PSI - Issue 38

Hugo Roirand et al. / Procedia Structural Integrity 38 (2022) 149–158 Author name / Structural Integrity Procedia 00 (2021) 000 – 000

152

4

Table 3. Constant additive manufacturing parameters V (mm/s) P (W)

t (µm)

h (µm)

D laser (µm)

E geom (J/mm

3 )

E

loc (J/mm

3 )

SLM

700

275 124

50 25

120

70-75

65 90

94 98

ORLAS

1000

55

40

By looking at the table 3, there are no significant differences for the local VED value between both machines; however, the geometric one shows higher VED for Orlas machine. This result is in line with other studies which underline that VED value is not enough to describe a set of laser parameter and predict microstructure or properties (Bertoli et al., 2017; Gunenthiram, 2018).

Table 4. Building strategies

Laser pattern

Rotation between layers

SLM_SR SLM_Rot

Meander Meander

No

Yes (33°) Yes (33°) Yes (45°)

SLM_Chess

Chessboard

Orlas

Meander

2.2. Methods For microstructural characterization (SEM and optical microscopy (OM)), samples have been polished with the following sequence of SiC paper: P300, P600, P1200, P2400; then semi-automatically polished with 3 µm and 1 µm diamond suspension and finally with an OPS solution. To reveal grain boundaries and micro-segregation cells, samples have been etched with aqua regia (HNO 3 + HCl) during 1 min. X-Ray computed tomographic analysis have been realized with an Easytom 130 tomograph on 4 mm diameter cylinders and with a voxel size of 6.7 µm. Only the central volume of 3.5 mm diameter was studied in order to avoid the perturbation of the surface in the defect distribution analysis. A Zeiss Axiovert A1m optical microscope and a Leo 435 VP SEM has been used for the microstructural characterizations and fracture surface analysis. Micro hardness tests were done on polished specimens on a Wilson Hardness Tukon 1202 machine with a Vickers indent. At least 20 measurement points have been realized on each sample. The variation of hardness with load has been investigated between 10 g and 5 kg and no variation has been found. So, 1 kg was chosen for the entire study. An Instron 5900R 100 kN machine was used to perform the tensile tests. They have been carried at a speed of 0.5 % / min until 0.5 % of strain, then 10 % / min until fracture. The displacements have been measured by an AVE2 Axial Instron video laser extensometer. At least three identical specimens have been used for each conditions. Fatigue specimens were machined from as-built cylinders. The geometry is given in Figure 2. After machining, they were heat treated under vacuum at 650°C during 2h in order to relief residual stresses (caused by both building and machining). Stress control fatigue tests has been realized at 25 Hz and R=-1 with an hydraulic fatigue machine (Instron 100 kN) on polished specimens (following the previous sequence, without OPS polishing).

Fig. 2. Fatigue specimen geometry