Optimizing recycled PET 3D printing using Taguchi method for improved mechanical properties and dimensional precision

Hiba Mohammad Hafiz; Ans Al Rashid; Muammer Koç

doi:10.1515/epoly-2025-0019

Artikel Open Access

Optimizing recycled PET 3D printing using Taguchi method for improved mechanical properties and dimensional precision

Hiba Mohammad Hafiz , Ans Al Rashid und Muammer Koç

Veröffentlicht/Copyright: 13. September 2025

Veröffentlicht von

Veröffentlichen auch Sie bei De Gruyter Brill

Manuskript einreichen Informationen für Autor*innen

Aus der Zeitschrift e-Polymers Band 25 Heft 1

Abstract

Rapid plastic consumption has increased global plastic waste, with polyethylene terephthalate (PET) widely used in packaging. Recycling PET is challenging, as conventional methods often fail to manage the growing waste stream. Additive manufacturing (AM) offers a way to repurpose PET waste into useful products. Preserving the mechanical and physical properties of recycled PET (rPET) during printing is essential. This study used a design of experiments approach to optimize AM parameters, including nozzle temperature, bed temperature, and print speed. Mechanical tests evaluated tensile strength, Young’s modulus, and failure strength. Results revealed optimal conditions that improve part quality and performance. The study identified optimal printing parameters that significantly enhance both dimensional accuracy and mechanical properties of rPET parts, validating their suitability for functional AM applications and advancing circular economy efforts.

Keywords: recycled PET; additive manufacturing; 3D printing; printing parameters; dimensional accuracy; mechanical properties

1 Introduction

In recent years, the extensive population growth has led to uncontrolled waste production associated with the use of plastics for packaging and disposable products, posing major issues for waste management and disposal (1–3). Waste management is particularly challenged by the unregulated waste stream originating from different sources (2,4,5). Most three-dimensional (3D) printer filaments are generated from virgin plastic, and the usage of 3D printing (3DP) materials is predicted to exceed 250 million pounds by 2020 (6). Apart from virgin plastic, a raw material for 3DP, the US discards 33.6 million tons of plastic waste annually, of which only 6.5% is recycled (7). The sustainability of plastic bottles from the waste stream is limited since it is impossible to reconstruct recycled plastic bottles into new bottles without using virgin material. Americans purchase 42.6 billion single-serving (1 L or fewer) plastic water bottles annually. A staggering 300 million are littered on highways, beaches, or waterways, and nearly eight out of ten end in landfills or incinerators (8).

Polyethylene terephthalate (PET) or glycol-modified PET (PETG) is a thermoplastic polymer that falls under the category of polyesters. It is widely used to make food and drink containers and synthetic fibers. Apart from its cost-effectiveness, PET polymer is valued for its combination of desirable properties such as strength, stiffness, lightweight nature, dependable resistance to deformation over time, and ability to withstand chemicals. These factors contribute significantly to its widespread utilization (2,9). PET exhibits a variable crystalline structure, ranging from amorphous to semi-crystalline, influenced by parameters such as cooling levels and stretching. While terephthalic acid and ethylene glycol make up the majority of PET, 1,4-cyclohexanedimethanol is added to replace around 30% of the diol moles to generate PETG. By reducing the crystallization rate and guaranteeing superior clarity, this alteration maintains the amorphous nature of PETG (10,11). There has been growing attention toward the extensive adoption of PET and PETG in beverage bottles, driven by their brief storage lifespan of about half a year and their common disposal after one-time use (12,13). Crystallinity plays a crucial role in shaping the physical and mechanical properties of PET. As a result, the recycling process for post-consumer PET (14,15) is not as straightforward as one might envision since it is a multidisciplinary activity that must include polymer chemistry, physics, processing, and manufacturing engineering (16). These characteristics, along with socioeconomic variables such as rapid consumption and insufficient recycling facilities at local scales, have resulted in PET and PETG containers being the most frequent plastic litter (17). The concept of a circular economy plays a critical role in addressing plastic pollution by shifting from a traditional linear model of “take, make, dispose” to a regenerative system aimed at minimizing waste and maximizing resource efficiency. In the context of PET waste, a circular economy approach involves designing recycling loops that allow materials to be reused multiple times in high-value applications. Additive manufacturing (AM), particularly through fused filament fabrication (FFF), aligns well with circular economy principles by enabling localized, on-demand production using recycled materials, thereby extending the lifecycle of plastics and reducing dependence on virgin resources. Plastic pollution is widely seen as a global hazard that crosses borders, motivating several governments to pledge to boost their regional recycling practices (18,19). The most cost-effective strategy to decrease PET waste is through recycling processes (20–22). In contrast, new and less expensive PET recycling techniques benefit the PET recycling industry by offering affordable PET, given that the price of virgin PET remains stable (14).

In recent years, introducing modern technologies such as AM has provided a possible alternative for recycling polymers. AM, often known as 3DP, is a production technology that uses the fusing of components to build items, generally stage by stage, from 3D modeling information (23,24). AM has various benefits over conventional production techniques, such as the capacity to build intricate multipurpose objects, reduce plastic waste, customize and personalize manufacturing, simplify distributing channels, and be cost-effective for short manufacturing cycles (25,26). Material extrusion, which comprises FFF, appears as a promising method among various manufacturing techniques since it allows for the direct use of post-consumer plastics in 3DP (27).

