3.1. Analysis of Powders
DTA-TGA characterized powder samples of all the tested materials to determine their thermal properties. This ensures that the correct parameters for 3D printing were used. The chamber temperature was set using the measured temperatures of the powders, where the chamber temperature was higher in comparison with the melting temperature for the good sintering of powders [
61]. The results of the thermal analysis can be found in
Table 3. The samples were characterized by melting point, maximum temperature, and additive content. The PA_2200 sample does not contain additives and its melting point was determined to be 176.5 °C, which correlates with the literature on polyamide 12 [
62]. The PA_640_GSL sample had a melting point of 184.2 °C and an 18.68 wt.% of the hollow glass and carbon fibers mix. Therefore, the additive content has an effect not only on physical properties but also on mechanical [
63,
64] and technological properties. PA_2210_FR powder contained 8.13 wt.% of halogen-free flame retardant and the melting point was determined at 186.9 °C. PA_3200_GF contained 38.35 wt.% of glass fibers, a melting temperature of 188.5 °C, and the last sample PA_Alumide contained 51.41 wt.% of alumina powder and a 184.7 °C melting temperature.
The results show the effect of additives on the specified melting temperature; after adding additives, it increased by about 8 °C. Furthermore, the samples were also characterized in terms of T
max. Most of the samples containing additives showed higher heat resistance when compared to pure PA_2200. Only the sample PA_2210_FR, which contained flame retardant, showed a lower T
max than the PA_2200. The courses of thermal behavior can be seen in
Figure 3. The content of additives in the polymer matrix has a great influence on T
max (it is a thermal characteristic about the temperature at which the sample loses its maximum weight), and since it is different, it is difficult to draw closer conclusions. A lot depends on the type of aid related to the possibility of applying different maximum amounts. However, from the application’s point of view, it is certainly important to know the processed materials in terms of their thermal properties. Another important characteristic in terms of thermal parameters is determining the so-called sintering window, which is the temperature interval between melting and crystallization onset points. As the DTA-TGA technique is not as sensitive as a differential scanning calorimetry (DSC), in our case, it was not possible to determine this important parameter in terms of LPBF technology. This parameter can be used, for example, to monitor the reuse of materials [
38].
An important process parameter is the size of the particles used. The larger the particles, the higher the likelihood of inhomogeneity in the 3D product. This can increase the potential for crack propagation and lower the mechanical and technological properties. The results of the size distribution analysis can be found in
Table 3. Particle size, shape, and distribution of powders are known to affect the surface roughness and porosity of components printed using LPBF technologies [
65]. In addition, the smaller the particle size, the larger the specific surface area of the particles was observed [
66].
In terms of the average particle size of the powders used, the results are shown in
Table 4, and the shape of powders can be seen in
Figure 4. The average particle diameter size for the tested sample is comparable from this point of view. The results for PA_2200 are correlated with the particle diameter results reported in the study [
67]. In general, the average particle size of polymer-based materials for LPBF processing is around 50–90 μm [
68]. The roughness will not be affected by the size of the particles but by their shape, where there is a prediction of a stronger influence of fiber-shaped particles (glass fiber [
69], carbon fiber [
70]) than those in the form of spherical particles [
71] (aluminum powder, flame retardant additive) on the surface properties of the 3D product.
3.2. Analysis of 3D Samples
The samples of PA_640_GSL and PA_Alumide were observed using a digital optical microscope as they were colored samples, and they had better visibility under the digital optical microscope. The samples of PA_2200, PA_2210_FR, and PA_3200_GF could not be observed due to white staining and poor visibility. The photo of PA_2200 was presented in the paper [
36]. However, the same distribution was assumed for all materials.
Figure 5 shows the distribution of the additives for the three print orientations.
Figure 5A–C show the distribution of additive particles in a sample prepared from PA_Alumide. The distribution appears uniform in all three. However, the A1 sample appears to be the most homogeneous. For the samples prepared from PA_640_GSL, the samples showed a comparably homogeneous distribution (
Figure 5D–F). From these observations, it could be argued that the homogeneous distribution of different types of additives was determined. The observation does not show the effect of the sample orientation during printing on the particle distributions in the printed samples with regard to the type of additive used.
Păcurar et al. [
37] was engaged in the study of PA2200 material processed using LPBF technology, where the influence of the fracture surface related to the production parameters displayed using a scanning electron microscope (SEM) was demonstrated. The influence of the energy density on the porosity of a 3D product made of PA_2200 was also studied when the structure was monitored by SEM. The samples built with medium energy density exhibited lower porosity, while at low and high energies, the porosity was higher [
72].
The samples were also characterized in terms of color measurement according to the CIE
L*
a*
b* system. CIE
L*
a*
b* is currently the most popular way of describing color and is the basis of modern color management systems [
73]. The difference between two colors in a CIE
L*
a*
b* space is the usual Euclidean distance Δ
E between two points in three-dimensional space. It is assumed that the standard observer notices a color difference as follows: 0 < Δ
E < 1—doesn’t notice the difference, 1 < Δ
E < 2—only an experienced observer notices the difference, 2 < Δ
E < 3.5—an inexperienced observer also notices the difference, 3.5 < Δ
E < 5—notices a clear color difference, 5 < Δ
E—the observer has the impression of two different colors. It can be seen from the color measurement results (
Table 5) that there is a slight scatter when comparing samples printed in different directions for the same material. However, for each powder, the variance of the results is within the maximum measured deviation. The results Δ
E* show how much the samples differ when compared to the reference (in our case with PA_2200). The higher the number, the more obvious the change.
