Influence of process instabilities and elevated temperatures on mechanical properties of PBF-LB-manufactured polyamide 12 Open Access

“Thermoplastic” originates from Greek and is translated as warm/hot and forming, describing the greatest property and simultaneously the biggest dilemma – such polymers become plastic and, thus, processible at higher temperatures, and with elevated temperatures they become pliable (Ehrenstein, 2001; Grellmann, 2011; Erhard, 2008; Menges et al., 2011). Considering the ongoing climatic changes with the increase in average temperature of a few degrees, more and more extreme heat waves can be expected (Perkins-Kirkpatrick and Lewis, 2020). This will also affect the performance of thermoplastic parts in the environment and most likely in a negative way. Aside from the thermoplastics, the other main polymer classes, namely, thermosets and elastomers, cannot be molten anymore, which complicates the recycling drastically (Domininghaus et al., 2012; Ehrenstein and Pongratz, 2007; Ehrenstein, 2001; van Krevelen and Nijenhuis, 2009). Thermoplastics are suitable for mass production, above all polyethylene, polypropylene, polyvinylchloride and polyethylene terephthalate. Another big group in this polymer class are polyamides, which are often applied in a reinforced state for the automotive industry (Ehrenstein, 2001; Grellmann, 2011; Domininghaus et al., 2012; Michler, 2016). Polyamides are not only used for mass-producing injection molding but also in additive manufacturing. For thermoplastic polymers, powder-based selective laser sintering (SLS) and material extrusion-based fused filament fabrication (MEX or FFF) are the most commonly applied techniques. The latter is based on a filament melted in a nozzle, which is then deposited according to a predefined path (Bandyopadhyay and Bose, 2015; Wimpenny et al., 2017). For the powder bed fusion (PBF)-based SLS process, the polymer has to be pulverized with certain requirements such as particle size and shape (Gebhardt and Hötter, 2016; Tofail et al., 2018; Fastermann, 2014; Schmidt et al., 2007; Guo and Moudgil, 2023; Goodridge et al., 2012). In industry, polyamides are the most favored powders for SLS processes besides thermoplastic polyurethanes and polypropylene. With a melting temperature in the order of 180 °C, polyamide 12 (PA12) requires the lowest temperature to be molten among the commonly used polyamides (Fischer et al., 2023).

During a build cycle of an SLS print job, the powder bed is kept at elevated temperatures just below the melting temperature of the used polymer, but above the crystallization temperature. Within this “sintering window,” the energy from a laser provides the remaining energy required to melt the powder in the desired locations. After one layer is sintered, the powder bed with the already sintered powder moves down, and a coater spreads a new layer of powder coming from the feed region over the previously sintered layer. The next step is the melting of the powder particles at the desired spots using the laser to create the parts. These two steps, local melting and reapplying powder, are repeated until the build cycle is finished (Gebhardt, 2003; Breuninger et al., 2013; Garret et al., 2017; Fischer et al., 2023; Gebhardt and Hötter, 2016; Schmidt et al., 2007). This simplified description of the SLS process does not reveal the high number of parameters that must be set correctly to achieve satisfying results. To mention a few, the process depends on the laser power, laser speed, powder bed temperature, layer thickness, hatch distance and delay time. For a successful print, the build cycle has to be continuous and without any interruptions or leakage of oxygen into the build chamber. Because of the elevated powder bed temperature, the build chamber of SLS machines is flooded with nitrogen during the build cycle. If oxygen enters the chamber, it results in thermo-oxidative degradation of the polymer chains (Stiller et al., 2022; Yang et al., 2023; Sanders et al., 2022; Stiller et al., 2025). But, what happens if the build cycle is interrupted? For conventional production methods such as injection molding or compression molding, such parts have to be discarded and the process repeated to produce “good” parts. In SLS printing and PBF in general, this results in the discard of all parts in the part cake and a significant amount of powder during the subsequent so-called powder refreshing, which is ecologically and economically not desirable or sustainable. Typically, 1 kg of PA12-powder suitable for SLS costs €50 to €100, which means the discarding of a full build cycle conducted on an industrial machine (volumetric capacities above 50 L) and the invested energy for the sintering lead to a significant financial loss.