The escalating plastic waste crisis, particularly driven by the extensive consumption of PET, highlights the urgent need for sustainable recycling strategies. In response to this crisis, there is a pressing need for innovative recycling solutions that can both reduce environmental impact and create value from waste materials. AM offers a viable pathway for converting post-consumer PET into functional products. However, a key challenge is ensuring that the mechanical and physical of recycled PET (rPET) remain robust after conversion. Additionally, a comprehensive investigation into the effects of AM process and process parameters on the quality and durability of rPET-printed products is essential to confirm their suitability for practical applications as shown in Figure 1. Despite the environmental imperative to recycle PET waste, existing recycling methods fail to fully recover material quality, and the integration of rPET in AM remains limited by unresolved challenges related to mechanical performance and print quality. This study aims to optimize AM printing parameters for rPET and assess the quality of the resulting components, addressing the pivotal question: Can rPET be effectively repurposed in AM while maintaining its durability and functionality?

Figure 1

Flow diagram illustrating a closed-loop circular economy model for plastic waste.

In this study, the main focus is on recycling discarded PET bottles using the FFF technique and optimized the 3DP process parameters. It is critical to analyze the influence of printing parameters on the overall performance of 3D-printed products. This research is part of our ongoing analysis, in which all components of the 3DP workflow were extensively analyzed, including material quality, printing parameters, and product design. This study investigates the effect of nozzle temperature, print bed temperature, and print speeds on the dimensional stability of rPET 3D-printed components. To demonstrate this, the design of experiments (DOE) strategy was used to determine the effects of certain process conditions, i.e., nozzle temperatures (255°C, 265°C, and 275°C) within PET’s printability window, as well as variable print bed temperatures (65°C, 70°C, and 75°C) and print speeds (50, 55, and 60 mm·s⁻¹). These parameters were then optimized to achieve ideal component quality, first by modifying the nozzle and print bed temperatures and then by fine-tuning print rates while maintaining the optimal temperatures. Dimensional accuracy and tensile properties are evaluated as part of the quality assessment of the 3DP process. The primary objectives of this study are as follows:

To investigate the feasibility of using rPET in AM through the FFF technique.
To optimize 3DP process parameters – specifically nozzle temperature, print bed temperature, and print speed – using a Taguchi-based DOE approach.
To evaluate the mechanical performance and dimensional accuracy of rPET-printed components under optimized conditions.

2 Materials and methods

2.1 Mechanical recycling of PET

The process of mechanical recycling of PET began with the initial phase of preparing the filament for extrusion. Locally sourced milk bottles (Figure 2) primarily consisting of PET were used as feedstock material. After removing the caps and labels, these bottles underwent a cleaning process involving rinsing with pressurized warm water to dislodge debris or impurities. Additionally, a mild detergent was used to ensure thorough cleaning. After rinsing, the bottles were dried to eliminate moisture and prepare them for further processing.

Figure 2

PET bottle waste and filament extrusion setup for 3DP.

Once cleaned and dried, the PET containers were sliced into uniform pieces with a 3D-printed strip-cutting device. This strip-cutting device precisely sliced the bottles into uniform strips of 12 mm width. The strip cutter ensured consistent dimensions for the subsequent extrusion process. The processed strips of rPET were subsequently introduced into a filament extrusion setup, as shown in Figure 2, consisting of a modified volcano nozzle for the hot end, an automated extrusion mechanism, and a winding filament tool. Based on the findings of Al Rashid and Koç (28), who established a temperature range of 190–220°C as ideal for optimizing filament quality, this study selected a similar temperature window to enhance filament consistency and uniformity. Continuous monitoring of filament diameter was done to ensure that the quality standards were maintained and suitable for the next stage of 3DP.

2.2 3D printing of rPET

The prepared filaments as seen in Figure 3 underwent vacuum drying to eliminate moistness and were kept in sealed packaging to prevent moisture absorption. As per ASTM D638 (29), tensile testing samples were 3D-printed using an Ultimaker 3 Extended 3D printer, following ASTM D638 (Type IV) criteria, with an overall length of 115 mm, a width of 19 mm, a gauge width of 6 mm, and a thickness of 3 mm. The test sample models were sliced using Cura software, defining the 3DP process parameters. The samples were 3D-printed using a 0.4 mm nozzle, 0.2 mm layer height, 0.8 mm wall thickness, and 50% infill density. Three parameters were altered, and each variation was repeated three times, as indicated in Table 1. These parameters included the nozzle temperatures (255°C, 265°C, and 275°C), print bed temperatures (65°C, 70°C, and 75°C), and printing speeds (50, 55, and 60 mm·s⁻¹). The experimental design (DOE) matrix was employed to identify the optimal settings for achieving the highest print quality, as shown in Table 1.

Figure 3

rPET filaments.

Table 1

DOE approach on process parameters optimization of rPET

DOE	Nozzle temperature (°C)	Print bed temperature (°C)	Speed (mm·s⁻¹)	Overall length (mm)	Grip width (mm)	Gauge width (mm)	Thickness (mm)
1	255	65	50	114.73 ± 0.04	18.76 ± 0.01	6.05 ± 0.01	3.23 ± 0.03
2	255	70	55	114.74 ± 0.07	18.85 ± 0.05	6.11 ± 0.17	3.2 ± 0.12
3	255	75	60	114.71 ± 0.07	18.87 ± 0.03	6.07 ± 0.04	3.14 ± 0.18
4	265	65	55	114.7 ± 0.11	18.95 ± 0.03	6.12 ± 0.17	3.16 ± 0.13
5	265	70	60	114.85 ± 0.13	19 ± 0.2	6.16 ± 0.16	3.2 ± 0.02
6	265	75	50	114.82 ± 0.08	18.91 ± 0.15	6.16 ± 0.07	3.25 ± 0.02
7	275	65	60	114.84 ± 0.24	18.93 ± 0.12	6.19 ± 0.17	3.24 ± 0.05
8	275	70	50	114.71 ± 0.13	18.84 ± 0.14	6.08 ± 0.04	3.25 ± 0.02
9	275	75	55	114.73 ± 0.04	18.93 ± 0.03	6.13 ± 0.04	3.22 ± 0.06

2.3 Dimensional accuracy analysis

Following 3D printing, the measurements of all the samples were taken using a digital caliper (0.01 mm accuracy) and illustrated in Table 1 to analyze the precision in measurements of the 3D-printed samples. The average of three repeated measurements is calculated for each row of the DOE matrix. These results will be compared with the ASTM D638 (Type IV) criteria to determine the best parameters for 3DP of rPET.

