Abstract

This study focuses on evaluating the material properties of recycled PET-CF15 when used for 3D printing. The study involves the shredding of used 3D printing prototypes and extrusion into new filament. The recycled material was both single and double recycled before samples were created from the different filaments. Material properties, such as tensile strength, notched impact strength, flexural modulus and temperature resistance, were analysed extensively. While in the case of tensile strength, the influence of the recycling passes has much less impact than, for example, the position of the model during printing, a significant deterioration was observed in the case of notched impact strength after each pass. Flexural modulus and temperature resistance also change significantly with each recycling pass. These findings offer valuable insights into the use of recycled PET CF15 filament in 3D printing and support targeted adaptation to specific requirements.

1. Introduction

The present research project aims to determine the optimised process parameters for the production of new filament from worn-out 3D printed prototypes made of PET-CF15. The investigation covers the entire recycling process from grinding to the finished filament, but also the 3D printing of samples from both original BASF PET-CF15 and thermomechanically recycled filament. With these parts, a detailed comparative analysis of the material properties is carried out. For this purpose, several mechanical tests are carried out, including notched bar impact, tensile and 3-point bending tests as well as a temperature test.

This study also provides some in-depth insights into the effects of the thermomechanical recycling process. Some of the material was processed twice to detect trends of changes in material properties after repeated recycling. Also, a variant of the recycling process is observed, in which the material is processed at an increased rate. This is to give an indication, whether a shortened exposure to heat and pressure can reduce the degeneration of the material properties after recycling.

The overall objective of this study is to provide sound knowledge on the sustainable recycling of PET-CF15 materials and to identify suitable areas of application based on the determined new material properties of the recycled material. Furthermore, potential optimisations of the process parameters are to be determined in order to achieve the best possible mechanical properties of the recycled filament.

Filament production - QiTech

2. Production of recycled filament

Production is carried out with the following equipment on a laboratory scale: First, the material is ground into granules with a side mill (JARVIS Pro Shredder). The laboratory drying unit "Digicolour Drywell 2540" is used for drying. The extruder (JARVIS Pro Extruder) then serves as the central unit by melting plastic granulate and pressing it through a nozzle to form the desired filament. The filament then passes through an air cooling section (2x JARVIS Airpath) to ensure rapid cooling and stabilization of the diameter.
‍
The pulling unit (JARVIS Winder) takes over the task of drawing off the cooled filament at a suitable speed to achieve the desired diameter, which is continuously recorded by the laser measuring unit (JARVIS DRE). After this step, the filament produced is wound onto spools or rolls using the winder to make it ready for use in the 3D printer and further material analysis.

2.1. Starting material

The original material used consisted of a collection of about 80 worn sliding pieces with different dimensions and filling patterns. The sizes of the individual parts ranged from 250x200x35 mm to smaller holders with dimensions of 100x30x30 mm. All parts had a plate shape with a slight elevation on one side, and some were equipped with through holes. The filling density varied from 20 % to 60 %. A few prototypes with dimensions slightly different from those mentioned above were also included.

sliding pieces of different sizes, on the right a larger prototype

2.1.1. Contamination

The material has minimal to negligible contamination. It was stored under dust-free conditions at Welser Profile GmbH. On about 20 % of the parts, there is a visible debris of metallic particles on the surface of the components, due to their use as sliding parts. Another visually visible contamination is the marking of the parts, which is either in the form of laser inscription or handwritten part number with white text marker.

impurities on 3D printed parts. Laser and highlighter marking (Right) and metallic abrasion (Left)

2.2. Grinding/Shredding

For the shredding, the JARVIS PRO Shredder was used, a granulator with a scissor-shear cutting principle. (in fact, a predecessor model of the current series device was used). As usual with every material change in the production chain, in order to avoid contamination of the material sample, the shredder was thoroughly cleaned using an industrial hoover and compressed air, and a 5 mm sieve was also placed in the mill. As an additional precaution, the first kilogram of shredded PET-CF was evacuated separately to ensure that there were no particles of the previous material (Qi-Tech PETG) in the grinding chamber.

