13th International Conference on Fracture June 16–21, 2013, Beijing, China -7- Table 1. Material properties of the adhesives aE [MPa] a [‐] c [MPa] [N/mm] Ref. AV138/HV998 4890 0.35 39.5 0.38 [2] Hysol EA 9321 3870 0.36 46.0 0.45 [7] Redux 326 4440 0.35 50.9 (0.3) [6] 4.2. Comparison to experimental results Several experimental studies have been selected for comparison to the failure load predictions. The experimental studies where chosen such that only well documented and repeated studies are used. Only adhesives with substantially brittle behavior were used for the comparison. The effect of the adhesive layer thickness was studied experimentally by Castagnetti et al. [3] for steel adherends with two different adhesives. The results of the sufficiently brittle adhesive Hysol EA 9514 are used for comparison. In the work by da Silva et al. from 2004 [4] the effect of the overlap length on the failure load of the steel joints with a bismaleimide adhesive Redux 326 was studied. In the study by da Silva et al. from 2006 [7] three different adhesives were tested with steel adherends. From these adhesives two adhesives, AV138 and Hysol EA 9321, were sufficiently brittle. The material data that were used are summarized in Table 1. The strength and the fracture toughness are not identified by means of the used experimental results. Material parameters from standard tests that are given in literature [2,6,7] are used. No specific value for the fracture toughness could be found in literature for the bismaleimide adhesive Redux 326. It is assumed that the fracture toughness attains values around 0.3 N/mm. The comparisons of the present failure model to the experimental results are shown in Fig. 5a-5c. Obviously, the failure load predictions agree well with the experimental results. Especially when the scattering of the experimental results is considered it becomes clear that the failure load prediction by the present model is of good quality. Another important feature is that the effects of the geometrical parameters are covered correctly by the present model. Especially the adhesive layer thickness effect has been subject to many studies that try to explain the effect on the failure load. If only stress criteria are used it appears that the strength of the adhesive reduces if thinner bondlines are considered. No physically sound explanation can be given for such a change of the adhesive strength. The present model can correctly predict the effect of the adhesive layer thickness by an additional energetic condition that must be satisfied simultaneously. 5. Summary A new failure model for adhesive joints has been given. It bases on a coupled stress and energy criterion in the framework of FFM. A closed-form analytical model for the mechanical behavior of a single lap joint is used to set up the failure model. The failure model requires two basic failure parameters, the strength and the fracture toughness of the adhesive. A study of the effect of the geometrical parameters on the failure load on adhesive joints is shown and discussed. The predicted effects are in good accordance to general knowledge on adhesive joints. Furthermore a comparison to experimental results is shown. In general it shows a good accordance of the prediction and the experimental results. All trends are covered very well. Only failure parameters as given in published results from standard tests are used in the comparison and no fitting of the input parameters needs to be performed. A remarkable outcome is the fact that the effect of the adhesive layer thickness is covered correctly.
RkJQdWJsaXNoZXIy MjM0NDE=