PSI - Issue 46

David Liović et al. / Procedia Structural Integrity 46 (2023) 42 – 48 D. Liovi ć et al. / Structural Integrity Procedia 00 (2019) 000–000

44

3

The annealing heat treatments were carried out under argon atmosphere at the heating rate of 3.5 °C/min until the temperature of 840 °C is reached. The temperature of 840 °C is maintained for 120 minutes, after which furnace cooling (FC) started down to 150 °C with corresponding Δ t 8/5 = 290 min. The tensile test and cubic specimens still attached to the build plate prior to heat treatment process are shown in the Fig. 1.

Fig. 1. Test specimens manufactured using SLM technology: a) Tensile test specimens; b) Cubic specimens.

2.2. Surface roughness measurements The surface roughness measurements were performed using Jenoptik Hommel Tester T1000 profilometer on two side surfaces and one upper surface of cubic specimens, as well as on one side surface of tensile test specimens. The surface roughness measurements are performed along two perpendicular directions on the side and upper surfaces. Each measurement is repeated five times on cubic specimens and three times on tensile test specimens, in order to make the average surface roughness ( R a ) data more reliable. 2.3. Porosity estimations Porosity measurements were performed on the bottom surfaces of the 9 cubic specimens (10 x 10 x 10 mm 3 ), where the support was previously removed, and the surface was prepared using abrasive papers with different granulations. Surface images were taken using a DINO - Lite Premier digital microscope. All captured images of cubic specimens’ surfaces were cropped from each side for 0.5 mm to avoid the influence of the geometry irregularities on the porosity estimate. 3. Results and discussion 3.1. Side surface - average surface roughness Based on 90 measurements of the average surface roughness values on each of the two adjacent side surfaces of the cubic specimens, and by applying the Kolmogorov-Smirnov normality test, it was found that it is not justified to reject the null hypothesis which states that data comes from normal distribution ( p > 0.15 in both cases, while level of significance is set to α = 0.05). Therefore, it can be concluded that the results of the average surface roughness measurements possess a normal distribution. It is important to note that the Shapiro-Wilik test also leads to the same conclusion. Furthermore, by using two sample t-tests, it was found that it is not justified to reject the null hypothesis which state that the difference in group means is zero ( p = 0.826). Thus, it can be concluded that there is no statistically significant difference in the mean values of the average surface roughness between two adjacent vertical surfaces of the cubic samples. Since cubic specimens were manufactured in the first batch and tensile test specimens in the second batch using the same SLM process parameters, it is possible to investigate the influence of the two manufacturing phases on the average surface roughness of the test specimens. Namely, the powder needs to be sieved and recycled back into the system, which can lead to certain changes in the distribution of powder granulation.