Higher Fatigue Threshold Research Articles

The effect of crack closure on the grain size dependence of fatigue crack growth threshold

Recent studies (I-7) clearly indicate a strong dependence of fatigue threshold parameter, A K on grain size in several alloy systems. Attempts to explain these observations on the basis of crat~tortuosity (1,8), fracture surface roughness (5,9) and crack closure (6) appear to present a fairly clear picture of the mechanisms that cause a reduction in crack growth rates at threshold. In general, it has been shown that coarse grained microstructures exhibit higher fatigue threshold in low carbon steels (1,5) aluminium alloys (7) and titanium alloys (6). In spite of these observations, there exists (10-1#) considerable uncertainity about the manner in which the AK~L depends on prior austenitic grain size in quenched and tempered steels. Studies in quenched and tempered steels demonstrating both an increase (3,12,14) as well as a decrease (11,12) in AKth with an increase in prior austenitic grain size can be sought to illustrate this point. Occasionally , the absence of any sensitivity of AKth to the variations in prior austenitJc grain size has also been reported (11,13). While a few investigators (5-7) comfortably rationalised the grain size effects on AK~L on the basis of crack closure by a comparison in terms of the closure-free component of the thresho~Ifc~, AK -f such an approach has yet to be extended to high strength steels, An attempt has been made in t~et ,pthrg sent study to explai. n the effect of pri, or austeniti.c grain size on &Kth on the basis of crack closure measurements in a high strength steel.

High frequency fatigue crack growth and threshold behavior in monocrystalline copper

Purushotaman and Tien(1) found that the crack growth rate in a monocrystalline Ni-base alloy did not exhibit threshold behavior at low stress intensities when crack propagation was performed by 20 kHz ultrasonic stressing. However, thresholds were found for polycrystalline materials (copper, iron, steel) tested in a noncorrosive silicon oil environment (2). Also the influence of corrosive salt water was studied (3). The results of these experiments suggested the use of monocrystals in order to exclude the influence of grain boundaries and to reveal the influence of orientation on crack growth behavior at very low stress intensities•

Creep and fatigue crack propagation in a directionally-solidified carbide eutectic alloy

Creep and fatigue crack growth rates and threshold stress intensity amplitudes have been measured for a directionally solidified carbide-eutectic alloy, C73. Fatigue testing temperatures have been ambient, 750 and 950°C for cracking perpendicular and parallel to the solidification direction. In the former cracking direction comparative propagation rates may be understood in terms of the properties of the matrix, which shows a phase transformation from hexagonal to cubic above ~900°C. A situation where crack growth rates decrease with increasing apparent stress intensity amplitudes (ΔK) has been found to exist for propagation parallel to the solidification direction at low ΔK values and high temperatures only. This phenomenon can be related to the occurrence of crack branching and multiple cracking of the carbide fibers. Considerations of plastic zone sizes and critical defect sizes for crack propagation are consistent with the conditions necessary for such crack deceleration to occur. Transformation of the M7C3 fibers, present in the as-cast condition, to M23C6 at cell boundaries of the solidification structure occurs at a temperature of 950°C. Although M23C6 carbides are easily cracked and therefore probably reduce propagation rates by causing secondary cracking, their presence is known to be detrimental to creep properties. High cycle fatigue threshold stress intensity amplitudes for C73 in either loading direction at room temperature, 750 and 950°C are ~20 pct lower than for the cast nickel-base alloy, EST 738LC,i.e. critical defect sizes are ~10 pct smaller in C73. Despite the known sensitivity of cracking rates and threshold values to factors such as minor fluctuations in loading amplitude it is believed that these differences are significant.