Effect of predamage from low cycle fatigue on high cycle fatigue strength of Ti-6Al-4V

Abstract

Effects of prior low cycle fatigue (LCF) cycling on the subsequent high cycle fatigue (HCF) limit stress corresponding to a life of 10⁷ cycles are investigated for Ti-6Al-4V at room temperature. Tests are conducted at 420 Hz on an electrodynamic shaker-based system at several different LCF maximum loads and under subsequent HCF at R=0.1, 0.5 and 0.8 using a step loading procedure. Under these load combinations, which include the possibility of overload or underload effects if cracks form, there is no statistically significant effect of the prior LCF on the subsequent HCF limit stress. While LCF loading at a high stress of 900 MPa is seen to result in strain ratcheting, no distinct features on the fracture surface and different mechanisms of crack propagation from those obtained at lower maximum loads were observed. LCF loading up to 50% of expected life did not produce any indications of crack formation from either the stress limit data or the fracture surfaces.

Introduction

Up until the 1980s, the demand for improved performance in the US Air Force gas turbine engines resulted in considerable attention and research activities in low cycle fatigue (LCF). This resulted in improved design and maintenance practices and has resulted in engines that are now more resistant to low cycle fatigue (LCF). In fact, the introduction of damage tolerance requirements through the ENSIP specification [1] has almost completely eliminated failures due to LCF. The decrease in failures related to LCF has resulted in increased attention to failures related to high cycle fatigue (HCF), i.e. failure under cyclic loading at 10⁷ or more cycles, as they have become a larger and larger percentage of the total number of engine failures. Currently, HCF is one of the major causes of turbine engine failures experienced by many military and civilian aircraft [2]. Engine components, such as airfoil blades and vanes, are subjected to HCF loading conditions involving high frequency, vibrational type fatigue loads, often superimposed on a high mean stress. The ‘damage tolerant’ approach, where the remaining life is predicted from the crack propagation rate of an initially inspectable flaw size to a critical size, has worked well for LCF. Its direct extension to HCF is a challenging task, since a large fraction of life during HCF is spent in crack nucleation and growth to a detectable size, while only a very small fraction of life is spent in crack propagation from an inspectable flaw size to a critical size. Hence, the concept of a fatigue limit or fatigue threshold is an attractive alternative for HCF design. However, the actual loading in many practical applications often involves both LCF and HCF loading conditions during typical operation of turbine engines in military and civilian aircraft. Further, loading history can involve many complexities, such as multiple overloads and underloads, different frequencies, variations in load ratios, and more. Present research activities are focused in this direction and, as a first step, the effects of predamage from LCF on the HCF strength/life are being studied. These studies are referred to as LCF/HCF interactions and seek to determine the detrimental effects of LCF loading on the subsequent HCF limit stress or threshold stress intensity.

There have been limited studies of LCF/HCF interactions on total life in the HCF regime. Nicholas and Maxwell [3] investigated the effect of prior LCF at stress ratio, R=−1 on the subsequent HCF fatigue limit at R=0.5 in Ti-6Al-4V plate using a step loading technique. The LCF life was approximately 10⁵ cycles for a maximum stress of 600 MPa at R=−1. They found that there does not appear to be any significant degradation of HCF fatigue strength due to prior LCF cycling, even up to 75% of life. Further investigation also showed no indication that the step loading produced coaxing [4] or any history effect on the subsequent high cycle fatigue limit stress. Similar results were obtained by Lanning et al. [5] on smooth specimens of both bar and plate product forms of Ti-6Al-4V, as well as on notched specimens of the plate material. In the case of notched specimens, consideration had to be given to the inelastic deformation in the form of plasticity during the LCF portion of the testing and strain ratcheting during HCF, the latter which has also been observed in smooth bars under HCF at stress ratios of R=0.8 or higher [6]. In the study by Lanning et al., several specimens showed indications of cracks formed during LCF testing from heat tinting prior to HCF testing. Cracks with depths ranging from 25 to 35 μm had no apparent effect on the subsequent HCF fatigue limit stress and were either below a threshold ΔK or arrested due to the notch stress field. Moshier et al. [7], in studying influence of LCF cracks on HCF strength in notched specimens, showed that such small cracks, formed during LCF, have little or no effect of the fatigue limit stress when they fall in the small crack regime of a Kitagawa diagram. Further, overload and underload effects may influence the fatigue limit stress if cracks are formed during LCF testing. Thus, the importance of stress ratio, R, has to be considered in studying LCF/HCF interactions.

In the current study, LCF at various stress ratios is used to introduce the predamage, if any, which is subsequently characterized at a similar range of stress ratios to ascertain the fatigue limit stress corresponding to a life of 10⁷ cycles. The range of values for stress ratio is from R=0.1 to R=0.8 which covers cases where the LCF load is both below and above the subsequent HCF load. Further, the deformation behavior of the material is elastic for low stress ratios, but involves stress ratcheting at the highest value of R=0.8. The behavior is characterized in titanium alloy, Ti-6Al-4V, identical to that used in prior investigations [3], [4], [5], [6], [7], [8] and is a representative material of fan airfoils in many military engines.