Therefore, this study analyzes the effect of interruptions during the build cycle without temperature drop or oxygen interference and how they affect the mechanical properties of SLS-printed PA12 samples. Furthermore, the effective mechanical behavior of SLS-printed PA12 in dependence on the application temperature is investigated over a broad range of temperatures.

All SLS print jobs for this study were performed on a Farsoon laser sintering machine of the type “FS402P” (FARSOON Europe GmbH, Stuttgart, Germany) modified by LSS (LSS Laser Sinter Service GmbH, Holzwickede, Germany). They were conducted by the company RPD (Rapid Product Development GmbH, Kapfenberg, Austria). The investigated material was “ALM PA 650,” a polyamide 12 (PA12) powder resin from ALM (Advanced Laser Materials, Temple, Texas, USA). As it is standard in the industry, refreshed powder (a powder blend from used and virgin powder) was chosen for this study. Each build cycle, during which a defined set of specimens was produced, had a powder bed temperature between 173°C and 175°C and a layer thickness of 0.1 mm. If not stated otherwise, the applied laser energy input was equal to an Andrew number of 0.017 J/mm² (Schmid, 2018b; Pilipović et al., 2018). The Andrew number represents the introduced energy by the laser divided by the speed of the laser and the layer thickness (Schmid, 2018b). The overview for the parameters is given in Table 1.

In total, three different build cycles were conducted, as shown in Figure 1: a continuous build cycle and two interrupted build cycles to produce different sets of specimens. The positions of continuously sintered tensile specimens in the powder cake, which were used to study the effect of elevated temperatures on SLS-printed parts, are shown in Figure 1(a). Contrary to common SLS build cycles, also discontinuously printed samples were produced. For this, the two corresponding build cycles were intentionally interrupted for 3 min and afterward continued by applying the next layer of powder. This was done to simulate short build cycle interruptions in the production with neither leakage of oxygen, which would lead to thermo-oxidative degradation (Stiller et al., 2022; Yang et al., 2023; Sanders et al., 2022), nor with a decrease in the chamber and powder temperatures. For the tensile and impact specimens in Figure 1(b) and (c), the printing interruption was located at the half height of the specimens. For the flexural specimens in Figure 1(c), a more complex interruption scheme was analyzed, which is described below and illustrated in Figure 2. The influence of the critical interruption layer on the overall mechanical performance was later studied using tensile, flexural and impact tests.

Additionally, for the discontinuously sintered tensile specimens, the laser power was either kept constant [Figure 1(b)] or locally increased by 10%, 15% or 20% to study the effect of higher laser power after the printing interruption [Figure 1(c)]. In the latter case, the resulting maximum Andrew number was 0.02 J/mm². In all printing jobs, tensile specimens according to ISO527 type 1A with a total length of 170 mm and a testing area of 10 × 4 × 80 mm³ (width × height × length) were produced (ISO527, 2012).

To study the combined effects of the printing orientation, printing interruption and direction of loading, impact samples were produced in vertical (V), edgewise (E) and flat (F) orientations. The positions of these impact specimens during the interrupted printing job are shown in Figure 1(b). To produce specimens without printing interruption, the corresponding specimen positions were vertically shifted in the powder cake so that their geometry did not coincide with the interrupted printing layer. It has to be noted that the powder bed alignment of the impact specimens was inconsistent with respect to the moving direction of the powder coater. However, it was indicated by Khudiakova (Khudiakova et al., 2020) that this has a marginal and therefore, negligible impact on unfilled powders. The dimensions of the impact specimens 10 × 4 × 80 mm³ – width × height × length) were according to ISO179 (ISO 179, 2010).

