Experimental characterization of the inter-ply shear behavior of dry and prepreg woven fabrics: Significance of mixed lubrication mode during thermoset composites processing

https://doi.org/10.1016/j.compositesa.2019.105725Get rights and content

Abstract

A custom-built test device was developed to measure the inter-ply shear resistance of dry carbon woven fabrics as well as prepregs at different temperatures, consolidation pressures, and forming rates typical to those used in autoclave processing. A quantitative understanding of the nature of the shear forces encountered during fabric consolidation was established, suggesting the strong dependence of these forces on the process conditions along with the dominance of a mixed lubrication regime. Among the most influential state variables, the resin viscosity and distribution of the resin on the prepreg surface, as well as the surface roughness of the reinforcing fibers could be identified.

Introduction

Composite materials offer reduced weight while maintaining high strength and stiffness, as well as improved fatigue resistance when compared to a range of metallic counterparts [1]. Over recent decades, a wide range of forming techniques have been developed for composites. There is a fairly large number of reinforcement materials available, both in the dry or pre-impregnated forms, while the fabric architecture taking various forms such as unidirectional, woven, or non-crimp. The core motivation for the introduction of the composite processes in the industry is increased automation, reduced production cycle cost/time, and increased part quality. Yet, there still exists a range of difficulties in reaching high-quality, reliable manufacture of fibrous composite structures, which is partly associated with the continuous and complex nature of the reinforcement, particularly when they are in the textile form [2], [3]. The forming operation becomes more problematic when structures of a certain geometrical complexity, e.g., doubly curved parts, need to be fabricated. In particular, the inter-ply shear (also known as inter-ply slip or inter-ply friction), which is defined based on the net displacement between the two sliding prepreg surfaces of uncured composite prepreg materials, has been reported as a key process parameter behind the formation of major defects in composite components. These defects include geometrical distortions [4], [5], [6], core crushing during autoclave processing of honeycomb sandwich components [7] and out-of-plane wrinkles due to an excessive resistance of the plies during forming [8], [9]. Above all, wrinkling is regarded as a critical quality issue by designers, as it can lead to the reduced mechanical properties and lifespan of the final structure [10]. The severity of wrinkling can be affected by various manufacturing factors, including the mold selection, composite material characteristics, lay-up configuration, the interaction between plies, and quality of processing equipment [11], [12], [13], [14]. The stresses developed during deformation can be partly or entirely relaxed by inter-ply shear [15]. Upon consolidation, the composite laminate is composed of alternating fiber-rich layers and resin-reach inter-layers. The inter-ply shear is often desirable since otherwise an initially flat stack of plies cannot be formed acceptably to a single or double curvature part [15], [16]. A good understanding of each step of the manufacturing process and the forming behavior of a given composite material system is key to mitigate such defects and control the properties of the final part. In the area of unidirectional (UD) composites, the causes for process-induced defects and the underlying mechanisms have been fairly identified across various bodies of the published work, particularly in the area of thermoplastic composite materials which was thoroughly investigated as a benchmark study by Sachs et al. [17]. Nevertheless, sporadic research has been reported to thoroughly characterize the behavior of uncured thermoset woven fabric prepregs and understand the influence of the processing parameters, which is also vital for implementation of material models via numerical methods to simulate forming processes.

Similar to the intra-ply shear characterization, no standardized method for the determination of friction between plies of reinforcing textiles has been prevailed thus far [18], [19]. Nonetheless, several attempts have been made at adapting methods used for the UD materials. The main difficulty is to determine what portion of the prepreg’s response is in charge of the perceived behavior; i.e., whether the response is resin-dominated (viscoelastic), or fiber-dominated, and therefore purely elastic. This is exacerbated as current forming techniques often involve several modes of deformation at once, not to mention the existence of e.g. thermoplastic tougheners in the recent generations of prepregs which make the material composition even more complex [20]. Inter-ply shear is often treated as a friction/lubrication phenomenon. As the majority of regular friction test instruments are tribological by design [21], [22], [23], a variety of plate-type setups have been employed to characterize this mode of deformation on woven fabrics [24], [25], [26], [27], [28], [29]; namely, to describe the ply-ply and tool/ply friction behaviors where a slider moves over a test bed while the resulting frictional force is recorded through a load cell [30]. Additionally, as a recent pull-out method, there are configurations with vertical traversing directions, where a carrier plate with the spanned sample is pulled between two pressure plates [31], in which a single ply is drawn from a small laminate whilst under some degree of pressure. The latter test has also been evolved from the single fiber pullout testing to testing cured composite materials where failure mechanisms, and whereby the interface around the fiber cracks, can be studied during inter-ply characterization [32]. For an uncured prepreg, however, the problem is more complex, as the behavior is heavily influenced by the uncured resin, which results in some viscoelastic yield, rather than a comparatively simple Columb-type frictional behavior [7]. There is also a problem with the continuity of the sample. As the ply is drawn from the stack, the sample area is effectively reducing, which in turn lessens the resistance to movement. In the original form of the test, this is not problematic, as it accurately represents the failure mechanism. When considering forming, however, movement of the ply in one area of the laminate can result in drawing material from elsewhere; i.e., no void would form [33]. Therefore, for this test to be representative of forming or consolidation processes, it is suggested that the sample area be maintained constant. This issue was addressed using the rigs developed by Murtagh [28], Wilks [34] and Larberg et al. [35], in which two small plates can be drawn into contact with a larger, central plate, with the use of a pressure cylinder. The small side plates can then be drawn in the central plate and the frictional resistance can be measured, while the region of interest (ROI) remains unchanged.

