Sage Journals: Discover world-class research

Abstract

In the past two decades, polyether ether ketone (PEEK) has been vastly used in different industries such as automotive, aerospace, and medical sectors due to its mechanical strength, high heat stability, and excellent chemical resistance. Since this semi-crystalline polymer has a high melting temperature, the standard procedure to select extrusion-based 3D printing working parameters has many challenges to ensure each bespoke product with the best quality. Slight shift from optimum process parameters in fused deposition modeling (FDM) of PEEK has a remarkable impact on the final product quality. The present research addresses the long-standing problem of selecting a working setup quickly and accurately which is not dependent on the user's knowledge. In this article, to identify the best arrangement of FDM process parameters, an orthogonal matrix of Taguchi Design of Experiments (DoEs) is applied. Different values of bed, chamber, and nozzle temperatures, print speed, and layer thickness were thoroughly examined for the printing of standard tensile specimens. The mechanical properties and the fracture morphology of samples are examined in different printing conditions to explore a roadmap tool for designing better process parameters which satisfy mechanical requirements of products. In this study, the interpretation of the experimental results is investigated based on analysis of variance (ANOVA). The obtained results show that the best working set up for nozzle temperature, bed temperature, print speed, and layer thickness are 415°C, 160°C, 50 mm s⁻¹, and 0.2 mm, respectively. This article consists of five sections. After introduction in the first section, the effects of working parameters on the strength of final fused products are investigated in Experimental setup section. The results and the confirmation test are considered in seperate sections, respectively. The research is summarized in the conclusion section.

Keywords

3D printing PEEK FDM fused filament fabrication additive manufacturing process parameters

Introduction

In recent years, the 3D printing market has recorded an upward trend in production of components in aerospace, automotive industries, biomedical, and energy sectors along with increasing attention for personalized equipment.¹ Fused deposition modeling (FDM) or fused filament fabrication (FFF) is a widespread process in additive manufacturing (AM) technologies. Low initial investment in equipment and material with high safety and comfort use of filament are the main advantages of FDM. Although acrylonitrile butadiene styrene (ABS), poly lactic acid (PLA), and nylon are the most common filaments for FDM, polyaryletherketone (PAEK) and polyether ether ketone (PEEK) are highly requested recently for FFF technique.² PAEK and PEEK are a family of semi-crystalline thermoplastics which have attractive attributes such as high-temperature stability and high mechanical strength. PEEK's biocompatibility, resistance to chemical damage, compatibility with different reinforcing agents (such as glass and carbon fibers), and radiolucency make PEEK a suitable biomaterial for orthopedic, spine and dental implants.^3–13 In recent years, the limitations and challenges in 3D printing of PEEK through the FDM process has been the scope of many researchers. Some prevalent problems would occur in PEEK 3D printing such as weak interlayer bonding, distortion of initial layers, delamination, internal pores, high porosity, and internal crack propagation.^14–16

Although there are some parameters with tight tolerance that have no impact on the final printing sample, there are some important parameters which can directly affect the mechanical strength and the perfection of PEEK printed products. The best selection of working parameters could improve the print condition and eliminate the probable defects. The effective parameters include print speed, nozzle, bed and chamber temperatures, layer thickness, infill type, infill density, nozzle diameter, PEEK filament quality, desiccation of initial PEEK filament, and post heat treatment.¹⁴ Figure 1 shows the classification of effective parameters in PEEK extrusion-based 3D printing. Nozzle diameter and PEEK filament quality are not adjustable parameters in the FDM process and the engineer should change the hardware of the machine to implement a new setup. The feedstock, filament moisture level, and post heat treatment are also obvious that they cannot be considered as adjustable parameters. The rest of parameters are working setup which can be changed during the print process.

Figure 1.

Classification of effective parameters in fused deposition modeling (FDM) of polyether ether ketone (PEEK).

