FDM Process Parameters Impact on Roughness and Dimensional Accuracy of PLA Parts ^†

Abstract

Interest in research on FDM systems using inexpensive materials like PLA and ABS is constantly increasing. In this regard, the scope of this study is narrowed to exclusively focus on PLA. To improve the surface finish of PLA printed products, it is important to have optimal values of the most important process parameters, notably layer height, temperature, and printing speed. The surface roughness is a critical aspect of additive manufacturing that directly impacts the functionality, aesthetics, and overall performance of printed parts. To accomplish the improvement of surface quality, the statistical method ANOVA (Analysis of Variance) is used to analyze data and identify the most relevant process parameters that impact roughness and dimensional precision. The response variables are identified during this study in order to define the optimal printing parameters for improving part quality and ensuring the best surface finishes. Additionally, the dimensional accuracy of the parts is analyzed in order to check the reliability and effectiveness of the optimum parameters. The results are validated through this additional assessment, which also provides insight into the capabilities and limitations of inexpensive FDM machines when the optimized parameters are used. In conclusion, this study emphasizes the significance of enhancing parameters to improve the performance of 3D printed components, providing insightful information about the potential of PLA as an inexpensive material for applications that need both high surface quality and precise dimensional control. According to the analysis, the thickness of the layers and printing speed have a significant role in the roughness for a better desired surface quality.

1. Introduction

Additive manufacturing, also called 3D printing, appeared around 1980 [1]. One of the most used additive manufacturing techniques is Fused Deposition Fabrication (FFF), often called Fused Deposition Modeling (FDM). This technique consists of depositing material layer by layer and permits the fabrication of complex parts in shapes that are difficult to make with conventional manufacturing technologies.

The technologies of additive manufacturing were initially principally used for prototyping and have evolved through the years, enabling the fabrication of moderate- to large-scale end-use parts while catering to individual customer needs more efficiently [2]. Over the past few decades, a larger interest in enhancing the surface quality of FDM parts has appeared. Research has been conducted using different methods to improve surface roughness, including the optimization of printing parameters, the use of chemical treatments, and the refinement of slicing techniques. For example, Anitha et al. [3] studied the influence of printing parameters such as layer thickness and printing speed on surface roughness using the Taguchi method. Galantucci et al. [4,5] studied the effects of a chemical dipping process using dimethyl ketone and water in order to improve the surface quality of ABS parts, finding that roughness could be significantly improved just by making minor changes in protype dimensions.

From another perspective, additive manufacturing technologies such as stereolithography (SLA), selective laser melting (SLM), and selective laser sintering (SLS) have transformed the fabrication of 3D objects. SLA technology uses photopolymerization to solidify liquid resin and produce parts using diverse materials with significant dimensional precision [6]. Huang et al. analyzed additive manufacturing techniques and their utility in various applications, such as medicine and small-scale pH sensor production [7].

SLM, by contrast, utilizes high-energy lasers to completely melt metal powders, allowing the production of complex metal components with enhanced mechanical properties [8]. Meanwhile, SLS employs a high-power laser to fuse powdered materials, such as metals and plastics, making it useful in various industries, including medicine, where it is applied in pharmaceutical dosage forms [9].

Surface roughness and dimensional accuracy in FDM printed parts remain areas for improvement, as these parts typically exhibit higher roughness and dimensional inaccuracies compared to those produced by alternative methods. Issues such as warping and dimensional deviations during printing are common, as noted in prior research [10]. Studies focusing on PLA printed parts using FDM for dental applications reveal that higher printing temperatures and lower layer thicknesses improve surface roughness, with 230 °C and a 0.16 mm layer thickness yielding better results [11]. Furthermore, other researchers found that reducing layer thickness minimizes roughness, and a lower flow rate reduces dimensional inaccuracies [12].

In the review literature, the importance of printing parameter optimization to achieve a better quality result is underlined. Depending on the used material, specific settings and improper parameter optimization lead to waste and increased energy use. Some studies focus on the optimization of FDM parameters to address sustainability issues by minimizing material waste and energy consumption [13]. While much research has focused on inexpensive FDM systems using widely used polymers like PLA and ABS, this study narrows its focus to PLA. The goal is to enhance surface roughness by fine-tuning parameters such as layer height, printing temperature, and printing speed. ANOVA is used to analyze the data, identifying key response variables, which are then used to determine optimal printing conditions. Dimensional accuracy evaluations will confirm the results and demonstrate the quality of inexpensive FDM machines using the optimized parameters.

