PSI - Issue 79

Oleg Plekhov et al. / Procedia Structural Integrity 79 (2026) 168–175

174

However, processing does not always lead to an improvement in fatigue life. If the balancing tension residual stresses generate the crack initiation region, the fatigue life may decrease dramatically. One of the good example of this is fatigue testing in gigacycle fatigue regime. The experiments are carried out using an ultrasonic test machine, Shimadzu USF-2000, which is able to operate in a loading frequency range of 19.5 – 20.5 kHz. In order to meet the resonant condition and ensure maximum stress value in the centre, the length of the specimen (Fig. 1b) corresponds to half wavelength of the oscillation propagating in the test machine horn. In experiments a fully reversal axial cyclic loading (R = -1) is realized.

Fig. 8. S-N curves for treated and untreated specimens.

The experimental points obtained in fatigue tests with reference and treated specimens are presented in figure 8. It is clearly seen that the endurance limit of the referent material at 10¹⁰ cycles is approximately 405 MPa. Concurrently, the number of cycles to failure of the specimens after LSP exhibits a critical reduction, measuring approximately three orders of magnitude. 5. Conclusions The developed experimental setup enables real-time measurement of pressure pulse profiles acting on the material and subsequent reconstruction of through-thickness CRS distributions. Experimental results demonstrate that properly optimized LSP parameters can increase the fatigue life of notched specimens by several times. Conversely, improper CRS may lead to catastrophic reductions in fatigue. As a case study, we present gigacycle fatigue test results, where significant tensile residual stresses generated in the specimen volume decreased the fatigue limit by three orders of magnitude. It was experimentally shown that the real duration of the stress pulse is significantly longer than the duration of the laser pulse. With a laser pulse duration of 10 ns, the duration of stress wavefront at a depth of 0.8 mm was 20 ns. Due to the plasma extension process limited by the resistance of the water layer, the stress equal to half the height of the pulse remained for more than 50 ns. This result should be taken into account when assessing the depth of the compressive residual stress layer. Compressive residual stresses can significantly increase the fatigue life of parts with stress concentrations. However, this LSP cannot completely replace traditional surface treatment technologies for metals. This technology can be effective when it is necessary to stop or prevent the formation of a fatigue crack in a specific location of a structure (for example, in a stress concentration). However, significant tensile stresses may occur in another location of the structure. As an illustration of an unsuccessful example of the LSP application, the results of fatigue experiments in the regime of gigacycle fatigue are presented. The created compressive stresses in the surface layer led to the emergence of significant balancing tensile