Additive Manufacturing for Automotive Applications: Mechanical and Weathering Durability of Vat Photopolymerization Materials

Abstract

In this study, the effect of the post-curing process as well as the long-term weathering behavior were studied for three different three-dimensional printable, pigmented (black), nonstabilized, ultraviolet (UV) cure resin formulations (A and B being thiol–ene chemistry, and C being acrylate chemistry). To study the effect of the post-cure process, the printed parts were post-cured using one of five different processes: no post-cure, UV-only, heat-only, UV+heat, and electron beam (EB) post-curing. Bulk tensile properties and nanohardness were measured for each of the systems and post-cure conditions. For weathering studies, the parts were post-cured using the recommended UV-only process and exposed using the ASTM D7869 exterior weathering protocol. The results show that the post-cure process had a significant effect on the final mechanical properties of the resins and was dependent on the underlying resin chemistry. Thermal post-curing was not as effective as UV curing for Resin C compared with the two other resins, which could both undergo thermal polymerization. In addition, Resin B showed the smallest change in mechanical properties before and after post-curing, regardless of the type of post-curing process. EB post-curing, even at very low dosages, that is, from 0.05 to 1 Mrad, resulted in considerable post-cure cross-linking to the point of embrittlement and a significant drop in percent elongation at break for dosages above 0.5 Mrad. Although Resins A and C outperformed Resin B in photooxidation performance, all three resins demonstrated that promising results considering no hindered amine light stabilizers were used in the formulations.

Introduction

Broadly defined, three-dimensional (3D) printing is an additive manufacturing (AM) process in which layers of material are successively deposited on top of each other to form a 3D object. Thanks to recent improvements, including enhanced speed and performance of the materials and machinery, AM is receiving increased attention in various applications including the automotive industry.^1–4 Technological advancements in AM have enabled the fabrication of 3D objects from a wide range of raw materials, including metals,⁵ ceramics,⁶ fibers,⁷ as well as polymers,⁸ which are the focus of this study.

Photopolymerization is a polymerization mechanism used in many engineering applications, such as coatings^9–13 and dental restoration,¹⁴ due to advantages such as rapid curing, low to zero volatile organic compound-containing formulations, and low capital investment.¹⁵ Owing to chemistry-related innovations, photopolymerization-based 3D printing techniques, for example, vat photopolymerization (VPP), have captured great interest over the past decades. In addition to these advantages, photopolymerization offers other attractive benefits in 3D printing. Versatility and high spatial/temporal control over photopolymerization reactions can significantly enhance the printing resolution and speed. Moreover, higher interactions between the layers lead to increased cross-linking, and therefore, improved mechanical properties in VPP-printed parts.¹⁶ However, despite these advantages, the nonuniformity of mechanical properties across the layers and their interfaces is still a significant challenge in VPP-based 3D printing.^17–19

Materials used in VPP-based 3D printing are photosensitive liquid oligomers and monomers that can rapidly convert to a solid polymeric network upon exposure to radiation. In this mechanism, the polymeric network is formed by the successive addition of free-radical building blocks. During this process, the material properties of the cured material change dramatically. Examples of photosensitive chemistries include acrylates and thiol–enes. The latter proceeds via a step-growth polymerization and is reported to offer no inhibition to oxygen, delayed gel times, more homogenous networks, and lower shrinkage stress compared with the conventional acrylate systems that polymerize through a chain-growth mechanism.²⁰ These unique features make thiol–enes an ideal chemistry for VPP-based 3D printing methods, such as digital light processing (DLP).^21,22

To ensure an acceptable print resolution in VPP-based 3D printing, ultraviolet (UV) absorbers such as pigments or dyes are usually used in the formulations to limit the penetration depth of the UV light into the resin solution. As a result, complete conversion of the carbon double bonds usually does not occur during the printing process. This postprint conversion state is referred to as the “green state,” in which the printed parts typically contain significant amounts of unreacted species that can undergo additional cross-linking when further exposed to UV light. If this UV exposure occurs during service, it can change the part's mechanical properties over time, which is not desirable in the automotive industry.²³ Exposure to UV during service can also induce photooxidation, that is, the degradation of material in the presence of oxygen, which can cause changes in the chemical composition and negatively affect the part performance.^23,24