2.4 Mechanical characterization

After dimensional accuracy analysis, rPET specimens (Figure 4) were mechanically tested according to ASTM (D638) mechanical testing at ambient temperature and a crosshead speed of 10 mm·min⁻¹. Specimen overall dimensions were around 114 mm × 19 mm, and thickness was in the range of 3.1–3.3 mm.

Figure 4

3D printed samples arranged according to DOE process parameters.

3 Results and discussion

3.1 Filament quality

Filament extrusion was performed at different extrusion temperature ranges between 180°C and 210°C. It was determined that an extruder temperature of 210°C was ideal for uniform extrusion, culminating in filaments with uniform diameters of 2.85 mm. This temperature setting facilitated smooth material flow through the nozzle and minimized clogging to ensure uniform filament formation. The consistency in filament diameter was crucial in maintaining the print quality and dimensional accuracy during 3DP. Conversely, the material tended to overheat at higher temperatures, such as 220°C or above, leading to excessive melting and deformation, adversely affecting print quality.

3.2 Dimensional analysis

The optimal printing conditions for rPET were determined using a statistical experimental approach known as the Taguchi orthogonal array DOE, as shown in Table 1. This approach utilized a carefully chosen fractional subset of configurations of various factors, resulting in nine experimental trials, each repeated three times. The dimensional analysis results were compared with the ASTM standard for printing rPET specimens, with an overall length of 115 mm, a width of 19 mm, a gauge width of 6 mm, and a thickness of 3 mm. In this section, the influence of printing parameters on each dimension will be discussed.

3.2.1 Overall length

The interaction plots for the overall length of 115 mm of rPET specimens are reported in Figure 5. The targeted overall length as designed in the CAD model was 115 mm. The nozzle temperature of 265°C, print bed temperature of 70°C, and printing speed of 60 mm·s⁻¹ were found optimum for the overall length (114.85 mm) to be the closest to the design length. In contrast, it was revealed that the nozzle temperature of 265°C, print bed temperature of 65°C, and printing speed of 55 mm·s⁻¹ resulted in the highest deviation in the overall length from the target value due to shrinkage, i.e., 114.7 mm.

Figure 5

Interaction plot for overall length.

From Figure 5, it is observed that the nozzle temperature of 275°C and print bed temperature of 65°C showed comparable results to the results shown by the optimum parameters. Ideal results of an overall length close to 115 mm can be achieved by balancing the temperature settings of nozzle temperature and print bed temperature. If the nozzle temperature is higher (275°C), then the print bed temperature should be less (65°C), if the print bed temperature is higher (70°C), then the nozzle temperature should be kept relatively lower (265°C). Printing speeds of 60 and 50 mm·s⁻¹ give comparable results when nozzle temperature is varied from 255°C to 265°C. However, after reaching 265°C, the printing speed of 50 mm·s⁻¹ drastically gives deviations in the overall length. Printing speed of 60 mm·s⁻¹ remains ideal if the print bed temperature is kept constant at 70°C.

3.2.2 Grip width

The interaction plots for the grip width of rPET specimens are reported in Figure 6. As designed in the CAD model, the targeted overall grip width was 19 mm. Nozzle temperature, print bed temperature, and printing speed (265°C, 70°C, and 60 mm·s⁻¹) resulted in the grip width of the exact design dimension (19 mm). In contrast, nozzle temperature, print bed temperature, and speed (255°C, 65°C, and 50 mm·s⁻¹) are printing parameters that gave the least optimum results (18.76 mm). A general trend that is observed for all nozzle temperatures (255°C, 265°C, and 275°C) is that with the increase in printing speed, the dimensional error in grip width is reduced. The reduction in dimensional error with the increase in print speed likely occurs because faster speeds limit excessive material deposition and reduce thermal distortion. Similarly, at each constant speed (50, 55, and 60 mm·s⁻¹), there is a negative grip width deviation observed when the nozzle temperature exceeds 265°C.

Figure 6

Interaction plot for grip width.

3.2.3 Gauge width

The interaction plots for the gauge width of rPET specimens are reported in Figure 7. As designed in the CAD model, the targeted overall gauge width was 6 mm. The nozzle temperature of 255°C, print bed temperature of 65°C, and printing speed of 50 mm·s⁻¹ were found optimum for the gauge width (6.05 mm) to be the closest to the design dimension (6 mm). In contrast, the nozzle temperature of 275°C, print bed temperature of 65°C, and printing speed of 60 mm·s⁻¹ revealed higher deviations from the targeted gauge width of around 6.19 mm·s⁻¹. With increasing nozzle temperature from 255°C to 265°C, the dimensional error in gauge width increases. At the highest speed of 60 mm·s⁻¹ and increasing print bed temperature from 65°C to 75°C, the accuracy of gauge width dimension is observed to improve from 6.15 to 6.07 mm whereas, with the lowest speed of 50 mm·s⁻¹, and increasing print bed temperature from 65°C to 75°C, the accuracy of gauge width dimension is observed to decrease from 6.05 to 6.16 mm. Higher nozzle temperatures increase material fluidity, which can cause over-extrusion and dimensional deviations, while increasing print bed temperature at higher speeds improves adhesion and reduces warping, leading to better dimensional accuracy.

Figure 7

Interaction plot for gauge width.

