Experimental assessment of Fully-Uncoupled Multi-Directional specimens for mode I delamination tests

Abstract

This paper proposes an experimental study whose goal is to validate the concept of Fully-Uncoupled Multi-Directional (FUMD) specimens for delamination tests. In order to reduce the likelihood of delamination jump, a glass/epoxy UD-fabric composite was used to fabricate double cantilever beam specimens. Mode I tests were performed with standard UD specimens and with FUMD specimens having the following delamination interfaces: 0 °//0 °, $0°$// $15°$, 0 $°$//30 $°$, 0 $°$//45 $°$ and −45 $°$//45 $°$. Delamination fronts after the tests were observed by means of ultrasonic C-scans. By comparison with the UD specimens, it is shown that FUMD sequences are able to promote a fairly symmetric and flat delamination front. Analysis of the rotations of such specimens during the test confirms that they do exhibit an uncoupled mechanical behaviour. Critical energy release rate is found to be dependent on interface plies mismatch angle, but not on global stiffness of the specimens.

Introduction

In the field of structural design, composite materials represent an appealing alternative to metals. Sectors like the aerospace and the automotive have adopted more and more these materials, due to their exceptional specific properties, that allow the design of efficient lightweight structures. In particular, long-fibres composite laminates are widely adopted in structural design, even for safety-critical components. Therefore, understanding the damage processes of laminates is of paramount importance.

Among others, delamination is one of the most critical damage modes, as it may significantly reduce mechanical properties of the structure, while being very difficult to detect. Hence, the design of safe structures demands a thorough knowledge of the delamination resistance of materials [1], typically in terms of critical strain Energy Release Rate (ERR). For these reasons, delamination has been long studied [2], and standards for evaluation of interlaminar properties of composite laminates have been developed, in particular for mode I [3], mode II [4] and mixed mode I/II [5]. For mode III, no standard exists yet. One of the most promising solutions proposed so far is the Edge-Crack Torsion (ECT) test [6]. However, research is still ongoing [7] and mode III toughness values obtained with this test may not be accurate and reliable [8].

Despite the fact that existing standards are recommended for non-woven UD materials only (due to the limited experience gained in round robin tests), extensive experience has been collected on their use, both for UD tape and woven materials [9]. However, this concerned mostly laminates with UD stacking sequences with the main reinforcement direction parallel to the specimen longitudinal direction ($0°$). As a consequence, only the interface between equally oriented plies at $0°$ (UD interface hereafter) is usually characterised. Instead, applicability of standards to laminates with multidirectional (MD) stacking sequences is still questionable. The reason is that UD laminates have an uncoupled thermoelastic behaviour allowing optimal test conditions and accurate data reduction techniques. MD laminates, on the contrary, have a more complex behaviour that lead to a certain number of problems that prevent the attainment of consistent and reliable results [10]. Consequently, delamination testing of non-UD interfaces (MD interfaces hereafter) is a complex task. Real structures, however, are built using MD laminates, and delamination may appear at any interface.

One first issue during tests on MD laminates is the appearance of additional energy dissipation mechanisms. In standard tests involving bending of the specimen, off-axis plies may experience matrix plasticity and intralaminar damage [11]. Intralaminar damage often leads to the delamination jump (or migration) [9,[12], [13], [14], [15], [16], [17]]. Strategies to model such phenomenon are currently being investigated [18,19].

Another relevant problem with MD laminates is the presence of thermal residual stresses [20]. It was shown that such stresses may promote matrix damage and thus facilitate delamination jump [21]. Moreover, they may greatly affect the evaluation of interlaminar fracture toughness, as demonstrated by Nairn [[22], [23], [24]] and Yokozeki et al. [25]. The importance of thermal residual stresses depends on the laminate layup [23].

