Figure 2

(a) The stress-range versus number of cycles to failure (S-N curve) data of (Zr₅₅Cu₃₀Ni₅ Al₁₀)₉₈Er₂ BMGs. The open circle is the fatigue life of the same material with a fixed L/D ratio (L/D = 1.67) and a frequency of 10 Hz at various stress ranges. The colored symbols are the data from this study. (b) The fatigue life of samples, A, B and C. The 1^st-run (in the as-cast condition) data is shown in red and the 2^nd run data is presented in blue. The number of cycles-to-failure of the 2^nd-run test starts from zero and does not include the cycles made in the 1^st-run test. The fatigue life of the “pre-fatigued” sample (leftover) is equal or longer, compared to the as-cast sample. The results suggest that the fatigue damage in the specimen is mainly localized and cyclic loading has no globalized effect on the sample. When the fractured part is removed, the rest of the material acts just like an as-cast material. (c) The number of cycles to failure as a function of the L/D ratio of the specimen. The results showed no correlation between the L/D ratio and the fatigue life.