PSI - Issue 33

Aprianur Fajri et al. / Procedia Structural Integrity 33 (2021) 11–18 Fajri et al. / Structural Integrity Procedia 00 (2019) 000–000

17

7

The results of the re-simulation above have been compared with studies conducted by N. Sinan Köksal (2013) which can be seen in Fig. 10a and Table 2. Fig. 10b shows the simulation results with variations of the mean stress theory used. a b

Fig. 10. (a) The results of benchmarking values with previous research; and (b) benchmark results with different mean stress theory approaches.

Table 2. Correlation and error analysis. Parameter value

N. Sinan Köksal (2013)

Current analysis

Error (%)

Min. fatigue life

90,700.00 11,025.00

91,794.00 10,894.00

1.20 1.10 0.20 0.20

Max. damage

Min. safety factor

0.60824 141.72

0.60686 142.04

Max. equivalent alternating stress

7. Conclusions In the present study, a simulation program based on the Finite Element Method to validate the fatigue analysis of a notched cantilever beam is presented. The local stress life approach based on the S-N curve is applied to the analysis using ANSYS Workbench. The mean stress correction theory being compared includes Goodman, Gerber, Soderberg, and ASME Elliptical. Goodman criterion are usually used for brittle materials while Gerber is usually used for ductile materials. The Soderberg criterion are permanently keeping the material below the yield stress. ASME Elliptical takes yield points into calculation but is not strictly conservative. Several parameters have been analyzed and the results show that in the notch area, there is a high-stress concentration which results in the smallest fatigue life and safety factor. Based on the results of the mesh convergence analysis, not only the value of the parameter but the computation time required needs to be considered, especially if this benchmark is applied to more complex geometry. The smaller the mesh size used, the more elements that will perform, which results in an increase in the computation time required. Areas, where stress concentration is suspected need to be refined (improving mesh quality) so that the results are more accurate. To obtain a more complete benchmark evaluation for future applications, more variables need to be included in the simulation. Recommendations for future research, this study should be extended and the results should be compared using other methods and approaches. The material used can be varied and must be validated using direct testing methods.