Fatigue Crack Behaviors under Asynchronous Biaxial Loading
Duan Min'ge,Li Yazhi,Liu Huan,Shu Yixiu
DOI: https://doi.org/10.1016/j.ijfatigue.2019.05.006
IF: 5.489
2019-01-01
International Journal of Fatigue
Abstract:Crack behavior of aluminum alloy 2A12-T4 is investigated experimentally and numerically using cruciform specimens subjected to in-plane biaxial loadings applied in horizontal (x axis) and vertical (y axis) directions. Different cruciform configuration plans, with variant sizes of slots in loading arms and different transitional fillet radius between adjacent arms, are analyzed by means of FEM to confirm a final specimen geometry with relative uniform stress distribution in its central area. The tests for fatigue crack growth behavior under cyclic biaxial sinusoidal loadings with phase differences of 0 degrees or 60 degrees between two orthogonal fatigue loads are conducted by using a biaxial fatigue test system. The biaxial load ratio, P-max(x)/P-max(y), varies from -0.5 to 1.0. Initial central cracks are inclining in angles of 0 degrees, 45 degrees and 60 degrees to x-axis. Crack growth path and lives are obtained. It is observed that the crack growth behavior under biaxial loading differs from that with only uniaxial loading. A widely accepted fatigue crack growth model, which combines Paris equation, maximum tangential stress (MTS) criterion and an equivalent stress intensity factor (SIF) proposed by Tanaka, is used for mixed mode-I/II crack subjected biaxial loading. The combined model is evaluated by the SIFs analysis based on analytical and numerical solution. The mixed mode SIFs are calculated based on finite element analysis with mesh updating. The simulated crack growth path and growth lives correlate well with the experiments in most cases. Both the experiments and the simulations demonstrate that the tensile load parallel to crack path behaves as the retarder to the crack growth, and the compressive load parallel to crack path behaves as the promoter to the crack growth. It is found that equivalent SIF used in the model overestimates the effect of mode II SIF on the fatigue crack growth rate by Paris equation. It is questionable to use the combined model for the cases with asynchronous biaxial loadings or those leading to crack surface contact.