3.2.4 Thickness

The interaction plots for the thickness of rPET specimens are reported in Figure 8. The targeted thickness dimension, designed in the CAD model, was 3.2 mm. The nozzle, print bed temperatures, and print speed (265°C, 70°C, and 60 mm·s⁻¹) were found optimum for the thickness to be exact as the design dimension of 3.2 mm. In contrast, nozzle temperature of 255°C, print bed temperature of 75°C, and printing speed of 60 mm·s⁻¹ are printing parameters that showed the least optimum results of 3.14 mm thickness. The printing speed range of 55–60 mm·s⁻¹ does not affect the dimensional accuracy of thickness when the print bed temperature remains constant at 70°C. The low printing speed of 50 mm·s⁻¹ increases the thickness of the specimen with increasing temperature ranges of the print bed and nozzle temperatures. Optimal nozzle and bed temperatures at higher speeds promote uniform melting and deposition, ensuring dimensional accuracy, whereas low speeds combined with higher temperatures cause excessive material flow and slight thickness variations due to over-extrusion.

Figure 8

Interaction plot for thickness.

3.2.5 Discussion

This section examines the observed trends in how varying process parameters influence the key dimensions of the 3D-printed rPET samples and highlights the correlations. The impact of nozzle temperature can be observed by analyzing the variations in other parameters while keeping the print bed temperature constant. When the print bed temperature is constant at 65°C, and the nozzle temperature increases from 255°C to 275°C, an increasing trend is noticed in the gauge width and thickness of the samples. Similarly, when the print bed temperature is fixed at 70°C, an increasing trend in the thickness of the samples is observed. When the print bed temperature is fixed at 75°C, the overall length of the samples increases as well. This indicates that at a constant print bed temperature, higher nozzle temperatures increase the gauge width and thickness of the printed samples. When the printing speed is constant at 50 mm·s⁻¹, and the nozzle temperature increases from 255°C to 275°C, an increasing trend is noticed with thickness of the samples. When the printing speed is constant at 55 mm·s⁻¹, and the nozzle temperature increases from 255°C to 275°C, an increasing trend is noticed with the gauge width and thickness of the samples. This suggests that a moderate printing speed combined with a higher nozzle temperature may lead to better control over the width and thickness of the printed components.

The print bed temperature appears to have a moderate effect on the grip width and gauge width of the samples. For instance, comparing runs 1, 2, and 3, where the print bed temperature increases from 65°C to 75°C while keeping nozzle temperature constant, there is a noticeable increase in both grip width and gauge width. Similarly, at a constant nozzle temperature of 265°C and increasing print bed temperature, an increasing trend is noticed in the overall length, gauge width, and thickness of the samples. Similarly, in experiments 1, 6, and 8, when the printing speed is constant at 50 mm·s⁻¹, and print bed temperature increases from 65°C to 75°C, an increase in the overall thickness, gauge width, and grip width of the samples is observed. These results suggest that higher print bed temperatures lead to an increase in the overall thickness, grip width, and gauge width of the samples, highlighting its moderate but consistent influence on dimensional properties.

The printing speed has a significant impact on the overall dimensions of the samples. As seen in experiments 7, 8, and 9, with a constant nozzle temperature of 275°C, the overall length, grip width, and gauge width increase with an increase in printing speed. Similarly, in experiments 4, 5, and 6 with a constant nozzle temperature of 265°C, the grip width of the samples increases with an increase in printing speed. Similarly, when the print bed temperature is held constant at 65°C, as seen in experiments 1, 4, and 7, the overall length and gauge width of the samples increases with an increase in printing speed. Based on these observations, it can be concluded that increasing the printing speed consistently leads to an expansion in the overall length, grip width, and gauge width of the polymer samples. This trend is evident regardless of variations in nozzle and print bed temperatures, indicating that printing speed plays a key factor in deciding the dimensional accuracy and consistency of the printed polymer samples.

3.3 Optimization of dimensional accuracy

This section presents a comparative analysis of dimensional accuracy based on the measured deviations in overall length, grip width, gauge width, and thickness for each experimental run. As summarized in Table 2, the results show a range of dimensional deviation from standard dimensions in percentage, across the nine DOE configurations, influenced by variations in process parameters such as nozzle temperature, print bed temperature, and printing speed. The percentage error deviation was calculated as the absolute difference between the measured value and the target value, divided by the target value, and multiplied by 100. Most dimensional deviations across the experiments remained relatively low, with overall length errors consistently under 0.26%, demonstrating that all process settings maintained acceptable longitudinal precision. Grip width deviations were generally under 1%, except in DOE 1 (1.26%), indicating a mostly stable lateral extrusion behavior. However, gauge width and thickness deviations varied more substantially, suggesting that fine-tuning is especially critical for these dimensions to ensure optimal material flow and layer consistency.

Table 2

Average dimensional deviation (%) calculated for each DOE

DOE	Overall length (mm)	Grip width (mm)	Gauge width (mm)	Thickness (mm)	Overall length (%)	Grip width (%)	Gauge width (%)	Thickness (%)
1	114.73	18.76	6.05	3.23	0.23	1.26	0.83	0.94
2	114.74	18.85	6.11	3.2	0.23	0.79	1.83	0.00
3	114.71	18.87	6.07	3.14	0.25	0.68	1.17	1.88
4	114.7	18.95	6.12	3.16	0.26	0.26	2.00	1.25
5	114.85	19	6.16	3.2	0.13	0.00	2.67	0.00
6	114.82	18.91	6.16	3.25	0.16	0.47	2.67	1.56
7	114.84	18.93	6.19	3.24	0.14	0.37	3.17	1.25
8	114.71	18.84	6.08	3.25	0.25	0.84	1.33	1.56
9	114.73	18.93	6.13	3.22	0.23	0.37	2.17	0.63

Among all experiments, DOE 5 emerged as the most dimensionally accurate configuration, with 0.13% error in overall length, 0.00% in grip width, and 0.00% in thickness. These values reflect a highly controlled and optimized thermal environment, resulting in excellent dimensional accuracy. Although DOE 5 recorded a slightly higher gauge width deviation (2.67%), its consistency across the other three critical dimensions confirms its superiority over other settings.

