Aging of 3D Printed Polymers under Sterilizing UV-C Radiation

2.1. Sample 3D Printing and UV-C Exposure Protocol

As mentioned in Section 1, standards for UV-C accelerated aging are not well defined. An appendix of the ISO 4892-2 standard mentions the possibility to modify standard testing conditions to allow the usage of a mercury lamp that generates 10 W/m² of 254 nm wavelength radiation. Another testing protocol is found in documentation provided by the Business and Institutional Furniture Manufacturers Association (BIFMA) for healthcare furniture design and specifies a sample should be subjected to a radiative energy of 291 kJ/m² between 12 and 24 h [55]. However, these are not specific to the medical industry. For the purpose of this study, an irradiation power of 10 W/m² and a backpanel temperature of 50 °C were used. The 3D printed samples (Figure 1a) were subjected to UV-C irradiation (Figure 1b) for 24 h and then left to cool down to room temperature for 4 h. The equipment used to perform the accelerated aging process is a BS-02 irradiation chamber made by Opsytec Dr. Grobel (Ettlingen, Germany).

The effect of accelerated aging was assessed in compliance with the ISO 4892–1:2016 standard, which offers guidance on how to analyze the data resulting from artificial accelerated irradiation exposures [56]. Tests have been designed according to ISO 4582:2017 [57] and include tensile strength testing and compressive strength tests.

The materials analyzed in this study are two of the most common materials used with filament-based 3D printing, PLA (polylactic acid) and PETG (polyethylene terephthalate, glycol-modified). The PLA material used was PLA Extrafill filament (opaque, blue), 1.75 mm in diameter, manufactured by Fillamentum (Hulin, Czech Republic). For this material, the manufacturer specifies a glass transition temperature of 55–60 °C and a melting temperature of 140–165 °C. The other 3D printing material used for fabricating test specimens is PrimaSelect transparent PETG filament, 1.75 mm in diameter, manufactured by Prima Printer Nordic AB (Malmö, Sweden). For this material, the manufacturer specifies a glass transition temperature of 86 °C and a melting temperature of 255 °C. The slicer software used to section the 3D models into horizontal layers is Cura 4.5 (Ultimaker, Utrecht, The Netherlands). All specimens were 3D printed on a Creality Ender 3 3D printer (Shenzhen Creality 3D Technology, Shenzen, China). The stock 3D printer came equipped with a 0.4 mm nozzle, a size commonly used by other 3D printer manufacturers as well. For layer settings, a width equal to 100% of the nozzle diameter was used (0.40 mm) while layer height was set at 50% of the nozzle diameter (0.20 mm). In MEX, the number of outside perimeters (contours) influences the mechanical properties of the fabricated part due to the inherent anisotropic properties [58]. This effect is more pronounced in narrow parts, such as the samples fabricated for tensile tests [59]. Two contours were used in order to balance the influence of the outside perimeter with that of the infill. By using 100% infill with a grid pattern at a 45° raster angle, the performance of the tested samples should be closer to that of a typical functional part. Extrusion temperatures were selected based on manufacturer recommendations and were set at 205 °C for all PLA specimens and 235 °C for PETG. A heated bed was used, set at 45 °C for PLA and 65 °C for PETG. Printing process parameters are shown in Table 1.

For each of the test types (tensile strength, compressive strength, tensile creep, compression creep), one group of 5 specimens was subjected to accelerated aging using UV-C radiation using the parameters highlighted previously. The second group of 5 specimens formed the control group. Visual and dimensional inspections were performed before and after the aging treatment.

2.2. Tensile Strength

For tensile strength tests, 10 specimens were 3D printed from each material according to ASTM type I dimensions [60]. It is worth mentioning that work done by Laureto et al. shows a difference between type I and type IV samples when testing MEX/FDM (Material Extrusion/Fused Deposition Modelling) 3D printed parts [61]. The samples where then randomly divided into 2 groups consisting of 5 specimens each.

The 3D printed specimens were tested for tensile strength and stiffness on an Instron 8872 (Norwood, MA, USA) universal testing machine, at a loading speed of 1 mm/min and a preload force of 5 N. Elongation of the sample was measured using an electronic extensometer. The results of tensile strength and stiffness testing (stress-strain charts) are presented in Section 3.2.

2.3. Compressive Strength

For compressive strength tests, 10 cubic specimens of 15 mm × 15 mm × 15 mm in dimensions were 3D printed from each material. The samples where then randomly divided in 2 groups consisting of 5 specimens each.

Compressive strength testing of the cubic samples was done on an Instron 8801 (Norwood, MA, USA) universal testing machine. During testing, the compressive load was applied along the Z-axis of the cubic samples. A preload force of 5 N was used. The results of compressive strength testing are discussed in Section 3.3.

2.4. Creep Characteristics

For tensile creep testing, 10 specimens were 3D printed from each material, in accordance with methods found in standard ASTM D2990-17 [62]. The narrow section of the part is 5.2 mm × 3 mm.

