An Investigation of Printing Parameters of Independent Extrusion Type 3D Print Continuous Carbon Fiber-Reinforced PLA

Abstract

Fused filament fabrication (FFF) technology is highly favored by various industries as the simplest and most commonly used technology in additive manufacturing. The embedding of continuous fiber-reinforced thermoplastic composites (CFRTC) is a great help to compensate for the mechanical properties of FFF-printed specimens. In this paper, the optimal printing parameters of printed specimens containing continuous carbon fiber-reinforced PLA were investigated by the Taguchi method, full factorial analysis, and the tensile test. Fiber printing layer thickness and fiber printing speed are significant factors. After excluding the influence of fiber overlap, the optimal printing parameters were obtained. When the thickness of the fiber printing layer is 0.05 mm, the speed of the fiber printing nozzle is 250 mm/min, and the temperature of the fiber printing nozzle is 210 °C, the maximum tensile stress of the sample is 189.52 MPa. In this paper, the maximum tensile stress of the specimen printed by different printing parameters can be doubled, which shows the influence of printing parameters on the mechanical properties of the specimen. Compared with the specimen using pure PLA printing, the increase was 703.5%. Then the failure mechanism of 3D-printed CFRTC specimens with different layer thicknesses was investigated by using microstructural morphology and tensile fracture interfacial property analysis. The influence of layer thickness parameters on the interfacial bonding force was revealed. Through analysis, it is found that the lower the thickness of the specimen printing layer, the better the interface bonding force of the specimen, and the minimum layer thickness suitable for FFF independent extrusion printer is found.

1. Introduction

Additive manufacturing technology is a process that uses a discrete-stacking approach to stack materials layer by layer to eventually form a three-dimensional solid [1,2]. It can generate parts directly from 3D solid models without process planning, which can greatly reduce the difficulty of machining complex parts. Additive manufacturing technology has many advantages; firstly, it can manufacture extremely complex parts that are difficult to produce by traditional processing methods. Secondly, it saves materials, and some materials of additive manufacturing technology can be recycled multiple times, etc. Because of its advantages, additive manufacturing technology is now being used in the medical, aerospace, automotive, food, and engineering industries. For example, the medical industry has started to experiment with the use of 3D printing technology to develop tissues, organs, and cellular structures [3,4]. The use of 3D printing in the automotive industry has reduced the time required to process molds and reduced the cost of manufacturing prototypes [5,6]. The aerospace industry has used 3D printing to reduce fuel consumption and material waste by lightweight traditional structures or replacing them with new geometries that are structurally complex but more practical [7,8].

Additive manufacturing technologies include photopolymer molding technology, selective laser sintering, 3D printing technology, and fused deposition manufacturing [9]. Among them, FFF is the simplest and most commonly used technique in additive manufacturing technology, which has simple equipment and is easy to use. In this paper, we choose the FFF 3D printing technique, which prints pre-prepared fiber filaments through a nozzle [10]. After passing through a nozzle with a heated section, the filament is driven by the nozzle to lay the filament in a pre-set path, and the layers are stacked to finally obtain the required prototype part. The choice of filament material depends on the application and the desired properties. Today, commonly used materials include polylactic acid (PLA), PETG, HIPS, and ABS. Among them, PLA is widely used for printing in FFF technology due to its good mechanical properties and printing performance. However, the tensile properties and strength of pure PLA are not enough, and in order to improve the strength most of the current methods are to add nanofibers, [11] short fibers [12,13], or continuous fibers [14] to it. There are many researches on the printing process parameters of FDM matrix printing and short fiber reinforced composite printing. Kantaros Antreas et al. [15] studied the pure matrix FFF printing process parameters, and they believed that layer height, infill percentage, printing speed, shell thickness, printing temperature, retraction, and supports existence have immense contribution towards achieving desired printing results, and the range of setting process parameters is proposed. Altan et al. [16] investigated the effect of process parameters on the surface roughness and tensile strength of PLA samples. The samples were made according to ASTM standards and the Taguchi L16 experimental design. The results showed that the print layer thickness and print head speed were the main factors affecting the surface roughness. Mercedes Pérez et al. [17] analyzed the effect of print parameters on the sample quality using statistical and graphical methods and performed an analysis of variance (ANOVA). It can be seen that the layer height and wall thickness are the factors that affect the surface roughness. The layer height was the most significant. For a wall thickness of 1 mm, a layer height of 0.15 mm resulted in the lowest surface roughness. Ala’aldin Alafaghani et al. [18] investigated the main effects of printing parameters on dimensional accuracy, dimensional repeatability, and mechanical properties in the FFF process using Taguchi’s DOE. It was found that maximum strength and Young’s modulus can be obtained when choosing 210 °C printing temperature, 0.30 mm layer thickness, triangular filling pattern, and 100% filling. To reduce dimensional errors, 190 °C printing temperature, 0.20 mm layer thickness, hexagonal fill pattern, and 20% fill percentage can be selected as printing parameters. Arit Das et al. [19] used the DOE method to evaluate the effects of layer thickness, raster angle, and infill pattern on the tensile properties of PA6 parts reinforced with short CF processed using FFF. They found that the layer thickness and print pattern have a profound impact on the maximum tensile strength and tensile modulus.

