Evaluation of energy and failure parameters in composite structures via a Component-Wise approach

摘要

This paper deals with the static analysis of fiber reinforced composites via the Component-Wise approach (CW). The main aim of this work is the investigation of the CW capabilities for the evaluation of integral quantities such as the strain energy, or integral failure indexes. Such quantities are evaluated in the global structures and local volumes. The integral failure indexes, in particular, are proposed as alternatives to point-wise failure indexes. The CW approach has been recently developed as an extension of the 1D Carrera Unified Formulation (CUF). The CUF provides hierarchical higher-order structural models with arbitrary expansion orders. In this work, Lagrange-type polynomials are used to interpolate the displacement field over the element cross-sections. The CW makes use of the 1D CUF finite elements to model simultaneously different scale components (fiber, matrix, laminae and laminates) with a reduced computational cost. CW models do not require the homogenization of the material characteristics nor the definition of mathematical lines or surfaces. In other words, the material characteristics of each component, e.g. fibers and matrix, are employed, and the problem unknowns are placed above the physical surface of the body. In the perspective of failure analyses, the integral evaluation of failure parameters is introduced to determine critical portions of the structure where failure could take place. Integral quantities are evaluated using 3D integration sub-domains that may cover macro- and micro-volumes of the structure. The integral quantities can be evaluated directly on fiber and matrix portions. Numerical results are provided for different configurations and compared with solid finite element models. The results prove the accuracy of the CW approach and its computational efficiency. In particular, 3D local effects can be detected. The use of the integral failure index provides qualitatively reliable results; however, experimental campaigns should be carried out to relate such indexes to the failure occurrence. (C) 2016 Elsevier Ltd. All rights reserved.