For VPP-based 3D printing, postprocessing is one of the most important factors affecting the mechanical properties of the printed parts.¹⁶ Post-curing is often employed after printing to cross-link some of the unreacted species before the parts are put into service. However, differences in resin formulation, part geometry, pigmentation, and stabilization can make it difficult to utilize a universal “one-size-fits-all” post-cure process. For instance, Zguris's studies clearly demonstrated that different 3D photopolymerizable resins require different radiation wavelengths, annealing temperatures, and time conditions to reach full-cure and show proper chemical performance.²⁵ Nevertheless, most of the former studies on mechanical properties of 3D printed parts are focused on the effect of various processing parameters,^26–30 while a detailed study of the effects of the post-curing process on the performance of 3D printed automotive parts is lacking. However, although VPP-based 3D printing methods have been used extensively for prototyping of visual aids and for validating new designs, its ability to produce durable functional parts for real-world applications, including automotive parts, is still in the early stages of adoption. As a result, the long-term weathering behavior of specimens printed through VPP is largely unknown. In our previous study, the effect of long-term interior weathering, that is, using dry cycles only, on the macro- and micromechanical properties and photooxidation rate of the same set of UV-cured resins was investigated.³¹ The results indicated that Resin B showed promising retention of both tensile properties and nanohardness after the interior weathering exposure. However, Resin B did not perform as well as the two other resins in photooxidation resistance.

In the current study, we investigated the effect of the post-cure process on the performance of 3D printed materials based on acrylate and thiol–ene chemistries. Tensile bars were printed using DLP printer per ASTM D638 Type IV and post-cured using one of five different processes: no post-cure, UV-only, heat-only, UV+heat, and electron beam (EB) curing. In addition to the effect of the post-curing process on the macro- and micromechanical properties of the resins, the effect of long-term exterior weathering on photooxidation was also studied. Photooxidation behavior, which provides a good indication of how quickly or slowly a material undergoes chemical degradation, was measured via the Photoacoustic Fourier Transformation Infrared Spectroscopy (FTIR-PAS) method and was further compared with previous results from interior weathering.

Experimental

Materials

In this study, three different UV-curable resin formulations, supplied by Colorado Photopolymer Solutions (Boulder, CO) were used to prepare the specimens. Resins A and B were based on thiol–ene chemistry, whereas Resin C was formulated using conventional acrylates. The alkene source in Resin B contains a low-molecular-weight aromatic heterocycle structure, whereas Resin A contains high-molecular-weight aliphatic urethane acrylates. Therefore, Resin B contains more functional groups, which, by nature, makes it more reactive than Resin A. The reactivity parameters of these resins, that is, Critical Energy (Ec) and depth of penetration (Dp), are as follows:

• Resin A: Dp: 7.6 mils; Ec: 9.7 mJ/cm²

• Resin B: Dp: 5.1 mils; Ec: 5.1 mJ/cm²

• Resin C: Dp: 7.0 mils; Ec: 31.9 mJ/cm²

According to these values, Resin B needs the lowest amount of energy to form a solid layer among the three formulations. One should note while these resins did not include any hindered amine light stabilizers (HALS) in their formulations for light stabilization, they do contain carbon black, which could perform as a strong UV absorber to ensure a good z-resolution during printing. The concentration of carbon black is the same across the three resins.

