3D Printing with Polyethylene (PE) Vitrimers

Christina Harmon, Christopher Lawrence, Justin Tan, Oliyad Dibisa, Jiachen (Tommii) Zeng

Abstract

Polyethylene (PE) is one of the most used polymers in traditional manufacturing polymer processes. However, it remains a challenge to use PE in 3D printing. Some of the challenges this material presents are poor self-adhesion and poor adhesion to traditional printing beds. Additionally, this material undergoes considerable warpage and shrinkage during the printing process. Vitrimers are a new class of polymers that have a dynamic cross-linked network which allows the material to be reprocessed and enhances the dimensional stability of PE at elevated temperatures. This research aims to understand if PE vitrimers can reduce mechanical anisotropy and warpage. HDPE and HDPE vitrimers achieved the best printing results on a 1/16’’ thick Polyethene sheet with 220°C nozzle temperature and 60°C bed temperature. HDPE required a printing speed of 800mm/min while HDPE vitrimer required 500mm/min. It was demonstrated that HDPE vitrimers were able to successfully reduce warpage and shrinkage during the printing process. Dynamic mechanical analysis showed that HDPE vitrimer resulted in higher mechanical anisotropy compared to HDPE. This is most likely due to the high viscosity of the vitrimer which resulted in lower chain diffusion between layers. Annealing printed HDPE Vitrimers led to improved mechanical properties as well as decreased mechanical anisotropy. This annealing process is only possible in HDPE vitrimers due to their superior melt strength at elevated temperatures. Future experiments will include finding better annealing parameters to improve the dimensional accuracy of 3D-printed HDPE Vitrimer parts.

Introduction

Polyethylene (PE) is a thermoplastic polymer with a variable crystalline structure and a vast range of applications depending on the type. PE is a popular material due to its affordability, machinability, and compatibility with other materials. High-density polyethylene (HDPE) and linear low-density polyethylene (LLDPE) are two types of PE. HDPE has molecules that are packed together tightly, have a linear structure, and are opaque. HDPE is strong and has a high tensile strength, rigidity, and impact resistance. LLDPE is also strong and is a substantially linear PE with numerous short branches. However, PE has been shown to be quite challenging to use when it comes to 3D printing. Specifically, when printing with HDPE and LLDPE, the material is unable to stick to the printing bed. Additionally, PE suffers unacceptable warpage and shrinkage [1].
Vitrimers are a new class of polymers that combine the properties of thermoplastics and thermosetting materials: they behave like cross-linked materials at service temperature while being processable at elevated temperature due to the presence of dynamic cross-links. Vitrimers feature dynamic covalent crosslinks that do not dissociate during the exchange reaction. However, despite a constant degree of crosslinking, these materials can flow under the action of heat depending on the kinetics of the exchange reaction [2]. Vitrimers are beneficial as they are able to achieve a compelling balance of robust mechanical properties and processability [3]. At high temperatures, they can flow like viscoelastic liquids, while at low temperatures the bond-exchange reactions are immeasurably slow (frozen) and the vitrimers behave like classical cross-linked materials [3].
Vitrimers are a promising material and combining the benefits of these vitrimers with PE, by utilizing PE vitrimers, in the 3D printing process could potentially resolve the challenges that occur when 3D printing with PE alone.
Extrusion additive manufacturing (EAM) is a 3D printer with screw-assisted systems that can be fed directly with printing materials in pellet form. The development of EAM has enabled such technology to be applied on 3D printers, which expands the range of 3D printing materials, reduces the cost associated with feedstock fabrication, and increases the material deposition rate compared to traditional fused filament fabrication (FFF) [4].
Therefore, the first objective of this project was to develop a successful method to 3D print with HDPE and HDPE vitrimers. Tuning of the printer was performed with HDPE. The second objective was to 3D print rectangular specimens of each material to use for warpage tests, shrinkage tests, annealing tests, and Dynamic Mechanical Analyzer (DMA) tests (printed at 0°, 45°, and 90°) so that the properties of each material could be compared and characterized.

Material Choice

The material choice for HDPE was HDPE F04660 supplied by SABIC. The HDPE vitrimers were prepared using reactive blending of maleic anhydride-grafted-polyethylene (PE-MAH) and 4,4’ dithiodianiline (DTA) as the crosslinker.  The material of choice for HDPE vitrimer was HDPE-V0.3 which represents 0.3% wt. of MAH and a molar ratio of MAH:DTA of 1:0.5. Parameter optimization was done using Simplify3D, and printing was performed with Cosine, a screw-based additive manufacturing machine. HDPE and HDPE vitrimers have similar printing properties, such as melt temperature and crystallization temperature. Due to the high price of manufacturing HDPE vitrimer, HDPE was chosen as the tuning material.

