PSI - Issue 75

Aijia Li et al. / Procedia Structural Integrity 75 (2025) 318–333

319

2

Aijia Li, Christian Garnier,, Marie-Laetitia Pastor, Xiaojing Gong, Clément Keller/ Structural Integrity Procedia (2025)

1. Introduction Thanks to their great mechanical performances and lightweight, multidirectional carbon fiber reinforced composites have been widely used in various industrial applications, including aeronautics (Katnam et al. (2013)), automobiles (Wazeer et al. (2023)), wind turbines (Brøndsted et al. (2005) and Kensche (2006)), and sports goods (Grande et al. (2018)). During the service of these materials, biaxial fatigue loads are quite common and critical, but traditional uniaxial fatigue testing methods are inadequate for studying biaxial fatigue behavior. Consequently, conducting biaxial fatigue research is of significant importance (Chowdhury et al. (2014)). Because comprehensive fatigue research requires sufficient and reliable experimental results, researchers and engineers have developed a variety of multiaxial fatigue testing methods over the past 60 years (Quaresimin (2015)): tension-tension biaxial loading, tension-torsion biaxial loading, bending-torsion biaxial loading, tension biaxial loading, and triaxial loading, as shown in Fig. 1. (a). Due to the willingness to understand damage mechanics and acquire variables such as temperature and displacement fields (Schmidt (2012) and Huang et al. (2020)), this paper will focus on tension-tension biaxial fatigue loading on the cruciform specimen, which allows a uniform and in-plane field for observation (Quaresimin (2010) and Quaresimin (2015)). Several issues regarding the specimen design pose difficulties for the tension-tension biaxial fatigue test. The first problem is the lack of a standard for the specimen geometry, resulting in plenty of work and costs in the specimen design and preliminary tests (Chowdhury et al. (2014), Busca (2014), Baptista et al. (2016), and Baptista et al. (2015)). Secondly, unexpected local damages and failures outside the gauge region, like delamination and arm breakage, can greatly affect the experimental results and even lead to a failed test. Thirdly, specimen design needs to take into account the requirements of various non-contact measurements, such as DIC (Digital Image Correlation) and IRT (InfRared Thermography), for further damage analysis. For instance, a relatively planar gauge region is favorable for temperature observation and calculation by IRT, while the displacement measurement by DIC has to involve a visible, bright, and representative region. Moreover, considering a more complex specimen geometry than that of uniaxial tests, manufacturing methods must be considered as a key factor in specimen design to avoid manufacturing defects like voids, cracks, and ply folds, which can lead to errors and deviation in the experimental results (Talreja (2013), Hörrmann et al. (2016), and Persson et al. (1997)). In the last decade, based on the cruciform shape, researchers have tried to optimize the specimen geometry mainly by justifying the gauge region, where a designed biaxial stress state can be observed and analyzed. As illustrated in Fig. 1. (b), circular hole (Baptist et al. (2016), Satapathy et al. (2013) and Skinner et al. (2019)), elliptical hole (Skinner et al. (2020) and Pham et al. (2021)), and reduced area (Moncy et al. (2021), Lua et al. (2022) and Moncy et al. (2023)) are the most used configurations for the gauge region, leading to a stress concentration that will be more critical than that around the transition region. Apparently, the continuity of the displacement and temperature cannot be guaranteed for a gauge region with a hole, and the dimension of the reduced area can be further improved to obtain a more ideal stress, strain, and temperature field for observation.