Methods

Printing method

A DLP 3D printing method was adopted in this study to print the test specimens. DLP is a bottom-up type of VPP that typically uses a masked digital light projector screen with UV radiation to photopolymerize a UV-curable resin³² by projecting from the bottom up. In addition to the projector, the DLP printer is equipped with a motorized platform that travels vertically along the z-axis, as well as a resin vat with a transparent bottom. The 3D printing process using DLP is as follows: First, the geometry of the sample is modeled using a computer-aided design program. Next, the 3D model is sliced into a stack of two-dimensional images, that is, a series of thin layers that together form the whole sample. Afterward, the motor-powered build platform is moved down until the gap between the vat bottom and the platform is equal to the intended layer thickness. The image of the first layer is then projected through the vat bottom onto the build platform, using UV light for a specified exposure time period. Finally, the UV light is turned off, the platform rises by a layer thickness, and the cured resin detaches from the window. Photopolymer resin in the vat then flows back between the previous layer and the vat bottom, and the former steps are repeated until the sample is completely printed.^33,34

A series of type IV dumbbell-shape tensile bars, with dimensions as described in ASTM D638, were printed with an early version of the Origin One printer, which did not have active heating (Origin, Printing equipment and supplies, San Francisco, CA) using a 50–75 μm layer thickness. The UV source in the printer, a projector, emitted light at 385 nm, with an intensity of 5 mW/cm². The area of UV light illumination was 144 × 81 mm. The layers were irradiated as follows: 25 s exposure for the first layer, 10 s exposure for layers 2 through 6, and 3 s exposure for all subsequent layers. The transparent window in the Origin One is composed of polytetrafluoroethylene, which flexes after each print layer to release the cured resin from the vat bottom. After printing, the samples were sonicated in an isopropyl alcohol bath for 3 min to clean off any uncured resin, followed by drying in an oven at 25°C for 1 h. The samples were then post-cured using various processes, wrapped in aluminum foil, and kept in a dark place to prevent any additional light-induced polymerization before testing.

Post-curing process

The printed samples were post-cured using one of the following processes: no post-cure, UV-only, heat-only, and UV+heat. As a complementary study, the effect of EB post-curing was also studied for Resin B only, which demonstrated the most desirable mechanical properties among all the resins. Thermal post-curing was conducted by placing the samples in an oven for 1 h at 150°C. An ELC-4001 UV Flood System (Electro-Lite) with a broad-spectrum lamp is used for UV post-curing. UV post-curing was performed for 4 min and 30 s on each side of the samples for a total curing time of 9 min. EB curing of the samples was conducted using an EB accelerator equipped with a variable speed, fiberglass carrier web (Broad Beam EP Series, PCT E-beam and Integration, LLC, Davenport, IA). The penetration depth of EB is dependent on the mass density and thickness of the materials.³⁵ In this study, to achieve a uniform penetration dosage across the entire thickness of the samples, they were flipped over during the EB post-curing process.

Accelerated weathering

To simulate exterior weathering conditions in the automotive industry, the printed samples, which were cured by standard UV-only post-cure method, were exposed in an Atlas Ci-35 Weather-Ometer running test method ASTM D7869 for up to 2500 h. The samples were periodically removed for analysis.

Testing and evaluation

Mechanical properties

The macromechanical properties of the specimens were evaluated by testing on an Instron 3366 tensile machine, employing a 30-kN load cell. According to ASTM D638, the initial distance between the grips was adjusted to 65 mm, and a 25-mm extensometer was used. Five test specimens were tested and averaged for each post-cure condition. The testing was carried out at a constant crosshead speed of 5 mm/min in ambient conditions.

The micromechanical properties of the specimens were examined utilizing nanoindentation to study the material nanohardness as a function of depth to assess the possible nonuniformities across the print layers. Strips of each sample, cut using a manual cutter, were prepared such that the sample cross-sectional surfaces were normal to the exposure direction, as shown in Figure 1. The resulting indentation load was applied perpendicular to the compaction axis. The strips were cold-mounted in epoxy resin to be polished with sandpapers of decreasing roughness starting from 600 down to 2400 grit #, followed by fine polishing using abrasive slurries, with a characteristic abrasive size starting from 5 μm down to 0.3 μm.