For discontinuous fiber-reinforced composites, it is still the matrix material that dominates the mechanical properties, and adding natural fibers even reduces the tensile and flexural strength of the composite. Compared with short fiber-reinforced composites, the tensile and flexural strengths of continuous fiber-reinforced composites can be seven and five times higher, respectively.

A large number of studies have been published to demonstrate that adding continuous fibers to a pure matrix material can substantially improve the mechanical properties of the material, but it is still not as good as the mechanical properties of parts produced using traditional processing methods. To further improve the mechanical properties of continuous fiber-reinforced composite parts, some scholars have optimized the process parameters; some scholars have changed the printing method; and some scholars have applied pressure to the printed parts during printing. All these methods can improve the mechanical properties of the printed parts.

For the study of improving mechanical properties by optimizing printing parameters, Tian et al. [20] developed their own co-extrusion printer to find out the appropriate process parameters for printing continuous carbon fiber-reinforced PLA. The printing temperature was 200–230 °C, the recommended layer thickness was 0.4 to 0.6 mm, etc. The flexural strength of the printed composite specimens was 335 MPa, and the modulus was 30 GPa when the fiber content reached 27%. Dong et al. [21] used continuous Kevlar fiber-reinforced nylon material for FFF printing and investigated different process parameters such as fiber orientation, number of fiber layers, fiber position, and nylon raster angle. The results showed that the modulus of elasticity of continuous fiber-reinforced nylon could reach 27 GPa, and the strength was 333 MPa, and the mechanical properties of continuous fibers were improved much more than those of short fibers.

Andrew N. Dickson et al. [22] used a Markforged Mark One 3D printer to print continuous carbon fiber, continuous aramid fiber, and continuous glass fiber-reinforced nylon composites, respectively, to investigate the effects of fiber orientation, fiber type, and volume fraction on mechanical properties. Tensile and bending experiments were also performed. The results showed that, among the studied fibers, the specimens made using carbon fibers showed the largest increase in mechanical strength per unit fiber volume. Their tensile strength values were 6.3 times higher than those using non-reinforced nylon polymers. As the volume content of carbon and glass fibers increases, the amount of air entrapment in the composite matrix also increases, which affects the mechanical properties. The maximum tensile strength was observed in glass specimens when the fiber content was close to 22.5%. Ali Bin Naveed et al. [23] used Kevlar fibers as reinforcing fibers for ABS, PLA, PLA- c, and PLA- cu thermoplastic materials with Taguchi orthogonal tests followed by tensile and flexural tests, which increased the tensile strength up to 3 times and flexural strength up to 1.6 times compared to the pure matrix material. Yesong Wang et al. [24] used continuous glass fiber-reinforced PLA and investigated the effects of process parameters such as printing temperature, speed, layer height, and fiber volume fraction. The mechanical properties were explored using tensile and three-point bending tests. The experimental results showed that the mechanical properties of continuous glass fiber-reinforced PLA were better than those of pure PLA. When the fiber volume fraction was about 5.21% and 6.24%, the tensile and flexural strengths of the specimens were increased by 400% and 204%, respectively.