Finally, the flexural specimens from Figure 1(c) were produced to study the influence of the interruption layer position and the effect of voids filled with unmolten powder on the (bending) mechanical properties. Such voids can be emptied during the build cycle to embed an radio frequency identification (RFID) tag or small printed circuit board. These specimens had a global geometry of 10 × 5 × 80 mm³ (width × height × length) and were divided into the following five categories, which are illustrated in Figure 2(a):

1. continuously sintered sample plus engraved notch on one side as illustrated in Figure 2(b);

2. continuously sintered sample with a chamber of unmolten powder plus engraved notch on one side as illustrated in Figure 2(b);

3. sample with printing interruption in the middle of the height;

4. sample with printing interruption in the middle and a chamber of unmolten powder; and

5. sample with printing interruption at the top of the unmolten chamber [interruption layer during flexural test either loaded in tension (“t”) or compression (“c”)].

The flexural specimens were slightly thicker than recommended in ISO178 (ISO178, 2010) (5 mm instead of 4 mm) to properly enclose the chamber of the unmolten powder, which had a size of 4 × 3 × 8 mm³ (width × height × length). This size of the chamber is big enough for a Hitachi UHF tag with a booster antenna. The sample Categories I and II were tested in two ways in the flexural tests, namely, with the smooth surface and with the notched surface in the tensile regime of the bending (facing down during the flexural test). The interruption layer in sample Category V was shifted by slightly adapting the alignment of these samples in the corresponding powder cake [Figure 1(c)]. In the same way, sample Categories I and II were produced without printing interruption by shifting the specimen positions in the powder cake accordingly.

All tensile tests were conducted according to ISO527-2 (ISO527, 2012) on a Zwick Z250 (Zwick Roell, Ulm, Germany) with mechanical clamping to determine Young’s modulus (elastic modulus), tensile strength and strain at break. A macro extensometer determined the strain during testing. The specimens for the study at elevated and reduced temperatures were dried at 80°C for 14 days in a vacuum drying oven and stored afterward in a desiccator. Depending on the chosen temperature, the previously dried samples were stored in the temperature chamber for at least 15 min to ensure homogenously tempered samples. It was expected that the PA12 absorbed only a marginal amount of humidity during this period because it generally absorbs less humidity than polyamides with a shorter CH₂-backbone chain, namely, a maximum of around 1% (Domininghaus et al., 2012). Flexural tests were done on the same Zwick Z250 following ISO178 (ISO178:2010) but with the bending strains exceeding the recommended 5% strain limit. A macro extensometer measured the bending deflection. The impact tests were conducted according to ISO179 (ISO 179:2010) on unnotched samples using a Zwick Roell striking pendulum HIIT25P (Zwick Roell, Ulm, Germany) with an instrumented 2 J pendulum. For all tensile five samples were tested to study the temperature influence as well as the impact of the interrupted layer. Flexural tests were done five times for each sample variation, and the impact tests were conducted ten times to study the influence of interrupted layer position. From those results, the arithmetic mean value and its standard deviation were calculated. After the mechanical testing, the samples were optically investigated on a light microscope of the type “Axioscope 5/7” (Carl Zeiss GmbH, Jena, Germany).

As SLS-printed PA12 behaves mechanically slightly differently than injection-molded or extruded PA12 and because only limited information in the literature (Dooher et al., 2021; Neugebauer et al., 2015) and from the (powder) material suppliers is available on this topic, a tensile test study over a broad temperature range from −20°C up to 140°C with dried tensile specimens was carried out. Within this temperature range examined, the glass transition temperature at around 40°C is passed by, which results in a drastic change in the mechanical behavior.

In Figure 3, the determined Young’s modulus is compared between the different temperatures and printing orientations [refer to Figure 1(a) for further details]. Not surprisingly, Young’s modulus drops with rising temperatures. The significant drop between 23°C to 50°C can be explained by the glass transition temperature (≈ 40°C). The Young’s moduli for temperatures from −20°C up to 23°C are in a similar range, but show slightly different trends for the different orientations. With temperatures rising over 23°C, both, the moduli values and their decreasing trend, are in a comparable range for all orientations (including the vertical).

