Mode I fracture of epoxy bonded composite joints: 1. Quasi-static loading

https://doi.org/10.1016/S0143-7496(00)00038-5Get rights and content

Abstract

Mode I constant displacement rate tests were conducted on epoxy-bonded CFRP joints at –50, 22 and 90°C. A comparison of experimental compliance and different beam theory approaches indicated that care needs to be taken when applying beam theory approaches across a wide temperature range. Temperature was seen to influence the mode of fracture which progressed from stable, brittle fracture at low temperatures to slip-stick fracture at room temperature and finally to stable ductile behaviour at elevated temperatures. This behaviour has been attributed to the dependence of critical strain energy release rate on crack velocity for epoxy adhesives and a model for the fracture behaviour of viscoelastic materials has been used to explain these results. The critical strain energy release rate was seen to increase with temperature and the failure locus transferred from predominantly in the composite substrate to predominantly in the adhesive.

Introduction

The use of composite materials in aerospace applications is becoming more widespread due to the improved performance that can be achieved by a reduction in weight. Adhesive bonding is seen as a desirable joining technique for these materials as it can offer substantial benefits over other joining methods such as mechanical fastening. Adhesive bonding offers improved fatigue performance, potential reductions in life-cycle maintenance costs, and also allows for greater flexibility in design. However, greater understanding of the fatigue performance of such bonded structures and, in particular, any degradation of that performance due to adverse environments, is needed for the benefits of adhesive bonding to be fully achieved.

Previous work at DERA in this area has centred on the mixed-mode loaded joints, such as the double-lap and lap-strap joints, that are representative of loading in real aerospace structures [1], [2], [3]. In this work it has been shown that both fracture mechanics and strain based failure criteria can be used to predict the fatigue thresholds in long overlap joints in a variety of environments. The total strain energy release rate was found to be a useful failure criterion when applied to joints with a similar mode mix. We are now expanding the range of geometries tested and assessing the approach used by Mall et al. [4], [5], [6] and Kinloch and co-workers [7], [8], [9], [10] in using mode I fracture data to predict failure in uncracked mixed-mode joints.

The double cantilever beam (DCB) sample has been chosen to generate the mode I fracture data. This is a popular test because of the ease of sample manufacture and testing, coupled with simple analysis methods. Ripling et al. [11], [12], [13] first adapted the DCB test for testing structural adhesives and suggested a theory based on a built in beam which neglected the contribution of the adhesive. They suggested that rotation at the assumed built in end could be corrected for by using an empirically derived rotation factor that was added to the measured crack length. Later workers [14], [15], [16], [17], [18] modelled the DCB as a beam on an elastic foundation in order to theoretically account for the contribution of the adhesive to the compliance.

The DCB has been used to generate fracture data for composites [19], [20], [21], [22], bonded composites [4], [5], [6], [7], [8], [23], [24], [25] and bonded metal joints [9], [10], [11], [12], [13], [14], [26], [27], [28]. The DCB is frequently used to generate fatigue data as well as quasi-static data. This aspect of our work is covered in more detail in Part 2 of this paper.

In this investigation DCBs composed of carbon fibre-reinforced polymer (CFRP) adherends bonded with an epoxy adhesive were tested in the temperature range −50 to 90°C. This temperature range was selected as typical of the limits to be expected in a supersonic aircraft. The effect of temperature on the fracture energy, the mode of fracture and the locus of failure was evaluated using automated crack measuring techniques and microscopy techniques. A number of analysis methods have been assessed with respect to their applicability across this temperature range and the results have been explained with reference to a model for fracture in a viscoelastic material. Part I of the paper is concerned with quasi-static loading of the DCBs and in Part 2 crack propagation for samples subjected to cyclic loading is described.

Section snippets

Theory

The energy criterion for crack growth is based on the work of Griffith [29]. The driving force for crack growth is that the stored elastic strain energy released when the crack grows must be at least as great as the energy required to create the new surfaces. The following equation can be derived for the strain energy release rate, G, in a plate with a through thickness crack, assuming linear elastic behaviour:G=P22bdCda,where P is the applied load, b the specimen width, a the crack length and C

Sample preparation

Samples were produced by adhesive bonding cured panels of CFRP. The composite was prepared from unidirectional pre-preg consisting of intermediate modulus graphite fibres in a BMI/epoxy matrix. A unidirectional lay-up of 16 plies at 0° was used to produce 2 mm thick panels. These were autoclave cured at 180°C for 60 min. The mechanical properties of the composite used in the analyses were those supplied by the manufacturer and are shown in Table 1. The adhesive used was a toughened epoxy film

Room-temperature testing

Fig. 4 shows the results of a typical constant displacement rate test conducted at room temperature. As expected, the load is seen to decrease with increasing crack length during the test. However, it is apparent from this plot that growth is not continuous, but instead proceeds as a succession of rapid growth and arrest phases. This is commonly referred to as stick-slip growth. It can also be seen that there is an increase in the size of the steps as the crack progresses. We can calculate two

Comparison of analysis methods

All three analysis methods gave similar results at 22°C, but at 90°C there was some difference between the values. The experimental compliance (Berry) method and the beam theory displacement method (CBT2) showed a reasonable correlation with one another, but the beam theory load method (CBT1) gave considerably lower values of GIc. This may be due to the simplification of assuming the longitudinal stiffness of the composite does not change in the test temperature range. If the stiffness of the

Conclusions

The various methods available for calculating mode I strain energy release rates from double cantilever beam joint geometries can result in different values. This is particularly true for beam theory related procedures when uncertainty may exist as to appropriate material properties, or if there is likelihood of significant joint non-linearity. Under such circumstances an experimental compliance method is the recommended option.

In the joints studied, modes of fracture, locus of failure and

References (43)

  • Ashcroft IA, Hughes DJ. Fatigue of epoxy adhesives. DERA Technical Report DERA/MSS1/CR990289/1.0, March...
  • S. Mall et al.

    Cyclic debonding of adhesively bonded composite

  • Mall S, Johnson WS. Characterisation of mode I and mixed mode debond failure of adhesive bonds between composite...
  • Mall S, Rezaizdeh MA, Ramamurthy G. Interaction of mixed mode loading on cyclic debonding in adhesively bonded...
  • A.J. Kinloch et al.

    Predicting the fatigue life of adhesively bonded joints

    J Adhesion

    (1993)
  • M.N. Charalambides et al.

    Strength prediction of bonded joints. 83rd Meeting of the AGARD Structures and Materials Panel

    (1996)
  • J.K. Jethwa et al.

    The fatigue and durability behaviour of automotive adhesives. Part IIIpredicting the service life

    J Adhesion

    (1997)
  • Little MSG. The durability of structural adhesive joints. Ph.D. thesis, Imperial College, London,...
  • E.J. Ripling et al.

    Measuring fracture toughness of adhesive joints

    Mater Res Stands

    (1964)
  • S. Mostovoy et al.

    Use of crack-line-loaded specimens for measuring plane strain fracture toughness

    J Mater

    (1967)
  • E.J. Ripling et al.

    Fracture mechanicsa tool for evaluating structural adhesives

    J Adhesion

    (1971)
  • Cited by (126)

    • Fatigue behavior and life prediction of wind turbine blade composite materials

      2023, Advances in Wind Turbine Blade Design and Materials
    • Characterization of adhesive joints under ballistic shear impact

      2022, International Journal of Adhesion and Adhesives
    View all citing articles on Scopus
    View full text