Small-Scale Component Tests
In order to derive reliable design methods to predict the adhesives’ strength behavior via continuum mechanical analyzing methods, various small-scale push-out tests were carried out. For the standardized tests (Table 1, IV), the load carrying behavior is treated separately for shear and tension loading (Table 1, VI). The tests take into account different adhesives and additionally the geometric influence of different joining geometries (Table 1, II). The general transfer of characteristic values determined by standardized characteristic tests (Table 1, IV) on bulk material or simple shear connections to small-scale components also has been investigated.
When the force-deflection behaviors of different joining geometries for the same load case (either shear or tension) are compared, it becomes evident that bonded connections with U – and L-profiles and butt splice joints and channel bondings have very similar structural behavior in terms of stiffness, strength, and ductility. Figure 3 shows results for the two-component polyurethane (K05) using the four joining geometries in Table 3, which are representative of the tested flexible adhesives. Connections with U – and L-profiles hereby show a considerable increase of strength and stiffness in comparison with butt splice and
FIG. 3—Force-displacement behavior for two-component polyurethane (K05) subjected to shear loading using different adhesive geometries (Table 3).
Copyright by ASTM Int’l (all rights reserved); Tue May 6 12:07:08 EDT 2014 Downloaded/printed by
channel bonded connections. The bonding geometry (U-shaped joint with three flanks versus butt joint) substantially determines the constraint of the lateral expansion, which, for the U – and L-shaped geometries, leads to a visible stiffening effect. The higher carrying capacity for the U-shaped joints results from the enlarged adhesive surface.
For simple shear loading, Fig. 4 shows that the stiffness of the push-out specimen can be reproduced by the standardized tests very well, whereas for simple tension loading this is not successful (see Fig. 5). In both figures the maximum loads of small-scale tests are generally higher due to scale effects.
Thus it is explicable that the results of the standardized block shear and tensions tests, which in the beginning of the INNOGLAST project are expected to be mechanical simple tests, cannot be transferred to large-scale component tests for every type of loading. The block shear test with significant bending effects in particular does not present a simple shear test.
In fact, most adhesives show a quasi-isotropic material behavior for small strains in the elastic region, but depending on the type of adhesive the state of stress and the following yielding and failure—in particular for elasto-plastic
Copyright by ASTM Int’l (all rights reserved); Tue May 6 12:07:08 EDT 2014
FIG. 5—Stress-strain behavior of standardized specimen and component tests subjected to tension loading.
CopyrightbyASTMInt’l(allrightsreserved);Tue May 6 12:07:08 EDT 2014 Downloaded/printed by
adhesives—is predominantly determined by the hydrostatic stress portion (see Eq(1))
To characterize this hydrostatic stress effect, simple tension tests, simple shear tests, and butt joint tests with constraint lateral expansion are performed; see the section “FE Modeling" and Fig. 9.
Due to the existing compressibility of elasto-plastic, stiff adhesives, the influence of the hydrostatic state of stress for these adhesives is larger than that for flexible, visco-elastic adhesives with approximately incompressible behavior (volume constancy), such as silicones or visco-elastic polyurethanes.
Based on the pronounced restraint of the lateral expansion, the push-out tests reveal a multiaxial state of stress (caused by a uniaxial state of strain), whereas the standardized tension tests offer a state of uniaxial stress. This fact must be considered for the FE calculations, choosing appropriate material laws and flow rules, e. g., with Drucker-Prager or Schlimmer-Mahnken .
For these mechanical reasons, and also for other important reasons, following the evaluation in Table3, not all of the four joining geometries are quite applicable for all adhesives. Major differences have been found in the area of the adhesive surface, the quality of work, the controllability, and the aesthetic appearance of the bonded joint. Indeed, bonded joints realized with U – or L – profiles proved to be more sustainable and more ductile due to the redistribution capacity from the frontal part to the flanks of the bonded joint. However, for these two geometries the disadvantages regarding production predominate, especially when they are being applied to large components. Here only casting with low viscosity adhesives is applicable. Focusing in particular on simplicity while taking into account all relevant criteria gained within this project, the butt joints turn out to be the most promising ones.