on the left, the granulator used, the predecessor of the JARVIS Pro Shredder

2.2.1 Comminution process and challenges

The shredding took about 1 hour. This was due, among other things, to the simultaneous creation of image material. The plate-shaped sliding pieces with a depth of 30 mm could be crushed effortlessly, so that they would be suitable for a fully automatic recycling process without any problems. One of the last pieces, a sliding piece with a depth of 35 mm, required manual intervention. It came to a mill stop due to overloading. After the grinding chamber was opened, the cut material could be removed. The increased toughness of this component could be due to the higher interior filling. After restarting the mill and reinserting the component, it could be further crushed without any problems.

the blocking part

2.2.2. Granate quality

After the first comminution, the resulting granulate was additionally passed through the mill twice. This served to increase homogeneity and to further comminute longer pieces. When comparing samples of the material that were comminuted once with those that were comminuted three times, there is a clear difference in size and homogeneity. Repeated comminution takes less than 10 minutes and achieves comparable results to using a smaller 3 mm special sieve.

material: 1x crushed - 2x crushed - 3x crushed

2.3. Drying

PETG has hygroscopic properties and attracts moisture from the ambient air. Even small amounts of moisture can lead to irregularities in the filament (bubble formation) due to evaporation in the production process.

2.3.1 Structure

the pellet dryer used - here mounted directly on the extruder

For drying, the laboratory drying unit "Digicolour Drywell 2540" is used to precisely control the drying time, temperature and dew point. Digicolour's material drying overview contains precise drying settings for optimal processing of different materials, here we have chosen the parameters recommended for the new production of PETG to ensure that we achieve at least the necessary degree of drying. Accordingly, we choose 6 hours at 75 degrees Celsius (dew point -20 degrees Celsius). The dried material is then filled into the extruder hopper. To simulate a completely rewetting-free production line, the hopper was always closed with a protective bonnet during the tests. The visual inspection of the resulting filament suggests that the safeguarding of the processing process against moisture from the ambient air was sufficiently ensured in our environment.

extruder hopper with cover

drying parameters overview

2.4. Extrusion

The JARVIS Filament Line is used to extrude the material into 3D printing filament. In preparation for filament production, the extrusion unit is rinsed with 500g of cleaning material.

cleaning the extruder

2.4.1. Experiment 1: 8 mm nozzle

most important parameters of the 1st test

In the first experiment, extrusion was carried out using an 8 mm nozzle at elevated temperatures. While a nozzle with a diameter of 115 % to 170 % of the filament diameter is common in the production of filament from homogeneous plastics, an 8 mm diameter nozzle was initially used here to counter the risk of the nozzle becoming clogged with fibres, which experience has shown to be the case with recycled granulate from various materials.

The filament produced was brittle and the extruder output very uneven. This observation suggests that the selected temperatures may not have been sufficient to achieve optimal melt and material homogeneity. The conclusion from this experiment points to the need to increase the extrusion temperature.

Output of the material using the 8 mm nozzle

2.4.2. Experiment 2: Extrusion with 8 mm nozzle and high temperatures

most important parameters of the 2nd test

In the second experiment, extrusion was carried out using an 8 mm die at elevated temperatures. Despite an improved tensile strength and some stabilisation of the melt flow, the resulting filament turned out to be even more brittle. On closer inspection, a continuous central hole was observed in the structure when the filament was cut. This observation indicates internal stresses or too rapid cooling, but requires further investigation to determine the exact cause.

Central hole in filament

2.4.3. Experiment 3: Extrusion with lower temperatures

most important parameters of the 3rd experiment

The third trial involved extrusion using an 8 mm die at low temperatures. The resulting extruder output was very uniform, but had an almost stiff consistency already at the die exit and could not be pulled to the 1.75 mm diameter as usual. This finding suggests that reducing the nozzle diameter could optimise the process. This would require less stretching of the filament after exiting the nozzle, which in turn could reduce the need for higher temperatures.

result with lower temperatures

2.4.4. Experiment 4: 3 mm nozzle & heating the nozzle holder

Most important parameters of the 4th experiment

In the fifth experiment, extrusion was carried out using a smaller 3 mm nozzle at moderate temperatures. The smaller nozzle proves to be significantly more effective as it allows for increased pressure build-up and, unlike the larger nozzle, is less susceptible to severe cooling. Previously, higher temperatures were required to keep the material in a molten state until it reached the nozzle. By implementing the new heating of the Jarvis extruder's die holder, an overall reduction in the temperature of the rear heating elements was achieved. This resulted in an overall homogeneous output and texture of the material. The filament produced still stood out as brittle compared to conventional PLA/ABS, but now to a comparable extent to the new material. The diameter remained very consistent at +/- 3% around 1.75 mm, and a total of 4 kilos were produced with these parameters.

