and sandwich panels over a wide range of water impulses. Furthermore, the model revealed the importance of the foam core in enhancing the energy absorption for sandwich composite panels. However, because the model did not include strain-rate effects on the composite laminate failure, a discrepancy was found for both monolithic and sandwich composite panels when predicting spatial distribution and magnitude of the fiber and matrix damage compared to experimental measurements. Therefore, in order to improve model predictive capabilities and gain a better understanding of the behavior of composite panels under high strain-rate conditions requires the formulation of a more accurate and easy-to-implement model in which material rate-dependence is taken into account.

In this paper, we briefly review the previous numerical model developed to simulate the FSI experiments for monolithic and sandwich composite panels sustaining under-water blast loadings. We then discuss the newly formulated rate-dependent damage model for unidirectional fiber-reinforced composite laminates, including a new failure criterion for orthotropic composite materials based on a modified fracture toughness function for fiber tensile failure. Next, we apply the new model to simulate scaled FSI lab experiments to predict monolithic and sandwich panel deformation and failure. We also compare the new predictions to those obtained in Wei et al. (in press) using Hashin’s rate independent model. We close with a discussion of possible future applications of the newly formulated numerical model.