Study on Effect of 3D Printing Parameters on Surface Roughness and Tensile Strength Using Analysis of Variance

Fused deposition modeling of 3D printing is the process of making workpieces or parts by adding filaments to each layer. Some indicators of a high-quality product of 3D printing are the precisions dimensions, the surface roughness, and tensile strength. This research aims to find the parameters most affecting surface roughness and tensile strength. The research design used an experimental method with input parameters: (1) print speed (15-35 mm/s), (2) print temperature (200-210  C), (3) layer height (0.1 – 0.3 mm), (4) infill line directions (0-90  ), and dependent variables were surface roughness and tensile strength. The data distribution used the L9 orthogonal array, and the statistic analysis used ANOVA. Material uses nanographite-reinforced polylactic acid (PLA) filament. The results indicate that print parameters that significantly affect surface roughness are layer height and infill line directions. The best surface roughness on the layer height parameter is 0.1 mm, and the infill line directions parameter is 90  . Based on ANOVA analysis, print speed, print temperature, and layer height do not significantly affect tensile strength, but infill line directions significantly affect tensile strength. The best tensile strength on infill line directions is 90  . The best average tensile strength with nanographite-reinforced PLA filament is 38.56 N/mm 2 , with 35 m/s print speed, 205  C print temperature, 0.1 mm layer height, and 90  infill line direction parameter. The best average surface roughness with nanographite-reinforced PLA filament is 0.66 µm, with 35 m/s print speed, 205  C print temperature, 0.1 mm layer height, and 90  infill line direction parameter. Copyright © 2023. Journal of Mechanical Engineering Science and Technology.


I. Introduction
The flow chart of making a product generally consists of ideas, designs, prototypes, performance tests, and implementation.Prototypes aim to evaluate the products before they are implemented and manufactured in mass production.Prototypes are made in small quantities so that the additive manufacturing process is prioritized over other manufacturing processes.Additive manufacturing is efficient and effective for small amounts of products [1].In addition, creating complex models using additive manufacturing can eliminate jigs and fixtures.
Additive manufacturing has several types, one of which is FDM (fused deposition modeling).Type FDM of additive manufacturing is the process of making workpieces or parts by adding filaments to each layer.Additive manufacture is appropriate if applied to prototypes that make the manufacture of varied parts and small quantities.Additive manufacturing has a cheaper and more consistent process price.As an illustration, additive manufacturing can make parts cheaper than injection molding processes in the range of 4000 to 12000 parts with a production cost of 2.1 €/part, in injection molding cheaper with parts above 12000 with a production price range below 2 €/part [2].
Additive manufacturing does not require a longer process.Additive manufacturing can make a simple process so that it requires two methods (raw material and component manufacturing), compared to traditional manufacturing, which requires three methods (raw materials, part manufacturing, and assembly parts).Besides that, in making parts, it is necessary to combine several machines for complicated shapes [2].Two hundred million users predicted in 2026, 3D printing is predicted to grow from 18% to 32% (2018 to 2026) with USD 7-23 billion to USD 51.77 billion [3].
Factors affecting the print result are the material, machine, and setting parameters.The quality of 3D print objects is affected by setting parameters, setting the distance of the reference point, and choosing a filament with the appropriate adhesion [4].Setting parameters of 3D printing greatly affects print quality.The lower layer height has an impact on increasing tensile strength, smoothness, and dimensional accuracy of 3D print objects, but affects the long print time [5].Some parameters like print speed (PS) and print temperature (PT) need to be tested.
Surface roughness and topography are the main parameters that indicate the accuracy of components.However, the average surface roughness (Ra) of arithmetic samples made by material extrusion varies between 9 and 40 μm, which can be categorized as poor surface roughness [6].Layer height (LH) or thickness affects the surface quality and dimensions of the workpiece more than other parameters such as PS and PT [7].In previous studies, researchers discussed the effect of PS, PT, and LH on surface roughness.It is necessary to show the contribution of print speed, printing temperature, layer thickness, and infill line directions to the tensile strength and surface roughness.The goal of this research is to find the parameters that most affect surface roughness and tensile strength.

