G-code generation for deposition of continuous glass fibers on curved surfaces using material extrusion-based 3D printing (2024)

With the rapid progress in additive manufacturing (AM) processes and the growing interest from various industries, substantial research is underway in this domain [1–4]. A key focus of this research is on manufacturing continuous fiber-reinforced composites with thermoplastic matrices, incorporating materials such as glass, carbon, and Kevlar [5]. In response to this demand, specialized 3D printers tailored for this purpose have been introduced to the market [6]. Researchers, exemplified by Huang et al [7], have concentrated on optimizing the 3D printing of polymer composites with variable fiber orientations. Zhang et al [8] delved into the forming characteristics and damage analysis of thermosetting composites with variable hardness, incorporating continuous fibers and 3D printed sheets. Liu et al [9] employed a stimulus-driven filling mapping approach for 3D printing composites with adjustable fiber density and filler morphology, suggesting integration with an optimized topology structure to enhance the fiber placement path and structural configuration. Results showcased the considerable advantages of the proposed pattern in generating adaptive filler patterns with intricate geometries. Further contributions to the field include Zhang et al's [10] introduction of a novel fabrication method for 3D printing curved shell fiber-reinforced composite sheets, Li et al's [11] presentation of a unique approach for 3D printing continuous carbon fiber-reinforced composites, and Zhu et al's [12] investigation into the effect of continuous fiber orientations on the quasi-static fracture properties (hardness) of 3D printed hybrid fiber-reinforced composites with carbon/Kevlar continuous fibers. Yan et al [13] achieved successful 3D printing of a strain sensor based on an acoustically reinforced fiber structure, while Zia et al [14] explored the mechanical behaviors and energy absorption of 3D printed poly lactic acid composites reinforced with continuous carbon/Kevlar fibers. Parker et al's [15] study delved into the strength and variability of 3D printed polymer composites reinforced with continuous fibers, and Karthik et al [16] investigated the impact of stacking sequence and silicon carbide nanoparticles on the properties of hybrid carbon/glass/Kevlar-reinforced polymer composites. Additionally, Liu and Sing [17] explored emerging materials for AM methods, including polyetherketone, polyamide, high-entropy alloys, and composites. Cuan-Urquizo and Guerra Silva [18] evaluated the mechanical properties of cellular, lattice, and porous biomaterials 3D printed using material extrusion-based 3D printers.

The exploration of 3D printing on curved surfaces and removing constraints associated with layer-by-layer printing has consistently captivated researchers. For instance, Lim et al [19], for instance, pioneered the production of curved concrete panels through 3D printing technology, utilizing a grid of threaded rods known as an adaptable membrane formwork. By adjusting the height of the threaded rod set, they established a support structure that enabled the 3D printing of curved panels. Their approach exhibited significant potential for 3D printing freeform and curved architectural surfaces. In a related study, Park et al [20] extended the application of 3D mesh printing to protective pads with curved surfaces. They initiated the process by 3D scanning various body parts, transforming the 3D model into a flat mesh structure, and subsequently printing it in a honeycomb-like 3D mesh structure. The resulting mesh displayed flexibility after 3D printing, allowing it to conform seamlessly to curved body surfaces. Tan et al [21] also employed an adaptable 3D mesh printing method to directly generate inherently curved 3D surfaces without relying on a support structure or mold. Using a calcium alginate microgel support matrix and silicone ink, they executed the freeform 3D printing process. Customized G-code, precisely matching the specifications of the scanned human body surface, dictated the freeform printing paths. They asserted that their approach held high potential for epidermal or soft robotic devices. Zhang et al [22] harnessed a five-axis material extrusion-based 3D printer to print spatially curved surfaces, comparing the mechanical properties of the printed products with those produced through layer-by-layer printing. The samples reinforced with continuous linen fibers exhibited a 29% increase in compressive strength and a remarkable 522% increase in modulus compared to the layer-by-layer printing method (flat slicing method). In a distinct approach, Zhang et al [23] introduced a process for producing parts or structures by printing along curved surfaces using six-axis industrial robots. They developed software to generate G-code for 3D printing, emphasizing that this robotic construction path, deviating from traditional vertical layer-by-layer 3D printing, enhances mechanical, thermal, electrical, and biomedical properties. This utilization of a six-axis robot in 3D printing design along curved surfaces signifies a stride towards freeform construction. Shang et al [24] explored the z-axis performance and fracture behavior of 3D printed sinusoidal reinforced composite materials, employing a seven-axis 3D printing robot for sample printing.