Many studies present the complication of selecting process parameters for printing PEEK specimens. Vaezi and Yang¹⁷ published the first report on successful PEEK 3D printing through FDM process. Rahman et al.¹⁸ investigated the effect of print's raster angle on the mechanical strength. Three raster directions were considered in their experiments, namely 0° (longitudinal), 90° (perpendicular), and 0°/90° alternating. They tested samples thoroughly with uniaxial, tension, flexural, compression, and impact tests. Their results revealed that printing in longitudinal direction presents higher mechanical strength. Zhao et al.¹⁵ run a set of experiments and assessment of mechanical loading tests under different printing conditions such as raster angle (0°, 45°, and 90°), post-treatment along with nozzle and ambient temperatures. Their study investigated individually for finding the best mechanical strength for each parameter to produce high strength biocompatible PEEK scaffold cranial structure. Wang et al.¹⁶ applied a finite element analysis (FEA) and experimental setup simultaneously to investigate the effect of various printing parameters such as nozzle diameter, printing temperature, print speed, printing layer thickness on the mechanical properties, and surface quality of FDM printed parts. Geng et al.¹⁹ studied the effects of the extrusion and printing speeds on the microstructure and dimensions of an extruded PEEK filament. Wang et al.²⁰ studied the effects of different process parameters such as temperatures of nozzle and plate, printing speed, and layer thickness on the mechanical strength. Song et al.²¹ considered the effects of variable process parameters on the porosity and mechanical strength of PEEK. Deng et al.²² have conducted an experimental research on the effects of PEEK printing parameters such as nozzle temperature, infill speed, and layer thickness on tensile strength. Although there are several research in the literature to investigate the impact of FDM process parameters on mechanical strength, most of them have considered the effect of relevant parameters partially while assuming some items as constant variables. Also, some other researchers did not apply a commercial FDM machine and they could not consider the whole range of all parameters. In this article, the process parameters of fused filament printing including print speed, nozzle temperature, bed temperature, and layer thickness are thoroughly investigated. Three samples per each experimental setup are fabricated and tested for eliminating the stochastic effects on results. The comprehensive data are collected to provide a new guideline to enhance the quality of PEEK products and reduce the fabrication failures. The article is organized as follows: in Experimental setup section, the domain of process parameters is described in detail. The experimental setup is explained in Results and discussion section, followed by morphology observation of the fracture section and in the next section, the confirmation test is done.

Experimental setup

Domain of process parameters

The next step after determining the effective process parameters is to find the best working range of each parameter. Many papers published to consider the effects of process parameters on mechanical strength.^14–39 Table 1 summarizes the process parameters domains in the literature for the highest mechanical strength. The common mechanical tests to analyze mechanical strength are tensile, bending, impact, and compressive tests. In the present article the uniaxial tension test has been applied for mechanical strength.

Table 1.

The proposed process parameters.

Author	Process parameters
Zhao et al.¹⁵	Nozzle temperature = 380, 400, 420°C Chamber's temperature = 50, 65, 80°C
Wang et al.¹⁶	Nozzle temperature = 380, 400, 420, 440°C Layer thickness = 0.10, 0.15, 0.20, 0.25 mm
Wang et al.²⁰	Nozzle temperature = 400, 410, 420, 430, 440°C Bed's temperature = 240, 250, 260, 270, 280°C
Deng et al.²²	Nozzle temperature = 350, 360, 370°C Layer thickness = 0.20, 0.25, 0.30 mm
Yang et al.²³	Nozzle temperature = 360, 380, 400, 420, 440, 460, 480°C Chamber's temperature = 25, 50, 100, 150, 200°C
Wu et al.²⁴	Layer thickness = 0.2, 0.3, 0.4 mm
Hu et al.²⁷	Chamber's temperature = 25, 60°C
Tafaoli-Masoule et al.³²	Nozzle temperature = 375 until 440°C Bed's temperature = 130 until 160°C
Rehekampff et al.³³	Print speed = 7.5, 15, 30 mm s⁻¹
Li et al.³⁵	Nozzle temperature = 445, 485, 525°C Layer thickness = 0.1, 0.2, 0.3 mm

Selection of working parameters

Ambient or chamber temperature has a key effect on mechanical properties of printed PEEK products. Increasing the chamber temperature improves the print condition. By considering the melting point of PEEK (343°C), increasing the chamber temperature close to melting point makes a lower cooling rate that finally results in high strength samples. The printed components at relatively low chamber temperature will have inappropriate interlayered bonding that creates delamination on the fracture surface¹⁵ while with the increase of the chamber temperature, the delamination is clearly inhibited. Zhao et al.¹⁵ reported that at a temperature higher than 80°C, improvement of the interlayered bonding eliminates the delamination. In the present study, due to the 3D printing machine setup, it is impossible to consider a higher ambient temperature of 90°C and the maximum chamber temperature is fixed to 90^oC. Furthermore, the infill type parameter is not considered in present study and all samples are printed in zigzag type. The effective parameters studied in this research are nozzle temperature, bed or platform temperature, feed rate, and layer thickness.

Design of experiments (DoEs) using Taguchi method

It is necessary to carefully design the experimental workflow. In the present study, 4 parameters are considered with 3 levels each, at least 81 experiments (i.e. 3⁴) need to be run for understanding the effects of each variable. Moreover, repetitive tests are needed for elimination of experimental error. In fact, if three sets of repetitive samples are used without implementation of Taguchi method the total experiments reach to 243 states that make the experimental activities time-consuming and costly.

