L ewis Milne

Abstract

In recent years, Ultrasonic Fatigue Testing (UFT) has become a common method of carrying out rapid fatigue tests at a frequency of 20kHz, thereby enabling the cost-effective production of fatigue data into the Very High Cycle Regime (>10⁷ cycles). It has been observed, however, that the increased test frequency can have significant influence on the fatigue results, depending on the material being tested. In particular, structural steels exhibit a phenomenon known as the “frequency effect”, in which the increased test frequency leads to an increase in the apparent strength of the material, and a corresponding improved fatigue performance. Until this effect can be quantified, the use of UFT to produce practical design data for structural steels remains limited.

Therefore, this investigation aims to quantify the severity of the frequency effect in UFT of structural steels. The fatigue performance of several structural steel grades will be evaluated at 20Hz and 20kHz and correction factors will be applied to results in order to relate the SN curves at the two frequencies. A cyclic correction factor will be applied by evaluating the cyclic yield behaviour of the materials and a strain rate correction factor will be applied following a range of high strain rate tensile tests.

Finally, a frequency sensitivity parameter will be defined by evaluating the discrepancy between the SN curves at the two frequencies. This parameter will be compared against several key material properties for the tested steels in order to evaluate their influence on the frequency sensitivity.

Session