2.4.5. Experiment 4: 3 mm nozzle with pelletized material (2 times recycled)

most important parameters of the 5th experiment

In the fourth experiment, extrusion was carried out using a 3 mm die with material that had previously been recycled once. The aim was to investigate whether the rate of change of the material properties remains constant after recycling twice.

Usually, an improved diameter consistency can be observed in the extrusion process with pellets, as pellets are drawn in more homogeneously by the extruder screw. In contrast, no clear improvement in t h e diameter consistency of the resulting filament could be observed in this test. It cannot be ruled out that this was a counteracting effect, as the pellets themselves were not quite uniform in size. We had to fall back on the non-uniform filament from the parameterisation phase for pelletizing here, as only a limited amount of material to be recycled was available and the filament suitable for the printer was needed in sufficient quantities for the production of the measuring parts.

2.4.6. Experiment 6: 3 mm nozzle with increased speed

most important parameters of the 6th experiment

It was investigated whether a shortened exposure to heat has a positive effect on the material properties after recycling. A successful process could also be established here. Especially during winding, the material was warmer because it was exposed to air cooling for less time. This reduced the risk of rapid breakage of the filament strand during threading into the spool.

2.5 Summary results of the filament production trials

The result of the series of tests carried out includes several samples obtained during the investigation. Two spools, each weighing 4.5 kilograms, were produced from recycled PET-CF. Another 1 kilogram spool contained double recycled PET-CF. In addition, a 1 kilogram spool of recycled PET-CF was produced at an increased speed to explore the effects of exposing the material to higher speeds. The collected samples and coils provide a basis for the detailed analysis of the material properties and their changes during the recycling process.

Professional filament analysis - Jantec

3. Analysis of material properties

In order to determine to what extent the components made from the recycled materials can be loaded later and what influence the recycling process has on the mechanical properties, various tests were carried out under constant conditions.

For this purpose, matching test specimens were printed on a Caribou MK3S 3D printer.

specifications of the used 3D printer

settings for the pressure of the test samples

3.1. Tensile test

A tensile test can be used to investigate the strength of plastics. It is crucial to evaluate the performance of plastics in applications that are subjected to tensile loads.

3.1.1. Structure (Tensile test)

setup of the tensile test

For this purpose, a manual crank test stand was equipped with a load cell in order to be able to measure the tensile force exerted on the sample in newtons. The measurement is made via an HX711 signal amplifier on an Arduino UNO. This plots the measured tensile force over time. At the high point of this curve, the maximum tensile force until breakage can be read off. Divided by the fracture area of the sample used, the tensile strength is calculated in MPa.

The specimens used for this test have two holes to clamp them in the tensile test. The force is directed onto a bar that has a cross-sectional area of 4 mm * 4 mm (= 16 mm^2). This specimen can then either lie flat (in the case of the printer "in XY direction") printed in order to exert the force along the printed lines, which is then decisive for the tensile strength of the material. Or the sample can be printed upright (i.e. "in the Z-direction"), whereby the layers stacked by the printer are torn apart again during the test. In this case, the layer adhesion determines the tensile strength and is an important material property, especially in FDM 3D printing.

construction drawing sample tensile test

3.1.2. Results Tensile test XY

results of the tensile test in XY direction as bar chart

The measurement data shows that the deterioration in tensile strength for simply recycled material is within the measurement error tolerance of the new BASF PET CF. . The tensile strength of the twice-recycled PET CF, on the other hand, decreases by about a third - but is still far above the tensile strength of workpieces where layer adhesion plays a role.

3.1.3. Results Tensile test Z

results of the tensile test in Z direction as bar chart

No significant differences can be found in the generally lower tensile strength in the Z-direction, as a measure of layer adhesion.

3.2. Notched bar impact test

A notched impact test is used to evaluate the toughness and impact strength of a material. To do this, a notched specimen is cut with a sudden impact load.

3.2.1. Structure (Notched bar impact test)

In the set-up used, a hammer (mass 300 g) swings. This is dropped at a fixed height and smashes the clamped sample in the process. This absorbs part of the potential energy of the hammer. The tougher the material the sample is made of, the more energy is absorbed.