Taguchi technique is an effective tool for the DoEs which reduces the number of required experiments while retaining the impact of each variable on the final output. The steps of Taguchi design are shown in Figure 2.^40–51 After running experimental tests, signal-to-noise ratio (SNR) and analysis of variance (ANOVA) were applied to analyze the experimental results. Process parameters for experiments are nozzle temperature (A), bed temperature (B), print speed (C), and layer thickness (D) with three levels of nozzle temperature and two levels of the rest parameters. The selected levels were obtained according to the literature and 3D printing machine limitations. The number of experiments is determined by the orthogonal array (OA) which is the minimum possible matrix for combination of design parameters. In the present study, the total degree of freedom (DOF) is five which should be smaller or at maximum equal to an OA's DOF. Therefore, a Taguchi L8 OA was applied with 7 DOFs. Table 2 shows the L8 OA design to improve experimental efficiency.

Figure 2.

The work flow of investigation.

Table 2.

L8 orthogonal array (OA) design for process parameters of polyether ether ketone (PEEK).

ExperimentNo.	Factor
ExperimentNo.	ANozzle temperature(°C)	BPrint bed temperature(°C)	CPrinting speed(mm s⁻¹)	DLayer thickness (mm)
1	390	120	10	0.1
2	390	160	50	0.2
3	415	120	10	0.2
4	415	160	50	0.1
5	440	120	50	0.1
6	440	160	10	0.2
7	415	120	50	0.2
8	440	160	10	0.1

FDM machine and material

The specimens are printed using a commercially available 3D printer namely 3DGence INDUSTRY F340. The FIRE WIRE^®PEEK filament with average diameter of 1.75 mm and density of 1.29 g cm⁻³ (manufactured by 3DXTECH, USA) is used as the feed stock.

Samples preparation

The CAD model of the tensile test was created in CATIA V5. The slicer software 3DGence Slicer 4.0 is used to slice the STL file. Eight dumbbell-shaped specimens with 50% infill density (filling ratio) for uniaxial tension test were printed based on standard test method for tensile properties of plastic (D638-14 type IV).⁵² To eliminate the interference of experimental error, increase the precision and gaining higher reliability three repetitive samples were printed with the same process parameters.

Porosity evaluation

Each specimen is fully investigated in terms of porosity. The porosity is measured by the following relation.¹⁷

Porosity (%) = \frac{V_{e} - V_{r}}{V_{e}} \times 100; V_{r} = \frac{m}{ρ}

(1)

where

V_{e}

represents the external volume includes the content of pores and full material.

V_{r}

is the real volume shows the content of printed material without any pore. In the present study,

V_{e}

is determined by a 3D scanner and

V_{r}

is calculated by the ratio of mass (

m

) per PEEK density (

ρ

). All samples’ masses are measured by an A&D Weighing micro balance with a readability of 0.1 mg.

Uniaxial tension test

A fully computerized universal testing machine STM-250 (SANTAM Engineering Design Co. Ltd) with servo electromechanical operation is used for tension test. Testing speed for all samples were fixed at 3.9 mm min⁻¹ that is equivalent to strain rate of 10⁻³S⁻¹ based on standard test method for tensile properties of plastic (D638-14 type IV).⁵²

Results and discussion

Tensile test

It is worth to note that for each combination of process parameter, three specimens are printed and the mean value is calculated. Table 3 demonstrates the average tensile strength, the SD for three repetitive samples, elastic modulus, elongation, and average porosity for the proposed experiment. In Table 3m the maximum tensile strength is 44.2 MPa.

Table 3.

Tensile properties of L8 orthogonal array (OA) design.

ExperimentNo.	Factor				Average tensile strength (MPa)	The SD (MPa)	Peak elongation (%)	Young's modulus (MPa)	Average porosity (%)	SNR
ExperimentNo.	A	B	C	D	Average tensile strength (MPa)	The SD (MPa)	Peak elongation (%)	Young's modulus (MPa)	Average porosity (%)	SNR
1	1	1	1	1	23.7	0.5	6	379.2	18	27.49
2	1	2	2	2	33.6	0.0	10	326.4	22	30.53
3	2	1	1	2	35.4	0.9	8	407.6	19	30.96
4	2	2	2	1	43.3	1.1	12	345.4	20	32.71
5	3	1	2	1	27.7	1.3	5	472.8	19	28.81
6	3	2	1	2	42.9	0.4	12	331.9	20	32.65
7	2	1	2	2	44.2	0.7	13	326.2	18	32.90
8	3	2	1	1	39.9	0.7	11	352.6	21	32.01

SNR: signal-to-noise ratio.

SNR and ANOVA

In the Taguchi approach, SNR measures the effect of design parameters on the final output. The three models of SNR are used for Taguchi are (a) nominal is better (to achieve the target with minimum deviation), (b) larger is better (to maximize the response), and (c) smaller is better (to minimize the response).⁴⁰

In the present study, the design parameters are process parameters, the final output is tensile strength, and the objective is selected to be the best selection of input parameters which provides the highest tensile strength. The second type of SNR expressed by equation (2) was applied.