In order to improve the quality of FFF parts, many research works have been carried out by researchers in the last decade. It has been confirmed that process parameters do have a significant impact on the quality of the manufactured product. However, there are many process parameters of FFF that affect the mechanical properties and dimensional accuracy of a part. In order to fully understand the influence of various processing parameters on the parts printed by using FFF technology, it is essential to select a suitable experimental design and analysis method to take full advantage of all parameter combinations, besides an extensive analysis process. In the past, there has been a lot of research on the effect of printing parameters on the surface roughness, dimensional accuracy, and mechanical properties of parts printed using FFF. Commonly used DOE significant experimental methods that can optimize process parameters are the Taguchi method [13,15,25] full factorial design [26,27], analysis of variance [28], bacterial forging techniques [29], and fuzzy logic [30].

In this paper, an independent extrusion type 3D printer is selected. The Taguchi method was used to scientifically select the factors investigated, and then full factorial analysis was used to investigate the ultimate printing parameters of three factors: fiber printing temperature, fiber printing layer thickness, and fiber printing speed, as well as the effects of their interactions on the strength of the printed parts. The findings were verified again in supplementary tests, and the effect of fiber overlap was excluded. Finally, the interfacial morphology of printed parts with different layer thicknesses was investigated using ultra-deep field microscopy and specimen fracture forms.

2. Experimental Method

2.1. Taguchi Method

The Taguchi method is a low-cost, high-efficiency quality engineering method that emphasizes that product quality is improved not through inspection but through design. In the product design process, the Taguchi method uses the functional relationship of quality, cost, and benefit to develop high-quality products at low cost.

In this study, the modified Taguchi method of Xuesong Jiang was used [31]. The evaluation method of fuzzy set theory and correlation matrix was combined to establish a factor selection model for the selection of Taguchi method factors. The three evaluation indicators of severity (S), detection degree (D), and controllability (C) of the factors were considered comprehensively, and the corresponding fuzzy linguistic term sets established by Xuesong Jiang were referred to. Finally, they are quantified using fuzzy numbers, which are determined based on the recommendations of published articles in the field and the experience of practical experiments. The fuzzy language term description level is divided into five levels, and the corresponding fuzzy numbers are from 1 to 5 points, respectively. The language terminology is shown in

Table 1. The language terminology.

Level	Severity (S)	Detection Degree (D)	Controllability (C)	Fuzzy Number
Very low	Almost zero effect on output items	The pattern and magnitude of impact is almost impossible to determine	Almost no human control is possible	1
low	Slight impact on output items	Low probability of detecting impact patterns and sizes	Human control is difficult	2
average	Affects product appearance and some functions	Probability of impact pattern and size detection in general	Human control is possible, but additional costs required	3
high	High impact on product use, function; customer dissatisfaction	Impact pattern and size can be detected	Easy human control, no additional cost	4
very high	Serious impact on product use, function, and even endanger personal safety	Impact patterns and magnitudes can be specified	Easy human control without any cost	5

.

The formula for assessing the importance of an impact factor X_i by using the correlation matrix method is as follows:

$$V_i=W_S{V}_{iS}+W_D{V}_{iD}+W_C{V}_{iC}$$

(1)

In the equation, X_i (i = 1, 2, …, n) represents the impact factor to be evaluated. W_S, W_D, and W_C represent the weights of severity (S), detection (D), and controllability (C). V_ij represents the fuzzy number of each dimension under different impact factors, (i = 1, 2, …, n; j is severity, detection, and controllability). V_i represents the combined assessment value of the impact factor X_i under the three dimensions of severity (S), detection (D), and controllability (C). The higher the value of V_i, the higher the importance of the factor.

The factors of the research were first selected by using the Taguchi method. Where the fuzzy number was determined based on the recommendations of published articles in the field and the experience of the actual experiments. The specific scores obtained by referring to the language terminology table for different print parameters are shown in

Table 2. Impact factor evaluation correlation matrix table.

Evaluation Indicators	Severity (S)	Detection Degree (D)	Controllability (C)	Vi
Weights	0.3	0.2	0.5	-
Fiber printing temperature	4	5	5	4.7
Fiber printing layer thickness	4	5	5	4.7
Print environment temperature	3	4	4	3.7
Fiber printing speed	4	5	4	4.2
Specimen forming direction	3	3	5	4
Number of continuous fiber layers	4	5	4	4.2
Filling pattern	3	4	4	3.7
Wall thickness	2	4	4	3.4

.

According to the influence factor assessment correlation matrix table, combined with the purpose of this paper, the printing temperature, layer thickness, and printing speed of the reinforced fiber were selected as the study process parameters. Initially selected parameter ranges are reinforced fiber print temperature 190–230 °C, reinforced fiber print layer thickness 0.2–0.4 mm, and reinforced fiber print speed 120–400 mm/min.