Methodology

Print Design and Optimization

The first objective of this project was to develop a successful method to 3D print with HDPE and HDPE vitrimers. The Cosine Pellet Feed Extruder is a screw-assisted 3D printer and was utilized in this project to produce test samples. HDPE and HDPE vitrimers have similar printing properties, such as melt temperature and crystallization temperature. Due to the high price of manufacturing HDPE vitrimer, HDPE was chosen as the tuning material.

The test sample model was drawn using SOLIDWORKS. The dimension of the test sample is shown in Figure 1. The geometry was designed specifically to fit the DMA machine. 

Figure 1. Test specimen dimensions for both HDPE and HDPE vitrimer

Parameter optimization was done using the software Simplify3D. When loading the model into the software, printing parameters could be adjusted and the geometry could then be translated into Gcode to communicate with the 3D printer. Parameters that were altered and tested included extrusion multiplier, print bed material, nozzle temperature, bed temperature, printing speed, layer height, first layer settings, brim, and the usage of an external heat source. Test samples were printed at 0°, 45°, and 90°.

Shrinkage Testing 

Shrinkage tests were performed using Ohaus Explorer Analytical Balance in which the weight of the sample in air and the buoyancy of the sample when fully immersed in water were measured. The density of the samples was given by the balance based on the measured values, and the volume of the sample was calculated. The calculated sample volume was then compared with the CAD volume to generate a volumetric difference. 

Warpage Testing

To perform warpage tests, two samples were printed on a flat printing bed with the Cosine Pellet Feed Extruder for HDPE and HDPE vitrimers using their optimized printing parameters. The samples were allowed to cool down fully and were then scraped off the bed. Pictures were taken with both samples laying on a flat surface and the warpage was compared visually. 

Dynamic Mechanical Analysis(DMA)

DMA strain sweeps and frequency sweeps in tension were performed on compression-molded, 0° and 45° 3D-printed HDPE and HDPE vitrimer samples using a Netzsch Eplexor 500N DMA at room temperature. Due to time constraints, DMA tests were not performed on the 90° 3D-printed samples. Strain sweeps using 0.001% to 0.5% strains were first performed to locate the linear viscoelastic region (LVR). Following the strain sweeps, a single dynamic strain of 0.03% within the LVR was used to conduct the frequency sweeps from 0.5 to 100 Hz on all samples. Anisotropy in 3D-printed HDPE and HDPE vitrimers was evaluated by comparing their complex modulus (E*) at 100 Hz by printing orientations. 

Annealing

0°, 45°, and 90° 3D-printed HDPE and HDPE vitrimer samples were heat-treated for two hours at 160 °C for two hours. The post-cured 0° and 45° 3D-printed HDPE vitrimer samples underwent the DMA frequency sweep using the same parameters. The post-cured 90° 3D-printed HDPE vitrimer sample could not be tested due to time constraints. The post-cured 3D-printed HDPE could not be examined by DMA as the samples melted during the post-curing process. 

Results and Discussion

Print Design and Optimization

An initial printing parameter was needed as a reference for optimization. Initial printing parameters were referenced from filaments.ca, where individuals share and discuss the optimal settings for 3D printing with HDPE filaments. The nozzle diameter was controlled at 1mm throughout the experiment. The first batch of successfully printed test samples of HDPE was printed using the parameters shown in Table 1. Defects were observed and presented in Table 2.

Table 1. HDPE initial printing parameters (0° orientation)
Table 2. HDPE initial print trail defects observation

Optimizations were made to address the defects observed when printing with HDPE. The melting and crystallization temperature for HDPE ??F04660 – SABIC is 138°C and 115°C, respectively. A bed temperature of 135°C would ensure that the material was above the crystallization temperature so the material would stick to the bed with no warpage during the printing process. However, such a high print bed temperature created a melt pool in the first layer of printed material. Even though a lower temperature difference between the nozzle and print bed is generally recommended, lowering the bed temperature to 60°C was more suitable for HDPE printing. Observations are presented in Table 3.