DOE 5 was printed at high nozzle temperature and print bed temperature. These observations are consistent with findings by Aslani and Kamil (30), who recommended high nozzle temperature to reduce dimensional error. Similarly, according to Aslani et al. (31), the nozzle temperature was found to be the most significant factor for the precision of polymer components by analysis of variance. The highest temperature was found to be optimum (230°C) for the rectangular specimen’s X and Y dimensions. A certain nozzle temperature range and print bed temperature range that ensures correct printing and flow rate is recommended to decrease the dimensional error. Furthermore, it is also observable that DOE 5 was printed at the highest speed of 60 mm·s⁻¹ among all the different configurations. These findings align with a study by Agarwal et al. (32), which showed a 4.5% variance at 65 mm·s⁻¹ print speed and an 8.57% variation at 30 mm·s⁻¹. Higher print speeds lead to better dimensional accuracy, with the lowest variance recorded at 65 mm·s⁻¹. This was an unforeseen conclusion; however, it was also discovered by Akbaş et al. (33).

3.4 Mechanical performance evaluation

This section presents the mechanical testing results of the 3D-printed specimens, focusing on stress–strain behavior, interaction effects, failure strength, and Young’s modulus. This analysis provides a detailed understanding of how the selected process factors influence the material’s mechanical characteristics.

3.4.1 Tensile strength

For each DOE, the optimal result from the three tests conducted was selected for comparison. Only one sample result from each DOE is shown in the stress–strain graph in Figure 9.

Figure 9

Stress vs strain curve for each DOE.

The experimental design (DOE) strategy was utilized to systematically evaluate the influence of various process factors on tensile strength. Among the different configurations tested, DOE 7 achieved the highest tensile strength at 32.26 MPa (Figure 10), closely followed by DOE 5, which recorded a tensile strength of 30.31 MPa. Both of these experiments were conducted at a printing speed of 60 mm·s⁻¹ and high nozzle temperatures in the range of 265–275°C. The results indicate that optimal tensile strength is obtained with nozzle temperatures ranging from 265°C to275°C, a printing speed of 60 mm·s⁻¹, and a print bed temperature of 65–70°C.

Figure 10

Comparative tensile strength of rPET specimens.

Similar results were reported previously by many researchers such as Alafaghani et al. (34) who carried out a study to examine the influence of nozzle temperature on the mechanical properties of polylactic acid (PLA) within the temperature range of 175–205°C. Their findings demonstrate that increasing the nozzle temperature from 175°C to 205°C results in a significant improvement of tensile strength, with an observed increase of over 53%, rising from 28.6 to 43.8 MPa. While Alafaghani et al. explored the impact of temperatures between 175°C and 205°C, this study extended this range to 265–275°C due to changes in the material. Despite the difference in temperature ranges, the observed trend in both studies suggests that higher nozzle temperatures contribute to enhanced tensile strength in polymers, reinforcing the critical role of thermal parameters in optimizing the mechanical characteristics of 3D-printed materials. In contrast, DOE 1 and DOE 2 (Table 3), characterized by a lower nozzle temperature of 255°C and printing speeds ranging between 50 and 55 mm·s⁻¹, demonstrated the poorest tensile strengths among all the tested configurations. These lower nozzle and print bed temperatures and speeds likely contributed to insufficient bonding between layers, resulting in weaker mechanical properties.

Table 3

Mechanical properties summarized for each DOE

DOE	Nozzle temperature (°C)	Print bed temperature (°C)	Printing speed (mm·s⁻¹)	Tensile strength (MPa)	Failure strength (MPa)	Young’s modulus (MPa)
1	255	65	50	17.48	0.07	272.65
2	255	70	55	19.21	0.06	401.3
3	255	75	60	25.47	0.06	520.95
4	265	65	55	26.97	0.06	526.51
5	265	70	60	30.31	0.07	526.57
6	265	75	50	28.57	0.06	592.77
7	275	65	60	32.26	0.06	683.73
8	275	70	50	23.38	0.05	509.44
9	275	75	55	24.98	0.14	211.63

The interaction plot analysis as shown in Figure 11 reveals that high nozzle temperatures (275°C) coupled with low print bed temperatures (65°C) yield tensile strengths exceeding 30 MPa.

Figure 11

Interaction plot for tensile strength.

However, as the print bed temperature increases, tensile strength decreases from 32 to 24 MPa as seen in DOE 7, 8, and 9, highlighting the importance of maintaining stable print bed temperature with high nozzle temperatures. Additionally, for nozzle temperatures of 255°C, 265°C, and 275°C, a general trend is observed where increasing the printing speed results in improved tensile strength.

Notably, a print bed temperature of 65°C emerges as an optimal option because tensile strength tends to increase with increasing nozzle temperature and print bed temperature. To achieve maximum tensile strength, a printing speed of 60 mm·s⁻¹ is ideal, provided that the print bed temperature is maintained at 65°C and the nozzle temperature ranges from 255°C to 275°C. In contrast, lower printing speeds, such as 55 mm·s⁻¹, result in inconsistent tensile strength, particularly at nozzle temperatures above 265°C.