S N R = - 10 l o g (\frac{1}{n} \sum_{j = 1}^{n} \frac{1}{y_{i}^{2}})

(2)

In equation (2), i stands the experiment number, j is the trial number, n indicates the total number of trials, and

y_{i}

is the output value. The last column of Table 3 shows SNR value. The maximum SNR that is related to experiment No. 7 is the best combination among L8 matrix. Figure 3 illustrates the plots of SNR that were calculated for maximum tensile strength. Using Figure 3, the variables that will be utilized to achieve a larger tensile value and their respective levels are concluded as A2, B2, C2, and D2. They are corresponding to the nozzle temperature of 415°C, bed temperature of 160°C, print speed of 50 mm s⁻¹, and layer thickness of 0.2 mm.

Figure 3.

The main effects plot for signal-to-noise (SNR).

The interactions between different parameters are not considered in this study. In Table 4, range analysis of L8 orthogonal experiments is listed. In equation (3), K_ij (i = 1, 2, 3) is the sum of experimental results in terms of level i, k_ij the mean of K_ij, and R represents the range.

R = max (k_{i j}) - min (k_{i j})

(3)

Table 4.

Range analysis of tensile strength.

Range	A	B	C	D
K_1j	172.0	392.8	425.6	403.6
K_2j	368.4	479.0	446.2	468.2
K_3j	331.4	-	-	-
k_1j	28.7	32.7	35.5	33.6
k_2j	40.9	39.9	37.2	39.0
k_3j	36.8	-	-	-
R	8.2	7.2	1.7	5.4
Optimum level	A₂	B₂	C₂	D₂
Optimum combination	A₂B₂C₂D₂
Order of priority	ABDC

According to the order of priority (ABDC), the nozzle temperature has the highest impact on the tensile strength. Although the impact order of process parameters can be detected in the analysis, the influence of each parameter individually is a challenging subject.⁴⁰

ANOVA as a statistical technique was applied to interpret the experimental data and complementary analysis. ANOVA results are shown in Table 5 with two important parameters F and P, which indicate variance ratio and percent of contribution, respectively. F-value is a ratio of the mean of squared deviations to the mean of squared errors and p-value is the percentage contribution of each parameter.⁴⁰ These values could be determined directly using equations (4) and (5).

F_{Factor} = \frac{V_{Factor}}{V_{Error}}; V_{Factor} = \frac{S_{Factor}}{f_{Factor}}

(4)

P_{Factor} = 100 \times [\frac{{S^{'}}_{Factor}}{S_{Total}}]; S_{Total} = \sum_{i = 1}^{n} y_{i}^{2} - \frac{{(\sum_{i = 1}^{n} y_{i})}^{2}}{n}

(5)

Table 5.

Analysis of variance (ANOVA) result for tensile strength.

Source	Degree of freedomf (DOF)	Sum of squaresS	Variance (mean square)V	Variance ratioF	Pure sum of squaresS’	Percent contributionP
A	2	545.88	272.94	25.63	524.58	42.33
B	1	309.96	309.96	29.11	299.31	24.15
C	1	17.70	17.70	1.66	7.05	0.56
D	1	174.15	174.15	16.35	163.50	13.20
Error	18	191.68	10.65	1.00	244.92	19.76
Total	23	-	-	-	-	100

In equaiton (4), $V_{Factor}$ is variance or mean square of each factor, $S_{Factor}$ indicates sum of squares, and f stands for DOF. $y_{i}$ and S’ in equation (5) indicate output result of each experiment and pure sum of squares, respectively while n is the total number of experiments. More details for ANOVA calculation exist in.⁵³ Table 5 indicates that the print speed (C) has the lowest p-value among other variables which shows that print speed is not a significant factor in the process while nozzle temperature (A) has the highest p-value which indicates the most important factor.

Influence of process parameters on tensile property

Nozzle and bed temperatures

The nozzle temperature has the most effectiveness (42%) on tensile strength. Figure 4 indicates the mean experimental results in nozzle temperature of 390, 415, and 430°C. The nozzle heat has the direct effect on adhesion of layers. The higher temperature of extruded material provides better adhesion to the previous printed layer which is a solid state. Furthermore, the low viscosity and high fluidity at high temperature leads to being well printed without blocking the nozzle.³⁵ Higher temperature transfers more energy to PEEK filament so the internal microvoids decrease and the strength of specimens would increase. On the other side, higher printed temperature creates better crystallization because the longer time would be allocated between the printed temperature and the ambient temperature.^21,23

Figure 4.

Tensile strength for different levels of nozzle temperature (a) and bed's temperature (b).