Table 3. Observation of different print bed temperatures

However, at a bed temperature of 60°C, adhesion became a significant issue. Print bed material optimization was performed by selecting various substrate materials and testing them under the same bed temperature. Materials were taped onto the aluminum print bed and heated to 60°C. Adhesion of the bed material with HDPE has been documented in Table 4.

Table 4. Print bed material optimizations at 60°C

A polypropylene (PP) sheet with a thickness of 1/16’’ was chosen as the best substrate material for HDPE 3D printing. Unlike LLDPE, PP provides just enough adhesion, allowing HDPE to stick to it but not bond to it. A thickness of 1/16’’ ensures better heat dissipation so that it does not warp whereas a thickness of  0.03’’ was too thin and resulted in warpage. When printing with PP substrate, a heat gun should not be used since it can quickly deform the material.

At the final tuning stage, brim was added to the printing design to further eliminate warpage and increase the adhesion of HDPE samples. It was found that having at least five brim layers significantly reduced the warpage at the edge of the sample. A direct comparison of samples with and without brim is shown in Figure 2. 

Figure 2. 0° orientation HDPE samples warpage comparison (side view and front view.) Left: No brim; Right: with brim

The final printing parameters determined for HDPE 3D printing are shown in Table 5.

Table 5. HDPE optimized printing parameters (0° orientation)

The same parameters were applied when printing HDPE vitrimers. However, due to the higher viscosity of the material, not enough material extruded out of the nozzle, causing gaps between each bead. A higher extrusion multiplier was applied to compensate for the insufficient extrusion but this introduced a new defect, melt fracture, shown in Figure 3.

Figure 3. Melt fracture with HDPE vitrimer

Melt fracture is usually observed when the residence time (processing time) is lower than the relaxation time. Therefore, the extrusion multiplier was decreased to lower the residence time. Further experiments will be conducted to measure the relaxation time of the HDPE vitrimer in the melt state. The residence time in the 3D printer will also be quantified with a color tracer. Trials of vitrimer printing are displayed in Figure 4. HDPE-V0.3 has a yellow to brown appearance due to the cross-linked chemistry.

The final printing parameters used for HDPE vitrimers are shown in Table 6. Test specimens with orientations of 0°, 45°, and 90° were printed with the printing parameters shown in Table 5 and Table 6 for HDPE and HDPE vitrimer respectively (Figure 5). 

Figure 4. HDPE vitrimer print samples (Left: top; Right: bottom) that show improvement in printing quality (M: extrusion multiplier, S: printing speed)
Table 6. HDPE vitrimer optimized printing parameters (0° orientation)
Figure 5. Printed samples with 0°, 45°, and 90° orientations from left to right. HDPE on the top row, HDPE vitrimer on the bottom row.

Samples were trimmed to desired dimensions in order to exclude the build-up of material on the perimeter of each sample and to meet the DMA machine size requirement. Final samples for DMA testing are displayed in Figure 6.

Figure 6. 0° orientation HDPE samples. Left: Trimed samples for DMA tensile testing with dimensions. Right: Untrimmed sample. (Samples are not to scale)

Shrinkage Testing

The Ohaus Explorer Analytical Balance was used to perform shrinkage tests. After careful comparison, 45° orientation printed samples for both HDPE and HDPE vitrimers were selected to perform the test due to having the best surface finishes. The testing equipment was set up to measure the weight of the samples in air and the buoyancy of the samples when submerged in water (the mesh pushed the sample down under the water to ensure maximum buoyancy.) Then the density of each material was displayed on the balance so that the volume could be calculated by hand. Each measurement was repeated three times to obtain an average value for each parameter. Results are shown in Table 7.

Table 7. Average mass, buoyancy, density, sample volume, volume difference, and theoretical simplified volume of CAD for printed 45° orientation HDPE and HDPE vitrimers

It is worth noting that in order to achieve a good bounding on the first layer of each printing, the first layer height was manually reduced to 0.05mm above the print bed. Such action applied high pressure which enhanced the adhesion between the first layer material and the print bed. This means, the actually printed sample height was one layer thinner than the drawing, in this case, it was 0.6mm. The theoretical volume calculation is shown in Equation 1.

Equation 1. Theoretical Simplified Volume of CAD in cubic centimeters

 Thus, the height and width for each sample were measured, and then volumetric differences between the simplified volume of CAD and the sample volumes are shown in Table 7 where the volumetric differences were calculated using Equation 2.

