Prediction method for local failure probability based on weakest-link theory

The work aims to calculate the fatigue limit and the local failure probability of surface-reinforced steel. A smooth specimen and two notched specimens were designed for carburized CrMn steel, carburized CrNiW steel and nitrided CrNiW steel, and very high cycle fatigue tests were carried out for the three steels. The fracture surfaces were observed using scanning electron microscopy to analyze the S-N curves of the three steels and determine their failure modes. At high stresses, surface failures were mainly induced by surface machining defects and surface inclu-sions, while at low stresses, internal failures were mainly induced by internal inclusions, inhomogeneous matrix area and fine granular area. The micro-hardness of three steels from the surface to interior was measured by nano-indentation. The depth of the carburized layer for carburized CrMn steels and carburized CrNiW steels is 500 μm and 900 μm respectively, and the depth of the nitrided layer for nitrided CrNiW steels is 280 μm. A prediction model for the local failure probability of surface-reinforced steel was constructed based on the weakest-link theory, the local equivalent force was calculated by using the method of establishing a finite element model, the parameter calculation method of this prediction model was optimized, and the fatigue limit and local failure probability of three steels were calculated. The results show that, the calculated fatigue limit by the model ties well with the test results, the maximum error between the two is 5. 37%. The calculated local surface failure probability by the model is closer to the experimental results, and the deviations between calculated local interior failure probability and experimental results were small.