A third inconvenient is the difficulty in controlling the loading mode. When pure mode tests are performed, no other ERR contributions should exist. When performing mixed-mode tests, knowledge of the exact mode-mix is required, in order to meaningfully reduce and exploit experimental data. While standard delamination test procedures [[3], [4], [5]] are devised to address these issues for UD specimens, it is not guaranteed that the same results are achieved when using MD specimens, due to the presence of elastic couplings that modify the kinematics of the specimen and that may induce unwanted rotations and parasite modes contributions to the ERR. In such cases ERR modal contributions could be evaluated numerically [26], but not without difficulties [27].

Besides modal partition, another relevant issue concerning ERR is its distribution along the width of the delamination specimen. Common data reduction procedures for delamination tests are based on 2D theories: a straight delamination front and a uniform ERR distribution are assumed, even though this has long been proven to be an idealisation [28]. Nowadays, tools to predict delamination growth direction have been developed, following geometrical considerations [29], or within the framework of Cohesive Zone Model (CZM) [30].

It was shown that in MD laminates 3D effects can become important and affect the shape of the front and of ERR distribution [[31], [32], [33], [34]]. In particular, the parameter $D_c$ [31,32]:

$$D_c=\frac{D_{12}^2}{D_{11}{D}_{22}},$$

was found to be related to curvature of ERR distribution and delamination front. Terms $D_{ij}$ in Eq. (1) are the components of the laminate stiffness matrix, as obtained by Classical Laminated Plate Theory (CLPT). On the other hand, the asymmetry of ERR distribution and of delamination front was found to be somewhat related to the parameter $B_t$ [33,34]:

$$B_t=\left|\frac{D_{16}}{D_{11}}\right|.$$

In order to obtain reliable evaluations of critical ERR using common data reduction techniques, both $D_c$ and $B_t$ should be kept as small as possible [35]. According to [36], the value of $D_c$ should be much lower than 1, while in [37] it was suggested that it be lower than 0.25 (referring however to End Notched Flexure specimens).

To sum up, delamination testing of MD interfaces requires an extremely careful design of the stacking sequence for the specimens. As observed in [38,39], the ideal stacking sequence should result in specimens that avoid as much as possible additional damage mechanism, that eliminate (or at least reduce) mechanical couplings, and that are not affected by thermal residual stresses. One approach suggested is to use stacking sequences containing as much $0°$-oriented plies as possible [12,33]. This approach reduces, but does not eliminate, couplings and thermal stresses. Other authors used FE simulations to try and find suitable stacking sequences [16,21]. Eventually some authors adopted an innovative approach by using quasi-trivial (QT) solutions [40] to obtain uncoupled specimens [[41], [42], [43]]. While only specific sequences were developed (thus with limitations on possible interfaces and on the capability to investigate other aspects), this nonetheless allowed to build specimens with interesting uncoupling properties.

This same approach was pushed further in a previous work [44], where the authors presented a method based on QT solutions and special superposition rules [45,46] to conceptually design novel Fully-Uncoupled Multi-Directional (FUMD) stacking sequences. This design approach aims at generating layups for delamination specimens that are not affected by thermal residual stresses, do not have any elastic coupling, and allow to test any desired delamination interface. Such results were obtained in closed-form solution within the framework of CLPT and verified by means of Finite Element analyses in [44]. This preliminary assessment suggested that FUMD specimens could represent an extremely appealing solution for delamination testing of MD interfaces.

To confirm such approach, this paper shows the results of the very first experimental campaign carried out using FUMD sequences and taking into account different delamination interfaces and different global stiffness properties of the specimens. The main goal of such activity was to experimentally assess the capability of FUMD sequences to deliver uncoupled behaviour, to produce the expected delamination fronts, to avoid undesired rotations and to allow an easy evaluation of the critical ERR with simple procedures proposed by standards.

The paper is organised as follows. Firstly, Section 2 recalls the fundamentals of FUMD stacking sequences and explains in detail how they were used to design the delamination specimens conceived for this study. Then, the material system and all the experimental procedures adopted, from specimen fabrication to data reduction, are described. Section 4 presents and discusses the results of the study. Eventually Section 5 ends the paper with conclusions and perspectives.