Figure 5(a) and (b) show the comparison between the specimen No. 1 from L8 OA which is printed in 390°C as the worst experiment (the lowest tensile strength) and the specimen No. 7 which is printed in 415°C as the best one (the highest tensile strength). Figure 5(b) indicates the appropriate filament-to-filament and interlayer bonding and inversely Figure 5(a) shows the weak bonding. Figure 5(b) indicates regular and proper bonding of sequence layers.

Figure 5.

Microstructure of fracture surface. (a) Specimen No. 1 (390°C), (b) specimen No. 7 (415°C), and (c) specimen No. 8 (440°C).

Figure 6 focuses on fusion line between infill filaments. In Figure 6(b), the layer connectivity shows good fusion for specimen No. 7 compare to Figure 6(a) which is related to specimen No. 1. The uniform wire fusion is clear during the FDM process. Based on Figure 6, the lower print speed and lower temperature lead to un-uniform deposition.²² Although the high nozzle temperature leads to better adhesion, the excessive heating can make it worse. Finding an optimum temperature is not straightforward and depends on each case. The optimum value of nozzle temperature is 415°C. In fact, the increase in the nozzle temperature than optimum value leads to the decrease in the strength (Figure 4). The specimen No. 8 with highest selected temperature of 440°C has created lower tensile strength. Based on Table 4, the A3 (mean strength of specimens which is printed in 440°C) amount of 4.1 MPa is the lower compared to A2 (mean strength of specimens which is printed in 415°C).

Figure 6.

Representative scanning electron micrographs of fracture surfaces. (a) Specimen No. 1 (390°C) and (b) specimen No. 7 (415°C).

With the increase of nozzle temperature than optimum value, three prevalent issues will happen. First, the diluted deposited filament with high energy has a deep penetrative effect. Each layer might be affected several times during printing the top layers and therefore, the solidification process is not uniform. Figure 5(c) represents the micrographs of fracture surface for specimen No. 8 with 440°C. Layers crush on each other in white rectangles in Figure 5(c) compared to Figure 5(b) which indicates regular bonding between layers. In Figure 5(b), layers are formed disorderly. Second, the degradation of PEEK will produce gas which will emerge in the deposited layer as micropores. The layer pores are shown in Figure 7. Micropores in the layer are resulted of the entrapped gases. They are generated by excessive heating from optimum temperature. Higher temperature will affect the volume and numbers of these micropores. When the porosity increases, the strength of the specimen will decrease. At micrographs of present experiments, this type of pore was not observed (it might emerge at higher temperatures). Third, low viscosity of melted filament does not create a whole defined thickness. Therefore, the distance between printed layers will occur. The next layer will print with air gaps between layers. The interlayer pores in Figure 7 demonstrate these gaps schematically between deposited layers.

Figure 7.

The layer pores and interlayer pores in fused deposition modeling (FDM) at higher temperatures than optimum value.

The nozzle temperature directly determines the viscosity of melted filament which plays an important role in the form of structure. While the appropriate viscosity of deposited filament creates a proper thickness and, the next layer is deposited uniformly without any gap, and extra dilution of melted filament cannot create a proper and entire thickness. Figure 8(a) and (b) show two adjacent layers of specimen No. 7 with 415°C and specimen No. 8 with 440°C, respectively.

Figure 8.

The observation of interlayer micropores. (a) Specimen No. 7 (415°C) and (b) specimen No. 8 (440°C).

The interlayer pores are specified with red rectangle in Figure 8(b). Un-uniform surfaces along with distance and gaps between two subsequent printed layers make high porosity which can decrease the strength of samples. According to the order of priority in Table 4 (ABDC), the second priority is the bed temperature with a p-value of 24%. The higher bed temperature provides higher adhesion. At constant low chamber temperature of the present experiment (90°C), higher temperature of bed or platform plays an important role in warming up the initial layers which are adjacent to the bed. Figure 4 shows the mean experimental results in bed temperature. If the limitation of 3D printer machine, allows the chamber temperature to increase up to 250°C (such as Wang et al.¹⁴), the bed temperature will shift from the second priority to the lower importance.

Print speed and layer thickness

The third and fourth priorities are layer thickness and print speed, respectively. Figure 9 shows the tensile strength for different levels of these factors.

Figure 9.

Tensile strength for different levels of layer thickness and print speed.

Although increasing the layer thickness provides low surface quality,¹⁴ thick layer has more preservation of heat and creates low cooling rate, and therefore, the subsequent layer takes higher adhesion. More preservation of heat will happen at a fast speed for merging the two adjacent layers. On the other side, when PEEK prints quickly, the material does not have sufficient time to form and orient. Weak bonding strength between the two adjacent printing layers will occur at the downside of rapid printing. The obtained experimental results indicate that the effect of heat preservation overcomes the orientation factor. The higher print speed leads to a higher strength specimen.