There is a fair body of literature focusing on the friction of dry tows and woven fabrics, whereas very limited work has been reported on the behavior of uncured fabric prepregs. The term dry in this context refers to the absence of a lubricating resin film between the two interacting plies, or the surface finish (known as sizing) applied by the tow manufacturer. The selection of the semi-finished product, the textile architecture, and the yarn density are deemed to be significant factors influencing the inter-ply shear behavior of the material system. The frictional mechanics developed between the fabric plies also depend on the interactions arising from the mesostructure. Cornelissen et al. [26] observed that the textile architecture prevents the fibers in their spreading and that the real contact area is limited to the peaks and valleys of the undulated tows. The normal force maxima, therefore, occur in those locations, showing strongly oscillating force curve progressions [36], [37]. In contrast with unidirectional reinforcements, inconsistencies in the frictional behavior could result from the superimposed and offset positioning of the layers to each other [38]. It has been experimentally shown that fabric forming defects depend on the tool geometry, the process parameters, and relative fabric orientation and significantly on inter-ply shear [39], [40]. The interaction between two fabric plies was found to be a mixture of yarn-to-yarn friction and interlocking caused by the interaction of individual yarns [41]. Nezami et al. [42] recently studied the interaction between the plies with high shear angles. They noticed increased friction values due to the higher structural integrity of sheared plies.

Thermoplastic and thermoset resins substantially alter the inter-ply shear behavior of the composite reinforcements by introducing a new dimension into the system, namely the viscous state between the plies. Inter-ply shear in thermoset prepregs is partly accompanied by shearing of the thin, resin inter-layers. This resin layer has high mobility and flow characteristics, whereby temperature and time dependence of the viscosity can be a complicating factor due to its lubrication effect on the interface, making these class of resins complex to investigate in terms of frictional regimes at different processing conditions. Thermoplastic resins, on the other hand, are highly viscous and rubberized and hence cannot be easily removed from the interface, providing a more stable lubricating effect at higher temperatures [21].

In early studies on inter-ply shear behavior, Murtagh [28] and Schrer et al. [16] measured static friction values to perform thermoplastic forming simulations. Thermoforming of thermoplastics occurs at temperatures greater than the melting point of the plastic matrix, hence the resin is essentially fluid during processing. In the case of thermosets, however, the viscosity profile of the resin changes significantly over the course of the cure cycle [21]. It is initially high, followed by a decrease to an almost inviscid liquid (depending on if the resin is toughened or not). Eventually, at the gelation point, it dramatically increases to infinity at the cure cycle’s hold period. In the context of defect generation, inter-ply shear measurements need to be performed throughout the heat-up period as the resin is still soft and fluid-like. Larberg [35] determined the prepreg-prepreg friction coefficients for four carbon fiber epoxy systems, examining a number of material property-related drivers, showing that there is a significant variation of friction coefficients among aerospace-grade prepreg systems with similar viscosities. Interestingly, different prepreg systems also responded differently to the displacement rate in terms of the friction coefficient. Material consolidation and prepreg surface roughness were also found to be drivers behind inter-ply shear [43]. Initial modeling of this behavior has been typical to represent the yield point with a simple Coulomb friction law [44]; however, it has been noted that the system is clearly much more complex. In the results obtained by Scherer et al. [16], the interface yield point was taken to be the y-intersect; however, the load noticeably had a two-stage response, with a transition from an initial gradient to an apparently frictional behavior. In the inter-ply shear test, a similar initial response but with some residual post-yield stiffness can be perceived [45], assuming the yield point to be at the transition of the two gradients. This has major implications for forming processes that involve large amounts of shear strain, such as double diaphragm forming (DDF) [46].

