PSI - Issue 75

Jan Papuga et al. / Procedia Structural Integrity 75 (2025) 289–298 Author name / Structural Integrity Procedia (2025)

292

4

configurations. An additional set of specimens with unnotched dog-bone configuration of minimum diameter of 8 mm, length of the central cylindrical section of 14 mm and fillets to gripping ends with radius of 60 mm was manufactured and tested. The evaluated load modes were replenished compared to [2] and included push-pull, torsion and plane bending this time. Only single load modes were applied; no truly multiaxial load cases were imposed. In all cases presented here, fully reversed load cycles with load ratio R = -1 were induced. All fatigue tests were run at VŠB -TU Ostrava. For testing of the push-pull case, INOVA FU-63-930-V1 pulsator with a load control was used. Torsion and bending load cases were managed by using modified an electromechanical Schenck PWXN machine. A scheme of the test setup to convert the torque driven machine to plane bending cases can be found in [9]. Both machines in all configurations were able to run the tests at a frequency of 25 Hz approximately. The criterion to end the test was set either by reaching the limit of 10 million cycles or forming a crack of approximately 5 mm length. Typically, each S-N curve is formed from at least 8 tested specimens, one from which may reach the run-out condition. The total number of test cases is thus formed by three load modes per seven types of specimens. However, the original hollow unnotched specimens will not be used for RSG and TCD models.

Fig. 2. S-N curves and data for (a) push-pull; (b) torsion and (c) plane bending modes.

The fatigue tests results are depicted in Fig. 2 for all tested configurations. To show the fatigue data and the S-N curve trends, the Kohout- Věchet curve [10] is chosen instead of the common power law. Its formula: = ( ++ ) , (1)