FIG. 1.

Illustration of (a) sample preparation and (b) indentation geometry for nanoindentation studies. Color images are available online.

The nanoindentation tests were carried out at room temperature using an Anton Paar (Graz, Austria) nanoindentation tester (NHT³) equipped with a Berkovich indenter, an optical microscope attached to a video camera, and a sample holder rigidly fixed to an x–y–z motorized table. The testing process is described in more detail in our previous studies.^31,36 The indents were made along a path of 50 μm in the z-direction (or the height of each layer) starting from the outside of the print (depth = 0 μm) to a total length of ∼350 μm (depth = 350 μm). One should note that depth of indentation varied from 3 to 10 μm across the different layers due the inherent heterogeneity. Eight indents per depth in the x-direction, separated from each other by a 50 μm gap, were used at each depth to calculate the average hardness as a function of depth (Fig. 1a). The force–displacement data were used to determine the point of contact. After the sample was contacted, the force was linearly increased, and the tip indented into the surface of the sample. A short dwell time occurred at the maximum load of 50 mN, and then, the sample was unloaded. In this study, Oliver and Phaar's method²² was used to calculate the hardness by dividing the maximum load, P_max, by the contact area of indent (A), as described in Equation (1).

Studying the nanohardness profile at various depths, as an approximation of the cure state at each depth, was conducted with two main purposes in this research: First, to investigate the effect of each post-curing process, and second, to monitor any possible difference between the front side (directly weathered) and the back side (indirectly weathered) of the weathered samples. To be able to readily distinguish these two sides, the backside was marked using a red tape before the cold mounting (Fig. 1b).

Photoacoustic Fourier transform infrared spectroscopy

Exposure to UV, heat, and/or moisture in the weathering process can cause chemical composition changes to the surface of the printed samples through photooxidation, leading to a growth in (−OH, −NH) region of the Infrared (IR) spectrum. This growth can be monitored using IR spectroscopy methods. In the current study, Fourier transform infrared (FTIR-PAS) spectroscopy was used to monitor the changes in the chemical composition of punched 3D printed samples as a function of exposure dosage (kJ/m²) in the case of interior weathering or exposure time (hours) in the case of exterior weathering. Photooxidation was quantified by measuring the ratio of the (−OH, −NH) area (4000–2500 cm⁻¹) normalized to the (−CH) area (3200–2800 cm⁻¹) in the IR spectra of the samples using Equation (2).^23,31,37

The FTIR-PAS spectra were recorded using a Thermo Scientific iZ10 spectrometer equipped with an MTEC model 300 photoacoustic detector. A helium gas purging flow was used to reduce the noise. The samples were punched and packed in small ring holders 10 mm in diameter. For each sample, 64 scans between 4000 and 400 cm⁻¹ at a resolution of 4 cm⁻¹ were recorded.

UV-Vis spectroscopy

To study the absorbance per micron () of the resins, solutions with 1 wt.% concentration in dichloromethane (DCM) solvent were prepared. The absorbance spectra of the solutions were measured using a PerkinElmer Lambda 18 UV-Vis spectrometer. DCM was used as the background, and the range scanned was 250–500 nm, with an average time of 0.3 s and data interval of 5 nm. Knowing that the length of the quartz cuvette used was 1 cm, the absorbance per micron for each sample was calculated using Equation (3).

Confocal Raman spectroscopy

Raman spectra of the printed specimens were taken to determine the extent of carbon double bond conversion. Measurements were made using a Renishaw inVia microscope that utilized a 532 nm excitation from a frequency-doubled Nd:YVO₄ laser from Coherent and a 50 × objective. This instrument uses a holographic notch filter as a beam splitter to couple the laser light into the microscope and onto the sample and to reject laser light from the backscattered signal that goes to the spectrometer while passing the Raman-shifted spectrum. A second notch filter further reduces the backscattered laser intensity. Spectra could be recorded in either a static mode with a relatively short frequency range or in an extended mode in which the grating is stepped synchronously with the shifting of the charge in the charge-coupled device array detector, to produce long continuous scans. Daily calibrations were taken using the silicon peak at 523 cm⁻¹.

