Search

Adhesive bonding has been increasingly employed in multiple industrial applications. This has led to a large industrial demand for faster, simpler, and cheaper characterization methods that allow engineers to predict the mechanical behavior of an adhesive with numerical models. Currently, these characterization processes feature a wide variety of distinct standards, specimen configurations, and testing procedures and require deep knowhow of complex data-reduction schemes. By suggesting the creation of a new and integrated experimental tool for adhesive characterization, it becomes possible to address this problem in a faster and unified manner. In this work, following a previous numerical study, the mode I and II components of fracture-toughness characterization were validated experimentally in two different configurations, Balanced and Unbalanced. For mode I, it was demonstrated that both configurations presented similar numerical and experimental R-curves. The relative error against standard tests was lower than ±5% for the Balanced specimen; the Unbalanced system showed higher variations, which were predicted by the numerical results. Under mode II, the Balanced specimen displayed plastic deformation due to high deflections. On the contrary, the Unbalanced specimen did not show this effect and presented a relative error of approximately ±2%. Nonetheless, it was proven that this approach to obtain such data by using a single unified specimen is still feasible but needs further development to obtain with similar precision of standard tests. In the end, a conceptual change is proposed to solve the current mode II issues.

The use of adhesive bonding in diverse industries such as the automotive and aerospace sectors has grown considerably. In structural construction, adhesive joints provide a unique combination of low structural weight, high strength and stiffness, combined with a relatively simple and easily automated manufacturing method, characteristics that are ideal for the development of modern and highly efficient vehicles. In these applications, ensuring that the failure mode of a bonded joint is cohesive rather than adhesive is important since this failure mode is more controlled and easier to model and to predict. This work presents a numerical technique that enables the precise prediction of the bonded joint's behavior regarding not only its failure mode, but also the joint's strength, when inorganic fillers are added to the adhesive. To that end, hollow glass particles were introduced into an epoxy adhesive in different amounts, and a numerical study was carried out to simulate their influence on single lap joint specimens. The numerical results were compared against experimental ones, not only in terms of joint strength, but also their failure pattern. The neat adhesive, which showed 9% and 20% variations in terms of failure load and displacement, respectively. However, looking at the doped configurations, these presented smaller variations of about 2% and 10% for each respective variable. In all cases, by adding glass beads, crack initiation tended to change from adhesive to cohesive but with lower strength and ductility, correctly modeling the general experimental behavior as intended.

Adhesives are increasingly being employed in industrial applications as a replacement for traditional mechanical joining methods, since they enable improvements in the strength-to-weight ratio and lower the cost of the overall structures. This has led to a need for adhesive mechanical characterisation techniques that can provide the data needed to build advanced numerical models, allowing structural designers to expedite the adhesive selection process and grant precise optimisation of bonded connection performance. However, mechanically mapping the behaviour of an adhesive involves numerous different standards resulting in a complex network of various specimens, testing procedures and data reduction methods that concern techniques which are exceedingly complex, time-consuming, and expensive. As such, and to address this problem, a novel fully integrated experimental characterisation tool is being developed to significantly reduce all the issues associated with adhesive characterisation. In this work, a numerical optimisation of the unified specimen's fracture toughness components, comprising the combined mode I (modified double cantilever beam) and II (end-loaded split) test, was performed. This was achieved by computing the desired behaviour as a function of the apparatus' and specimens' geometries, through several dimensional parameters, and by testing different adhesives, widening the range of applications of this tool. In the end, a custom data reduction scheme was deduced and set of design guidelines was defined.

Adhesives are extensively used in the automotive and aeronautical industries as they enable the creation of durable and light weight joints, with exceptional strength to weight ratios. The constant search for the means of adapting the mechanical performance of adhesives to each application has led to the use of several types of fillers to change their properties. Following a study on the effect of inorganic fillers, i.e., hollow glass beads, in the failure mechanisms of single lap joint's (SLJ), this work focuses on the response of the strength and fracture properties of structural adhesives to this filler. To this end, their tensile strength and mode I fracture properties were thoroughly analyzed by performing bulk tensile and double-cantilever beam (DCB) tests, at a quasi-static speed. The specimens were manufactured by adding different %v/v of filler to two epoxy-based crash resistant adhesives. Both adhesives have shown a negligible effect on the tensile strength, a decrease in strain at failure and critical energy release rate in mode I, as well as an increase of the Young's modulus, for higher % in volume of hollow glass beads. These phenomena were further analyzed recurring to scanning electron microscopy, and the concept of rule of mixtures.