PSI - Issue 38

240 Hamza Abbad El Andaloussi et al. / Procedia Structural Integrity 38 (2022) 238–250 Hamza Abbad El Andaloussi, Luc Mouton, Firas Sayed Ahmad, Xabier Errotabehere, Stéphanie Mahérault-Mougin, Stéphane Paboeuf/ Structural Integrity Procedia 00 (2021) 000 – 000 3

The design/geometry of the tested specimens is unique and rarely exactly in line with the actual geometry to be build. However, the objective of a fatigue test campaign is to define the fatigue behavior of a typical design details that will be derived in several application cases with adapted thicknesses, size and configuration. Reading the stress state in the fatigue specimen allows the designer to translate the results from the tests to a similar fatigue detail but adapted to the specificity of a project. The fatigue test results are thus presented in Stress to Number of cycles curve, the well-known and broadly used S-N curve. It is a recognized engineering practice adapted to assess the adequacy of a design to the fatigue loading that will be withstood in service. Several fatigue steel details include a geometrical discontinuity which induce a stress singularity: theoretically there is a stress concentration that drives the stress to infinite numbers in the very close vicinity of the notch. It threats the S-N curve approach as the stress level is intimately related to the distance of evaluation from the notch. The S-N curve could lead to any results depending on the distance of stress reading. To counter this problem, a stress reading methodology is associated to the fatigue results, method of stress calculation, distance from the notch, mesh size or other are part of the standard. To come back to bonded assembly, there are few references about bonded assembly fatigue strength in the literature showing the establishment of an S-N curve, see Broughton et al. [6] and Mazumba et al. [7]. The data commonly presented is the average stress along the bondline. This figure (average stress along bondline) is directly interesting for comparisons of adhesive, surface preparation, ageing conditions etc... on identical specimens geometry. However, for an actual repair on an offshore unit, the geometry will always be different from a test case. For instance, the bondline length will be adapted to the industrial repair geometry. Indeed, the increase of the bondline length, will cause a dramatical decrease of the average stress while the strength will increase up to a certain limit which is dictated by the stress concentration at the edge of the bondline [1]. Hence such data and test campaigns cannot be generalized and used for design in other configurations: longer overlaps, substrate thickn esses… Indeed, the singularity is systematical on the edge of the bondline for standard bonded assemblies. The use of techniques such as distances from the notch, mesh size etc... is facing difficulties as the stress profile is highly dependent on the substrate stiffnesses, overlap length, adhesive thickness, edge profile, type of loading.…Additionally, most adhesives present highly non -linear behavior then defining a stress to number of cycles fatigue curve becomes very difficult. It is generally considered that fracture mechanics and crack propagation techniques based on initial default [8] are preferable. Those techniques used in the aeronautical industry are quite complex and not common in the marine and offshore industry. They are based on a frequent inspection survey, far from what can be done on a large offshore unit. Thus, a pillar of reliability cannot be demonstrated, and the development of bonded assembly is hindered. As already detailed elsewhere [1] [2] , the design of the studied new bonded reinforcement allows to limit the stress concentration to a finite value instead of having a stress singularity and therefore, the design allows to use a stress-based approach for strength criteria. So, it was decided to carry out a large fatigue tests campaign for the studied bonded reinforcement to take into account accurately the real fatigue phenomenon for such type of assembly. Moreover, in order to have a fatigue strength representativeness, it was decided to carry out fatigue test campaign on full-scale bonded specimens in order to remove the scale effect. The objective of this campaign is to represent a bonded reinforcement installed on a ship’s deck and subject to cyclic WBM and SWBM. The present paper details the treatment of this problem on a specific industrial product. 3. Fatigue tests campaign In this section, the experimental tests campaign is detailed: test set-up and obtained results are presented. The purpose is to collect a sufficient amount of mechanical cyclic tests on the studied bonded reinforcement that will constitute the basis for the construction of a design S-N curve. For each specimen tested with a given cyclic load level, the number of cycles at failure and the fracture mode is determined.