The absorbed energy is calculated by measuring the height of the hammer, first the height that the hammer reaches when there is no sample in its path (zero height, roughly corresponding to the height at which the hammer was released) and then the braked height after the sample has been shattered. The energy that can be calculated from this, divided by the fracture area of the sample, gives the impact strength in MPa.

notched bar impact test set-up

3.2.2. Results (Notched bar impact test)

results notched bar impact test as bar chart

It can be seen that the impact strength of the recycled materials decreases significantly.

3.3. 3-point bending test

With a 3-point bending test, the stiffness, in other words the resistance to loads, of the materials can be examined. The comparable value here is the bending modulus in MPa. The higher this value, the lower the deformability.

3.3.1. Structure (3-point bending test)

In this test, a bending beam (specimen construction, see figure 19) lies o n two points at a distance of 100 mm. Weights are attached in the centre via a ring. These exert a force on the specimen and thus bend it.
‍
The bend can then be measured very precisely using a precision dial gauge on the ring.

Since the force applied to the specimen is freely selectable and thus known, the flexural modulus can be easily calculated.

3-point bending test set-up

construction drawing of the bending beam

3.3.2. Results (3-point bending test)

results 3-point bending test as bar chart

The bending modulus decreases per recycling step.

3.4. Temperature test

Because plastics are thermoplastics, they start to soften at higher temperatures until they eventually melt - as they do during 3D printing and recycling. Especially for components that are used in warmer environments, such as outdoor or close to electronic components, it is important to know at what temperature they can become soft and change their shape.

3.4.1. Structure (Temperature test)

In this test, the sample (same construction as the sample in the 3-point bending test) lies on two points in the oven and is weighted down with nuts (9 g per sample). The temperature of the oven is measured. If the material now softens and starts to buckle due to the weight, the maximum temperature can be read at which the material becomes so soft that it changes shape - in this case the sample buckles downwards by 1 mm.

temperature test setup

3.4.2. Results (Temperature test)

temperature test results as bar chart

It can be seen that the temperature resistance of the material decreases per recycling pass.

3.5. Summary of the results

The results of this study provide important insights into the material properties of recycled PET CF15 filament:

The tensile strength of the material decreases by about 30 % in the XY direction during the second recycling pass, while it remains almost constant in the Z direction. The significant deterioration in the XY direction could be due to a change in the fibre structure (arrangement, quantity and size of the carbon fibres).

It is clear that the orientation of the layers has a significantly greater influence on tensile strength, with differences of up to 400%, compared to the influence of material thickness loss due to recycling. In applications where tensile strength is critical, this emphasises the importance of layer orientation, more specifically orientation on the printer, over the choice between virgin and recycled material.

The notched impact strength of the recycled filament decreases by about 40% with each recycling pass. This could again be attributed to changes in the fibre structure. For applications with short, intensive or impact loads, the suitability of the recycled material should be carefully considered.

Furthermore, there is a slight decrease in the flexural modulus, which leads to a certain flexibilisation of the material. This property could be advantageous in certain applications that require a certain elasticity. At the same time, it should be noted that the increased stiffness often desired from CF materials is no longer fully present in the recycled material.

The temperature resistance of the recycled PET CF15 filament drops by about 10 °C in each of the first two recycling passes. This is particularly relevant for temperature-sensitive applications. Nevertheless, the temperature resistance up to 70° of the recycled material still clearly exceeds that of conventional PLA or PETG.

The findings from this study have the potential to influence the design and application of recycled PET CF15 filament. They provide valuable information for the development of future material strategies and contribute to the ongoing optimisation and application of recycled materials in 3D printing.

4. Further questions  / Perspectives

In terms of future research and development, several key areas for further investigation emerge from the trials to date:

Furthermore, further parameters can be optimised through more in-depth and precise tests on the mechanical properties. One example is the nozzle temperature at which the material is melted by the 3D printer. This can, for example, have a significant influence on the layer adhesion and thus tensile strength in the Z-direction.

These future investigations will help to further optimise the process for manufacturing 3D printing filament and generate new findings for practical use as well as for scientific research.

5. Acknowledgement

Our sincere thanks go to Welser Profile GmbH for taking the initiative to strive for a more efficient use of materials and for funding this study. Special thanks go to Stefan (YouTube: CNC Kitchen), as a source of inspiration for the scientific discussion of 3D printing topics as well as for the training in the field of material testing and sample production with the 3D printer.