DEVELOPMENT AND FINITE ELEMENTS ASSESSMENT OF STACKING SEQUENCES FOR INTERLAMINAR FRACTURE TOUGHNESS TESTING OF ANGLE-PLY INTERFACES

Interlaminar fracture toughness of laminated composites may be characterised using standard test procedures (see, e.g., for mode I testing, [1]). All existing standards, however, recommend the use of unidirectional (UD) layups. This restricts the characterisation to 0°/0° delamination interfaces, while structures are typically built using multidirectional (MD) layups and delamination may appear in any interlaminar layer. Since fracture toughness may depend upon layers orientations, it is clear that its characterisation in angle-ply interfaces is of paramount importance. However, as confirmed by relevant literature, a certain number of problems emerges when multidirectional layups are used. Most of these problems are related to additional damage mechanisms that may appear during experimental tests and to the coupled nature of MD specimens. Mechanical couplings of the stacking sequences used to fabricate specimens may cause thermal residual stresses that may appear during the curing phase and that may considerably affect fracture toughness values obtained experimentally. Furthermore, pure modes delamination tests require that the considered mode be the dominant one, while others be null or negligible; if mechanical couplings exist in the specimen, they affect the kinematics of the specimen and may induce parasite modes contributions during tests. Lastly, data reduction techniques adopted to obtain fracture toughness from tests are usually based on 2D models: while this may be an acceptable hypothesis for UD specimen, it is to be questioned in the case of MD ones. For these reasons, characterisation of interlaminar fracture toughness in angle-ply interfaces is still an open problem and layups able to prevent such issues are sought. The present study describes the conception of a novel type of specimen for testing interlaminar fracture toughness in angle-ply interfaces of laminated composite materials. While maintaining standard geometric characteristics (see, e.g., [1]), the novelty of the specimen lies in its stacking sequence. Such sequence is developed using Quasi-Trivial (QT) solutions [2,3]. They are a particular class of sequences satisfying the condition for membrane-bending uncoupling and/or homogeneity (i.e. equal normalised membrane and bending stiffness matrices) of a laminate, in the framework of Classical Laminated Plate Theory (CLPT). In more detail, QT quasi-homogeneous solutions (that is satisfying both membrane-bending uncoupling and homogeneity) are used for each of the two arms of a typical DCB delamination specimen. Additionally, such sequences are designed in order to have null in-plane couplings and bending-torsion coupling: thus, each arm of the specimen is a specially orthotropic laminate. Furthermore, sequences for the two arms are chosen in order to comply with superposition rules for QT solutions presented in [4]. As a result, the stacking sequence of the entire specimen is a QT quasi-homogeneous solution too. It has null in-plane and bending-torsion coupling terms: thus, the entire specimen is a specially orthotropic laminate, too. To sum up, such a specimen avoids all coupling problems, for both its arms and for its entire sequence. It’s worth mentioning that QT solutions are defined solely by specification of orientations positions in the sequence, regardless of orientations values. In other words, orientations values can be chosen freely without affecting the quasi-triviality of the sequence. Hence, from a unique QT solution multiple sequences, with different orientations, may be generated. Moreover, many different QT solutions may be used with the approach described. This allows generation of MD specimens with all kind of delamination interfaces (and maintaining all the described properties). To validate the proposed approach, a sequence with a 0°/45° delamination interface has been developed and compared to other sequences taken from relevant literature and having the same delamination interface [5]. A Finite Elements (FE) model of the DCB test [1] has been developed and a revised VCCT formulation [6] has been used to evaluate ERR partition and distributions at the initial straight delamination front. Results obtained show that the proposed design strategy could allow fabrication of delamination specimen having significantly reduced parasite modes contributions during delamination tests. Eventually, other sequences have been developed and studied by means of FE analysis in order to investigate some effects of stacking sequence on ERR modal partition and distribution. REFERENCES [1] A. D5528-13, Standard test method for mode I interlaminar fracture toughness of unidirectional fiber-reinforced polymer matrix composites, ASTM, International West Conshohocken (PA), URL: www.astm.org. [2] P. Vannucci, G. Verchery, A special class of uncoupled and quasi-homogeneous laminates, Composites Science and Technology, 61, 2011, p. 1465-1473. [3] T. Garulli, A. Catapano, M. Montemurro, J. Jumel, D. Fanteria, Quasi-trivial solutions for uncoupled, homogeneous and quasi-homogeneous laminates with high number of plies, in: R. Owen, de Borst, Pearce (Eds.), ECCM VI, International Center for Numerical Methods in Engineering (CIMNE), pp. 255-265. [4] T. Garulli, A. Catapano, M. Montemurro, J. Jumel, D. Fanteria, Quasi-trivial stacking sequences for the design of thick laminates, Composite Structures 200, 2018, 614-623. [5] T. Garulli, A. Catapano, D. Fanteria, J. Jumel, E. Martin, Design and finite element assessment of a Fully-Uncoupled Multi-Directional (FUMD) specimen for delamination tests, Composites Part B, paper submitted. [6] P. S. Valvo, A revised virtual crack closure technique for physically consistent fracture mode partitioning, International Journal of Fracture 173 (1), 2012, p. 1-20.