Equation 2. Volumetric Difference (VD) calculation in percentage

The volumetric differences presented in Table 7 cannot be understood as volumetric shrinkage. The volume of the model after slicing (Simplify3D) will be reduced compared with the volume of the CAD, but this would be  really hard to quantify. The theoretical simplified volume used in this calculation was based on the CAD model in Figure 1 which provided a fair comparison between two samples but dramatically increased the volumetric difference. As a reference, the volumetric shrinkage of HDPE is usually 2.5%[5], which indicates that other factors such as machine inaccuracy should be dominant. Based on the volumetric differences presented in Table 7, the conclusion can be drawn that HDPE vitrimer would have more geometric accuracy during printing compared to HDPE. The result coincides with the material properties shown in Figure 7. The shallow region of the HDPE vitrimer was smaller than the one of HDPE, indicating that the HDPE vitrimer will have lower crystallinity under the same cooling condition. 

Figure 7. DSC graph for HDPE and HDPE vitrimer tested.

Warpage Testing

Warpage of the samples was harder to quantify than shrinkage. A comparison of the warpage between HDPE and HDPE vitrimers was performed by visualization. Pictures were taken at the same angle to provide a direct comparison between the two materials as shown in Figure 8. From Figure 8, the HDPE vitrimer sample is observed to have less warpage compared with HDPE samples. The result coincides with the material properties discussed earlier.

Figure 8. 0° orientation HDPE and HDPE vitrimer samples warpage comparison (Picture sequence: side view with brim, front view with brim, side view without brim, front view without brim)

DMA Testing 

Results from the DMA strain sweeps are shown in Figure 9. The LVR is located between 0.003% and 0.1% strain and is consistent among all samples regardless of the material and fabrication method. A dynamic strain of 0.03%, which is within the LVR, was used for the subsequent frequency sweeps.

Figure 9: Dynamic Strain Curves for Compression-molded, and 0° and 45° 3D-printed HDPE and HDPE Vitrimer Samples.

Results from the DMA frequency sweeps are shown in Figure 10. As seen in Figure 10, for all samples, the complex modulus (E*) increases with increasing frequency. This is consistent with the time-dependent behavior of polymers in response to deformation. At higher frequencies or smaller time scales, polymers behave more solid-like, as characterized by higher E*; whereas, at lower frequencies or smaller time scales, the samples behave more fluid-like, as shown by lower E*.

Comparing different materials prepared by different processes, the compression-molded HDPE shows the highest E* across the entire frequency range, followed by compression-molded HDPE vitrimer, which shows an average drop of 60 MPa in E* across all frequencies. HDPE samples printed with 0° and 45° orientation show a drop of roughly 250 MPa in E* relative to compression molded HDPE for all frequencies; the printing orientation does not have a significant effect on the E* of HDPE. HDPE vitrimer printed with 45° orientation show a drop of roughly 320 MPa and 380MPa in E* relative to compression-molded HDPE vitrimer and HDPE across the entire frequency range. And HDPE Vitrimer printed with 0° orientation show a drop of roughly 450 MPa and 510 MPa in E* relative to compression molded HDPE vitrimer and HDPE across the entire frequency range. 

Figure 10: Frequency Sweep Curves for Compression-molded, and 0° and 45° 3D-printed HDPE and HDPE Vitrimer Samples.

Figure 11 shows the comparison of E* of HDPE and HDPE vitrimer at 100 Hz by preparation method. Both 3D-printed HDPE and HDPE vitrimers show significantly decreased E* compared to their compression-molded counterparts. For HDPE, the difference in E* between the 0° and 45° orientation is only at 4MPa. Whereas for HDPE vitrimer, the difference in E* between the two printing orientations is 188 MPa. The large difference in E* between the two printing orientations indicates that 3D-printed HDPE vitrimers are more anisotropic than 3D-printed HDPE. 

Figure 11: Comparison of E* of HDPE and HDPE Vitrimer at 100 Hz by Preparation Method.

The Frequency sweeps and anisotropy test show results against what was expected: based on literature in the field, vitrimers in general have superior mechanical properties and isotropy relative to their parent materials due to the presence of dynamic crosslinks. The lesser mechanical properties and isotropy of the 3D-printed HDPE vitrimer may be explained by the low degree of crosslinking (0.3% wt of crosslinkers) and decreased crystallinity due to the dynamic crosslinks, or decreased interdiffusion of polymer chains at the printed layers. However, further experiments are needed to confirm these theories. 