Raman spectra were taken of the cross section of the specimens. Spectra were recorded using a static scan mode, with an exposure time of 5 min and the grating centered in the polymer fingerprint region at 1450 cm⁻¹. Laser power to the sample was set to 10 mW and focused to a spot size of 1–2 μm. Depth resolution was ∼2–3 μm and spectra resolution about 4 cm⁻¹.

To determine acrylate double bond conversion from the Raman spectra, a second derivative transform was applied to each spectrum. Peak heights from these smoothed spectra were then used to quantify the concentration of acrylate double bonds. Details of this technique have been previously reported.^38,39 The strength of the acrylate C = C bond peak at 1636 cm⁻¹ was recorded and ratioed to a constant reference peak at 1765 cm⁻¹.

Results and Discussion

Effect of post-curing process on the mechanical properties

Macromechanical properties

The effects of the post-curing method on the macromechanical properties, that is, tensile strength, Young's modulus, and percent elongation at break (%E), were studied for all three resins. As shown in Figure 2, for Resin A, the less reactive thiol–ene-based resin, post-curing improved both tensile strength and Young's modulus regardless of the post-cure method, while reducing %E. For thermal-only post-cure method, tensile strength increased by Σ200% from 11.20 ± 0.64 MPa to 33.70 ± 3.43 MPa and Young's modulus increased by 111% from 758 ± 103 MPa to 1560 ± 169 MPa. However, %E showed more than an 85% reduction from 27.30 ± 9.27 to 3.38 ± 0.19. These values were not significantly different for UV-only and UV+heat post-cure methods. Both the UV and thermal post-cure methods resulted in similar tensile properties because thiol–ene systems can undergo polymerization via both radiation¹¹ and heating²³ processes. It should be noted that the drastic reduction in %E caused by post-curing is not desirable in automotive parts that require a high amount of %E retention during their service life. Therefore, Resin A may be a good option for parts that need high flexibility but are not exposed to mechanical forces that can induce dimensional changes. Examples of these elements include decorative parts or badges.

FIG. 2.

Effect of post-curing process on tensile properties of Resin A: (a) tensile strength, (b) Young's modulus, and (c) %E. %E, percent elongation at break; UV, ultraviolet.

As depicted in Figure 3, post-curing did not have a significant effect on the tensile properties of Resin B, the more reactive thiol–ene-based system. Post-curing slightly improved the tensile strength and Young's modulus and did not result in considerable reduction of %E, unlike Resin A. While Resins A and B are thiol–ene-based systems, they showed a different behavior upon post-curing. This is because, as mentioned before, Resin B includes a heterocyclic core structure with lower molecular weight, and in turn, higher functionalities than Resin A. For Resin B, the UV radiation during the printing process likely resulted in a vitrified network with limited segmental mobility that hindered further conversion of unreacted groups. This vitrification effect was slightly mitigated by heat treatment in case of UV+heat post-curing. Moreover, higher functionalities in Resin B can prevent oxygen inhibition more effectively, which, in turn, can induce higher conversions. It seems that ample cross-linking of the free radicals occurs during the printing process due to its higher reactivity. Therefore, post-curing does not seem to be necessary for Resin B, making it a beneficial option for use in high-volume manufacturing environments. Considering these points, Resin B could be a promising option for the automotive parts that require good retention of tensile properties, including %E. One should note that while the chemical structure may result in more complete conversion during printing and not require post-cure, the final elongation values are still in the same range as the degraded elongations for the other materials and still need to be improved for parts that need to flex more.

FIG. 3.

Effect of post-curing process on tensile properties of Resin B: (a) tensile strength, (b) Young's modulus, and (c) %E.