A similar trend as for Young’s modulus is given for the tensile strength, as visible in Figure 3(a) and (b): an increase in temperature leads to a decreasing tensile strength. In contrast to Young’s modulus, the strength does not show such a significant drop between 23°C and 50°C. Overall, the samples printed in vertical orientation show a lower tensile strength, especially for the lower temperatures. This results from the combined effects of the sintering direction (the print layer interface is loaded perpendicular and thus torn apart by the tensile load) and the brittle behavior at these temperatures (Ehrenstein, 2001; van Krevelen and Nijenhuis, 2009; Hofland et al., 2017; Rodríguez et al., 2023). For the higher temperatures, the already small standard deviation is even reduced. As expected, the strain at break rises significantly with higher testing temperatures, which is visible in the stress–strain curves in Figure 4.

For temperatures up to 23°C, the strain at break does not change notably, whereas starting with 50°C, a remarkable increase is observed. As mentioned before, this is due to the changes in the glass transition temperature, at which the amorphous regions in the polymer become soft and deformable. At the highest temperature (140°C), the edgewise and flat-printed samples revealed such a high elongation that no fracture could be created within the machine limits. For the other conditions without fracture in Figure 4, fracture occurred at strains above the maximum strain scaling limit of the diagrams. With rising testing temperatures, the standard deviations for the strain at break enlarged due to the greater sensitivity to defects in the elongated state.

In Figure 5, the representative side views of the broken specimens clearly show the material softening with rising temperatures. From a smooth brittle fracture between −20°C and 23°C to an increased ductility and, therefore, plastic deformation from 50°C on up to 140°C. Samples printed in vertical orientation show a straight horizontal fracture surface in Figure 5, resulting from the separation between two printing layers due to anisotropy in the sintered part. In general, the bonding between the stacked layers (interlayer bonding) is weaker than the bonding inside these layers (intralayer bonding). For the other two printing orientations (edgewise and flat), the fracture layer is inclined also because of the beforementioned anisotropy (Hofland et al., 2017; Khudiakova et al., 2020; Obst et al., 2018). By increasing the laser power, reducing the layer thickness or changing the scanning pattern (either slower or more often), the anisotropy during the sintering process can be reduced. Further post-annealing unifies the properties over the whole part.

In total, five different sets of tensile specimens were tested to analyze the effect of the printing interruption. One set was continuously sintered [Figure 1(a)], one was sintered without change in the laser energy input after the printing interruption [Figure 1(b)] and three were sintered with an increased laser power of 10%, 15% and 20% [Figure 1(c)], respectively. In Table 2, the results of the corresponding tensile tests are documented. For each condition, the values represent the average and standard deviation of five tensile tests.

Specimens with a higher laser power after the printing interruption show comparable Young’s moduli between 1,720 MPa and 1,750 MPa, which is slightly lower than that of the specimens without a change in laser power (around 1,820 MPa). For the continuously produced sample, the Young’s modulus is also in a comparable range with no significant difference. For the tensile strength as well as the strain at break, higher values were determined for the specimens with increased laser power and energy input (Andrew number). It is expected that by raising the laser energy input, the resulting density of the part becomes higher and therefore enhances the tensile strength (Goodridge et al., 2012). Regarding fracture, all samples broke, as expected, in the vicinity of the layer in which the printing interruption was located. A further increase in laser energy input did neither further improve the tensile strength nor the strain at break. To one part, this is attributed to a change in the specimens’ cross-section area after the printing interruption. This change is shown in Table 2 in the form of the cross-section ratio A_{upper half}/A_{lower half}, which is greater than 1 when the half of the tensile specimen printed after the interruption (upper half) had a larger cross-section area. The change in cross-section area with higher laser energies is caused by more surrounding particles (partially) melted and adhering to the sample. Based on this change in area, a small notch is formed, resulting in a stress concentration. This stress concentration in the vicinity of the printing interruption layer might compensate for potential improvements in the mechanical performance achieved by the increased part density. Additionally, considering the sensitivity of the polymer powder to high (local) energy input, a rise in the laser power can potentially also initiate degradation of the powder, especially because the energy density (Andrew number) was not kept constant for the increased laser powers. Pilipović et al. (Pilipović et al., 2018) showed in a detailed study how the single parameters of the Andrew Number affected the tensile properties of SLS-printed PA12 parts. With increasing energy density, the mechanical properties were enhanced up to a certain maximum value, after which they declined again. Considering the currently studied low-temperature sintering (Schlicht et al., 2024; Yang et al., 2024) or semi-sintering techniques (Gruber et al., 2024; Kobayashi et al., 2024) for polymers, the laser power can be increased based on the overall lower powder bed temperature, which prevents thermal degradation up to a certain energy input.

