Buckling and postbuckling behaviour of Glare laminates containing splices and doublers. Part 2: Numerical modelling

Abstract

A 3D finite element model using cohesive elements and continuum (bulk) material damage models was developed to examine the progressive damage and failure behaviour of Glare R Fibre Metal Laminate (FML) specimens subjected to in-plane compressive loading. The specimens contained internal 'splice' and 'doubler' features and were either pristine or contained simulated manufacturing defects in the form of artificial delaminations. The initiation and growth of delaminations at the inter-laminar interfaces, damage in the glass fibre reinforced polymer (GFRP) plies, ductile damage in the resin pockets (FM94 epoxy) and the onset of plasticity in the metal layers were examined. Geometric imperfections and load eccentricity were incorporated in an explicit dynamic nonlinear analysis implemented in the software Abaqus/Explicit. A series of buckling tests on specimens with and without artificial delaminations were conducted for validation, which are described in detail in a companion paper. Tests were monitored using Digital Image Correlation (DIC) for visualisation of full-field displacements and strains whilst Acoustic Emission (AE) monitoring enabled detection and localisation of the onset and progression of damage. Results for 'Glare 4B' specimens incorporating longitudinal and transverse delaminations into both splice and doubler geometries are presented. These results revealed that in order for the finite element analyses to be validated, all the damage and plasticity mechanisms * Corresponding author. Email address: Al-AzzawiAS@cardiff.ac.uk Preprint submitted to Composite Structures May 9, 2017 described above need to be accounted for, as well as load eccentricity and geometry imperfections.

Buckling and postbuckling behaviour of Glare laminates containing splices and doublers. Part 1: Instrumented tests

“…This is supported by the fact that the increase in cumulative energy is more gradual for the doubler specimen than for the splice specimen due to the smaller number of damage mechanisms which are active. This is in agreement with the Finite Element analysis presented in [37].…”

“…This suggests that the transverse doubler joint investigated here does not promote large scale delaminations as observed for the longitudinal splice joint. This is supported by the detailed Finite Element analysis presented in[37]. Contours of out-of-plane displacement (left) against location of AE events (right) for the doubler specimen (with an artificial defect) at different in-plane displacements ∆x (dashed lines indicate position of doubler).…”

“…Using the concept of uniform strains, and utilising the material strengths presented in the accompanying paper [37], a simplified stress analysis shows that fibre damage will not occur in the 90 • GFRP plies outside the joint at Figure 18. No delamination initiation or propagation from the embedded defect was noticed in the SEM results.…”

“…This is followed by a gradual increase in energy which corresponds to delamination initiation and growth along the splice feature during postbuckling. Again, this interpretation is backed by the analysis of detailed nonlinear Finite Element models in[37]. SEM of splice specimens along section 1 (side of specimen), with magnifications of (a) 100× and (b) 1000×.…”

“…Approximate 'lamina stresses' can be recovered during the initial elastic regime via the assumption of uniformity of in-plane strains. This will be discussed later in this paper, based on the material strengths provided by the accompanying paper[37].The in-plane displacements were obtained from DIC data instead of the machine cross-head displacement as the former is believed to be more accurate (since it is not affected by the compliance of the machine). In terms of the ultimate compressive loads for the splice coupons the experimental values Axial force versus in-plane displacement for splice specimens; (a) pristine and(b) with an artificial defect.…”