For Resin C, the acrylate-based resin system, thermal post-curing was not as effective as UV post-curing. As shown in Figure 4, tensile strength increased by 285% from 27.10 ± 2.29 MPa to 104 ± 1.04 MPa, and Young's modulus increased by 247% from 1174.30 ± 135.30 MPa to 4389.78 ± 757.67 MPa, as a result of UV post-curing. These values were about Σ45% from 27.10 ± 2.29 MPa to 39.50 ± 2.40 MPa for thermal post-curing and Σ109.6% from 1174.30 ± 135.30 MPa to 2460.72 ± 417.03 MPa for UV post-curing. Although it is known that acrylate groups can undergo thermal polymerization,²⁴ this resin did not contain any thermal initiators, so it is unlikely any significant cross-linking occurred as a result of the 150°C/1 h “heat-only” post-cure step. As a result, UV post-curing seems to provide the best tensile properties among all post-curing processes for Resin C. However, Resin C exhibited a similar post-cure reduction in %E to that of Resin A, that is, UV post-curing decreased %E by 58% from 7.95 ± 3.66 to 3.28 ± 0.62. This again implies that Resin C should not be used in applications that may significantly stress or load the part. Employment of both UV and thermal post-curing together negatively affected the mechanical properties of this resin for unknown reasons. One possible speculation could be increased curing of the sample leading to embrittlement after being heated at 150°C for 1 h.

FIG. 4.

Effect of post-curing process on tensile properties of Resin C: (a) tensile strength, (b) Young's modulus, and (c) %E.

To evaluate the effectiveness of an EB post-cure process, Resin B samples, which showed the highest consistency in the tensile properties as a result of post-curing process, were exposed to varying dosages of EB radiation. EB curing through free radical polymerization differs from UV curing, mainly in the initiation process. Photoinitiators are not required in the EB process because the high energy intensity of EB is enough to form the initiative species by cleavage of the carbon–carbon double bonds.²⁵ Since the initiating radical is formed from the resin itself, rather than from an added initiator, EB curing allows for a small amount of additional cross-linking compared with UV curing.²⁶ One of the downsides of EB cure systems is inhibition by oxygen, and therefore, EB curing is usually conducted in an inert atmosphere, such as nitrogen. This drawback could be overcome by the utilization of thiol–ene chemistry, and because these systems are already partially cross-linked, limited oxygen can permeate into the material to quench radicals. All these advantages make the post-curing of thiol–ene-based resin systems via EB an interesting option for DLP 3D printing.

As shown in Figure 5, higher dosages of EB post-cure (i.e., one-step 5, 10, and 20 Mrad) resulted in significant embrittlement of samples of Resin B compared with those that underwent standard UV post-curing. %E values were reduced by Σ50%, regardless of the dosage. Moreover, increasing the EB dosage from 5 to 20 Mrad decreased the tensile strength by Σ38% from 50.60 ± 7.10 MPa to 31.20 ± 0.86 MPa, while it increased the Young's modulus by Σ27% from 2732.8 ± 56.73 MPa to 3480 ± 250.73 MPa. It was unknown if this was simply due to an excess of EB dosage or if the samples underwent secondary heating during the EB post-curing process.

FIG. 5.

Effect of one-step versus step-by-step EB post-curing on tensile properties of Resin B: (a) tensile strength, (b) Young's modulus, and (c) %E. EB, electron beam.

Therefore, two changes were made to the EB post-curing process. First, EB curing was conducted in successive 2.5 Mrad steps for dosages >2.5 Mrad, with a pause between successive exposures to cool the samples in ambient conditions. This adjustment was made to reduce any secondary sample heating as much as possible (denoted as step-by-step curing in Fig. 5). Although step-by-step EB curing in higher dosages improved tensile strength and Young's modulus to some extent, %E was still far from the values achieved using the standard UV post-cure method.