2.2. DOE Full Factorial Test

The effect of interactions was not considered in the Taguchi method orthogonal test analysis. However, considering that there may be interactions between the printing parameters, the DOE full factorial orthogonal experimental analysis was used in the subsequent. A full factorial experimental design is defined as a design in which all combinations of all levels of all factors are tested at least once. However, when the number of factors is higher than 2, the number of trials increases at an exponential rate with the number of factors. Thus, only full factorial tests at level 2 are commonly used. In engineering or experiments, it is usually considered that a two-level test design with the addition of center points can replace a three-level test in a considerable extent. Also, the analysis is simple and easy to conduct. Usually, if the curve is bent after the orthogonal analysis, additional climbing experiments or response surface methods are used to obtain more detailed and comprehensive results.

3. Experiment Procedure

3.1. Materials, Equipment, and Test Preparation

The matrix fiber is China Eason PLA+. The continuous carbon fiber-reinforced PLA fiber is from Suzhou Bofeiyicheng Mechanical & Electrical Co., Zhangjiagang, China. The relevant performance parameters are shown in

Table 3. Print material parameter.

Material	Parameter	Value	Dimension
PLA+	Diameter	1.75	mm
PLA+	Density	1.3	g/cm3
PLA+	Tensile Strength	63	MPa
PLA+	Bending Strength	74	MPa
PLA+	Elongation at break	20	%
continuous carbon fiber-reinforced PLA	Diameter	0.38	mm
continuous carbon fiber-reinforced PLA	Tensile Strength	621.8	MPa
continuous carbon fiber-reinforced PLA	Bending Strength	594.9	MPa

.

The independent extrusion type FFF 3D printer used in this experiment is from Suzhou Bofei Yi Cheng Mechanical & Electrical Co., Zhangjiagang, China. It has two independent printheads in total, as shown in Figure 1a. One printhead prints pure matrix material and another printhead prints continuous fiber-reinforced composite material, as shown in Figure 1b. The reinforced fiber print nozzle of this printer, shown in Figure 1b, has a unique shape that can apply pressure to the printed interface. This experiment uses the universal testing machine of the Yangzhou Xinhong Testing Machine Factory to carry out the tensile test on the printed specimen, shown in Figure 1c. Finally, The ultra depth-of-field microscope of Leica Microsystems (Switzerland) Ltd., Heerbrugg, Switzerland was used to analyze the interlayer interface of specimens with different thickness.

Because the FFF technique makes parts layer by layer, the volume fraction and location of the carbon fibers can be controlled by the layer configuration. The same fiber content of the specimen is guaranteed by ensuring that each tensile specimen is layered with two reinforcement layers, each reinforcement layer embedded with 9 prepreg filaments and each fiber is length of 250 mm. The continuous fibers raster is 1.3 mm. Tensile tests were developed by using the ASTM D3039 standard. According to the standard tensile test, 250 × 15 × 1 mm specimens were made. To ensure the accuracy of the test results, five samples were printed and tensile tested for each set of parameters. The ultimate tensile stress of the samples was considered as the judging standard. The loading rate of the tensile test was 5 mm/min. The side view of the printed specimen with different thickness prepreg filament layers is shown in Figure 2.

During printing, the line width of continuous carbon fiber-reinforced PLA filament printing is 1.3 mm, the direction of continuous carbon fiber-reinforced PLA filament laying is 0°, the layer thickness of matrix filament printing is 0.15 mm, the line width of matrix filament printing is 0.5 mm, the speed of matrix filament printing is 3000 mm/min, the temperature of matrix filament printing is 220°, and the angle of matrix filament printing is ±45° alternating filament laying.

The flowchart of the experiment is shown in Figure 3.

3.2. Pre-Tests

The initial selection of the print parameters was obtained from the recommended parameters in the published papers. However, the printer that was used for the test was different, so the print parameters need to be determined again. In the parameter range selected by the Taguchi method, the edge limit print parameter is selected for test printing.