2. Methodology and Equipment

2.1. Software and Equipment

The following tools were used during this study:

-

CATIA v5: A 3D computer-aided design (CAD) software used for product design. It is used in this work to create the model and generate the STL file.

-

Minitab 21: Statistical and data analysis software. It is used to perform the ANOVA analysis in this study.

-

Mitutoyo Surf test SJ-410: A highly accurate, high-performance portable roughness measuring device.

-

Mitutoyo 500-160-30: A digital caliper that is accurate, reliable, and easy-to-use for precision measurements, specific to the 500 series. It offers a measuring range of 0 to 150 mm (or 0 to 6 inches) and has 0.01 mm of resolution. In this work, it is used for dimensional accuracy measurements.

2.2. Procedures and Setup

This work investigates the roughness and dimensional precision of FDM printed parts technology. A 3D model, created in CATIA V5 (an engineering software) and converted to an STL file, was printed using affordable 3D printers. The material used for these tests was PLA. The printing machine used in this experiment is a Creality Ender 3 V2 desktop 3D printer with bed dimensions of 220 mm × 220 mm × 250 mm. The experiments utilized a 0.4 mm nozzle diameter and maintained a constant bed temperature of 60 °C. Printing parameters were set with three contours and an infill flow rate of 100% with a line infill pattern. Printed parts were evaluated for surface roughness (Ra) using the MITUTOYO Surftest SJ-410 (Mitutoyo Corporation, Kawasaki, Japan), which provided average Ra values. Dimensional accuracy was assessed by comparing features against their nominal dimensions, including measurements of height (mm), width (mm), and thickness (mm). Measurements were recorded using MITUTOYO 500-160-30 (Mitutoyo Corporation, Kawasaki, Japan) Vernier calipers with a resolution of 0.01 mm. The collected results were analyzed to draw conclusions about the impact of the printing parameters on the components.

2.3. Model Preparation

For the tests on dimensional precision and surface roughness, the model was designed in CATIA V5 according to straightness and flatness values outlined in ISO 12780 [14] and ISO 12781 [15]. The model’s dimensions were set to 20.0 mm × 20.0 mm × 3.2 mm, as illustrated in Figure 1 and Figure 2. This research examined various printing parameters, including printing speed, layer thickness, and printing temperature. Printing speeds were configured at 40, 60, and 80 mm/s, while layer thicknesses were adjusted to 0.12, 0.16, and 0.2 mm. Printing temperatures were controlled at 190 °C, 200 °C, and 210 °C. Surface roughness was assessed in the direction perpendicular to the build orientation, and the average values were determined for each specimen. Figure 3 presents the orientation used for measuring surface roughness of the specimens.

The study evaluated three parameters: layer thickness (mm), printing speed (mm/s), and printing temperature (°C). The experimental design utilized Minitab 21 software and followed the Taguchi method with a 3³ design matrix. Nine samples were produced according to this setup.

Table 1. The orientation used for measuring surface roughness of the specimens.

Level	Layer Thickness (mm)	Printing Speed (mm/s)	Temperature (°C)
1	0.12	40	190
2	0.16	60	200
3	0.2	80	210

presents an overview of the parameters and their corresponding levels.

3. Results

3.1. Roughness Test

In this test, we measured the roughness (Ra) values of the parts. The arithmetic averages of Ra values for different printing parameters are indicated in

Table 2. Measured Roughness values (Ra) of the printed parts.

Test No.	Layer Thickness (mm)	Printing Speed (mm/s)	Temperature (°C)	Average Ra Value (μm)
1	0.12	40	190	3.202
2	0.12	60	200	3.795
3	0.12	80	210	3.979
4	0.16	40	200	4.601
5	0.16	60	210	4.388
6	0.16	80	190	5.306
7	0.2	40	210	5.852
8	0.2	60	190	6.230
9	0.2	80	200	7.092

. The maximum Ra value was 7.092 μm and corresponds to the 9th sample, and the minimum Ra value was 3.202 μm and corresponds to the 1st sample. The results show the important impact of layer height, printing process speed, and printing temperature on the roughness Ra results.

In order to clarify perceptions of these effects, an ANOVA analysis was conducted using Minitab software. The results are presented in

Table 3. ANOVA of the transformed response.