Second, additional samples were EB post-cured with significantly reduced dosages of 0.05, 0.1, 0.5, 1, and 2.5 Mrad (Fig. 6). The reduced EB dosages resulted in higher %E values than those measured for the higher EB dosages. Dosages <2.5 Mrad resulted in Young's modulus values close to that of standard UV post-cure, and tensile strengths more than that of the UV post-cure samples. These results indicate that higher dosages of EB likely induce excessive cross-linking of the resin, which will be investigated in our future studies using Raman spectroscopy.

FIG. 6.

Effect of EB post-curing with low dosages on tensile properties of Resin B: (a) tensile strength, (b) Young's modulus, and (c) %E.

Micromechanical properties

To study the nanohardness profile, as an approximation of the cure state at various exposure depths, a strip of each sample was cold-mounted in epoxy resin to be polished. Depth profiling was conducted to determine if any gradient in cure state exists for different post-curing methods. When post-cured by UV-only, Resin A showed a gradient in hardness as a function of depth, that is, the hardness decreased as the depth increased. This might be due to the higher UV absorption of this resin, which would limit the depth of cure. Therefore, UV-Vis spectroscopy studies were conducted, as will be discussed in the UV-Vis Spectroscopy section to investigate if this is the case.

For Resin B, nanohardness results were in good agreement with the tensile properties, that is, the nanohardness was slightly improved, regardless of the post-curing process (Fig. 7). Similarly, nanohardness was observed to be independent of depth, showing a uniform conversion as a function of depth. A similar trend was observed in higher EB dosages.

FIG. 7.

Effect of post-curing process on nanohardness of Resin B.

Figure 8 shows that the surface nanohardness of Resin C increased by >200% after post-curing (from ∼50 to >150 MPa), regardless of the process used. Moreover, no gradient in hardness as a function of depth was observed, which is particularly interesting in the case of UV curing. Heat curing showed a similar trend.

FIG. 8.

Effect of post-curing process on nanohardness of Resin C.

Effect of weathering on photooxidation and durability

The change in the [(−OH, −NH)/(−CH)] as a function exterior weathering time is demonstrated in Figure 9. As previously discussed, the [(−OH, −NH)/(−CH)] growth rates were obtained from FTIR-PAS spectra taken from punched 3D printed samples, which were cured using standard UV-only post-cure process.

FIG. 9.

Δ[(-OH, -NH)/(-CH)] versus weathering time.

According to the results, Resin B showed a higher rate of photooxidation compared with the two other resins, which demonstrated nearly identical results. This is likely due to the chemical degradation of thiol–ester linkages in the samples. In addition, the final values of photooxidation in exterior weathering are less than interior weathering presented in our previous study,³¹ likely because parts of the photooxidation products accumulated on the surface are washed away when water is sprayed in the former process. Overall, considering that the resins used in this study do not include any UV stabilizers such as UV absorbers or HALS, these results are encouraging, and it is expected that the addition of HALS could enhance the photooxidation resistance significantly.

Additionally, as discussed in our previous study,³¹ the effect of interior weathering on the tensile properties of Resin B was not as considerable as the other two resins, particularly on %E. Because Resin B is a very reactive thiol–ene-based system, the UV radiation during the printing process likely resulted in a vitrified network with limited segmental mobility that hindered further conversion of unreacted groups. Moreover, the fact that thiol–ene-based systems are minimally inhibited by oxygen can also contribute to increased photopolymerization of this resin during the printing process. It seems that ample cross-linking of the free radicals occurs during the printing process, limiting further cross-linking and embrittlement during weathering. However, this was likely not the case in Resins A and C, which have much less reactivity. As a result, interior weathering caused excess cross-linking in these two resins, and therefore, a significant reduction in their %E.

Knowing that typically exterior automotive parts need to flex less than interior ones, which might get more heated and be pushed on by users, Resin B is likely a better option for automotive interior applications. However, Resins A and C that demonstrated better weathering performance could be proper choices for exterior decorative automotive parts.

