Failure Progression and Mesh Sensitivity Analyses by the Plate Element-Failure Method

Abstract

Despite the fact that fiber-reinforced composite materials have been extensively used in aerospace, defense and marine structures for many years, the high stiffness and strength properties of these materials have not yet been fully exploited. In many instances, overly rigid adherence to conservative design rules based on ‘black aluminum’ (symmetric quasi-isotropic graphite-epoxy laminates) has eroded these advantages significantly. Efforts to better understand the failure mechanisms and develop efficient and robust computational tools will lend more confidence to designers and engineers to realize better weight savings and improve damage tolerance. The search for efficient computational tools for the modeling of progressive damage in composites should be coupled with sufficient fidelity to ensure that the mechanisms of the complex failure modes are accurately represented. In an effort towards this goal, the authors have proposed the element-failure method (EFM) using 3D solid finite elements, which has been successfully applied to failure analysis of composite laminates. The EFM has advantages over the traditional material property degradation method (MPDM) because the stiffness matrix of the element is not changed, thereby improving computational convergence and saving in computational efforts. In this article, a plate element-based EFM (PEFM) is presented for modeling progressive damage in fiber-reinforced composites. The implementation in plate elements further improves the efficiency of computation, when compared to analyses with 3D elements. In conjunction with several different failure theories, the plate EFM is applied to model progressive damage in quasi-isotropic composite laminates with open-holes, subjected to remote tensile loading. A mesh sensitivity study shows that the EFM results are relatively independent of mesh size effects. When compared with experimental data, the predicted ultimate strengths and damage patterns show good agreement.