ICF10B

figure 3b, the ratio of n/n' has changed to half that used for 3a, and the growth rate is reduced by about a factor of two. However the growth rate overall is accelerated relative to the linear summation by 1.5-2.2 for all sequences at R*=0.9 ( Table 2). Figure 2: Fatigue crack growth rates measured experimentally (points), and curve fitted (lines) TABLE 2 ACCELERATION FACTOR γ OBSERVED IN VARIOUS TESTS Test condition n = 20 n' = 1 n = 1000 n' = 50 n = 2000 n' = 50 k = 100 k'' = 9 k = 1000 k'' = 99 k = 2000 k'' = 199 Acceleration factor γ=∆at/∆af 1.66 2.19 1.52 2.55 2.67 6.32 Crack growth rates recorded for SVAL-2 tests are presented in figure 4. Data for linear summation calculations of growth rate are also shown. The data for the test with k = 104 are not shown as the growth rate increment was too small to be accurately measured. The value of γ = 6.3 recorded with k =2,000 is also subject to considerable error, as the recorded growth rate was only 10-10 m/cycle. The γ value calculated for each test is shown in table 2. For SVAL-2, there were fewer large cycles than in SVAL-1. Although the fraction of underloads was greater, it included the small R=0.7 cycles. γ values were 2.5-6.5 for SVAL-2, bigger than γ values of 1.5-2.2 calculated for SVAL-1. Figure 5a shows the γ factor for all tests plotted against the total growth increment obtained for each sequence. There is no clear relationship between γ and the total increment of growth obtained in 106 cycles. Figure 5b shows γ plotted against underload cycles expressed as a fraction of total cycles in the sequence. It shows a trend of increasing γ from 1.6 at a cycles fraction of 0.025, for SVAL-1 to between 2.6-6 for SVAL-2 at a cycles fraction of 0.1. DISCUSSION Under SVAL-1, only sequences with small cycles having R*=0.9 produced γ values greater than unity. No acceleration was detected for SVAL-1 with small cycles of 0.7, despite the fact that the ∆K of the large underloads (6.3 MPa m1/2 ), produced growth increments which were accurately measurable. It is concluded that acceleration effects under SVAL-1 is confined to small cycle R ratios above R* = 0.7. Testing at R* = 0.9 produced much larger growth increments, and γ increased from 1.6 to 2.2 as the fraction of under loads increased. Testing under SVAL-2 produced larger values of γ; despite the number of large underload cycles

RkJQdWJsaXNoZXIy MjM0NDE=