Search

Strength prediction of bonded joints remains a challenging task made even further complicated for hybrid joints in which adhesive is supplemented by mechanical fasteners. The present paper complements a series of experimental investigations on hybrid joints combining adhesive bonding with pre-tensioned bolts on galvanised and coated steel substrates, and considering two adhesives. The paper starts with modelling the strength of the bonded connection under the simultaneous action of tensile and compressive normal stresses, and shear. It then presents how these data are moulded into appropriate failure criteria and used the latter as input data for a subsequent full probabilistic finite element analysis. Following the paths of Weibull and subsequent researchers, joint capacity is predicted with reasonable accuracy for a dozen different combinations of substrates, adhesive, and pre-tension level – for which, on average, an accuracy slightly below 10% has been achieved. The results deliver meaningful insights into the subtle relationship between adhesive bonding, mechanical fastening, failure criteria, and resulting predictions methods. [ABSTRACT FROM AUTHOR]

Accelerating the polymerisation of adhesives has been a long-established field of research in context of adhesive bonding technology as long curing times – sometimes up to days – represent a decisive disadvantage in contrast to most mechanical joining techniques like riveting or screwing. In addition to, e.g., curing via UV, microwave or IR radiation, electromagnetic induction represents a promising solution for speeding up the cure of polymers. With this method, adhesively bonded components are subjected to an alternating electromagnetic field (EMF) that induces heat in EMF-sensitive materials, e.g., steel or aluminium. If intrinsic EMF-sensitivity of the materials to be bonded is not given, different types of susceptors like meshes, fibres or particles are added to the adhesive in order to be heated inductively and thus cure the adhesive from the inside. Recent research has focused on a special type of susceptors, so-called Curie particles (CP), which can only be heated inductively up to their specific Curie temperature (Tc) at which CP-induced heating automatically stops. As a result, a curing process is created that eliminates the need for external temperature monitoring while simultaneously preventing overheating of the bond. As underlying curing kinetics – most importantly polymerisation enthalpy and curing time – significantly differ depending on the considered adhesive, induction times needed to achieve full cure must currently be determined through costly preliminary investigations on an experimental level. Thus, to contribute for a more efficient and controllable bonding process, the present study aimed at developing a numerical model based on the Finite Element Method (FEM) and capable of predicting the curing degree α in dependency of curing temperature profiles TCure(t) and the CP content ccp. For this purpose, the curing kinetics of two fundamentally different 2K epoxy adhesives were linked to a transient heat flow simulation based upon experimentally determined heat loads. The validation of the developed FEA technique was successfully carried out using the example application of inductively cured large-scale Glued-in rod (GiR) specimens, whereby experimentally determined temperature profiles showed excellent agreement with the numerical predictions. The present paper focused on presenting preliminary experimental work as well as all analytical methods implemented for the numerical modelling. [ABSTRACT FROM AUTHOR]

This last paper of a series of three concludes the analysis of a series of adhesively bonded tube-in-tube (TiT) joints composed of glass-fibre reinforced polymers (G-FRP) for which induction heating with Curie-particles (CP) was used to accelerate curing. It focuses on the prediction of their load capacity using a probabilistic method (PM). First, the G-FRP material was characterised with regard to transverse (tensile) and shear stresses, which allowed to determine a valid failure criterion, including the related statistical description subsequently required for the PM. Then, finite element analysis (FEA) was used to determine the stresses within the TiTs, for not only a perfect geometry, but also considering a series of possible geometrical imperfections (offset and misalignment of the centrelines, and ovalisation) with three degrees of severity. The influence of aforesaid imperfections on stresses was determined, and discussed. Based thereupon, load capacity was numerically determined, compared to the experimental values, and the influence of imperfections highlighted. Predicted joint capacities agreed well with the experimental ones. The influence of centreline offset and ovalisation was found to be of minor importance, if compared to that of the misalignment. By an extension of the procedure, scattering of the experimental data was also predicted in form of upper and lower quantiles, which also agreed well. [ABSTRACT FROM AUTHOR]