PSI - Issue 38

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

151

3

Fig. 1. SEM image of Höganas powder

Table 2. Physical properties of 316L SS powder

Apparent density (g/cm3)

Flowability HALL (s)

D 10 (µm)

D 50 (µm)

D 90 (µm)

Measured

Höganas

Measured

Höganas

19.3

29.2

41.1

4.5

4.15

17.5

18.0

Cylinders of 96 mm height and a diameter of 13 mm have been built in a SLM125HL machine located at the Institute of Mines Telecom in Albi (France). Three plateau of 45 cylinders were printed with a pre-heating of the plateau at 100°C. The same laser parameters have been used for these three fabrications and are detailed in table 3. Only the scan strategies are different between the fabrications (table 4). With the chessboard strategy, the chessboard cases are filled with 20 scan vectors of 2.4 mm long, the recovery between islands is 100 µm and the total pattern is filled from an in-out spiral. In order to have a comparison point from another machine, two other plateau were manufactured on an Orlas Coherent Creator. There are many differences between these machines (chamber size, laser power, laser diameter, recoating method, etc…) but the most important are the following ones: - First, the plateau of Orlas machine are circular with a diameter of 110 mm whereas plateau of SLM125 are square of 125*125 mm². In consequence, only 9 cylinders have been built for each plateau in the Orlas machine. The number of cylinders built by plateau could influence the material by modifying the thermal history. - The argon flux is coming from the right of the plateau onto the left for SLM machine although there is a vertical argon flux (from bottom to top) for Orlas machine. It is known that the argon flux direction could modify the final microstructure of a material by orienting the metal vapor plume under the laser and so modifying the laser/powder interaction (Andreau, 2019). Two formula of volumetric energy density (VED) have been used in order to compare this study to others in literature. The first one (equation 1) is called “geometric” VED as it take s into account the power (P) provided by the laser which is divided by a geometric volume obtained from layer thickness (t), hatch distance (h) and laser velocity (V). The second one (equation 2) is more local as it uses laser diameter (D) in replacement of layer thickness and hatch distance. Both of them are frequently used in literature in order to use a unique scalar to correlate parameters and measured/observed properties (DebRoy et al., 2018; Gunenthiram, 2018). However, they have to been used very carefully because they don’t capture me lt pool physics and so they poorly predict track morphology or solidification conditions (Bertoli et al., 2017). = ℎ (1) = 4 2 (2)