Zhang et al [25] introduced an AM system utilizing a robot for 3D printing reinforced polymer structures with continuous fibers, specifically focusing on grid-stiffened shell structures. They highlighted that the multi-degree-of-freedom robotic motion facilitates rapid exploration for selective and spatially distributed strengthening, thereby enhancing the AM process for reinforced polymer structures as a durable production option. In a study by Liu and Liu [26], the mechanical properties of composites with an auxetic structure that were 3D printed and reinforced with continuous carbon fibers were investigated. The results demonstrated a significant improvement in strength and modulus with a slight increase in mass. The study emphasized that the proper allocation of reinforcing fibers can enhance the auxetic behavior, although the addition of continuous fibers leads to varied performances in auxetic structures. Kang et al [27] designed tunable metasurfaces based on chiral bi-material structures, employing four-dimensional carbon fiber printing technology for their production. The study explored the impact of structural and interfacial material parameters of metasurfaces on the effective thermal expansion coefficient. Additionally, it investigated the effect of changing the curvature parameters of the metasurface on electromagnetic wave transmission through theoretical analysis, finite element simulation, and experimentation. Marchal et al [28] introduced a laminate optimization method designed to improve the stiffness of 3D printed parts efficiently, employing low computational time. The method, relying on two-dimensional stress flow, optimized the orientation of fibers for each layer in the stacking direction. Experimental testing on printed wrenches revealed an 18% increase in stiffness, underscoring the effectiveness of their developed tool.

Previous studies have predominantly demonstrated that in (AM) processes, continuous fibers are primarily deposited on flat two-dimensional surfaces [29]. However, when the need arises to deposit fibers on curved surfaces, researchers often resort to utilizing robots with high degrees of freedom. In this particular study, a novel algorithm is proposed, drawing inspiration from conventional methods in computer-aided design and manufacturing. The algorithm involves modifications to the G-code and partial changes in the device, enabling the deposition of continuous fibers and the 3D printing of composite materials using a material extrusion-based 3D printer. Beyond accurately assessing the deposition process on curved surfaces, the printed sample undergoes 3D scanning and is subsequently compared with computer-generated 3D models. This comparative analysis facilitates the calculation of dimensional accuracy and the evaluation of surface form tolerance. The main novelty of this research lies in overcoming the limitation of the 2.5D printing method, transitioning to true 3D printing with simultaneous movement in three axes, accomplished using a conventional 3D printer.

In this study, a Quantum 2020 3D printer, manufactured in Iran and based on material extrusion technology, is employed. This 3D printer has the capability to produce specimens with dimensions measuring 200 ×200 × 200 mm. Initially, a specimen featuring a curved surface, as depicted in figure 1(a), is printed. The curves constituting the surface are extracted in two-dimensional forms, as illustrated in figure 1(b). For the printing of the neat polymer specimen and the simultaneous impregnation of continuous glass fibers, polylactic acid (PLA) filament with a diameter of 1.75 mm from Sizan, Iran, is utilized. The detailed procedure for impregnating polymer and glass fibers concurrently is thoroughly explained in reference [6]. Continuous E-glass fiber yarn sourced from HITEX, China, is employed in the study. This glass fiber yarn has a mass of 0.1 g per meter, a diameter of 0.22 mm, a tensile strength of 935 MPa, and a tensile modulus of 56.67 GPa, as detailed in reference [6].

2.1.How to extract G-code

2.2.Modification of G code

The generated G-code in the previous step needs another parameter called feed of filament [30], which should be added to each line of the G-code. The feed of filament is calculated and added to each line of the G-code according to equation (1), which is stated in references [6] and [29, 30].

Where is the feed of filament in mm, is the extrusion width in mm, is the layer height in mm, is the path length of the deposited raster in mm, and is the filament diameter in mm. The length of each path is calculated based on the values of and (columns 4 and 6 of table 1) from equation (2) (column 7 of table 1).

The feed of filament is initially calculated for each line of the G-code (column 8 of table 1). According to the fiber diameter, extrusion width, and volume percentage of fibers, the parameter of the extrusion multiplier is also calculated according to equation (3) [6] and its value is multiplied by the amount of feed of filament (column 9 of table 1).