From the test printing results, it can be seen that when the nozzle temperature is too high, the prepreg filament will be softened in advance, which will lead to the failure of the filament delivery and blockage in the filament delivery tube. When the temperature is too high and the printhead is moving too slowly, the matrix at the printhead is difficult to cool quickly, which will prevent the prepreg filament from quickly sticking to the previous layer of printed specimens. When the nozzle temperature is too low and the printing speed is fast, the prepreg filament will not be fully sticking to the upper layer before it is pulled away by the tensile force of the nozzle. When the layer thickness is too high, the printed fibers cannot be completely bonded to the printed specimen’s surface. When the layer thickness is too low, excess PLA will be extruded, resulting in large accuracy errors. When the layer thickness is too high, the fiber-reinforced filament cannot connect to the print layer effectively. In summary, the final choice of parameters is print nozzle temperature of 210–230 °C, layer thickness of 0.05–0.2 mm, and printing speed of 250–400 mm/min.

3.3. Sample Preparation and Experiment

3.3.1. Full Factorial Test

In this paper, an independent extrusion 3D printer was used to prepare tensile parts with continuous fiber-reinforced composites inside. The experimental design was carried out according to the full factorial design table of L₁₁ (2³) to investigate the effects of print temperature, print layer thickness, and print speed of reinforced fibers on the strength of the tensile specimens. The optimal parameters for printing were also derived. A three-center full factorial test table was generated using the Minitab 19 software, and the test was performed as designed. The tensile test was conducted using a universal testing machine, and the tensile results are shown in

Table 4. Experimental scheme and results.

Test Number	Fiber Print Temperature	Fiber Print Layer Thickness	Fiber Print Speed	Maximum Tensile Stress (Average)
unit	°C	Mm	mm/min	N
1	210	0.2	400	87.56
2	210	0.05	250	189.52
3	230	0.2	250	145.37
4	210	0.2	250	134.85
5	220	0.125	325	153.76
6	230	0.2	400	131.16
7	210	0.05	400	180.28
8	230	0.05	400	179.21
9	230	0.05	250	186.3
10	220	0.125	325	164.14
11	220	0.125	325	167.04
Maximum tensile stress at center point (average) (test 5, 10, 11)				161.65
Pure PLA printing				23.59

.

The results were analyzed and simulated using the Minitab software. The fitting results showed that p = 0.008 < 0.05 of the model. p = 0.258 > 0.05. p = 0.291 > 0.05 for bending. It shows that the model is suitable. The designed full factor test table and the maximum tensile stress obtained are shown in Table 4.

The results were analyzed and simulated using the Minitab software. The fitting results showed that model’s p = 0.008 < 0.05, misfit’s p = 0.258 > 0.05, and bending’s p = 0.291 > 0.05. It shows that the model is suitable. The variance analysis data are shown in

Table 5. Analysis of variance.

Source	DF	F-Value	p-Value
Model	6	17.66	0.008
Linear	3	32.00	0.003
Fiber print temperature	1	3.70	0.127
Fiber print layer thickness	1	83.26	0.001
Fiber print speed	1	9.030	0.040
2-Way Interactions	3	3.32	0.138
Fiber print temperature × Fiber print layer thickness	1	5.08	0.087
Fiber print temperature × Fiber print speed	1	1.85	0.246
Fiber print layer thickness × Fiber print speed	1	3.04	0.156
Error	4
Bending	1	1.63	0.291
Misfit	1	2.46	0.258
Pure Error	2

.

3.3.2. Supplementary Test and Introduction of Fiber Overlap

In the previous section, the effects of fiber print layer thickness, fiber print speed, and fiber print temperature parameters on the mechanical properties of specimens were discussed. The thickness of fiber print layer and fiber print speed are significant factors. The resulting curve is not curved and is linear. In order to improve the accuracy of the results, several groups of experiments were added for further analysis and verification.

Firstly, the data were interpolated in the significant factors, fiber print layer thickness and fiber print speed, for analysis. Secondly, considering that the most significant factor fiber print layer thickness, the lower the fiber print layer thickness, the better the mechanical properties of the tensile specimen. It is speculated that the difficulty in improving the mechanical properties of FFF-printed parts may be due to the bubbles distributed inside the specimen. These bubbles will lead to a poor bond between the matrix and the fiber, which will lead to uneven force and easier fracture. At present, many scholars are trying to solve this problem. Some people use manual intervention during specimen forming, such as printing under vacuum or low pressure or vibration to reduce the formation of bubbles. Others choose to add a roller to the printer to bind the reinforced fibers closer to the matrix and reduce bubble formation.