Tensile creep testing was performed by tensile loading 3D printed specimens with a load corresponding to 25% of the measured tensile strength for samples from the control group. Hence, 3D printed samples were placed in a purposefully designed rig with a variable length balance beam providing variable mechanical advantage (Figure 2a). One end of the beam was loaded with a known weight calculated based on tensile test results. Elongation of the specimen was measured after 2 h, 6 h, 12 h, and 24 h, then every 24 h for a total of 5 days.

For compressive creep testing, a more practically oriented method was chosen, as follows. A practical compressive creep test was performed by compression loading 3D printed specimens with a clamping force corresponding to the standard tightening torque for an M6 fastener (5.07 N·m). This was done by tightening a bolt and nut assembly on an 8 mm thick 3D printed plate. The clamping force was transmitted to the 3D printed plate through a metal washer with an outer diameter of 12 mm and an inner diameter of 7.5 mm. Tightening was performed using a wrench attached to a torque sensor, MR55-1000 (Mark-10 Corporation, Copiague, NY, USA). After 7 days, the bolt and nut assembly was undone, and the required untightening torque was measured using the same torque sensor (Figure 2b). Creep testing results are discussed in Section 3.4.

All mechanical tests described in this paper were performed at 24 °C and 50% humidity.

2.5. Scanning Electron Microscopy (SEM)

Scanning electron microscopy analysis (SEM) was performed in order to reveal any potential changes to material internal structure and to outer surfaces of the irradiated samples. For this analysis, samples were selected for investigation from each group and for each material (2 × PLA treated, 1 × PLA control, 2 × PETG treated, 1 × PETG control). The internal morphology of the samples was analyzed using a scanning electron microscope with 1.2 nm resolution (Quanta Inspect F50 from Thermo Fisher Scientific–Eindhoven, The Netherlands). Before imaging, all samples were sputter-coated with a thin layer of gold using a Q150 PlusSeries machine (Quorum Technologies, Lewes, UK). Sample coverage time was 90 s and the resolution of the energy-dispersive X-ray analyzer was set at 133 eV at MnKα. The results of this analysis are discussed in Section 3.5.

3.1. Visual and Dimensional Changes Following UV-C Exposure

Visual and dimensional analyses were performed on the parts following the UV-C aging process. Table 2 presents dimensions of samples manufactured for creep testing, from both the irradiated and control groups, measured with electronic calipers. This ensured that a dimensional check was performed on both thin and long parts (tensile creep specimen, 5.2 mm × 3 mm section) and on bulky parts (compression creep specimen, 25.2 mm × 8 mm section). Measurements show there were no significant changes in part dimensions following the 24-h UV-C exposure cycle.

Figure 3 shows a visual comparison between the appearance of parts from the control group versus those from the UV-C exposed group. There were no visible changes in the appearance of PLA parts post exposure to ultraviolet radiation. The PETG samples darkened and gained a yellow tint. This was accompanied by a loss in reflectivity.

3.2. Tensile Strength and Young’s Modulus

The fracture mechanism for tensile tested specimens was visually analyzed and it was found that the PLA samples ruptured in a zig-zag pattern, along the deposited polymer filaments, for both the irradiated and control groups (Figure 4a). This failure mechanism was expected due to the anisotropic aspect of 3D printed parts where tensile strength along the deposited filaments is higher than that between adjacent filaments [63]. In MEX, a heated extruder brings a filament of material near its melting point and then it extrudes it through a nozzle. Immediately after exiting the nozzle, the material begins cooling through convective interaction with air. Once the extruded material comes into contact with the 3D printed part, it also transfers heat through conduction [64]. Because the convective effect is the dominant thermal transfer effect, the outer surface of the deposited filament cools down faster than the center, creating a temperature gradient [65]. This temperature gradient produces non-uniform properties in the 3D printed part, as the bonds formed between adjacent filaments are weaker than those found in the midsection of filaments. For PETG parts from both groups, fractures were focused along a single plane due to the better interlayer adhesion characteristic of this material (Figure 4b). The fact that both the control group and the UV-C exposed group experienced the same failure modes points to an isotropic degradation of the mechanical properties in the investigated samples.

Figure 5a shows the stress–strain diagrams obtained from the tensile testing of PLA, with the results for PETG being shown in Figure 5b.

Parts made of PLA sustained a 9.1% loss in tensile strength following UV-C exposure compared to those in the control group (26.86 MPa vs. 29.54 MPa, F = 28.53, p = 6.9 × 10⁻⁴). On average, the stiffness of the parts in the irradiated group saw a small decrease of 1.9% (2721 MPa vs. 2775 MPa). After calculating the standard error for results of each group, it was found that the observed decrease in stiffness is not statistically significant (F = 1.98, p = 0.19).

The samples made of PETG that were subjected to accelerated aging treatment experienced a significantly higher tensile strength loss, losing 38.1% of their strength compared to PETG parts in the control group (19.48 MPa vs. 31.30 MPa, F = 522.27, p = 1.4 × 10⁻⁸). The Young’s modulus of the irradiated PETG changed slightly to the downside, lowering by 5.8% (1552 MPa vs. 1648 MPa, F = 15.55, p = 4.3 × 10⁻³). Table 3 shows the average tensile strength and the Young’s modulus of the tested parts with standard error.