Unnotched impact tests were conducted to investigate the influence of printing orientation and the respective orientation of the interrupted layer. For this, continuously printed and interrupted impact specimens were printed in vertical, edgewise and flat orientations, as illustrated in Figure 1(b). Each of the specimen sets was tested in flat and edgewise impact orientation. All of these results are summarized in Figure 6(a), with representative pictures after testing in Figure 6(b). Overall, the impact orientation during testing (flat vs edgewise) has only a small effect on the impact strength, especially when considering the standard deviation of the individual bars in Figure 6(a). For the continuously printed specimens, significant differences in the impact strength are only present for the vertically printed samples, which have a lower impact strength. Because of the weaker interlayer bonding (bonding between the printed layers) compared to the intralayer bonding (bonding of the particles within the layers), vertically printed parts in SLS are less resistant against loads in general and impact loads especially.

This trend becomes even more pronounced for the interrupted samples, with the lowest impact strength at around 2.5 kJ/m² for vertically printed samples and the highest values at around 20 kJ/m² for samples printed in the edgewise orientation. Considering the alignment of the interrupted layer with respect to the “loaded” cross-section during testing, this pronounced difference can be explained. For this, the interrupted layers (red line) and the impact planes (green area) are shown in Figure 6(a). For the vertically printed specimens, the interrupted layer completely coincided with the impact plane, while for both, edgewise and flat printed specimens, interrupted layers and impact planes were perpendicular to each other. Nevertheless, there is also a difference in the impact strength of edgewise and flat printed specimens, which is less obvious and only partially significant. Considering this and because it is the only difference between these two printing orientations, it is interpreted that the bigger overlap between the interrupted layer and impact plane is responsible for the decreased impact performance of the flat printed specimens. On the other hand, the smaller overlap between the two improves the impact performance of the edgewise-printed samples. For the edgewise printed sample, also the testing orientation has a significant influence on the impact strength, with smaller values for the flat-tested samples. In this direction, the impact bending deformation causes shear loading of the interruption layer. As illustrated in Figure 6(b), this in some cases even leads to delamination of the impact specimens. A similar loading condition is present for the flat-printed specimens tested edgewise. However, in this case, the effect seems to be smaller and did not significantly influence the corresponding impact strength. Overall, the presented results show that the presence of the interrupted layer strongly weakens the impact performance of SLS-printed PA12 specimens. Although this is not surprising from the general perspective, it was unexpected with regard to the tensile test results, which were not negatively influenced by the presence of the interrupted layer. A printing interruption has a different effect in the quasi-static and impact loading regime, which should be considered for part designs. During the design process and the selection of position in the printing chamber, the optimal orientation regarding later application has to be considered. To counterbalance this weakening effect for especially vertically printed parts, an optimized structure for energy absorption can be considered (Ashok et al., 2022a; Ashok et al., 2022b; Ashok et al., 2023a).

