Interacting Effects of Internal Defects and Corrosion Pits on the Stress Concentration of Hourglass-Shaped Specimens

Abstract

The interacting effects of internal defects and corrosion pits on the stress concentration factors in hourglass-shaped specimens were investigated through a finite element method investigation. Very high cycle fatigue tests were performed to confirm the shape of the internal defects and the depth of the corrosion pit. According to the experimental observations, a series of three-dimensional models containing internal voids and corrosion pits were established. Stress distribution around the internal void and the corrosion pit under uniaxial tensile loading were obtained. Based on the finite element analysis, the stress concentration factors around the defects were found fully dependent on the structure parameters of the internal void and the corrosion pit. Stress concentration factors increased with the depth-to-length ratios Rd for models with a single corrosion pit. For models with a single internal void, the stress concentration factors increased dramatically when the internal void is present near the surface. Parameters of distance-to-radius ratio (D2 − h)/a and depth-to-length ratio Rd play important roles in the stress distribution for a model with both an internal void and a corrosion pit. Severe stress concentration would be introduced when the parameter (D2 − h)/a < 2. For a deep corrosion pit (i.e., Rd ≥ 1), the stress concentration factors were determined by two parts which were dominated by parameters of (D2 − h)/a and Rd, separately. For a shallow corrosion pit (i.e., Rd < 1), the stress concentration factors were dominated by the interaction of (D2 − h)/a and Rd. Two very close corrosion pit and void were most likely to be an initiation position and gave rise to devastating effects on the very high cycle fatigue performance of the specimens. Empirical equations were proposed to evaluate the relationship between the stress concentration factors and the structural parameters of defects. The obtained quantitative information would be useful for understanding of crack initiation and fatigue life prediction of engineering components.