Elsevier

Composites Part B: Engineering

Volume 96, 1 July 2016, Pages 354-363
Composites Part B: Engineering

Effect of adhesive thickness on the Mode I and II strain energy release rates. Comparative study between different approaches for the calculation of Mode I & II SERR's

https://doi.org/10.1016/j.compositesb.2016.04.042Get rights and content

Abstract

The assembly of composite components using adhesives, is one of the principal methods used for the assembly process. Composite materials have lately been used extensively in different engineering domains, as a result their use in wind energy, or specifically wind turbine blades has also seen a significant increase. This paper deals with the problem of determining the resistance of bonding and the effect of adhesive thickness on the resistance based on the strain energy release rate in both Mode I (peeling) and Mode II (shearing) type failure. Different theories are used to measure the SERR's and the effect of thickness on the measured SERR can be seen pronounced with each theory. This has been attributed to the type of failure and also the deformation of the interface before failure in each thickness case.

Introduction

Composite materials are at the forefront of innovative materials for the manufacture of all sorts of mechanical components. From aeronautics to automotive to the energy sector, their usage is ever increasing. As is typical for large components, composite structures are assembled. Most of the assembly is done using adhesive bonding. Wind turbine blades used for the production of wind energy are no exception. Currently almost a 100% of wind turbine blade manufacturers (medium to large size) manufacture them with composite materials, in components later assembled to form complete structures. The composite blades being large in size are joined with adhesives that can reach tens of millimeters in thickness. Such thick adhesives cannot be considered as interfaces in the assembly design and the behavior of the massive adhesive should also be taken into account.

The delamination and debonding of composite assemblies is one of the main sources of failure [1], [2], [3], [4]. During service the composite components undergo solicitations which can be normal and parallel to the interface plane. These solicitations cause transverse and in plane stresses. These stresses therefore, when attain a certain value can cause the initiation of cracks. These cracks usually originate at internal defects present within the structure or singularities due to stress concentrations as a result of design. Furthermore the internal structure such as ply stacking and fiber configurations such as weaved mats can be a source of initiation as well [5].

The most common type of failure in composite is that of delamination and failure at the bonded interface [6]. This is usually due to the fact that the matrix and adhesive is a lot weaker in strength and stiffness than the fibers hence under the applied loads they tend to fail first. The same in tension matrix failures apply to in-plane compression; where the load is applied in the plane of the plies. Kink band formations cause tensile stresses in the matrix thus causing it to fail, eventually giving rise to damage nuclei [7], [8].

As explained earlier wind turbine blades [9] like most composite structures [10], [11] are bonded with adhesives [12], [13]. They are manufactured in several different components then they are assembled together with the help of adhesives. In the particular case for wind turbine blades it should be noted that, due to the quality of manufacturing, the adhesive thicknesses vary from position to position. Most of the design and research found in the literature bases the properties of the bonded joints as an interface and the effects of thickness of the adhesive on the measured SERR is ignored. This paper is an attempt to address this issue commonly ignored. Furthermore in this paper a number of analytical approaches are used to calculate the SERR's, although cohesive models and their different variants are also used for calculation of SERR [14]. The various studies show an effect of the shear stiffness of adhesive layers on the measured SERR thus rendering the effects of thickness an important design parameter [15].

Section snippets

Materials tested

The materials used in this study are the ones taken directly from a real wind turbine blade. They constitute of a 45° Bi-axial fiberglass mat of 0.286 mm thickness in a Polyester resin matrix. For bonding a NORPOL Polyvinylester adhesive of different thicknesses has been used. The specimen geometries are shown in the corresponding sections.

The composite has been prepared using hand layup and cured under vacuum bagging under atmospheric pressure at room temperature. The composite mechanical

Mode I opening mode: experimental setup

Mode I type interfacial failure also called opening or peeling is the type of failure which occurs when the applied force or stress is normal to the crack plane.

The specimens used for the characterization of adhesive bonds are shown schematically in the Fig. 1. The adhesive bonded specimens are made up of a 45° Biax type composite bonded together with varying thicknesses of NORPOL adhesive. The faces plates for the composite sandwich specimens are equally made up of 45° Biax type composites.

Observations on results

The results from experiments show some deviation. There are some sources of error which are introduced directly from the experiments. These are discussed here in this section.

  • 1.

    The Mode I SERR values for bonded specimens show good correlation and as expected for thinner specimens the deviation between analytical methods and the VCCT is not very high.

  • 2.

    The degree of deviation in the results can be accredited to the fact that, for bonded beams, since the specimens are thicker there is not lot of

Conclusions

The SERR calculated for the opening mode (Mode I) does not show any significant change in the measured value, when the thickness of the adhesive is changed. As long as the failure mode remains the same there is very little variation in the calculated SERR. However for the case of the plane shearing mode (Mode II) SERR there is some effect of the adhesive thickness on the calculated value. Although these differences in values have been seen to arise from asymmetry in the shearing plane,

References (28)

Cited by (25)

  • Estimating microscale DE parameters of brittle adhesive joints using genetic expression programming

    2022, International Journal of Adhesion and Adhesives
    Citation Excerpt :

    Popular adherends such as glass, stainless steel, timber, aluminium and various composites can be frequently seen in current experimental reports [8–11], as the selection of adherends will result in various failure modes. It is also found that the adhesive thickness yields significant influences on the fracture toughness of joints [12] as it leads to a variation of plastic zone or fracture process zone (FPZ) at the crack tip. In addition to the experimental approaches, in order to capture more insightful information of the failure mechanism of adhesive joints, numerical methods have been widely employed [13,14].

  • Ultrasonic reflection characteristics of FRP-to-FRP bonded joints with thick adhesive layers for bonding evaluation: Theoretical analysis

    2020, Composite Structures
    Citation Excerpt :

    It’s inevitable that bonding degradation will occur during their service life [1]. Among wind turbine blades, FRP plates are bonded together by adhesives, whose thickness ranges from several to dozens of millimeters in thickness [2–5]. For conventional bonded joints in engineering fields like aerospace, automobile, etc., the adhesive layer is sufficiently thin compared to wavelength and is often modeled by single interface with zero thickness.

  • Additive manufacturing of polymer composites: Processing and modeling approaches

    2019, Composites Part B: Engineering
    Citation Excerpt :

    This requires a multiscale analysis that is based on the fundamental laws of nature and links the electronic structure through the atomistic and mesoscale length to the macroscopic material behavior. Figs. 26 and 27 show some examples demonstrating the scale analysis of the polymer composites [172–178]. In each scale, a number of different physical phenomena are considered for the characterization of the printed part.

  • Mechanical performance evaluation of sandwich panels exposed to slamming impacts: Comparison between experimental and SPH results

    2019, Composite Structures
    Citation Excerpt :

    A fixture system made of steel 355S having a weight of 58.5 kg was used to fix these samples in the machine. Experimental characterization of these specimens under static loading has been presented in the previous works [11,12]. The material properties are presented in Table 1.

View all citing articles on Scopus
View full text