PSI - Issue 53

S. Senol et al. / Procedia Structural Integrity 53 (2024) 12–28

17

6

Author name / Structural Integrity Procedia 00 (2019) 000–000

2.4. Characterization The surface roughness profiles of all 3PBF specimens were collected using a tactile profilometer (Formtracer Mitutoyo CS3200, USA) equipped with a 60° conical probe and 2 µm radius tip. Five measurements were performed on the top surface of each sample along their curved region with a measuring length of 10 mm, indicated in Fig. 1(b). Surface roughness was then represented by the following parameters; i) Ra , as the arithmetical mean of the filtered data, using a Gaussian regression filter as stated in ISO 16610-31 (2016) and ii) Rv , as the maximum valley depth. The average Ra and Rv values were calculated, using the cut-off length ( λ c) of 2.5 mm as defined in ISO 4288, to compare the different surface conditions. Additionally, the critical stress concentration factors ( k t ) were evaluated as described in (Cutolo et al., 2022) with the general extreme value probability plot for critical notches with a probability of 95%. The internal porosity was visualized in 3D by X-ray microcomputed tomography (µCT) using a TESCAN UniTOM XL at the KU Leuven XCT Core Facility, with 225 kV, 15 W, 165 ms, 2400, as the scan voltage, power, exposure time, and number of projections, respectively, and a 1 mm thick aluminium filter. The resultant voxel size was 5.8 µm. Reconstructions were performed in the TESCAN Acquila reconstruction software and Avizo 2021.2 was used for 3D data analyses. Finally, polished cross-sections (YZ) parallel to the building direction (BD) were examined by optical microscopy (OM) (VHX 6000, Keyence, Belgium). Also, etching was applied (30 seconds in Keller’s reagent, 2.5 ml Nitric acid (HNO3) + 1.5 ml Hydrochloric acid (HCl) + 1 ml Hydrofluoric acid (HF) + 95 ml deionized water¬) to display the details of the sub-surface microstructure. X-ray diffractometry (XRD) (Bruker D8 Advance) was used to compare the Al phase-specific surface residual stress state for different surface conditions. The instrument was equipped with a LYNXEYE XE-T detector (1D mode), a Cu anode emitting K- α 1 radiation (40kV, 40mA, 1.54 Å), directed through a 1 mm diameter collimator. The (422) Al lattice plane with a 2 θ diffraction peak located approximately around 135.5° was used for the measurements. The residual stresses for three different phi ( ϕ ) orientations (0°, 45°, and 90°) were measured (ISO mode) to obtain the full stress tensor: 1) ϕ =0°, σ 11 in longitudinal direction (set along y axis in Fig. 1(b), perpendicular to the pulsed wave laser scan direction)), 2) ϕ =90°, σ 22 in transverse direction (set along x axis in Fig. 1(b), parallel to pulsed wave laser scan direction)), 3) ϕ =45°, intermediate direction. The sin 2 ψ method was applied using eleven psi ( ψ ) angles (- 45° to +45°). The calculated principal stresses ( σ I ) are reported, corresponding to the absolute maximum stress acting on the sample. The data processing was done via LEPTOS 7.9, with the Poisson ratio as 0.352 and the Young’s modulus as 70922 MPa input as material properties; corresponding to -4.970E-6 and 1.907E-5 as the elastic constants S1 and 1/2S2, respectively. The residual stress measurements were performed on the minimum cross sectional area of 13 mm x 6 mm x 5 mm cuboids with a curved top surface (half of the curved region in 3PBF geometry in Fig. 1(a). However, it should be noted that the exact residual stress values might differ for the 3PBF samples as compared to the cuboids, due to the geometry differences, especially in the case of in-process dL-PBF (powder removal and subsequent re-melting) treated condition, R. Therefore, the residual stress measurements were used only for qualitative comparison between the different surface conditions. Nanoindentation experiments were performed using KLA – iNano ® (KLA, Milpitas, USA) with a Berkovich pyramidal tip which was calibrated for area function using a fused silica sample of known modulus (72 GPa) and Poisson’s ratio (0.17) (Oliver & Pharr, 1992). High speed nanoindentation mapping based on enhanced surface detection criteria (Bajpai et al., 2021; Datye et al., 2020; Dong et al., 2022; Lai et al., 2021) was performed on the samples to generate hardness (H) and reduced modulus ( E ru ) (Hay & Corporation, 2018) distribution maps with multiple grid arrays typically 10 x 10 or 10 x 50 with a 10 µm spacing at peak load up to 2mN. The measurements were done at the YZ cross section (Fig. 1(b)) and the indents were made applying a step size of 10 µm and creating a grid of 400 µm (along the BD) x 200 µm (perpendicular to BD). The position of the indentation grid was adapted to start approximately 30 µm away from the top surface to ensure proper indentation, thus is referred as the near-surface hardness. Since the Poisson’s ratio of the sample is unknown due to various additions (alloying elements, ceramic reinforcements) to the aluminium matrix or the possibility of local anisotropy, the reduced modulus ( E ru ) is reported instead of the traditional modulus measurement. The average near-surface H and E ru values are reported for each condition.