For the independent extrusion 3D printer selected in this paper, the fiber printing nozzle is in contact with the hotbed by a plane with rounded corners, as shown in Figure 1b. This allows adjusting the layer height to exert a pressing effect on the fibers and parts that have been printed on the hotbed. Considering this reason, the smaller the thickness of the fiber printing layer, the better the mechanical properties of the print specimen.

When the thickness of the fiber layer is low, the carbon fiber may be flattened and dispersed, that is, the width of continuous carbon fiber-reinforced PLA will become larger. At some time, there will be a certain overlap with the adjacent reinforced fibers. When the thickness of the PLA filament is 0.05, overlap may occur. The effect of overlap degree on printing needs to be excluded or verified. In this section, the concept of “fiber overlap” is introduced in order to explore the reasons why small thickness of fiber printing layer leads to better mechanical properties.

To explore whether air bubbles can be reduced by the pressure of the print nozzle on the printed specimen or by the overlapping extrusion of two adjacent fibers. Using the print parameters in the previous section, the print temperature of continuous carbon fiber-reinforced PLA was set at a median value of 220 °C due to its non-significant factors. In this experiment, the overlap degree was adjusted by controlling the fiber print line width. When the fibers were required to be non-overlapping, the fiber print line width was 1.3 mm, and when the fibers were required to be overlapping, the fiber print line width was 0.9 mm. The test scheme and results are shown in

Table 6. Test table of fiber print layer thickness, fiber print speed, fiber overlap.

No.	Fiber Print Layer Thickness	Fiber Print Speed	Fiber Overlap	Maximum Tensile Stress
12	0.05	300	Yes	179.03
13	0.05	300	No	170.77
14	0.05	350	Yes	177.19
15	0.1	250	No	165.19
16	0.15	250	No	152.56
17	0.125	325	Yes	163.84
18	0.2	250	Yes	141.08

. The side view of the printed specimen of the 0.1mm prepreg filament layer and the printed specimen of the 0.15mm prepreg filament layer is shown in Figure 4.

4. Experimental Results and Discussion

4.1. Effect of Print Parameter on Mechanical Property of 3D-Printed CFRCs

The influence of the main effect of three factors and the second-order interaction effect on the stress of the specimen is analyzed using Minitab19. On the Pareto graph, the reference line for statistical significance is drawn at the dotted line, where it is the (1-α/2) fractional place of the T-distribution of the degree of freedom equal to the degree of freedom of the error term. When the standardized effect of the factor is greater than the value at the dotted line, it indicates that the factor is a significant factor. It can be seen from the normal and Pareto diagrams that the printing layer thickness and printing speed significantly affect the tensile stress of the print tensile specimen, as shown in Figure 5a,b. Secondly, the variance analysis of the factors affecting the mechanical properties of the tensile parts also shows that the printing layer thickness and printing speed of the fiber-reinforced nozzle can significantly affect the maximum tensile stress because their p values are less than 0.05.

Using the Origin 2021 software, the 3D color-mapped surface with projection is drawn using the fiber print layer thickness as the x-axis, the fiber print speed as the y-axis, and the maximum tensile stress as the z-axis, as shown in Figure 5c. It can be seen that within the parameters explored in the test, the slower the printing speed, the smaller the print layer thickness and the greater the maximum tensile stress the specimen can withstand. The maximum tensile stress obtained in the test at the slowest fiber printing speed of 250 mm/min and the minimum fiber printing layer thickness of 0.05 mm was 189.52 MPa. The printing temperature at this time was 210 °C. According to the above test results, the optimal process parameters of this independent extrusion 3D printer were determined: fiber printing layer thickness of 0.05 mm, fiber printing nozzle speed of 250 mm/min, and fiber printing nozzle temperature of 210 °C.

Take the data of significant factor as layer thickness and maximum tensile stress of the specimens to make scatter diagram, as shown in Figure 5d. It can be seen that the smaller the layer thickness, the better the mechanical properties of the specimen. It is again confirmed by test 15 and test 16 that the lower the layer thickness, the higher the maximum tensile stress can be at the same temperature and speed.