3.4.2 Failure strength

Figure 12 illustrates the interaction of failure strength with three varying process parameters: nozzle temperature, print bed temperature, and printing speed. All three parameters – nozzle temperature, print bed temperature, and printing speed – affect the failure strength of 3D-printed specimens. Interaction plots depict how these parameters interact to influence the desired outcome. From Table 3, it is evident that the highest failure strength is attained at a nozzle temperature of 275°C coupled with the highest print bed temperature of 75°C. Furthermore, a printing speed of 55 mm·s⁻¹ has been identified as optimal for maximizing failure strength. At this constant printing speed, failure strength shows a proportional increase with the escalation of nozzle and print bed temperatures, highlighting the significant impact of these parameters on the material’s structural performance.

Figure 12

Interaction plot for failure strength.

From the results of DOE 1, 5, and 9 (Figure 13), which exhibit the highest failure strengths, a trend emerges: increasing the print bed temperature leads to higher failure strength. These findings align closely with those reported in other studies. For instance, Pulipaka et al. (35) identified platform temperature as a critical parameter affecting ultimate tensile strength (UTS). Their study found that a platform temperature of 150°C produced an optimal UTS of 49.650 MPa, highlighting the role of platform temperature in enhancing layer adhesion and print stability, thereby reducing warping and delamination. Similarly, Xiaoyong et al. (36) reported a clear correlation between platform temperature and tensile strength in polyether ether ketone material, where an increase in platform temperature from 25°C to 130°C led to a significant rise in tensile strength, from 52.7 to 71.2 MPa. The consistency across these studies underscores the critical influence of thermal factors on the mechanical characteristics of 3D-printed materials, with greater temperatures generally leading to improved strength and stability.

Figure 13

Comparative failure strength of rPET specimens.

3.4.3 Young’s modulus

The analysis presented in Figure 14 demonstrates that all three parameters, nozzle temperature, print bed temperature, and printing speed, have a direct influence on Young’s modulus. The highest Young’s modulus value of 683.73 MPa is observed at the maximum nozzle temperature of 275°C, print bed temperature of 65°C, and highest printing speed of 60 mm·s⁻¹. From the experiments DOE 4 and DOE 5, both of which have identical Young’s modulus values of 526 MPa, it is evident that the nozzle temperature set at 265°C which is constant in both cases has a more significant impact on Young’s modulus compared to variations in print bed temperature or printing speed.

Figure 14

Interaction plot for Young’s modulus.

Similarly, as seen from DOE 1, 4, and 7, with an increase in nozzle temperature from 255°C to 275°C, an increase in Young’s modulus of the samples is observed from 272 to 683 MPa, provided the print bed temperature is constant at 65°C. This suggests that nozzle temperature plays a more important role in influencing the material’s rigidity and overall mechanical properties.

These results are aligned with a research study (34) conducted on PLA, in which Young’s modulus increased by 54% on increasing the temperature from 175°C to 205°C. The data also revealed that at a constant nozzle temperature of 255°C, Young’s modulus increases proportionally with both printing speed and print bed temperature. Similarly, at a fixed print bed temperature of 65°C, Young’s modulus increases with increasing nozzle temperature and printing speed. The increase in Young’s modulus with higher nozzle temperatures can be attributed to improved interlayer bonding and polymer chain diffusion at elevated temperatures, resulting in enhanced rigidity. Additionally, higher printing speeds and bed temperatures contribute to reduced internal stresses and better layer adhesion, further strengthening the material’s mechanical properties.

3.4.4 Discussion

As shown in Figure 4, the specimens, shaped like dog bone were produced via 3D printing, with each variation repeated three times to ensure consistency. Through iterative testing, a range of nozzle temperatures (255°C, 265°C, and 275°C), print bed temperatures (65°C, 70°C, and 75°C), and printing speeds (50, 55, and 60 mm·s⁻¹) were identified as optimal for achieving satisfactory print quality. These temperatures played a significant role in determining the overall quality of the 3D print, affecting the glass transition, melting, and crystallization points of the materials used. This underscores the importance of precise temperature control in achieving desired printing outcomes. For high tensile strength and young’s modulus the parameters of DOE 7 were the most optimum with printing parameters of 275°C, 65°C, and 60 mm·s⁻¹.

4 Conclusions

Based on the experimental findings, it is evident that optimizing printing parameters for rPET in AM is crucial for advancing sustainable circular economy practices. This study successfully demonstrated the feasibility of reusing post-consumer PET for filament fabrication and 3D printing through the FFF technique. The optimal extrusion temperature of 210°C ensured consistent filament diameter and printability. Using a Taguchi DOE approach, combination of nozzle temperature at 265°C, print bed temperature at 70°C, and print speed of 60 mm·s⁻¹ were identified that yielded the best dimensional accuracy across all four key dimensions: overall length, grip width, gauge width, and thickness. The dimensional deviation in optimized settings (DOE 5) remained well within acceptable limits, with negligible error in grip width and thickness, confirming the reliability of the process conditions. The findings support the notion that printing speed, nozzle temperature, and bed temperature each have interdependent effects on dimensional fidelity. For mechanical performance, DOE 7 (275°C nozzle, 65°C bed, and 60 mm·s⁻¹) yielded the highest tensile strength and Young’s modulus. These findings confirm that rPET can deliver both high dimensional precision and mechanical reliability under optimized conditions. This study ensures that rPET components possess the necessary mechanical integrity and durability for their “second life.” Beyond demonstrating feasibility, this research contributes a structured optimization approach using the Taguchi DOE method to enhance the print quality of recycled polymers. The findings validate rPET’s potential in AM, supporting its role in advancing the circular economy and addressing plastic waste challenges. In conclusion, the outcomes of this research offer compelling evidence for the viability of integrating rPET into AM processes, contributing to circular economy and sustainability.

Acknowledgments

Hiba Mohammad Hafiz would like to acknowledge Graduate Sponsorship Award (GSRA) for their support through the grant (GSRA10-L-2-0530-23062) for her graduate studies. Open Access funding provided by the Qatar National Library.

