ICF10B

Paper Ref No: ICF100764OR FATIGUE CRACK GROWTH PREDICTIONS IN AA 5083 AND AA 2024 USING A SIMPLE GEOMETRIC MODEL. Ton Riemslag(1), Kees van Kranenburg(1), Saskia Benedictus-de Vries(1), Fred Veer(2) and Jan Zuidema(1)(*) (1) Laboratory of Materials Science, Delft University of Technology (2) Faculty of Architecture, Delft University of Technology E-mail to: j.zuidema@tnw.tudelft.nl Abstract Constant amplitude fatigue crack growth tests conform ASTM E647 are performed on AA 5083-H321 for different R-values in air and seawater. Fatigue crack growth starts in the flat tensile mode, but after some crack advance shear lips start on both sides of the fracture surface. The shear lips grow in width on further crack growth until (almost) the total crack surface has become slanted. The start and growing in width of the shear lips is accompanied by a decrease in crack growth rate, resulting in a change to a lower slope of da/dN-∆K on double-logarithmic scale. When the shear lips cover the total fracture surface the original slope of da/dN versus ∆K is almost restored. Crack closure relations for several parts of da/dN versus ∆K have been found. A simple macroscopic geometrical model is made in order to predict fatigue crack growth behaviour in the shear lip crack growing area. The model is also applied to a situation with underloads in AA 2024-T351. Keywords crack closure, shear lips, underloads, crack growth prediction Introduction The fatigue crack growth properties of two aluminum alloys, AA 2024 and AA 5083, are compared. The AlCuMg Aluminium Alloy AA 2024 is widely used in both civil and military aircraft because of its superior damage tolerance and fatigue properties. The AlMgMn Aluminium Alloy AA 5083 is increasingly used in shipbuilding, including high-speed ferries, because of its good mechanical properties and excellent corrosion (fatigue) resistance. Both materials have been fatigued in laboratory air and AA 5083 also in seawater. The difference in fatigue fracture behaviour is described in terms of differences in macroscopic fracture surface and different corrosion-fatigue behaviour. Crack closure was accounted for by varying the stress ratio. Especially transitions in the crack growth rate, that are accompanied by developing shear lips, form a subject of study. The shear lips grow in width as the crack length increases, but there exists no direct correlation between start of shear lips and transition in crack growth rate. It will be shown that suppression of shear lip initiation and growth also shows the same transitions. For both materials the shear lip starts at about the same crack growth rate da/dN, at which value also the slope change in da/dN-∆K occurs, but both events are dependent on the environment. When the shear lips cover the total fracture surface the original slope of da/dN versus ∆K is changed again. Crack closure relations for AA 5083 for some parts of da/dN versus ∆K have been found. A simple macroscopic geometrical model is made in order to

RkJQdWJsaXNoZXIy MjM0NDE=