Page 154 - Biaxial Multiaxial Fatigue and Fracture

The Influence of Static Mean Stresses Applied Normal to the Maximum Shear Planes in ... 139

2-D Line Profile

Wall Thickness
Fig. 13. A 2-D representation of the fracture surface morphology from specimen A taken with
the CSLM. Sections 1 and 2 represent the fatigue crack growth and fast fracture regions,
respectively.

DISCUSSION

The present paper has shown that the fatigue life of specimens or components subjected to
alternating shear stresses decreases as the magnitude of the applied static mean stress increases
until it reaches the value when the crack faces are fully open. Confocal scanning laser
microscope measurements have shown that, for the hard SAE 1045 steel (BHN 456), the average
asperity height decreases with decreasing levels of tensile mean stress. For the soft steel, the
interference free stress level deduced from fatigue data (Sint), 76 MPa, was found to be very
close to the static mean stress level, 78 MPa, determined using the CSLM technique.
Current multiaxial static mean stress theories discussed in this paper predict a continuous
linear relationship in which the alternating shear stress decreases as the static mean stress
increases. However, the present results show that the decrease ends at the mean stress at which
the crack faces fully separate. The following modifications to Findley's parameter are suggested,
PFindley = rmax + kc, *
for on * = 6, for on < Sint or on * = Sint for on > Sint

For SAE 1045 steel, BHN 456 and 203, a k value of 0.5 reduced the experimental data into a
narrow band for the hard and soft steels. A k value of 0.5 was initially suggested by Findley.
The interference free stress (Slnt) is 500 MPa for the hard material and 76 MPa for the as received
material.
The modified Findley parameter is plotted versus fatigue life for the hard and soft steels in
Figs 14 and 15, respectively. The use of the suggested modification to Findley's parameter
results in the majority of the experimental parameter values falling within the 2x fatigue life
boundaries drawn about the mean curve. The modified Findley parameter was found to be the
best static mean stress parameter investigated during this study for condensing the experimental
data. Imposing a cut off level, above which further increases in tensile mean stress no longer
reduce fatigue life, accounted for the crack face interference free behavior found in this
investigation.