Confirmation test

The final step is to predict the best value for each parameter based on Taguchi approach and ANOVA in order to improve the tensile strength. According to the range analysis of tensile strength based on Taguchi results, the optimum combination is A₂B₂C₂D₂. Considering Taguchi approach, the best value for optimum combination is calculated. The maximum estimated tensile strength of Taguchi (T_max) is calculated by

T_{max} = T_{a v} + \sum_{j = 1}^{4} {Max (K_{i j}) - T_{a v}}

(6)

In equation (6), j refers to four parameters and T_av is average response of the tensile strength for all experiments. The maximum estimated tensile strength (T_max) is 48.9 MPa based on Taguchi approach with population parameters. More population of samples would improve the precision of the estimation. In statistics, it is common to represent the confidence interval (CI) parameter as a range for confidence with following equation:⁵³

C . I . = \pm \sqrt{\frac{F_{0.01} (f_{1}, f_{2}) \times V_{e}}{n_{e}}}

(7)

where

F_{0.01}

is the variance ratio for DOF of

f_{1}

and

f_{2}

at a confidence level of 99%,

f_{1}

indicates DOF of mean which is always 1,

f_{2}

is DOF of error term,

V_{e}

is the variance of error term, and

n_{e}

indicates number of equivalent replications which is given by:

n_{e} = (\frac{Number of trials}{DOF of mean + DOF of all factors used in the estimate})

(8)

The range of 2.7 MPa is obtained for the estimation of data. Therefore, it can be stated that there is a 99% probability for having the estimation value of 48.9 MPa

\pm

2.7 MPa (between 46.2 MPa and 51.6 MPa). For the validation of Taguchi result as well as complementing the experiment, the optimum combination (A₂B₂C₂D₂) was set up as a new experiment. Figure 10 compares the experimental stress–strain curves of Taguchi optimum combination (A₂B₂C₂D₂) and the best experiment of L8 OA which is the seventh experiment (A₂B₁C₂D₂). The maximum tensile strength (50.0 MPa) in the experiment has a good agreement with Taguchi estimated value (T_max = 48.9 MPa) and it is aligned with the CI range. The difference of 1.1 MPa is the impact of the interaction parameters. More trials will proivde a closer approximation. The result of this research directly depended on the level numbers of process parameters. Regarding the experimental limitations, two or three levels were selected. Although more levels lead to more precision, numerous levels increase the experiments exponentially which is time consuming and strictly costly.

Figure 10.

Stress–strain curves of optimum L8 combination and new optimized experiment.

Conclusion

In the present study, the uniaxial tension test of PEEK samples at different FDM printing conditions is thoroughly studied. The basic parameters which have major effect on mechanical strength are listed and the domain of each parameter and the best value of them are indicated. Nozzle temperature, print speed, platform temperature, and layer thickness were considered as the effective parameters for PEEK printing. Based on Taguchi, an L8 OA was applied in the experiments. The levels for nozzle temperature are selected in three different levels including 390°C, 415°C, and 440°C. Experiments revealed that the best value for nozzle temperature is 415°C, whereas lower values provide weak bonding connectivity and higher temperature provides disordered layer adhesion. The SEM microstructures of specimens with different nozzle temperatures were compared. The levels of platform temperature were selected 120°C and 160°C. The levels of print speed were selected as 10 mm s⁻¹ and 50 mm s⁻¹. The levels of layer thickness were selected as 0.1 mm and 0.2 mm. Based on L8 Taguchi result, the best value for platform temperature, print speed, and layer thickness are 160°C, 50 mm s⁻¹, and 0.2 mm, respectively. By applying range analysis for Taguchi, the best estimated process parameter was calculated. A new specimen with calculated process parameter was fabricated and then was tested experimentally. The comparison of Taguchi optimum condition and the new fabricated test demonstrated a good agreement.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors acknowledge the funding support of Babol Noshirvani University of Technology through Grant program No BNUT/370434/01.

ORCID iD

Mohsen Shakeri

References

Kumar

Chohan

. The role of additive manufacturing for biomedical applications: a critical review. J Manuf Process 2021; 64: 828–850.

Singh

Ramakrishna

Singh

. Material issues in additive manufacturing: a review. J Manuf Process 2017; 25: 185–200.

Kurtz

. An Overview of PEEK Biomaterials. PEEK Biomaterials Handbook 1–7 (Elsevier, 2012). DOI: 10.1016/B978-1-4377-4463-7.10001-6.

Oladapo

Zahedi

Ismail

, et al. 3D printing of PEEK–cHAp scaffold for medical implant. Bio-Des Manuf 2020; 2020: s42242–020-00098-0.

Zhao

Niu

, et al. 3D printing high interfacial bonding polyether ether ketone components via pyrolysis reactions. Mater Des 2021; 198: 109333.

Oladapo

Zahedi

. Improving bioactivity and strength of PEEK composite polymer for bone application. Mater Chem Phys 2021; 266: 124485.

