ICF10B

being reduced. The three SVAL-2 tests while having the same cycles ratio (0.1) differ in the frequency with which the largest cycle is applied. This appears to result in larger values of γ. a b Figure 3: Comparison of measured and predicted fatigue crack growth rates under SVAL-1 at R* = 0.9. Figure 4: Comparison of measured and predicted fatigue crack growth rates under SVAL-2 spectra. It is unlikely that these results are a consequence of errors in constant amplitude data. Figure 2 shows excellent agreement of curve fits with the growth rate data points. In addition, the observation of changes in γ with the same cycles fraction of underloads suggests that these effects are a consequence of load interaction between low and high R cycles. Underload effects on fatigue crack growth previously have been studied by a number of workers [5,9,10]. Acceleration effects were reported by Fleck [5], but under very different load conditions than those used in the present work. Dabeyeh & Topper [9], measured decreases in crack opening stresses as a consequence of transitions from compressive underloads to constant amplitude small cycles, up to R values of 0.8.

RkJQdWJsaXNoZXIy MjM0NDE=