Co-authors​ :

 Oliver VOELKERINK (GERMANY), Patrick Adrian MAKIELA (GERMANY), Christian HÜHNE (GERMANY) 

Adhesive bonding is a suitable joining technology for lightweight fibre composite structures. In contrast to bolted joints, no holes have to be drilled, therefore no load bearing fibres are cut and the cross section of the structure is not reduced. To add, stresses perpendicular to the load direction are distributed more evenly. In addition to strength, an important requirement for bonded joints in aircraft design is the desired failure mode. In case of overload a failure of the surrounding structure and not the adhesive bond itself is desired. This design philosophy ensures that the adhesive bond is not the weakest link in the structure. The substantiation can then be focused on the undisturbed composite structure. In order to reduce the experimental effort in the design and substantiation, modeling and simulation methods are required that can predict not only strength but also the failure mode of the structure. One approach are numerical mesoscale progressive damage analyses in which each ply of the composite as well as the adhesive bond are modeled individually.
However, there are several different approaches to model the fibre composite adherends in a mesoscale analysis. Shell, continuum shell or solid elements can be used in combination with 2D in-plane failure criteria and continuum damage mechanics models for shells or with fully 3D criteria and models for solids. Furthermore, delamination can either be explicitly modelled using methods such as cohesive zone modelling or modelled in a smeared way by capturing the effects of delamination in the 3D failure criterion and in the damage model. Two main questions arise: What is the best modelling strategy for the fibre composite ad- herend as trade-off between simulation runtime and predictive capability? The second related question is: Can the modelling strategy be determined on the basis of fibre composite coupon tests without any adhesive bonds loaded primarily in-plane, or must it be determined on the basis of bonded joint coupon specimens which capture the out-of-plane loading due to sec- ondary bending and thus are close to the target application?
To answer these questions, progressive damage analyses of open hole tension (OHT) as well as single lap shear (SLS) specimens are performed with different modelling approaches for the fibre composite. The results are then compared to own experimental data. First, the best modelling strategy is evaluated for each application. Based on this, the results for both applications are then compared in order to answer the question of whether the recommenda- tion from the OHT simulations also applies to the SLS simulations.
In this way, the question of the best modelling strategy as a compromise on speed and ac- curacy of fibre composite adherends in SLS joints with secondary bending is addressed. In addition, it is shown whether simulations of laminate-only coupons with in-plane loads are suitable for determining the modeling strategy in bonded joints.