It is known that the increased testing rate and corresponding reduced test duration inherent in Ultrasonic Fatigue (USF) testing can have a significant influence on the fatigue response of the tested material. Particularly, for Body Centered Cubic (BCC) materials such as ferritic steels, the dislocation glide mechanisms are known to be very sensitive to strain rate, leading to the material appearing to be much stronger when tested at ultrasonic frequencies. This results in the apparent fatigue resistance of structural steels being much higher in USF testing than at conventional frequencies. Until a method can be used to estimate and correct this frequency discrepancy, the resultant USF data for structural steels is of limited practical use.

In this experiment, the fatigue response of several structural steel grades: S355JR, Q355B, S275JR, S275J2, and C45 were evaluated at 20kHz frequency using a Shimadzu USF-2000A machine and at conventional 20-50Hz frequencies using an Instron Electropuls E3000 testing machine. Test parameters such as the specimen geometry and test temperature were kept consistent between the two frequencies to limit the difference in test conditions to just the test frequency. Additionally, high strain rate tensile tests were carried out to evaluate how the materials’ strength varies with strain rate.

The frequency sensitivity was evaluated for each of the tested materials based on both the finite life region and the fatigue limit value of the SN curves. Several previously proposed frequency sensitivity models from the literature were evaluated for the tested materials, along with empirical comparisons. The accuracy of these models applied to the tested materials is discussed. Additionally, by comparing the frequency sensitivity results for the tested materials, alongside similar materials in the literature, the influence of the material parameters on the model parameters is investigated.

Finally, the aforementioned models were also adapted and applied as a correction factor to the USF data to attempt to relate the USF response to the conventional frequency fatigue response, with the efficacy of this approach being discussed.