Schwitalla

Muller

. PEEK dental implants: a review of the literature. J Oral Implantology 2013; 6: 743–749.

Alam

Varadarajan

Koo

, et al. Additively manufactured polyetheretherketone (PEEK) with carbon nanostructure reinforcement for biomedical structural applications. Adv Eng Mater 2020; 22: 2000483.

Haleem

Javaid

. Polyether ether ketone (PEEK) and its 3D printed implants applications in medical field: an overview. Clin Epidemiol Glob Health 2019; 7: 571–577.

10.

Oladapo

Zahedi

Ismail

, et al. Recent advances in biopolymeric composite materials: future sustainability of bone-implant. Renewable Sustainable Energy Rev 2021; 150: 111505.

11.

Zhong

Vaezi

Mei

, et al. Strategy for controlling the properties of bioactive poly-ether ether-Ketone/Hydroxyapatite composites for bone tissue engineering scaffolds. ACS Omega 2019; 1–8. DOI: 10.1021/acsomega.9b02572.

12.

Oladapo

Zahedi

Ismail

. Mechanical performances of hip implant design and fabrication with PEEK composite. Polymer 2021; 227: 123865.

13.

Haleem

Javaid

. Polyether ether ketone (PEEK) and its manufacturing of customized 3D printed dentistry parts using additive manufacturing. Clin Epidemiol Glob Health 2019; 7: 654–660.

14.

Wang

Müller

Rumjahn

, et al. Parameters influencing the outcome of additive manufacturing of tiny medical devices based on PEEK. Mater (Basel) 2020; 13: 466.

15.

Zhao

, et al. Mechanical characterization of biocompatible PEEK by FDM. J Manuf Process 2020; 56: 28–42.

16.

Wang

Zou

Xiao

, et al. Effects of printing parameters of fused deposition modeling on mechanical properties, surface quality, and microstructure of PEEK. J Mater Process Technol 2019; 271: 62–74.

17.

Vaezi

Yang

. Extrusion-based additive manufacturing of PEEK for biomedical applications. Virtual Phys Prototyp 2015: 1–13.

18.

Rahman

Reese

Letcher

. Mechanical properties of additively manufactured PEEK components using fused filament fabrication. ASME 2015 International mechanical engineering congress and exposition 2015: 1–11.

19.

Geng

Zhao

, et al. Effects of extrusion speed and printing speed on the 3D printing stability of extruded PEEK filament. J Manuf Process 2019; 37: 266–273.

20.

Wang

Zou

Ding

, et al. Effects of FDM 3D printing parameters on mechanical properties and microstructure of CF/PEEK and GF/PEEK. Chin J Aeronaut 2020; 2020: 1–11.

21.

Song

Shi

Song

, et al.. Used deposition modeling of poly(ether ether ketone) scaffolds. High Temp Mater Processes 2021; 40: 1–11.

22.

Deng

Zeng

Peng

, et al. Mechanical properties optimization of poly-ether-ether-ketone via fused deposition modeling. Mater (Basel) 2018; 11: 216.

23.

Yang

Tian

, et al. Influence of thermal processing conditions in 3D printing on the crystallinity and mechanical properties of PEEK material. J Mater Process Technol 2017; 1–12. DOI: 10.1016/j.jmatprotec.2017.04.027.

24.

Geng

, et al. Influence of layer thickness and raster angle on the mechanical properties of 3D-printed PEEK and a comparative mechanical study between PEEK and ABS. Mater (Basel) 2015; 8: 5834–5846.

25.

Cicala

Latteri

Del Curto

, et al. Engineering thermoplastics for additive manufacturing: a critical perspective with experimental evidence to support functional applications. J Appl Biomater Funct Mater 2017; 15: 10–18.

26.

Han

Yang

, et al. Carbon fiber reinforced PEEK composites based on 3D-printing technology for orthopedic and dental applications. J Clin Med 2019; 8: 240.

27.

Duan

Xing

, et al. Improved design of fused deposition modeling equipment for 3D printing of high-performance PEEK parts. Mech Mater 2019; 137: 103–139.

28.

Basgul

MacDonald

, et al. Does annealing improve the interlayer adhesion and structural integrity of FFF 3D printed PEEK lumbar spinal cages. J Mech Behav Biomed Mater 2020; 102: 103455.

29.

Zhao

, et al. Flexural properties and fracture behavior of CF/PEEK in orthogonal building orientation by FDM: microstructure and mechanism. Polymers (Basel) 2019; 11: 656.

30.

Gebisa

Lemu

. Influence of 3D printing FDM process parameters on tensile property of ULTEM 9085. Procedia Manuf 2019; 30: 331–338.

31.

Motaparti

Taylor

Leu

, et al. Experimental investigation of effects of build parameters on flexural properties in fused deposition modelling parts. 2017. ISSN: 1745-2759 (Print) 1745-2767 (Online).

