PSI - Issue 37

5

R. Fernandes et al. / Procedia Structural Integrity 37 (2022) 462–468 R. Fernandes et al./ Structural Integrity Procedia 00 (2019) 000 – 000

466

400

400

As-built

 (MPa)

 (MPa)

300

300

As-built

Stress relief

200

200

Stress relief

100

100

0

0

-1.25 -0.75 -0.25 0.25 0.75 1.25

-1.25 -0.75 -0.25 0.25 0.75 1.25

 (%)

 (%)

-100

-100

T6 condition

T6 condition

-200

-200

Stress relief As-built T6 condition

stress relief as-built T6

-300

-300

-400

-400

(a)

(b)

Fig. 4. Mid-life hysteresis loops of the AlSi10Mg aluminium alloy manufactured by laser-beam powder bed fusion for the different manufacturing routes: (a)  a = 1.25%; and (b)  a = 0.5%.

strain-hardening for higher strain amplitudes and strain-softening for lower strain amplitudes. Either for higher strain amplitudes (see Fig. 3(a)) or lower strain amplitudes (see Fig. 3(b)), the stress relieved condition led to higher lives. In the other two cases, although the stress level is quite different, the life expectancy was rather similar. Not surprisingly, the as-built condition had the higher strain amplitudes during the tests, which is in line with Figure 2. Furthermore, we can conclude that the cyclic response has small variations at the first few cycles and then a saturated state is reached, irrespective of the strain amplitude or the manufacturing route. In the case of the stress-relieved condition, as already observed in Figure 2, the final stage is characterized by significant changes of stress amplitude until the total failure has been reached. Nevertheless, based on the previous observations and as usual for many metals, the mid-life cycle can be considered as representative of the cyclic stress-strain response. Figure 4 compares the mid-life hysteresis loops at two different strain amplitudes (  a = 1.25% and  a = 0.50%) for the three manufacturing routes. In this kind of representation, the differences between the shapes of the hysteresis loops are evident in both the linear and the plastic responses. Nevertheless, the shapes of the stress-relieved and T6 conditions tend to be similar as the strain amplitude decreases. We can infer that the plastic strain energy density per cycle is higher for the T6 condition than that of the stress-relieved. At higher strain amplitudes, due to differences in the shapes of the hysteresis loops, it is more difficult to establish general conclusions. Regarding the as-built condition, the linear portion of the hysteresis loops shapes is significantly higher than those of the other two cases, irrespective of the strain amplitude. The strain-life relationships for the tested manufacturing routes are presented in Figure 5. The dashed lines represent the fitted functions for the data points collected in the experiments. Overall, these functions agree well, in terms of total strain (  a,t ), plastic strain (  a,p ), or elastic strain (  a,e ) components. For the sake of clarity, the fitted functions of  a,t versus N f and  a,p versus N f for the as-built condition (Fig. 5(a)) are also plotted in Fig. 5(b) and Fig. 5(c). The plastic component versus life relationship, i.e. the well-known Coffin-Manson relationship shows important differences when we compare the three cases. In the two first cases, i.e. the as-built and the stress-relieved conditions, on a log-log scales, we can see bi-linear trends. However, on the contrary, in the range studied for the T6 condition, the data points can be successfully fitted via a linear function.