II. Material and Methods
This research was an experimental study, experimental data distribution used L9 (3 3 ) orthogonal arrays.Statistical analysis used the analysis of variance (ANOVA).The ANOVA method was utilized to understand the percentage of contribution of each parameter.ANOVA analysis was used to find the critical factor for a specified response [8].In this research, data distribution used L9 (3 3 ) orthogonal arrays because it was more cost-effective than the full factorial method [9].Variable independent and dependent is shown in Figure 1, and the level of the dependent variable is shown in Table 1.The object of this study was the ASTM D638 Type IV specimen with PLA material.The design of the L9 (3 4 ) orthogonal array with three replications as shown in Table 2.The study used nanographite-reinforced polylactic acid (PLA) filament with the specifications as shown in Table 3.The print process uses a 3D printer (Creality Ender 3 Prusa i3) with a diameter of a single nozzle is 0.4 mm.The surface roughness (Ra) was measured in a Surftest SJ-310 Series (Mitutoyo, Japan), and the tensile strength test was conducted in JTM-UTS210 Computer Servo Universal Testing Machine (2T) using the standard of ASTM D638 Type IV [10] as shown in Figure 2. The statistical analysis used in this study was a three-way ANOVA (three-lane ANOVA).A three-lane ANOVA is used to test the mean differences of three or more sample groups with three independent variables and one dependent variable.In this study, ANOVA analysis used Minitab software.
The hypotheses of this study are: H0 = there is no difference between the average n groups.H1 = there is a difference between the average n groups.The interpretation of c is: If the p-value is less than α = 0.05, so H1 is accepted, or H0 is rejected If the p-value is more than α = 0.05, so H0 is accepted, or H1 is rejected

III. Results and Discussions
The data of roughness and tensile strength of 3D-printed product is shown in Table 4.

Analysis of Surface Roughness
ANOVA analysis of the surface roughness of 3D-printed product is shown in Table 5.The interpretation of data from the ANOVA results is shown in Table 6.The analysis results in Table 6 show that the 3D printing parameter variables (PS, PT, LH, and ILD) that affect the surface roughness of 3D-printed product is the PS and LH [11] and ILD.From the summary model obtained R-Square by 70.88 %, this means that the value of the influence of PS, PT, LH, and ILD on surface topology is 70.88% while other variables influence the remaining 29.12%.
The grouping information results are shown in Table 7, and the simultant test results using the Tukey test are shown in Table 8.They show that PS of 15 mm/s, LH of 0.1 mm, and ILD of 90 have a small average value, so that is a smooth surface.Based on Figure 3, LH of 0.1 mm has a better surface than LH of 0.2 mm and 0.3 mm.LH of 0.1 mm has a good surface because each print has a small height, so the nozzle output is also small and the result smoother.The smaller the print height, the better the results obtained, but the longer the printing time.The LH, followed by the nozzle diameter, are the process parameters that greatly influence the arithmetical mean height (Ra) [12].

Analysis of Tensile Strength
The result of ANOVA of tensile strength of the 3D-printed product is shown in Table 10.The interpretation of data from the ANOVA results is shown in Table 11.From the summary model, it is obtained R-Square by 61.90%.This means that the value of the influence of print speed (PS), print temperature (PT), layer height (LH), and infill line direction (ILD) on tensile strength (Y) is 61.90% while other variables influence the remaining 38.1%.Based on ANOVA analysis, PS and PT have no significant effect on tensile strength.Other variables influence based on the remaining R-square (38.1%).Other possible influencing variables, such as material and filament diameter, need to be investigated.Based on ANOVA Analysis, layer height (LH) and infill line direction (ILD) significantly affect tensile strength.3D printing type FDM has the best tensile strength at the 90-angle print (parallel to the tensile axis) and has poor tensile strength when the print angle is below 50 [13].3D Print of FDM makes shapes by adding layer by layer with a pattern like arranging fibers, therefore the best tensile strength is a tensile force that is parallel with the fibers.Nanographite-reinforced polylactic acid filaments

Figure 3 Fig. 3 .
Figure3show the grouping of LH with topographic results.

Table 1 .
Variables and data distribution

Table 2 .
Design of experiment L9 orthogonal array

Table 3 .
Characteristic of nanographite-reinforced PLA filament for 3D Print

Table 4 .
Roughness and tensile strength of the 3D-printed product

Table 5 .
ANOVA analysis results for surface roughness of the 3D-printed product Fadillah et al. (Study on Effect of 3D Printing Parameters on Surface Roughness and Tensile Strength)

Table 6 .
Interpretation of ANOVA results for average surface roughness of the 3D-printed product

Table 7 .
Grouping information using the Tukey method and 95% confidence

Table 8 .
Tukey simultaneous tests for differences means

Table 9 .
Interprestasi data Tukey simultaneous for surface roughness of the 3D-printed product

Table 10 .
ANOVA analysis results for tensile strength of the 3D-printed product

Table 11 .
Interpretation of the ANOVA analysis for tensile strength data of 3D-printed product