Interestingly, only the presence of an interruption layer is sufficient to decrease the impact behavior significantly even if the layer is not aligned in the loading direction. This strong effect indicates tremendous chemical/morphological differences between the interruption layer and the rest of the SLS-printed PA12 specimen. As the build chamber of the machine was still under a nitrogen atmosphere during the interruption period, excessive thermo-oxidative degradation can be excluded as a reason for the impact performance drop. Although some additional thermal degradation can be expected by the prolonged 3 min exposure to the heaters (Schmid, 2018a; Schmid et al., 2014), it is assumed that the main reason is a polymer physical background. From the polymer physics point of view, high entanglement (and tie-molecule) densities are key parameters for tough and impact-resistant polymeric materials. Consequently, the loss of these entanglements is highly disadvantageous for the mechanical impact behavior. It is expected that for the interruption layer in its surrounding areas, molecular disentanglement in the laser-molten and consolidated regions through reptation movement of the molecules occurred (Gennes, 1971). A similar effect was experimentally demonstrated by Plummer et al. for high-temperature crystallization of polyoxymethylene (Plummer et al., 1995). In their study, they did not use the impact strength but the fracture toughness, a fracture mechanical parameter also used to quantify material toughness. Of course, the laser-molten and consolidated PA12 in the powder bed is kept within its sintering window during the process, also and especially during the interruption period, and thus, no crystallization should be expected to take place. Nevertheless, it cannot be excluded that some molecular pre-alignment occurs already at this elevated temperature, leading to the mentioned disentanglement through reptation of the polymer molecules and subsequently, to the poor impact performance of the interrupted specimens. To gain more detailed knowledge about the stress distribution, a digital image correlation and a finite element analysis can be conducted (Ashok et al., 2023b). The interruption layer can be modeled as a bonding layer with stiffer, but more brittle properties.

In addition to the basic printing interruption study conducted in quasi-static tensile and impact bending regimes, an additional study on printing interruption positioning and impact of unmolten powder chambers was carried out in bending mode. A chamber with unmolten, enclosed powder is undesirable and is created if part holes, material recess areas, etc. are produced without an extraction hole for the remaining powder. As a general statement, the bending tests were stopped at a maximum bending strain of 10%, because beyond this point, the specimens started to slip on the cushion and falsified the measurement results. For all sample categories, the corresponding arithmetic mean values and standard deviations of the bending tests are summarized in Table 3.

Sample Category I was printed without any interruption or chamber but had an engraved notch on the surface, which only slightly affected the stiffness of the specimens (flexural modulus around 1,300 MPa and 1,250 MPa), but more significantly, the strength (63 MPa and 58 MPa) and strain at break (> 10% and 8.1%), when put in the tensile regime of the bending deformation. A similar result was obtained for sample Category II, which had a powder-filled chamber in the middle of the specimens. As expected, the engraved notch on the surface acted as an (additional) defect and, therefore, further reduced the strain at break from 8.9% to 5.3% through a stress concentration on the tensile-loaded surface. Comparing the average values of sample Categories I and II, with flexural moduli of approximately 1,300 MPa and 1,400 MPa, with those of the sample Categories III, IV and V, with respective values of around 1,680 MPa (III), 1,500 MPa (IV), 1,510 MPa (V_t) and 1,600 MPa (V_c), the modulus values rose. Similar was the trend for the tensile strength, while the strain at the break did not change significantly. Hence, as was already observed for the tensile properties, the presence of a printing interruption did not negatively influence the mechanical performance of the quasi-static flexural tests either. The flexural modulus and strength were even improved. We assume this originates from the prolonged exposure of the last sintered layer before the interruption to the heaters. This was also shown by Nussbaum et al. but for shorter exposure times (a few seconds) for each layer in a different sintering approach (Nussbaum et al., 2021). In a continuous printing job, a new layer of powder is applied directly after finishing the previously sintered layer. This new powder comes from the feed region, which is also pre-heated in conventional SLS processes, but to a lower temperature than the powder bed (about 25°C lower). Thus, the new layer slightly cools down the powder bed and absorbs energy from the heaters to heat up. Due to the interruption of 3 min, however, the sintered layer has more time and energy (from the heaters) to consolidate and, therefore, the quasi-static mechanical properties are enhanced. Nevertheless, as observed for the impact test results, the impact behavior simultaneously decreases significantly, even if the interrupted layer was not aligned in the loading direction. Note that no (excessive) degradation and powder bed temperature change should have taken place within this period.