In another benchmark study [17], different factors were investigated for their influence on the friction characteristics of a TWINTEX fabric to a metal surface. For the same materials and test configuration, differences of more than 50% were found for obtained values of the dynamic friction coefficient, denoting the inadequate expressiveness of actual tests and their dependence on the test setup and the variability of the as-delivered material. Additionally, variables such as forming rate, temperature, or compaction pressure can widely differ among different stages of a manufacturing process, such as between performing and consolidation stages, and hence it is critical that the process parameters studied be representative of those used in the actual manufacturing floor. Table 1 summarizes a survey of past studied input variables for different forming and consolidation scenarios.

Understanding the various lubrication regimes provides insight into how processing parameters would influence the inter-ply shear behavior. Friction between two surfaces in contact may be purely hydrodynamic, as when the two surfaces in relative motion are completely separated by a fluid lubricating film [47]. In this case, the friction may be predicted in terms of the traction forces acting on the film. If no fluid separates the interface, the friction is governed by the force normal to the interface and generally described as Coulomb friction, governed according to Eq. (1) in which f is the frictional force, µ is the coefficient of friction and N is the normal force to the interface:f=μN

For hydrodynamic friction, as given by Eq. (2), shear stress is described by rate of movement of one surface over other, fluid viscosity and thickness of the fluid film [48], where τ is the shear stress, η is the viscosity of fluid, h is the thickness of fluid film and v is the velocity of one surface over other.τ=ηhv

Here, inspired by the design of journal bearings, friction and lubrication theories are employed to study the effects of friction in composite forming, as the resin matrix at certain temperatures and strain rates can act as a lubricant between prepreg plies [49]. The Stribeck curve provides an overall view of the friction variation in the entire range of lubrication [16]. As illustrated schematically in Fig. 1, the curve provides a graphical representation of the friction coefficient, μ as a function of the Hersey number, a non-dimensional quantity that is dependent on the dynamic viscosity of the lubricant film, η, velocity, ν, and the Normal force N. The curve is classified into three regions based upon the level of contact between the plies or surfaces of interest; namely, (a) Boundary lubrication region in which Coulomb friction is dominated, (b) Hydrodynamic lubrication, where shear deformation of the fluid film is dominant, and (c) Mixed lubrication which is a combination of the above two modes. A number of past studies have used the Stribeck curve to chart the frictional region as a function of composite process parameters [50], [51]. Studying the comingled TWINTEX fabric, Gorczyca et al. obtained the majority of the Hersey numbers within the hydrodynamic friction region, [24], whilst for thermoset prepregs, data points obtained by Larberg [20] adapted to the transition from the mixed to hydrodynamic region, although more experimental data were needed to fully confirm the behavior.

Remark: Recalling the literature, there appears to be a considerable variation in interpreting the frictional response of composite reinforcement materials, particularly at elevated temperatures. The differences between the magnitude of frictional force at various pressures, rates, and temperatures are still not fairly clear and the observed trends are usually fed to empirical models and master curves in order to be used in finite element simulations [24], [48]. The concept of tribology proposed above seems to be a promising route to delve into an improved understanding of this phenomenon. Ideally, a physics-based definition of inter-ply shear capable of handling the transition between boundary lubrication, mixed lubrication, and the full hydrodynamic state would be of interest by designers to cover the entire range of material processing windows. For thermoset prepregs, the onset of fiber entanglement is believed to be the state where the response falls in the mixed lubrication zone. Ersoy et al. [49] noted that the thinning of inter-layer resin due to low viscosity can increase the interfacial shear stress and reduce stress relaxation due to increased mechanical interlocking of the contacting fibers. A thorough understanding of this phenomenon is aimed in the sections to follow to chart not only hydrodynamic and boundary lubrication modes, but also the mixed lubrication mode during forming, as a function of processing parameters.

The goal of the present work is to gain an improved understanding of inter-ply shear in autoclave processing of thermoset woven fabric composites and thereby, move towards developing standard tests capable of capturing the appropriate mechanisms and to demonstrate how these tests can be applied to manufacture of flaw-less woven composite components. Previous investigations of the frictional behavior of fabrics (per Section 1.1) mostly served as tests for producing fitting data for forming simulations and were limited to static and dynamic friction coefficients. In this contribution, however, a mixed-lubrication mechanism leading to different levels of inter-ply shear in fabric composites forming is realized. Section 2 will present the employed test material, a customized test setup, and a characterization plan. Section 3 discusses the experimental results, followed by outlining the main conclusions and future prospects in Section 4. It will be realized throughout the work that extra measures should be taken into account as to accurately capture the inter-ply interaction of thermoset prepregs during forming and consolidation events.