32.

Tafaoli-Masoule

Shakeri

Zahedi S

, et al. 3D printing of PEEK-based medical devices. Addit Manuf Meets Med 2019; 1–2. DOI: 10.18416/AMMM.2019.1909S10T05.

33.

Rehekampff

Schroeffer

Irlinger

, et al. Influence of Infill Structures and Process Parameters on the Tensile Strength of 3D-printed PEEK Parts. 2019 IEEE International Conference.

34.

Gonabadi

Yadav

Bull

. The effect of processing parameters on the mechanical characteristics of PLA produced by a 3D FFF printer. Int J Adv Manuf Technol 2020; 1–15. 10.1007/s00170-020-06138-4.

35.

Lou

. Tensile and bending strength improvements in PEEK parts using fused deposition modelling 3D printing considering multi-factor coupling. Polymers (Basel) 2020; 12: 2497.

36.

Maloch

Hnátková

Žaludek

, et al. Effect of processing parameters on mechanical properties of 3D printed samples. Mater Sci Forum 2018; 919: 230–235.

37.

Oladapo

Zahedi

Adeoye

AOM

. 3D Printing of bone scaffolds with hybrid biomaterials. Composites Part B 2019; 158: 428–436.

38.

Vaezi

Black

Gibbs

DMR

, et al. Characterization of new PEEK/HA composites with 3D HA network fabricated by extrusion freeforming. Molecules 2016; 21: 687.

39.

Geng

Zhao

, et al. Manufacture and thermal deformation analysis of semicrystalline polymer polyether ether ketone by 3D printing. Mater Res Innovations 2014; 18: 12–16.

40.

Ferdous

Manalo

Aravinthan

. Bond behaviour of composite sandwich panel and epoxy polymer matrix: taguchi design of experiments and theoretical predictions. Constr Build Mater 2017; 145: 76–87.

41.

Shailesh

Sundarrajan

Komaraiah

. Optimization of process parameters of Al-Si alloy by centrifugal casting technique using Taguchi design of experiments. Procedia Mater Sci 2014; 6: 812–820.

42.

Falamaki

Veysizadeh

. Taguchi design of experiments approach to the manufacture of one-step alumina microfilter/membrane supports by the centrifugal casting technique. Ceram Int 2008; 34: 1653–1659.

43.

Patyal

Modgil

Maddulety

. Application of Taguchi method of experimental design for chemical process optimisation: a case study. Asia-Pacific J Manage Res Innovation 2013; 9: 231–238.

44.

Pop

Titu

Oprean

, et al. Applying experimental research management to a technological process using Taguchi’s method. Procedia – Soc Behav Sci 2018; 238: 355–363.

45.

Jankovic

Chaudhary

Goia

. Designing the design of experiments (DOE) – an investigation on the influence of different factorial designs on the characterization of complex systems. Energy Build 2021; 250: 111298.

46.

Hsieh

Tai

. Enhanced stock price variation prediction via DOE and BPNN-based optimization. Expert Syst Appl 2011; 38: 14178–14184.

47.

Ghalme

Mankar

Bhalerao

. Integrated Taguchi-simulated annealing (SA) approach for analyzing wear behaviour of silicon nitride. J Appl Res Technol 2018; 15: 624–632.

48.

Modi

Desai

. Review of Taguchi method, design of experiment (Doe) & analysis of variance (ANOVA) for quality improvements through optimization in foundry. J Emerging Technol and Innovative Res (JETIR) 2018; 5: 184–194.

49.

Martowibowo

Khloeun

. Minimum warpage prediction in plastic injection process using Taguchi method and simulation. Manuf Technol 2019; 19: 469–476.

50.

Mohammad Khanlou

Chin Ang

Talebian

, et al. Electrospinning of polymethyl methacrylate nanofibers: optimization of processing parameters using the Taguchi design of experiments. Text Res J 2014: 1–13.

51.

Dutta

Debnath Das

. Optimization and partial characterization of intracellular anticandidal protein from Aspergillus giganteus MTCC 8408 using Taguchi DOE. BIOENGINEERED 2017; 1–13.

52.

Designation: D638 – 14. Standard test method for tensile properties of plastics. Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States, 2014.

53.

Ranjit

Experimental investigation of process parameters in polyether ether ketone 3D printing

Abstract

Keywords

Introduction

Experimental setup

Domain of process parameters

Selection of working parameters

Design of experiments (DoEs) using Taguchi method

FDM machine and material

Samples preparation

Porosity evaluation

Uniaxial tension test

Results and discussion

Tensile test

SNR and ANOVA

Influence of process parameters on tensile property

Nozzle and bed temperatures

Print speed and layer thickness

Confirmation test

Conclusion

Footnotes

Declaration of conflicting interests

Funding

ORCID iD

References