While the bending strain at break was not significantly affected by the printing interruption layer (compare Samples I and III), it was by the presence of the chamber defect (Samples III vs IV). The presence of the chamber with unsintered powder also partly compensated the before-mentioned stiffening and strengthening effect by the interruption layer, however not to the full extent. Furthermore, the stress distribution in a bending specimen results in a neutral axis, with no stress or deformation at the half height for this specimen. Based on this, an eccentrically positioned chamber or interruption layer is stressed more and is thus more critical (Grellmann and Altstädt, 2011; Brewis et al., 1999). Moreover, it has to be kept in mind that the chamber with unmolten powder does not exactly act as a hollow chamber (and thus as a void) because it is filled with powder. Upon deformation, the powder inside the chamber can also carry the load, for example when loaded in compression. Nevertheless, the presence of such unmolten powder chambers will still increase the normal and shear stresses inside the specimens (and parts) due to the smaller sintered cross-section areas.

In this work, the mechanical performance of selective laser-sintered PA12 specimens with respect to low and elevated temperatures as well as process instabilities was studied. On the one hand, the influence of a broad temperature range (−20°C up to 140°C) was investigated in the tensile regime. On the other hand, the consequences of printing interruptions and further defects in the mechanical properties were examined in the tensile, bending and impact bending regimes. In the temperature study, only small changes were observed below the glass transition temperature. However, above this temperature, the mechanical behavior changed significantly and was accompanied by an obvious increase in ductility. Consequently, especially for elevated temperatures, it is important to know the application temperatures of selective laser-sintered PA12 parts well enough already in the design stage. In this regime, small temperature differences have a comparatively big effect on the final mechanical performance of the part. Fiber-reinforced powders can provide a higher thermal stability but simultaneously result in a higher anisotropic mechanical behavior due to the fiber orientation. Alternatively, the material has to be changed to a high-performance polymer such as polyether ketones, which are at the same time significantly more expensive. If the application temperature exceeds a critical range, it has to be considered if SLS-based parts are still feasible.

The printing interruption study showed a significant decrease in impact strength with a strong dependence on the load direction of the printing interruption layer and a stiffening effect in quasi-static tensile and bending tests. By raising the laser energy input after the printing interruption, a slight increase in tensile strength could be achieved, while an increase in cross-section leads to a change in dimensional accuracy and a notching effect. With the results from this study on interrupted prints, it can be concluded that a short printing interruption of several minutes without temperature drop or oxygen leakage in the machine build chamber does not negatively affect the printed parts when used for applications with moderate loading speeds. Consequently, in the production of such parts, they do not necessarily have to be discarded, but a potential slight change of the mechanical properties has to be considered. This helps to prevent waste and, concerning the high SLS powder prices (€50/kg up to €100/kg), to save costs.

Chambers with unmelted polymer powder are usually not present in components, but they could work as a predetermined breaking point, which indicates overloads. By the fully enclosed chambers, this safety feature is not visible and for SLS prints easy to include. Such defects must be considered properly in the design stage to ensure the reliability of the component in later application.

Thinking about future generations of SLS printers, an implementation of a robotic arm inside the machine to position parts (e.g. RFID tags or small printed circuit boards) requires an interruption. Without a drop of temperature or interference of oxygen, the part quality and property do not suffer drastically. However, for high loading speeds in the application, such printing interruptions are deadly for the printed parts, and they then cannot be used for their intended purpose. In the future, further investigation of the interruption effects on the powder itself is aimed, especially with a focus on the print quality of the next build job.

The research work of this paper was performed at the Polymer Competence Center Leoben GmbH (PCCL, Austria) within the framework of the funding program “Production of the Future” of the Federal Ministry for Climate Action, Environment, Energy, Mobility, Innovation and Technology and the Federal Ministry for Digital and Economic Affairs in the projects “fun-3Dmanu,” “sustainableSLS” and “RFIDinSLS” with contributions by Joanneum Research Forschungsgesellschaft mbH, Materials Science and Testing of Polymers/Montanuniversitaet Leoben and RPD Rapid Product Development GmbH. The PCCL is funded by the Austrian Government and the State Governments of Styria, Lower Austria and Upper Austria.