PSI - Issue 75

Arthur THIBAULT et al. / Procedia Structural Integrity 75 (2025) 509–518 Arthur THIBAULT/ Structural Integrity Procedia (2025)

510

2

© 2025 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of the Fatigue Design 2025 organizers Keywords : Microstructure gradient, Stereoscopic digital image correlation, Multiaxial loading, FEM representative model, Welded structures, Behavior law identification Introduction Since the 2000s, active travel has seen significant growth, particularly in cities. This trend can largely be explained by the growing environmental awareness of the public and the resulting desire to find new, non-polluting methods of transport, mainly for everyday journeys in urban areas. Among the modes of transport experiencing significant growth, bicycles are at the forefront. Indeed, 37% of French people report using their bicycle regularly, according to the survey by the Ministry for the Ecological Transition and Territorial Cohesion (2023). The increasing prevalence of bicycles, as well as the aim to reduce their environmental impact, leads to increased geometric complexity, particularly for the purpose of material reduction. To achieve this, traditional round tubing is gradually disappearing in favour of tubes with variable geometry and thickness. The transition from steel to aluminium for many models has enabled a significant increase in bicycle performance. All these changes lead to increased complexity in the structural design of bicycle frames, particularly concerning the welds. Indeed, multiple issues converge in these zones, such as variations in mechanical properties and microstructures. Research by Ambriz et al. (2011) highlighted the presence of softened zones in the HAZ (Heat Affected Zone). Furthermore, the heat input from welding leads to microstructural changes and consequently strength changes in these same zones (Ambriz et al., 2009). Welding processes and associated heat treatments also induce residual stresses, which differ between the welded zones and the rest of the structure (Mueller, E. & Hermann, T., 2022). Moreover, the thin tubular structure of bicycle frames involves significant geometric variations at the welds toe, which induces stress concentrations detrimental to static and fatigue performance (Zettlemoyer, N., 1976). © 2025 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under the responsibility of Dr Fabien Lefebvre with at least 2 reviewers per paper 1.

For all these reasons, welds therefore constitute the critical points of bicycle frames. The complexity of the problem, as well as the drive for continuous optimisation, explain why the cracks observed in Decathlon's internal laboratory exhibit fractures at the weld toe, especially under severe tests conditions (Fig. 1). Welded zones therefore present a challenge in terms of their structural design, considering the various local phenomena present in their vicinity. Furthermore, the multiaxial nature of the differents loadings imposed during standardised tests induces a complex stress state in bicycle frames, particularly at the bottom bracket. For this study, in order to limit the number of uncertainties and sources of scatter, we will focus only on manual TIG welding. For the same reasons, only one aluminium alloy with a specific heat treatment will be studied: the 6061 alloy in the T6 condition, applied after bicycle frame assembly. This precipitation-hardening alloy, particularly when subjected to this T6 heat treatment, achieves high mechanical properties while maintaining a low density, making it highly valued in the transport sector (Vargel, C., 2010).

Fig. 1. fatigue crack during normative tests at Decathlon internal laboratory

It is therefore important to have the most accurate possible simulations for the welded zones, in order to avoid the need for multiple prototypes. The improvement of the simulations currently used at Decathlon is limited by a lack of small-scale experimental data in the welded zones. The overall objective of this project is to improve the accuracy of numerical fatigue simulations of aluminium bicycle frames by taking into account local phenomena. To achieve this, two experimental campaigns were established: one involving static testing to determine local properties in the weld region, and another involving multiaxial fatigue testing. This paper details the methodology and the results obtained from the first experimental campaign.