Influence of Motion Mechanism Change in 3D Printers on the Quality of Printed Models

The objective of this study is to compare the quality of 3D printed models manufactured using the Fused Filament Fabrication (FFF) method. The samples were fabricated using a robotic arm with 4 axes and conventional 3D printers. The focus of this experiment lies in assessing the influence of the devices’ structural design on the resulting quality of the 3D printed models. Additionally, the study aims to identify the strengths and limitations of each device and define their respective applicability. The 3D model designed for this investigation comprises intricate geometrical shapes specifically chosen to evaluate the precision and repeatability of layer deposition while establishing geometric tolerances and determining shape deviations. The samples were 3D printed under identical printing conditions and parameters, and subsequently, these produced samples will undergo 3D digitization through an optical scanner, namely ATOS II Triple Scan. The obtained data will then be subjected to a comparative analysis utilizing GOM Inspect software to determine the geometric tolerances. The findings from this analysis will be critically evaluated and serve as a basis for informing and guiding future research endeavors.


Introduction
Additive manufacturing is a term that encompasses a group of technologies enabling the production of three-dimensional objects through the melting and layer-by-layer deposition of thin material layers [1,2].Technological devices supporting additive manufacturing are allowing the fabrication of objects from various materials, most commonly plastic or metal [3,4].Production is carried out based on computer-aided designed geometries of components.[5].Additive manufacturing has found application in various industrial sectors such as the automotive industry, food industry, plane production or medicine, often being utilized in the production of components with complex geometries and intricate properties [6,7].Its significant utilization has been particularly prominent in prototype development and concept validation, significantly streamlining and speeding up designing and manufacturing new products [8].Additive manufacturing facilitates the production of components with enhanced heat transfer capabilities (lattice structures), porous/hollow sections, with more efficient mass distribution, or components with improved aesthetics [9], [10].A pivotal factor in the continuous advancement of additive manufacturing is the flexibility of design, ease of product adaptation, reduction in production times and a decrease in the overall number of final product parts.This, especially when combined with the aforementioned attributes, makes additive manufacturing stand out without competition [11].
It is imperative to note that the design of devices used in this type of production can have a substantial impact on the quality of the resulting products [12].Characteristics of these devices, such as the motion control system of the print head or platform in the XYZ axes, the material deposition method, and the control system, can significantly influence the surface and mechanical properties of the manufactured products [13].These factors can provide advantages in the production of specific types of products [14].The attributes of additive manufacturing are the primary drivers behind ongoing research and the emergence of new technologies and applications dedicated to additive production, including additive manufacturing using industrial robots [15].

Methodology
The aim of this experiment is to compare the impact of device construction design on the quality of 3D printed models, assess their advantages and disadvantages, and define potential applications for the compared devices.
To achieve the stated objective, a test sample was designed and printed using three devices with different construction designs.For the experiment, two conventional 3D printers were selected, both with the same construction design but differing material feeding mechanisms -"direct" material feeding and "indirect" material feeding.The third device utilized a 4-axis robotic arm with an "indirect" material feeding method.Samples were printed on all three devices, and subsequently, 3D digitization was performed on each sample, followed by measurements.The gathered data were then processed and analyzed to accomplish the following objectives: • To determine the impact of the construction design of the utilized devices on the quality of the printed samples.• To identify the advantages and disadvantages of each individual device.
• To define the potential applications of the compared devices.A more detailed description of each individual device and the steps involved in conducting the experiment are outlined in the following subsections.
The experiment compares the print quality of 3 different devices and focuses on the printing accuracy of defined geometric shapes using these devices.The main steps of the experiment are depicted in Figure 1.

Design of the test sample
In order to investigate the impact of the construction design of the utilized devices on the quality of the printed samples, to identify the advantages and disadvantages of each individual device, and to define the potential applications of the compared devices, a test sample model was designed (see Figure 2).This test sample model was used as a standardized representation to conduct the comparative analysis among the different printing devices.

Figure 2. 3D model of test sample
The model was designed with the aim of testing manufacturability, quality, and precision of individual components of this model.The measurement object is a square-shaped specimen with dimensions of 50x50x5 mm, which includes the following features: • Four circular holes with diameters of Ø4, Ø6, Ø8 and Ø10 mm.

3D printing of the test samples
The production of components was executed using three distinct devices, namely, the Prusa i3 MK2, Creality Ender 3 V2, and DOBOT Magician.These devices, along with their respective parameters, are presented in Table 1 for reference.The 3D printing was performed using the same printing parameters for all three devices, which are listed in the Table 2

3D digitization of samples
The 3D digitization of the printed samples was carried out using the optical 3D scanner GOM ATOS II Triple Scan.Prior to the actual 3D scanning of the printed samples, the specimens were cleaned and marked with reference points.Scanning was performed with the aid of a turntable, and the device captured images at each 30° rotation of the turntable, resulting in a sequence of 12 steps and a complete 360° rotation.Two scanning series were conducted on all samples, capturing both the top and bottom parts.The reference points were used to ensure a more precise alignment of individual images and scanning series.The scanning process generated point clouds for each sample, which were subsequently converted into 3D polygonal models in ".stl" format through triangulation.