It can be seen from test 12 and test 14 that for the same fiber print layer thickness, the same fiber print temperature, and the same fiber print line width, test 12 with a lower fiber print speed has a higher tensile stress. As can be seen from Table 4, the average maximum tensile stress at the three center points is 161.65 MPa. It has the same fiber printing temperature, fiber printing layer thickness, and fiber printing speed as test 17, and the difference only lies in whether there is overlap between adjacent fibers, and it can be seen that the experimental group with fiber overlap can take more tensile stress. It can also be seen from test 12 and test 13 that the experimental group with fiber overlap was subjected to greater fracture stress. It is tentatively determined that the degree of fiber overlap has an effect on the maximum tensile stress.

4.2. Failure Modes and Pore Distribution of 3D-Printed CFRCs

From the above analysis, it can be seen that layer thickness is the most significant factor affecting the maximum tensile stress. In the tensile test, we can see that the bonding between layers is better when the layer thickness is lower. With the decrease in layer thickness, the failure condition changes from inter-layer failure to fracture failure. This change is very obvious in the tensile test, as shown in Figure 6.

For the specimen with reinforced fiber thickness of 0.2 mm, the failure of the specimen is accompanied by the explosion of the matrix layer, as well as long fiber pull-out and fracture. In the specimen with 0.15 mm fiber print layer thickness, it can be seen that in the range of fracture section, the matrix surface layer and the fiber layer have interlayer failure. There are fibers pulled out but the matrix surface layer does not explode. The specimen with 0.125 mm reinforced fiber print layer thickness had partial interlayer failure and very short fiber pull-out in the fracture section. For the specimen with 0.10 mm fiber print layer, the matrix surface layer and reinforced fiber layer only failed between layers with the size of a few bubbles at the fracture, and only a few fibers were pulled out. The specimen with reinforced fiber print layer thickness of 0.05 mm was neatly fractured at the fracture without fiber pulling out, and there was no interlayer failure between the matrix layer and the fiber layer. It can be seen that the smaller the layer thickness, the better the bonding condition between the layers and the stronger the interface.

It can be seen using the ultra depth-of-field microscopy that the specimen with reinforced fiber layer thickness of 0.05 mm has a great interlayer interface, as shown in Figure 7. This relatively perfect infiltration and interlayer bonding ensures almost no cracking between layers even when it is fractured. At the same time, the 0.2 mm reinforced fiber layer thickness specimen has almost complete separation of the fiber layer from the matrix layer. The 0.125 mm reinforced fiber layer thickness specimen has cracking and gaps between the fiber layers and the matrix layer. In addition, with practical printing, if the fiber layer thickness is lower than 0.05 mm, there will be part of the excess matrix overflowing the fiber layer, seriously affecting the dimensional accuracy of the printed parts. Therefore, the layer thickness of 0.05 mm is the optimal printing layer thickness for independent extrusion printers.

5. Conclusions

In this paper, the optimal printing parameters of printed specimens containing continuous carbon fiber-reinforced PLA composites were investigated by using an independent extrusion 3D printer. The fiber printing layer thickness, fiber printing speed, and fiber printing temperature parameters were optimally selected using the Taguchi method. The effects of the printing parameters on the mechanical properties of the printed specimens were investigated by using full factorial tests and tensile tests. The effects of different layer thicknesses on the interlayer bonding and fracture failure forms were investigated after excluding the interference factor of overlap degree. The results show that the fiber printing layer thickness of 0.05 mm, fiber printing nozzle speed of 250 mm/min, and fiber printing nozzle temperature of 210 °C when the specimen can withstand greater tensile stress of 189.52 MPa. In this paper, the maximum tensile stress of the specimen printed by different printing parameters can be doubled, which shows the influence of printing parameters on the mechanical properties of the specimen. Compared with the specimen using pure PLA printing, the increase was 703.5%. Fiber overlap was investigated in this paper, and it was found that it had only a small effect and did not affect the results of the full factor test. Then the failure mechanism of 3D-printed CFRTC specimens with different layer thicknesses was investigated using microstructural morphology and tensile fracture interfacial property analysis. When the printing thickness of fiber is 0.2 mm, the failure mode of tensile specimen is interlayer failure accompanied by a large number of fibers pulled out. When the fiber thickness was 0.05 mm, the failure mode of the tensile specimen was neat fracture. The influence of layer thickness parameters on the interfacial bonding force was revealed. Through analysis, it is found that the lower the thickness of the specimen printing layer, the better the interface bonding force of the specimen, and the minimum layer thickness suitable for FFF independent extrusion printer is found, which is 0.05 mm.