The main objective of this study is to investigate the design S-N curve on the basis of the Hot Spot Stress (HSS) approach considering properly thickness effect for welded aluminum structures. The effect of main plate thickness is examined through results of fatigue tests in this study and references to fatigue data of previous literature. Published hot spot design S-N curves of standards provide un-conservative or over-conservative side assessments in some case of main plate thickness. We propose a more suitable hot spot design S-N curve for welded aluminum structures with various main plate thicknesses based on the research of the thickness effect.
The HSS includes the stress raising effect of structural discontinuities excluding stress concentration at weld toe. The HSS depends on macrostructural effects and loading conditions. The fatigue strength assessment at weld toe based on the HSS approach is typically used where nominal stress is not clearly defined due to complex structural effect, or where classification of welded joint is difficult. The HSS approach is widely used for fatigue strength assessments of welded steel structures.
Some standards describe the fatigue strength assessment based on the HSS approach for welded aluminum joints. To the best of the author's knowledge, two standards give design S-N curve on the basis of the HSS for welded aluminum joints. The two standards are the International Institute of Welding (IIW) recommendations (2016) and the guide for building and classing for liquefied gas carriers with independent tanks (American Bureau of Shipping (ABS), 2010). In the IIW recommendations, fatigue strength class of the hot spot design S-N curve for non-load-carrying joints is FAT 40 (FAT refers to the characteristic stress range at 2 million cycles).
The thickness effect is a decrease of the fatigue strength with an increase in the main plate thickness. The cause of thickness effect is raising local stress at weld toe with an increase in main plate thickness. In the IIW recommendation, the thickness effect is considered in case of main plate thickness over 25 mm, same as welded steel structures.
Partanen and Niemi (1999) reported fatigue test results of welded aluminum joints whose main plate thickness up to 6 mm, and they mentioned that the hot spot design S-N curve of the IIW recommendations is suitable. On the other hand, Tveiten, Xiaozhi and Berge (2007) indicated that the fatigue strength of main plate thickness 10 mm was lower than the hot spot design S-N curve of the IIW recommendations. Zamzami and Susmel (2017) investigated fatigue strength of welded aluminum joints, and concluded that the hot spot design S-N curve of the IIW recommendations is adequate level of conservatism; however, thickness effect for HSS approach is not considered. Maddox (1995) carried out fatigue tests of main plate thickness between 3 and 24 mm, and pointed out that consideration of thickness effect is necessary even if main plate thickness is less than 25 mm.