Annealing

   Figure 12 shows the 3D-printed HDPE vitrimer samples following the post-curing process. HDPE samples completely melted during the post-curing process; whereas HDPE vitrimer samples show better dimensional accuracy following the post-curing process, suggesting higher dimensional stability of the 3D printed HDPE vitrimers.

Figure 12: Post-cured 3D-Printed HDPE and HDPE Vitrimer samples. Top left to right: 0°,45°, and 90° HDPE. Bottom left to right: 0°,45°, 90°, and 90° HDPE Vitrimers.

Figure 13 shows the frequency sweep results of post-cured 3D-printed HDPE vitrimer samples in addition to those shown in Figure 10. It can be seen that post-curing raised E* of  3D-printed HDPE vitrimers to the same level as the compression-molded HDPE vitrimers across the entire frequency range. 

Figure 13: Frequency Sweep Curves for Compression-molded, 0° and 45° 3D-printed HDPE and HDPE Vitrimer Samples, and Post-cured 0° and 45° 3D-printed HDPE Vitrimer Samples.

As shown in Figure 14, the difference in E* between the two printing orientations at 100 Hz reduced to 43MPa following post-curing. Both printing orientations show similar values of E* to the compression-molded HDPE vitrimer. The 45° printing orientation even shows a 24 MPa increase in E* relative to the compression-molded sample.

Figure 14: Comparison of E* in Compression-molded and 3D-printed HDPE and HDPE Vitrimers, and post-cured 3D-printed HDPE Vitrimers at 100 Hz.

3D-printed HDPE vitrimers showed better dimensional stability,  improved mechanical properties, as well as decreased mechanical anisotropy following the post-curing process. This suggests that a combination of 3D printing and post-curing processes may be a promising route to produce functional 3D-printed HDPE vitrimer parts. However, better post-curing parameters are needed to improve the dimensional accuracy of post-cured HDPE vitrimer parts. 

Conclusions

In this work, a screw-assisted 3D printer was utilized to print HDPE and HDPE vitrimer samples that were used to quantify shrinkage, warpage, and viscoelastic properties of both materials.  The optimal printing parameters included 60°C bed temperature, 220°C nozzle temperature, print bed material of 1/16’’ thick PP sheet, and 800mm/min printing speed for HDPE or 500mm/min for HDPE vitrimers. It was demonstrated that HDPE vitrimer samples had 13.148% lower volumetric shrinkage and lower warpage compared to HDPE. Therefore, HDPE vitrimers proved to be the ideal material choice that will result in better geometric accuracy of printed parts. Finally, viscoelastic measurements show that HDPE vitrimers led to higher mechanical anisotropy. However, the annealing process improved the mechanical properties of HDPE vitrimers to levels comparable to the compression-molded samples and led to decreased mechanical anisotropy. The annealing process was not possible for HDPE due to their low dimensional stability at higher temperatures. Future work will include testing the samples in the 90-degree orientation, finding better annealing parameters as well as analytically calculating warpage, and obtaining a more accurate value of volumetric shrinkage.

References

[1] Spoerk, Martin, et al. “Material Extrusion?Based Additive Manufacturing of Polypropylene: A Review on Ho]w to Improve Dimensional Inaccuracy and Warpage.” Journal of Applied Polymer Science, vol. 137, no. 12, 2019, p. 48545., https://doi.org/10.1002/app.48545.

[2] Van Zee, Nathan J., and Renaud Nicolaÿ. “Vitrimers: Permanently Crosslinked Polymers with Dynamic Network Topology.” Progress in Polymer Science, vol. 104, 2020, p. 101233., https://doi.org/10.1016/j.progpolymsci.2020.101233.

[3] Jurowska, Anna & Jurowski, Kamil. (2015). Vitrimers – The miracle polymer materials combining the properties of glass and plastic?. CHEMIK nauka-technika-rynek. 69. 392-394.

[4] Justino Netto, J.M., Idogava, H.T., Frezzatto Santos, L.E. et al. Screw-assisted 3D printing with granulated materials: a systematic review. Int J Adv Manuf Technol 115, 2711–2727 (2021). https://doi.org/10.1007/s00170-021-07365-z

[5] Montoya Ospina, Maria Camila & Osswald, Tim. (2020). Comparison of additive and conventional tooling on injection molded part properties.