PSI - Issue 79

Daniela Neves et al. / Procedia Structural Integrity 79 (2026) 266–274

271

Table 4. Manufacturing Parameters (Border)

Manufacturing Parameters Up-skin In-skin Down-skin Core

Number of boarders

0

1

0

1

Laser Power [W]

-

90

-

90

500

-

500

Scan Speed [mm/sec]

-

0.06 -

0.06

Hatch Distance [mm]

-

0.04 -

0.04

Layer Thickness [mm]

-

Energy Density [J/mm 3 ]

-

75

-

75

2.1. Experimental methods Vickers microhardness tests were conducted in accordance with ASTM E384-22 (2022). For these tests, a load of 0.3 kgf was applied for 10 seconds, and the spacing between indentations was 0.5 mm, ensuring a minimum distance of 2.5 times the Vickers diagonal. Minitab software was used to assess the normality of the measurements. Anderson Darling tests were performed to generate probability plots, and results were considered significant when the p-value was at least α = 0.05. Roughness evaluation was carried out using a DIAVITE DH-8 roughness meter. In this study, two main roughness parameters were measured: the arithmetic mean roughness (Ra) and the total roughness height (Rt). To analyse the material microstructure, one specimen from each batch was cut along a plane containing the longitudinal axis and then gradually polished using silicon carbide papers of increasing fineness (180 to 1000 grit). Finally, diamond particles with a diameter of 3 μm were used to achieve a mirror-like surface. The polished surfaces were subsequently etched with a modified Keller' s reagent (4% HNO₃, 23% H₂O₂, 6% HF, 67% H₂O) and observed using an Olympus BX51M optical microscope. The fractured surfaces of the broken specimens were analysed with a scanning electron microscope (SEM), HITACHI S-2400. Finally, to analyse the manufacturing defects, nanotomography analysis was performed. Data acquisition was carried out using Bruker's SKYSCAN 2214 nanotomograph. A FlatPanel camera was used, with a voltage of 150 kV, a current intensity of 61 μA, and a Cu 1mm filter. The image pixel size was 3.5 μm, with an exposure of 2500 ms, 360° rotation with a 0.4° step and a frame averaging of 3 images per step. The images were processed using the NRecon reconstruction, CTvox visualisation, and CTAn analysis software, also supplied by Brucker. 3. Results and Discussion 3.1. Vickers Microhardness Tests Table 5 shows the Vickers microhardness values obtained in this study. Compared with the literature by Lekoadi et al. (2022) and Eshawish et al. (2021), the microhardness value is higher. This hardness value indicates bigger cooling rates, which might represent a less fragile microstructure. In the study by Lekoadi et al. (2022), the average hardness of the as-built samples was 389 ± 10 HV0.3, while Eshawish et al. (2021) reported a slightly higher value of 408 HV0.3 for similar SLM Ti-6Al-4V specimens.

Table 5. Vickers Microhardness results for each batch Batch Vickers Microhardness (HV0.3)