Where is the extrusion multiplier in percentage and is the diameter of continuous fibers in millimeters. Finally, the value of should be added cumulatively to each line of G-code (In each line of G-code, it is added with the value of the previous line of G-code: column 10 of table 1).

The next important point is that the ball nose tool is used in the finishing (milling) processes. Since the tool head is spherical and the goal is chip removal, at any time one point of the tool is tangent to the workpiece. This means that the center of the tool does not need to be exactly on the coordinates of the points of the path. However, in the material extrusion-based AM process, the center of the nozzle is placed exactly on the coordinates of the path. Therefore, another correction is needed so that the center of the nozzle with its rotation is always perpendicular to the direction of movement (figure 4(a)). Otherwise, the nozzle will interfere with the curved path (figure 4(b)). Therefore, it is necessary to determine the angle of the path at any point of the path and the amount of its angle to be given to the nozzle by a rotary axis. The path angle is calculated according to figure 4(c) from equation (4). Since this angle rotates around the Y axis, it comes in degrees in the G-code and in every line of the program after the letter B (columns 11 and 12 of table 1). In this case, it is necessary to make changes in the 3D printer and connect the nozzle to a stepper motor so that it can perform the required rotation. In this case, a negative angle means clockwise rotation and a positive angle means counterclockwise rotation.

In the post-processing phase, Excel software is employed to make modifications to the G-code generated by SolidCAM software. The resulting Excel file is saved with the .txt extension, and subsequently, by changing the file name, this extension is converted to the '.gcode' extension. This modified G-code is then executed after initially printing a neat polymeric specimen with a curved surface. The continuous fibers are deposited onto the 3D-printed specimen with the curved surface using this approach. For clarity, other printing parameters utilized in this process are listed in table 2, providing a comprehensive overview of the settings and configurations involved in the AM process with continuous fiber deposition.

Table 2.Printing parameters.

Parameter	Value
Nozzle Temperature	210 °C
Bed Temperature	50 °C
Fiber Diameter	0.22
Extrusion Width	0.5
Layer Height	0.22
Multiplier Extrusion	65%
Fiber Volume Percentage	35%
Printing Speed (Neat Polymeric Specimen)	40
Printing Speed (Fiber Deposition)	5

After the completion of the printing process, including the deposition of the continuous fibers onto the curved surface of the specimen, a 3D scanner, specifically the EinScan SE, is employed to scan the composite specimen. The resulting scanned file is then compared with the original 3D model to calculate deviations, dimensional differences, and geometrical tolerances. This comparison process is carried out using Geomagic Design X software, which facilitates accurate and detailed assessments of the printed composite specimen against the intended design. This comprehensive analysis helps evaluate the fidelity of the 3D-printed specimen, ensuring that it adheres to the specified dimensions and geometrical tolerances.

The modified G-code incorporates parameters for rotation angle and filament feed rate (figure 5(a)). To validate the accuracy of the code, a simulation is conducted using Simplify3D software, as illustrated in figure 5(b). It's important to note that this software lacks the capability to simulate rotational axes. Nevertheless, the proposed method enables the deposition of fibers on any surface using a 3D printer. Initially, a body with a curved surface is printed layer by layer using an extrusion-based 3D printer to assess the efficiency of the method and the accuracy of the modified G-code (figure 6(a)). Subsequently, by executing the modified G-code, continuous fibers are simultaneously impregnated with the polymer, deposited, and 3D printed onto the curved surface (figure 6(b)). The sample, with half of it deposited by continuous fibers, is shown in figure 6(c). The composite sample after the completion of the 3D printing process is depicted in figure 6(d), and the final composite part is presented in figure 6(e). These composite materials are versatile and can be utilized to manufacture complex geometric structures, such as grid-stiffened shell structures. As noted in previous research [6], the fiber volume percentage in this method can be adjusted from low percentages to 49%. The build time for a polymeric sample with a curved surface was 40 min, and the time for the deposition of continuous fibers on it was 15 min. This efficient and flexible approach demonstrates the potential for producing composite parts with intricate shapes and tailored fiber content.

The results of the 3D scanning and deviation determination are illustrated in figure 7. The depicted data reveals that the deviation falls within the range of 0.1 millimeters. According to the DIN 16901 standard, which is applicable for plastic component tolerances, this deviation value is deemed acceptable. This adherence to established standards reinforces the precision and reliability of the 3D printing process, validating its capability to produce components within the specified dimensional tolerances.