Issue34

C. Ronchei et alii, Frattura ed Integrità Strutturale, 34 (2015) 74-79; DOI: 10.3221/IGF-ESIS.34.07 77 The mean square-root error method [12] is applied to the statistical analysis of the fatigue lifetime results determined by using the modified C-S criterion. In particular, the value of the root mean square logarithmic error is computed as follows:   2 ,exp , 1 log n f f cal i RMS N N E n    (7) where n is the total number of data, ,exp f N is the experimental multiaxial fatigue life, and , f cal N is the theoretical multiaxial fatigue life determined by considering Eqs (1; 3-6) . The mean square error RMS T is given by: 10 RMS E RMS T  . For the different examined materials, Fig. 2 (a), (b) and (c) represents the mean square error obtained for the five expressions of angle  (from 1  to 5  ). 0 0.5 1 1.5 2 2.5 T RMS a) 30CrNiMo8 Steel  1 = 16.49°  2 = 21.71°  3 = 19.12°  4 = 13.91°  5 = 4.30° 1.5 1.75 2 T RMS (b) 6082-T6 Aluminium Alloy  1 = 45.17°  2 = 45.09°  3 = 45.12°  4 = 45.24°  5 = 45.47° 0 0.5 1 1.5 2 2.5 3 T RMS (c) S355J0 Alloy Steel  1 = 37.72°  2 = 40.77°  3 = 39.40°  4 = 35.76°  5 = 28.41° Figure 2 : Mean square error related to: (a) 30CrNiMo8 Steel; (b) 6082 - T6 Aluminum Alloy ; (c) S335J0 Alloy Steel. (a) (b) (c)