Measurement of samples
Afterward, the measurement of individual samples was carried out using the GOM Inspect software on 3D models in ".stl" format.The software's working environment was initially set up by importing the model obtained through 3D digitization and the 3D CAD model of the designed component.Using the "Prealignment" function, these two models were aligned, and a unified coordinate system was established for comparison.Subsequently, measurements were conducted by comparing these two models.For each sample, a colormap of deviations was initially generated, and an example of this colormap can be seen in Figure 3. Subsequently, measurements were conducted for the diameters and cylindricity of cylindrical holes and protrusions, along with the diameters of semi-spherical protrusions.Next, the dimensions of square protrusions and holes were measured, followed by the overall dimensions of the component and the flatness of individual surfaces.The measurement process initially focused on a single sample, and the measured values were graphically inserted into the measurement report.After completing the measurements and generating the measurement report for the first sample, a measurement stage was created.Using this stage, all remaining samples were loaded into the software, and the software automatically conducted measurements on all samples and generated measurement reports in PDF format for each of them.

Measured data and their processing
The measured data was processed and compiled into an Excel spreadsheet.For each investigated printing parameter, the data was analyzed, and arithmetic averages of the nominal values were calculated for each device.Subsequently, graphs were generated based on the measured and processed data for the investigated variables.This data processing approach allowed for a comparative analysis of the devices and enabled the identification of their deviations through the experiments.

Results
The processed data in the form of arithmetic averages of nominal values for each device can be seen in the following figures.As shown in Figure 5, after evaluating the overall dimensional deviation of the samples, it was found that the highest deviation values were observed in the samples produced by the "Dobot" device.In the X-axis, the deviation value is 0.276 mm, and in the Y-axis, it is 0.262 mm.The devices "Prusa" and "Dobot" have nearly identical deviations between the X and Y axes.In contrast, the "Creality Ender" device exhibits the lowest deviation in the X-axis, with a value of 0.037 mm, while in the Yaxis, it reaches a value of 0.156 mm.This difference may be attributed to the improper leveling of the print bed in the "Creality Ender" device, leading to likely layer shifting along the Y-axis and subsequent model stretching.The print bed leveling for the "Creality Ender" device is manual, which means there might be human errors during its setup before starting the 3D printing process.
The highest deviations in surface flatness are observed on surface A1 for the "Prusa" and "Dobot" devices.This surface represents the final layer, also known as the top layer, which significantly influences the surface quality of the printed model.Improper software settings can lead to higher surface flatness deviations on this layer.Important parameters that affect these values include material flow rate, movement speed, and material retraction when changing the printing head's path.

Discussion
The results of the experiment indicate that several factors can influence the quality of 3D printing.
Variations can be attributed to factors such as the precision of individual parts' movement in the devices, the design and software solutions of the devices, print bed leveling, or the trajectory generation algorithm for printing specific objects.The outcomes and insights from these conducted experiments are valuable for planning further research focused on selecting appropriate device designs in the field of additive manufacturing.The goal of the ongoing experimental series is to define suitable parameters for 3D printing and select the appropriate construction solutions for 3D printers and devices enabling 3D printing (e.g., Dobot) for specific industrial applications.Future planned experiments will focus on identifying potential limitations of selected 3D printing devices.

Conclusion
Additive manufacturing and 3D printing are becoming increasingly important tools in industry and research.Their flexibility, speed, and ability to create complex geometries with ease provide significant advantages for prototyping, product development, and solving specific manufacturing challenges.This article focuses on comparing the quality of 3D printed models produced using the Fused Filament Fabrication (FFF) method with different devices.The experiment aimed to compare the device's construction design and its impact on the quality of printed samples, determine the advantages and disadvantages of each device, and define their potential applications.For the experiment, a test sample was designed, comprising specific geometric shapes to verify the accuracy and repeatability of printing individual layers and define geometric tolerances.The samples were printed on three devices -two conventional 3D printers and a robotic arm with 4 axes.Subsequently, all the samples were digitized using an optical scanner and measured to obtain data on geometric tolerances.
The evaluation of the measured data revealed differences in the printing quality among the individual devices.It was found that the "Dobot" device exhibited the highest values of deviations and inaccuracies during printing, especially when dealing with small dimensions and circular shapes.On the other hand, the "Creality" device achieved the lowest values of deviations, while the "Prusa" device showed higher values of deviations, particularly with square shapes.These results suggest that the device's construction design and software settings significantly influence the quality of 3D printed models.The accuracy and repeatability of printing can be affected by the mechanical properties of the device and the adjustment of printing parameters.Furthermore, it was observed that the used devices have various advantages and disadvantages, and their suitability for specific applications may vary.These findings highlight the importance of carefully considering the selection of the appropriate 3D printing device based on the intended use and desired print quality.
The results of this experiment serve as a foundation for further planned experiments, which will continue to explore and optimize the printing process using these devices.The goal is to achieve better accuracy, repeatability, and print quality, as well as to define the ideal settings for each device based on the specific application requirements.These future experiments aim to enhance the capabilities of the devices and ensure their optimal performance in various applications.

Figure 1 .
Figure 1.The main steps of the experiment

Figure 3 .
Figure 3. Example of colormap generated in software GOM Inspect for specific sample

Figure 4 .
Figure 4. Results a) chart of permissible deviation limits for linear dimensions b) illustrational picture of measured parts

Figure 5 .
Figure 5. Results a) chart of permissible deviation limits for linear dimensions and cylindricity of circular holes, b) illustrational picture of circular holes, c) graph of absolute values of arithmetic averages of deviations for linear dimensions of circular protrusions, d) illustrational picture of circular protrusions

Table 1 .
The devices used in the experiment, along with their properties