UV-Vis spectroscopy

UV absorbance is an important parameter in achieving a proper z-resolution and ample curing of layers in DLP 3D-printing. UV-Vis spectroscopic studies were conducted to elucidate whether the depth of cure and extent of cure of the samples could be correlated with their resin absorbance per micron. Recall that the Resin A sample cured by a standard UV post-curing process demonstrated a gradient in hardness as a function of depth at time 0 (before weathering). The Resin C sample demonstrated a similar gradient as a function of depth from the front side after interior weathering. Figure 10 shows that the UV absorbance per micron () for Resin A ≥ Resin C > Resin B is 385–405 nm wavelength region, which is radiated by most of the DLP printers. This higher absorbance for Resins A and C could be a contributing factor in limiting the depth of curing after UV irradiation, and in turn, a reduction in the hardness as a function of depth.

FIG. 10.

UV-Vis spectra of 1 wt.% solutions of the three resins in dichloromethane.

Confocal Raman spectroscopy

Raman spectroscopy was conducted on Resin B, the sample that showed the highest retention of mechanical properties, to investigate whether correlation between the nanohardness profile and percent C = C conversion at each depth exists. Figure 11 shows the percent C = C conversion of Resin B as a function of depth when cured by UV-only and UV+heat processes. According to the results, using UV+heat slightly increases the double bond conversion compared with UV-only post-curing. The UV+heat post-cure also displays increased nanohardness compared with UV-only, as shown in Figure 7. Also in agreement is the commensurate gradient between C = C conversion and nanohardness, as shown in Figures 7 and 11. This means that nanohardness correlates well with the amount of double bond conversion. Unfortunately, we were not able to use Raman spectroscopy to compare the degree of cure for the green state and UV post-cured samples with the EB post-cure samples because the mechanisms of radical generation and cross-linking during EB exposure are different from the UV process, which resulted in significantly different spectra. However, it was observed that for all EB doses the 1636 cm⁻¹ C = C peak completely disappeared after EB post-curing.

FIG. 11.

Percent double bond conversion as a function of depth for Resin B.

Conclusions

The primary goal of this study was to investigate the effects of various post-curing processes on the mechanical performance of DLP 3D printed parts for automotive applications to provide good insight into selecting and optimizing the proper post-curing method per formulation/application. To this end, three resin systems with two different chemistries were used to print the bars used for tensile testing. The bars were then post-cured using the five following methods: no post-cure, UV-only, heat-only, UV+heat, and EB curing. Tensile properties and nanohardness results demonstrated that thermal curing was not as effective as UV for acrylate-based resins such as Resin C compared with the two other resins with thiol–ene chemistry, which could undergo thermal polymerization as well. However, EB curing, even at very low dosages, was very effective for cross-linking of one of the thiol–ene-based resins to the point of embrittlement for EB dosages above 0.5 Mrad. Regardless of the type of post-curing, the highly reactive thiol–ene-based Resin B showed the smallest change in mechanical properties as a result of post-curing. Therefore, post-curing does not seem to be necessary for Resin B, which could be a benefit in certain applications. Furthermore, Resin B could be a promising option for the automotive parts that require good retention of tensile properties including %E. Resins A and C did not show a very high %E retention. In addition, the results of our previous study showed that Resin B maintained promising consistency in both macro- and micromechanical properties before versus after interior weathering. However, Resins A and C outperformed Resin B in photooxidation resistance.

Considering the above-mentioned points, Resin B could be a promising option for the automotive parts that require good retention of tensile properties such as % elongation. These applications include those that have load-bearing requirements such as brake brackets. Resins A and C, however, might be good options for applications that are exposed to UV but are not exposed to mechanical forces that can induce dimensional changes, such as decorative parts or badges. It should be mentioned that both mechanical properties and weathering durability of these resin systems are expected to improve even further after addition of HALS to their formulations. The results indicate that the performance of 3D printed parts can be tailored by a combination of resin chemistry and formulation, print process, and post-cure process.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This research is funded by RadTech International—The UV&EB Technology Association.