Section snippets

Test materials

The materials selected to conduct inter-ply shear testing consisted of two sets of plain weave carbon fiber fabrics in dry and prepreg forms, respectively. The matrix was a 177 °C curing controlled-flow epoxy (CYCOM 970) resin system, that was already pre-impregnated onto the fabric material. The as-received material was a weight-balanced plain weave pattern, as shown in Fig. 2. Next to the prepreg state, including the dry fabric in the design of experiment (DOE) provided the opportunity to

Dry carbon fabric tests

The dry fabric samples were aimed to act as a baseline for subsequent analysis on the prepreg material. While these tests were conducted as a comparison of friction measurements, they could also provide insight into the Coulomb friction mechanisms of fibers prior to introducing resin onto the dry reinforcement. The absence of the partially-cured resin layer assists in the effective comparison of inter-ply shear behavior and surface roughness. A typical inter-ply shear response at pressures of

Concluding remarks and future prospects

A dedicated custom-built experimental setup and testing methodology were designed and manufactured which enabled the investigation of the inter-ply shear behavior of composite fabric reinforcements at varying interfacial temperatures, rates, and compaction pressures. The experimental parameters were selected in the ranges that mimic those encountered in the consolidation process in autoclave processing of laminates and sandwich panels. The materials investigated consisted of two different types

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgment

The authors wish to sincerely acknowledge the financial support from the Natural Sciences and Engineering Research Council (NSERC) of Canada. Prepregs were provided by the Boeing Canada Technology Ltd., Winnipeg. The support of colleagues at the Composites Research Network, and in particular, the stimulating discussions with Mr. Mark Shead, Mr. Kurtis Willden, and Dr. Robert Courdji from the Boeing Company, are greatly valued.

References (63)

  • K. Farnand et al.

    Micro-level mechanisms of fiber waviness and wrinkling during hot drape forming of unidirectional prepreg composites

    Compos A Appl Sci Manuf

    (2017)
  • R. Scherer et al.

    Inter-and intraply-slip flow processes during thermoforming of CF/PP-laminates

    Compos Manuf.

    (1991)
  • U. Sachs et al.

    Characterization of the dynamic friction of woven fabrics: experimental methods and benchmark results

    Compos A Appl Sci Manuf

    (2014)
  • M.H. Kashani et al.

    Analysis of a two-way tension-shear coupling in woven fabrics under combined loading tests: global to local transformation of non-orthogonal normalized forces and displacements

    Compos A Appl Sci Manuf

    (2016)
  • J. Cao et al.

    Characterization of mechanical behavior of woven fabrics: experimental methods and benchmark results

    Compos A Appl Sci Manuf

    (2008)
  • W. Stanley et al.

    Intraply shear characterisation of a fibre reinforced thermoplastic composite

    Compos A Appl Sci Manuf

    (2006)
  • L. Wang et al.

    Characterization of inter-ply slipping behaviors in hot diaphragm preforming: experiments and modelling

    Compos A Appl Sci Manuf

    (2019)
  • B. Cornelissen et al.

    Frictional behaviour of high performance fibrous tows: friction experiments

    Compos A Appl Sci Manuf

    (2013)
  • A.M. Murtagh et al.

    Surface friction effects related to pressforming of continuous fibre thermoplastic composites

    Compos Manuf

    (1995)
  • J. Page et al.

    Prediction of shear force and an analysis of yarn slippage for a plain-weave carbon fabric in a bias extension state

    Compos Sci Technol

    (2000)
  • B. Cornelissen et al.

    Dry friction characterisation of carbon fibre tow and satin weave fabric for composite applications

    Compos A Appl Sci Manuf

    (2014)
  • Y.R. Larberg et al.

    On the interply friction of different generations of carbon/epoxy prepreg systems

    Compos A Appl Sci Manuf

    (2011)
  • S. Allaoui et al.

    Effect of inter-ply sliding on the quality of multilayer interlock dry fabric preforms

    Compos A Appl Sci Manuf

    (2015)
  • D.M. Mulvihill et al.

    Effect of tool surface topography on friction with carbon fibre tows for composite fabric forming

    Compos A Appl Sci Manuf

    (2017)
  • L. Montero et al.

    Characterisation of the mesoscopic and macroscopic friction behaviours of glass plain weave reinforcement

    Compos A Appl Sci Manuf

    (2017)
  • S. Erland et al.

    Characterisation of inter-ply shear in uncured carbon fibre prepreg

    Compos A Appl Sci Manuf

    (2015)
  • N. Ersoy et al.

    An experimental method to study the frictional processes during composites manufacturing

    Compos A Appl Sci Manuf

    (2005)
  • A.M. Murtagh et al.

    Characterisation of shearing and frictional behaviour during sheet forming

    Compos Mater Ser Elsevier

    (1997)
  • L. Karaoğlan et al.

    Frictional contact/impact response of textile composite structures

    Compos Struct

    (1997)
  • E.J. Barbero

    Introduction to composite materials design

    (2017)
  • A. Levy et al.

    Corner consolidation in vacuum bag only processing of out-of-autoclave composite prepregs laminates

    Proceedings of SAMPE 2014 technical conference

    (2014)
  • Cited by (0)

    View full text