ΔE* values indicate that the highest difference compared to the reference is shown by samples with aluminum content, which were values around 10.5. The results of the characteristics for determining the color are important when choosing materials for a specific application with regard to the requirements of, for example, the customer.
From the theoretical assumption, the higher the surface roughness, the easier that cracks can propagate under mechanical stress following a fracture [
74]. Therefore, the surface roughness of the prepared 3D samples was determined. Results for all samples can be seen in
Table 6. According to the theoretical assumption, the highest roughness values were measured in all the G1 samples. Samples printed in the A1 and B1 orientations were comparable in terms of measurements. The results show the same trend with respect to the direction of printing, which corresponds to the results presented in [
75]. Roughness values are comparable for a 45 and 90° measuring angle; however, a 0° measuring angle gave a high surface roughness. Roughness significantly influences the contact angle measurement of a flat surface for various processes [
76].
A disadvantage of using polyamides in 3D printing is that they have a tendency to absorb water [
77]. For this reason, the contact angles of the printed samples were determined, which enables the materials to be classified as hydrophobic or hydrophilic. The results of the contact angle measurement can be found in
Table 5. A material is classified as hydrophilic when the contact angle is smaller than 90°; if it is greater than 90°, it is classified as hydrophobic [
78]. The results show that the print orientation has an effect on the contact angle. In general, G1 samples showed higher hydrophobicity when compared to the A1 and B1 samples. The most hydrophobic sample was the pure PA_2200 sample. Hydrophobicity decreased from print type G1 through B1 to A1 for all materials, except for the PA_640_GSL sample. This may be because the material contained large particles of additive, as can be seen in the microscope image (
Figure 4). A similar trend was also observed in the study by Modi et al. for 3D technology fused filament fabrication (FFF), where samples prepared with print orientation 90° showed the highest values of contact angles [
79].
From the point of view of the applicability of thermoplastics, polyamides are among the best structural polymeric materials which provide excellent mechanical properties. As these are water-absorbing materials, there is a drive to optimize their processing method as much as possible. The results, presented in
Table 5, indicate that with the addition of additives, absorbency increases. The highest absorbency value was found in the PA_2210_FR sample, which contains a flame retardant and the PA_640_GSL sample, which contains a combination of carbon and glass fibers with 1.08 wt.%. Research has shown that carbon fiber has high absorbency [
80]. On the contrary, the sample of PA_3200_GF had a low value of absorbency, which indicates that glass fibers themselves have a low absorbency value. The results were compared with the pure PA 12 sample.
In terms of mechanical testing, the worst results were expected for the G1 print orientation, while the A1 and B1 print orientation results were expected to be comparable. From the results presented in
Table 7, it can be seen that the results were as predicted. The highest maximum load was applied to the PA_2200 sample, and the tensile strength of the samples decreased with the addition of additives. Comparable values for the reference material PA12 (PA_2200) were measured in the tensile test, namely, the determination of Young’s modulus 1270 ± 71 MPa [
13], and tensile strength was about 43 MPa in orientations A1 and B1; results are comparable with a previous study [
81]. On the contrary, the toughness of the four samples containing additives was higher, which can be seen in the higher Young’s modulus values. Since the parameters supplied by the manufacturer for each material were used, the aim was not to monitor the effect of laser power, scanning speed, or scanning pitch. It was observed [
82] that the mentioned parameters affected tensile strength results. It was found for PA_2200 [
82] that the best parameter combination was laser power 40 W, scanning speed 3000 mm·s
−1, scanning pitch 0.4 mm, and tensile strength up to 46.42 MPa, from the perspective of tensile strength, processing cost, and processing cycle.
It is important to determine the ability of each material to produce a sample that conforms to the dimensions of the drawing (
Figure 2). The dimensional analysis results are presented in
Table 8. Positive values correspond to measured dimensions, which are above the drawing dimensions, while negative values indicate the opposite.
When comparing samples according to the print orientation, it was found that the samples printed with the A1 and B1 orientation are relatively comparable to G1. However, with the PA_2200 sample, the difference in thickness and width dimensions for the A1 and B1 print orientation was comparable, but G1 shows a smaller increase in size. With the addition of additives, a more significant effect on the dimensions of the samples was expected. The highest increase in dimensions can be seen in the PA_640_GSL sample, which contained a combination of glass fiber and carbon fibers. There was an increase in thickness of over 10% for the A1 and B1 orientations and an increase in width of over 2% for all three orientations. On the contrary, the dimensions of the PA_Alumide sample were smaller. A similar analysis was performed in [
81] where there were similar trends for the results of pure PA 12 (PA_2200), there was also an increase in thickness, as well as a small change in width.
The weights of the tested samples were measured, as this is an important parameter in terms of practice. The average weights of the samples are presented in
Table 8. As expected, the PA_Alumide sample with the aluminum-based additive has the highest weight, of 12.93 g ± 0.15 g.