PSI - Issue 17

C.A.R.P. Baptista et al. / Procedia Structural Integrity 17 (2019) 324–330 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

328

5

The surface residual stresses values (mean and standard deviation) can be seen in Table 2. Even though the most prominent feature of these results is their high variability, it does seem likely that a slight increase in compressive stresses occurred in the laser processed specimens. In any way, it was shown by Dorman et al. (2012) that residual stresses due to laser shock peening can be negligible at the surface and increase with depth. This occurs due to the relaxation of stresses at free surfaces.

Table 1. Roughness and surface hardness results. Material Condition Ra (  m)

HR15T

1.38  0.33 2.78  0.44 2.94  0.37

62.2  1.5 68.4  2.6 67.2  2.2

AR

LP50 LP75

Table 2. Residual stress results. Material Condition

 x (MPa) -16  12 -24  4 -20  15

 y (MPa) -28  16 -40  10 -27  17

AR

LP50 LP75

During the FCG tests, dozens of load vs. COD datasets were collected, each one comprising various loading cycles. Thus, many closure load values P cl were determined by the automated method using these data. The algorithm converges to an optimal value of the closure load corresponding to a minimal error. However, when the crack closure is small or negligible, imperfections in the raw data can lead to a wrong value of P cl . Therefore, in order to obtain a more realistic value of the closure load for each test, the median instead of the mean of the calculated values was considered. The median is known to be a much more robust and sensible measure of the central tendency, being not affected by very large numbers. Having stated that, the results of crack closure calculations for R = 0.2 tests are shown in Table 3 in terms of the median values obtained for each of the tested specimens. Bearing in mind that the reference (minimum load in one cycle) is P min = 370 N, the obtained values of P cl are in fact small. Nevertheless, it can be stated that both laser processing conditions promoted some crack closure in the tests with R = 0.2. On the other hand, the calculations for the R = 0.5 tests (in which the minimum load in one cycle was 1,000 N) resulted in negligible closure (P cl / P min < 1.05) for all of the tested material conditions.

Table 3. Median of crack closure calculations for specimens tested at R = 0.2. Specimen designation P cl (N) P cl / P min AR_1 384 1.04 LP50_1 403 1.09 LP75_1 453 1.22 AR_3 374 1.01 LP50_3 440 1.19

The FCG test results are summarized in Tables 4 and 5 in terms of the number of cycles spent in crack propagation between specific crack sizes, i.e., from 18 to 26 mm and from 26 to 34 mm. The crack length 18 mm corresponds to the first laser track (3 mm ahead of notch tip) and the 8 mm crack growth from this point spans roughly the first half of the laser processed region. The second stretch of crack growth (26 to 34 mm) ends 1 mm ahead of the last laser track. The FCG datasets for specimens AR_1 and LP50_1 are plotted in Figure 4, which shows a slower crack growth in the laser treated specimen starting at approximately the crack length of 23 mm. The