Options
2012
Conference Paper
Title
Seemingly contradictory: Influence of stress-reduction-methods on the strength of bonded joints composed of brittle adherends
Other Title
Scheinbar widersprüchlich: Einfluss von Spannungsreduktionsverfahren auf die Festigkeit von Klebeverbindungen aus spröden Fügeteilen
Abstract
Adhesively bonded joints are characterized by sharp stress peaks towards the ends of the overlaps, which trigger failure, thus defining strength. A repeatedly reported and at first sight intuitive method to increase the strength of these elements is to reduce the stress peaks by either acting on the geometry, e.g. chamfering the adherends or implementing fillets in the adhesive layer, or on the adhesive, e.g. using ductile adhesives or grading them. Regarding brittle adherends, i.e. FRP and timber, three commonly reported methods to reduce stress magnitudes, i.e. chamfering, ductile adhesives, and adhesive grading, were experimentally and numerically investigated. These investigations, pursued on a wide range of parameters resulting in five different experimental series, allowed for the following observations: (1) All of the investigated stress-reduction-methods do, according to FEA, lead to significant stress reductions, yet the so reduced stresses act on larger volumes of the adherends; (2) All experimentally investigated adhesively bonded joints failed in a brittle manner, so the materials they were constituted of; (3) None of the investigated stress-reduction-method, neither for FRP nor for timber, did significantly increase the experimentally determined joint strength. The latter listed observations did allow drawing the following conclusions: (4) There is no direct correlation between stress magnitude and failure, even considering a verified failure criterion, when dealing with brittle materials; (5) Thus, direct stress based approaches to predict the strength of adhesively bonded joints composed of brittle adherends are not suited for dimensioning purposes. Considering the aforementioned, the authors computationally developed a probabilistic strength prediction method, which implemented a statistical based formulation of size-effects. The probabilistic method not only offers a comprehensive a mechanically coherent for the seemingly contradictory relation between stress reduction and lack of corresponding joint strength increase, it also delivers consistently accurate joint strength estimates for all considered experimental series which makes it useful for their dimensioning.
Conference