Service Communautaire d'Information sur la Recherche et le Développement - CORDIS

FP5

BONDSHIP Résumé de rapport

Project ID: G3RD-CT-2000-00101
Financé au titre de: FP5-GROWTH
Pays: United Kingdom

Numerical and theoretical modelling bases and results explaining the mechanical behaviour

Superstructure modelling. The superstructure modelling has helped to investigate the general behaviour of the global superstructure and outline hot spots and load levels in specific bonded areas. This load response of the structure was performed according to DNV rules, thus defining a worst-case situation for the ship.
Adhesive modelling approach. As the adhesive chosen by VT presented non-linear behaviour, it was proposed to use a non-linear finite element model to study the strength of the different joints. Several approaches available with the ANSYS package were considered but only one was sustained: a multi-linear approach. This method presented good agreement between experimental and numerical modelling for butt strap tests with 1-, 3- and 5-mm adhesive thickness, whereas for the 10-mm adhesive thickness, divergence occurred when load-displacement curves were compared.
Stress-strain approach.

This approach assumes the adhesive joint to be perfectly made, without any flaws. To assess the stress state in the adhesive layer, 3 approaches were considered: analytical method derived from classical theories, linear elastic finite element method and non-linear elastic finite element method. Analytical methods proved to be quick to implement, giving a first estimate of the stress state in the adhesive although not as accurate as the finite element formulations. In the case of structural joints, the finite element formulation has enable to predict stress concentration in the bondline and give input for fracture mechanics study as cracks are likely to propagate in such areas of high stress.

Fracture mechanics / damage tolerance approach. The damage tolerance approach considers that a damaged structure is acceptable but to a certain level. Therefore, the investigation focused on the prediction of the maximum acceptable size of crack for a given load applied to the bonded structure. After analysis of DCB tests results, the latter were compared to numerical values obtained from FE models of structural joints with different crack sizes. It has been found that the fracture energy of the DCB tests was far beyond the fracture energy level reached in the numerical model of damaged structural joints. However, these results should be considered carefully since DCB tests were performed with a 1-mm adhesive thickness whereas the structural joints specification states a 3-mm adhesive thickness.

Potential applications for these results are for the design of large bonded structures. End-users of these results would be naval architects, structural designers and owners who would be interested in implementing this technology in the construction of new units. The main innovative features of this work is a validated new approach for assessing adhesively bonded structures. Success factors will be a gain in weight in the structures, the possibility to fasten different materials and cost savings.

Informations connexes

Reported by

University of Southampton
Highfield
SO9 5NH Southampton
United Kingdom
See on map