PSI - Issue 68

Amy Milne et al. / Procedia Structural Integrity 68 (2025) 666–673 Milne et. al. / Structural Integrity Procedia 00 (2025) 000–000

671

6

Figure 4: Post-test X-built (vertical) crack path at both surfaces of the sample with arrows indicating build and loading directions

4.2. CCG Rate Analysis The PD signal for these tests were found to be atypical of CCG. An example of the PD calibrated to give crack length is shown for example in Figure 5 for a (a) Y and (b) Z orientation sample. As can be seen in Figure 5(a), there are regions of sudden crack growth (around 650–700 hrs and 850–900 hrs). The crack growth rate in the other regions is relatively steady. This may indicate a region of crack jumping where discontinuous cracks suddenly link up.

Figure 5: (a) 2MY- and (b) 1LZ-built (horizontal) crack growth versus time graphs

The CCG rate was correlated with the C* parameter, however in general the trends shown were not sensible, hence not shown here. Due to the extent of discontinuous cracking shown above, there is no single creep crack growing through the sample as CCG theory suggests. In addition, it is recognised that the PD technique may not be accurate in characterising the CCG rate when there is significant discontinuous creep cracking according to Tarnowski (2016). In this instance, the ∗ power-law correlation was therefore deemed inappropriate to characterise the CCG behaviour. Therefore, to examine the influence of sample orientation on CCG, the test duration was plotted against the stress intensity factor, K , value at the initial crack length for all samples, as shown in Figure 6. The Z orientation specimens generally showed much higher CCG resistance than X and Y samples, though there is one anomalous result, indicating the potential variability in LPBF samples. This is expected to be caused by notable lack of fusion porosity around the crack tip. It has been recognised that building multiple parts of different heights on the same build plate can affect the porosity in a given sample according to Williams et al. (2019), therefore the distribution of pores can vary between all samples on the build plate. This effect was clearly illustrated on the single edge notch bend samples manufactured on the same build plate as the compact tension samples, as detailed in Tan et al. (2022).