3.3. Compressive Strength

During the compressive strength tests, all parts failed plastically without any visible shear penetrating the exterior contours. All the samples undergoing the accelerated aging process performed worse than their counterparts in the control groups. PLA samples lost 13.1% of their compressive strength following UV-C irradiation (67.69 MPa vs. 78.06 MPa, F = 142.73, p = 2.2 × 10⁻⁶) while PETG samples lost 33.9% of their compressive strength compared to the control group (43.57 MPa vs. 65.94 MPa, F = 6.17, p = 3.8 × 10⁻²). A stress–strain diagram for PLA is shown in Figure 6a while the same diagram for PETG is shown in Figure 6b.

Table 4 shows the average compressive strength found during testing compiled with standard errors.

3.4. Creep Characteristics of the Analyzed Materials

Following tensile testing, a 29.54 MPa tensile strength was found for PLA and a 31.3 MPa tensile strength was found for PETG. These values were reference points for selecting the loads applied during creep testing, with the creep stress level being set at 25% of the maximum tensile strength, i.e., 7.5 MPa stress induced in the PLA samples and 8 MPa stress induced in the PETG samples. The equivalent force loads for these levels of stress are shown in Table 5. In order to obtain these levels of stress, a mass of 1250 g was attached at one end of the beam of the testing rig, while the rotating support joint of the beam was positioned to produce a 9.4× force multiplication for PLA samples and a 10× multiplication for PETG samples.

The practical compression creep test was performed according to the conditions described in Section 2. The resulting stress from tightening the screw and nut assembly is proportional to the tightening torque. Considering a screw torque coefficient of 0.3, the resulting stress becomes 40.9 MPa, representing 52% of the compression strength of PLA and 62% of that of PETG.

Part elongation was calculated after measuring the resulting deformation with a micrometer after 2 h, 6 h, 12 h, 24 h, 48 h, 72 h, 96 h, and 120 h. The graphs of elongation vs. time (Figure 7) show that UV-C did not have a significant impact on creep compliance. For both materials, the majority of the elongation happened immediately after stressing the parts with the tensile force. The elongation rate of change follows a relatively similar pattern for PLA parts after irradiation, consistent with the small changes in mechanical properties, especially part stiffness. For PETG parts, the elongation rate of change is slower after irradiation, a change that can be attributed to the deteriorated mechanical properties. Measurements are available in Supplementary Table S1.

Following compression testing, it was found that all samples experienced a loss in clamping force after seven days, corresponding to a lower measurement of the untightening torque. The average torque reduction measured in PLA samples and attributed to UV-C radiation exposure was 3.2% (4.76 Nm vs. 4.92 Nm control group) and the largest reduction found in a single PLA sample was 5.2% (4.764 Nm vs. 5.031 Nm). For PETG, the average reduction was 8.8% (4.10 Nm vs. 4.49 Nm) and the largest reduction found in a single PETG sample was 9.6% (4.047 Nm vs. 4.478 Nm). The total reduction in torque was 6.4% for PLA and 23% for PETG. A chart of the untightening torque measurements is shown in Figure 8.

Average values for the untightening torque determined experimentally, with standard errors are shown in Table 6.

3.5. Scanning Electron Microscopy (SEM)

SEM was used to get a better understanding of the changes occurring in the analyzed parts following the accelerated aging UV-C process. Figure 9a shows a ruptured PLA sample from the control group. As seen previously in the macroscale photographic images (Figure 4a), the sample ruptured along the deposited filaments, forming 45° fracture planes, indicative of 3D-printing specific anisotropy. Filament necking is visible at the fracture interface. In contrast, the first imaged PLA sample from the UV-C exposed group shows less material necking occurring (Figure 9b). As shown in Figure 9c the irradiated PLA sample shows increased roughness on the surfaces of deposited filaments. This change is not seen in the PETG samples and could be attributed to the material opacity and dark pigment.

PETG samples from both groups ruptured along a more uniform transversal plane, indicating a less pronounced anisotropic character, most likely due to increased interlayer and interfilament adhesion. Here, following irradiation, the reduction in filament necking at the fracture interface is more severe than in the case of PLA parts. The extensive necking found in a control group sample (Figure 9d) is replaced with a more defined fracture plane (Figure 9e) and with the appearance of material flaking at the rupture interface (Figure 9f). At microscale, less filament necking at the fracture interface and the development of flaking indicates lower elongation at break when tensile stressed. These changes point to an increase in the brittleness of the PETG samples after irradiation. This increased brittleness is also observed at a macroscale during tensile strength testing, as the tensile strength of PETG samples decreased by 38.1% while average elongation at break dropped from 3.06% to 1.36%. A probable cause for the reduction in mechanical properties of PETG is molecular photodegradation. Under the action of ultraviolet light, polymer materials experience chemical degradation through oxidative reactions [66]. It is suspected that the exposure to UV-C caused the photolysis of PETG, a chemical process whereby the transfer of light energy breaks the chemical bonds. Such changes have been identified by other researchers in PET (polyethylene terephthalate) [67]. All investigated samples show signs of microvoids specific to the fabrication process [68,69].