Funding information: Authors state no funding involved.
Author contributions: Hiba Mohammad Hafiz: writing – review & editing, writing – original draft, methodology, formal analysis, data curation, experimental work, conceptualization. Ans Al Rashid: writing – review & editing, supervision, software, formal analysis. Muammer Koç: supervision, resources, project administration, funding acquisition.
Conflict of interest: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Data availability statement: The authors confirm that the data supporting the findings of this study are available within the article.

References

(1) Al Rashid A, Koç M. Additive manufacturing for sustainability and circular economy: needs, challenges, and opportunities for 3D printing of recycled polymeric waste. Mater Today Sustain. 2023;24:100529. 10.1016/J.MTSUST.2023.100529.Suche in Google Scholar

(2) Ferrari F, Corcione CE, Montagna F, Maffezzoli A. 3D printing of polymer waste for improving people’s awareness about marine litter. Polymers. 2020;12:1738. 10.3390/POLYM12081738.Suche in Google Scholar PubMed PubMed Central

(3) Ferrari F, Striani R, Minosi S, De Fazio R, Visconti P, Patrono L, et al. An innovative IoT-oriented prototype platform for the management and valorisation of the organic fraction of municipal solid waste. J Clean Prod. 2020;247:119618. 10.1016/J.JCLEPRO.2019.119618.Suche in Google Scholar

(4) Corcione CE, Ferrari F, Striani R, Visconti P, Greco A. An innovative green process for the stabilization and valorization of organic fraction of municipal solid waste (OFMSW): optimization of the curing process II part. Appl Sci. 2019;9:3702. 10.3390/APP9183702.Suche in Google Scholar

(5) Ferrari F, Striani R, Visconti P, Corcione CE, Greco A. Durability analysis of formaldehyde/solid urban waste blends. Polymers. 2019;11:1838. 10.3390/POLYM11111838.Suche in Google Scholar PubMed PubMed Central

(6) Laird K New filament aims to bring sustainability to 3D printing 2015. https://www.plasticstoday.com/3d-printing/new-filament-aims-to-bring-sustainability-to-3d-printing# (accessed May 13, 2024).Suche in Google Scholar

(7) Cohen D, Sargeant M, Somers K 3-D printing takes shape | McKinsey 2014. https://www.mckinsey.com/capabilities/operations/our-insights/3-d-printing-takes-shape# (accessed May 13, 2024).Suche in Google Scholar

(8) Container Recycling Institute. Plastic 2023. https://www.container-recycling.org/index.php/factsstatistics/plastic (accessed May 13, 2024).Suche in Google Scholar

(9) Wang CQ, Wang H, Liu YN. Separation of polyethylene terephthalate from municipal waste plastics by froth flotation for recycling industry. Waste Manag. 2015;35:42–7. 10.1016/J.WASMAN.2014.09.025.Suche in Google Scholar PubMed

(10) Bhagia S, Kore S, Wasti S, Ďurkovič J, Kováč J, Zhao X, et al. 3D printing of a recycled copolyester of terephthalic acid, cyclohexanedimethanol and tetramethylcyclobutanediol. Polym Test. 2023;118:107916. 10.1016/J.POLYMERTESTING.2022.107916.Suche in Google Scholar

(11) Bex GJP, Ingenhut BLJ, ten Cate T, Sezen M, Ozkoc G. Sustainable approach to produce 3D-printed continuous carbon fiber composites: “a comparison of virgin and recycled PETG”. Polym Compos. 2021;42:4253–64. 10.1002/PC.26143.Suche in Google Scholar

(12) Ros-Chumillas M, Belissario Y, Iguaz A, López A. Quality and shelf life of orange juice aseptically packaged in PET bottles. J Food Eng. 2007;79:234–42. 10.1016/J.JFOODENG.2006.01.048.Suche in Google Scholar

(13) Gerassimidou S, Lanska P, Hahladakis JN, Lovat E, Vanzetto S, Geueke B, et al. Unpacking the complexity of the PET drink bottles value chain: a chemicals perspective. J Hazard Mater. 2022;430:128410. 10.1016/J.JHAZMAT.2022.128410.Suche in Google Scholar PubMed

(14) Awaja F, Pavel D. Recycling of PET. Eur Polym J. 2005;41:1453–77. 10.1016/J.EURPOLYMJ.2005.02.005.Suche in Google Scholar

(15) Raheem AB, Noor ZZ, Hassan A, Abd Hamid MK, Samsudin SA, Sabeen AH. Current developments in chemical recycling of post-consumer polyethylene terephthalate wastes for new materials production: a review. J Clean Prod. 2019;225:1052–64. 10.1016/J.JCLEPRO.2019.04.019.Suche in Google Scholar

(16) Demirel B, Yaraș A, Elçiçek H. Crystallization behavior of PET materials. BAÜ Fen Bil Enst Derg Cilt. 2011;13:26–35.Suche in Google Scholar

(17) Hardesty BD, Roman L, Leonard GH, Mallos N, Pragnell-Raasch H, Campbell I, et al. Socioeconomics effects on global hotspots of common debris items on land and the seafloor. Glob Environ Change. 2021;71:102360. 10.1016/J.GLOENVCHA.2021.102360.Suche in Google Scholar

(18) Geyer R, Jambeck JR, Law KL. Production, use, and fate of all plastics ever made. Sci Adv. 2017;3. 10.1126/SCIADV.1700782.Suche in Google Scholar PubMed PubMed Central

(19) Arp HPH, Kühnel D, Rummel C, Macleod M, Potthoff A, Reichelt S, et al. Weathering plastics as a planetary boundary threat: exposure, fate, and hazards. Env Sci Technol. 2021;55:7246–55. 10.1021/ACS.EST.1C01512/ASSET/IMAGES/LARGE/ES1C01512_0005.JPEG.Suche in Google Scholar

(20) Yoda N. Corporate management in the age of global environment awareness – a case study of PET bottle recycling issues in Japan. Science and Technology of Polymers and Advanced Materials. Boston, MA: Springer; 1998. p. 613–27. 10.1007/978-1-4899-0112-5_54.Suche in Google Scholar

(21) Throne JL. Effect of recycle on properties and profits: algorithms. Adv Polym Technol. 1987;7:347–60. 10.1002/ADV.1987.060070401.Suche in Google Scholar

(22) Farahat MS, Abdel-Azim AAA, Abdel-Raowf ME. Modified unsaturated polyester resins synthesized from poly(ethylene terephthalate) waste, 1: synthesis and curing characteristics. Macromol Mater Eng. 2000;283:1–6. 10.1002/1439-2054(20001101)283:1<1::AID-MAME1>3.0.CO;2-V.Suche in Google Scholar

(23) Alarifi IM. PETG/carbon fiber composites with different structures produced by 3D printing. Polym Test. 2023;120:107949. 10.1016/J.POLYMERTESTING.2023.107949.Suche in Google Scholar

(24) Almutairi MD, Mascarenhas TA, Alnahdi SS, He F, Khan MA. Modal response of hybrid raster orientation on material extrusion printed acrylonitrile butadiene styrene and polyethylene terephthalate glycol under thermo-mechanical loads. Polym Test. 2023;120:107953. 10.1016/J.POLYMERTESTING.2023.107953.Suche in Google Scholar

(25) Ahmad MN, Ishak MR, Mohammad Taha M, Mustapha F, Leman Z, Irianto. Mechanical, thermal and physical characteristics of oil palm (Elaeis guineensis) fiber reinforced thermoplastic composites for FDM – type 3D printer. Polym Test. 2023;120:107972. 10.1016/J.POLYMERTESTING.2023.107972.Suche in Google Scholar

(26) Guan L, Fan J, Chan XY, Le Ferrand H. Continuous 3D printing of microstructured multifunctional materials. Addit Manuf. 2023;62:103373. 10.1016/J.ADDMA.2022.103373.Suche in Google Scholar

(27) Nguyen PQK, Panta J, Famakinwa T, Yang R, (Chunhui), Ahmed A, et al. Influences of printing parameters on mechanical properties of recycled PET and PETG using fused granular fabrication technique. Polym Test. 2024;132:108390. 10.1016/J.POLYMERTESTING.2024.108390.Suche in Google Scholar

(28) Al Rashid A, Koç M. Additive manufacturing for sustainability, circularity and zero-waste: 3DP products from waste plastic bottles. Compos Part C: Open Access. 2024;14:100463. 10.1016/J.JCOMC.2024.100463.Suche in Google Scholar

(29) Miller K ASTM International: Standard test method for tensile. – Google Scholar 2004. https://axelproducts.com/wp-content/uploads/2024/02/TestingPlasticForFEA.pdf (accessed May 14, 2024).Suche in Google Scholar

(30) Ansari AA, Kamil M. Effect of print speed and extrusion temperature on properties of 3D printed PLA using fused deposition modeling process. Mater Today Proc. 2021;45:5462–8. Elsevier Ltd. 10.1016/j.matpr.2021.02.137.Suche in Google Scholar

(31) Aslani KE, Kitsakis K, Kechagias JD, Vaxevanidis NM, Manolakos DE. On the application of grey Taguchi method for benchmarking the dimensional accuracy of the PLA fused filament fabrication process. SN Appl Sci. 2020;2:2823. 10.1007/s42452-020-2823-z.Suche in Google Scholar

(32) Agarwal KM, Shubham P, Bhatia D, Sharma P, Vaid H, Vajpeyi R. Analyzing the impact of print parameters on dimensional variation of ABS specimens printed using fused deposition modelling (FDM). Sens Int. 2022;3:100149. 10.1016/J.SINTL.2021.100149.Suche in Google Scholar

(33) Akbaş OE, Hıra O, Hervan SZ, Samankan S, Altınkaynak A. Dimensional accuracy of FDM-printed polymer parts. Rapid Prototyp J. 2020;26:288–98. 10.1108/RPJ-04-2019-0115.Suche in Google Scholar

(34) Alafaghani A, Qattawi A, Alrawi B, Guzman A. Experimental optimization of fused deposition modelling processing parameters: a design-for-manufacturing approach. Proc Manuf. 2017;10:791–803. 10.1016/J.PROMFG.2017.07.079.Suche in Google Scholar

(35) Pulipaka A, Gide KM, Beheshti A, Bagheri ZS. Effect of 3D printing process parameters on surface and mechanical properties of FFF-printed PEEK. J Manuf Process. 2023;85:368–86. 10.1016/J.JMAPRO.2022.11.057.Suche in Google Scholar

(36) Xiaoyong S, Liangcheng C, Honglin M, Peng G, Zhanwei B, Cheng L. Experimental analysis of high temperature PEEK materials on 3D printing test. Proceedings – 9th International Conference on Measuring Technology and Mechatronics Automation, ICMTMA 2017; 2017. p. 13–6. 10.1109/ICMTMA.2017.0012.Suche in Google Scholar

Received: 2025-03-17

Revised: 2025-05-24

Accepted: 2025-06-23

Published Online: 2025-09-13

This work is licensed under the Creative Commons Attribution 4.0 International License.

Artikel in diesem Heft

https://doi.org/10.1515/epoly-2025-0019

Schlagwörter für diesen Artikel

recycled PET; additive manufacturing; 3D printing; printing parameters; dimensional accuracy; mechanical properties

Creative Commons

BY 4.0