Source	DegFre a	Adjus SSqu b	Adju MSqu c	F-Value	p-Value
Layer thickness	2	11.3361	5.66803	898.72	0.001
Printing Speed	2	1.3159	0.65796	104.33	0.009
Printing Temperature	2	0.2701	0.13506	21.41	0.045
Error	2	0.0126	0.00631
Total	8	12.9347

^a Degree of freedom. ^b Adjusted sum of squares. ^c Adjusted mean squares.

. Figure 4 shows the correlation between the process parameters. The relevant printing parameters that influence the roughness value Ra were revealed by analyzing the mean response values.

A layer thickness of 0.12 mm gave us the minimum roughness value. The roughness value was augmented significantly by increasing the layer height to 0.2 mm. This result highlights the influence of layer thickness on the finish of the printed parts; here, a small layer thickness led to a smooth surface and finer resolution compared to a large layer height. These results are consistent with previous research that showed that reducing layer thickness results in a lower Ra value [16].

For printing temperature, a minimum Ra value was found at 190 °C, and the maximum at 210 °C. Even if temperature did not show a significant impact on ANOVA analysis, the trend suggests that a higher temperature (210 °C) may facilitate the diffusion of the molecules and bonding among filaments, leading to a smooth surface. However, the effect is less pronounced compared to printing speed. The lower temperature (190 °C) restricts the movement of molecules and results in weaker bonding and a rough surface. This analysis is consistent with prior research [17] that showed that a higher nozzle temperature improves filament bonding and surface quality, but speed effects remain more impactful.

For the printing speed, the lowest (Ra) value was detected at 40 mm/s, while the highest occurred at 80 mm/s. The difference in Ra values between the two speeds is related to the movement of the print head during layer deposition. At a higher speed, the extruded material cannot be deposited correctly because of the rapid movement, causing an unstable material flow within the extruder and rougher surfaces. On the contrary, at a slower speed (40 mm/s), the material takes enough time to be deposited accurately, leading to a smoother surface. This effect is consistent with previous studies that linked higher printing speeds to increased surface roughness [18].

The results in Table 2 show that layer thickness represents the most important parameter that influences Ra, followed by printing speed and printing temperature. These results show that adjusting the printing process parameters can improve the surface quality of the parts. An ANOVA table was generated to study the experimental results and identify the parameters that influence surface roughness. ANOVA analysis showed that both layer thickness and printing speed have a significant effect on surface roughness. The results clearly show that layer thickness and printing speed are the most influential parameters on the roughness value Ra, with p-values of 0.001 and 0.009. On the other hand, printing temperature was deemed insignificant, as its p-value is close to 0.05.

Therefore, concretely, a very thin layer thickness value will produce good surface quality without defects since the layers will be well adhered to each other. In addition, a low printing speed will allow these layers enough time to be deposited gently and to cool properly, thus guaranteeing good surface quality.

3.2. Dimensional Accuracy Test

The dimensional precision of the printed specimens was assessed. The assessment was conducted by comparing the measurements to the dimensions in the design. For this accuracy test, variable parameters were used, as listed in Table 1. These parameters were adjusted to analyze their influence on dimensional accuracy.

The overall dimensional accuracy of the 3D printed samples produced using the Creality Ender 3 V2 3D Printer (manufactured in Shenzhen, China by Creality3D) is presented in

Table 4. Dimensional accuracy of printed samples.

a. Dimensional Accuracy for Length in mm.
Design Dimension	Sample No.	Reading	Difference	Measuring Tools
20	1	20.18	0.18	Vernier Caliper
	2	20.36	0.36	Vernier Caliper
	3	20.61	0.61	Vernier Caliper
	4	20.52	0.52	Vernier Caliper
	5	20.54	0.54	Vernier Caliper
	6	20.43	0.43	Vernier Caliper
	7	20.64	0.64	Vernier Caliper
	8	20.33	0.33	Vernier Caliper
	9	20.44	0.44	Vernier Caliper
b. Dimensional Accuracy for Thickness in mm.
Design Dimension	Sample No.	Reading	Difference	Measuring Tools
3.2	1	3.23	0.03	Vernier Caliper
	2	3.23	0.03	Vernier Caliper
	3	3.24	0.04	Vernier Caliper
	4	3.33	0.13	Vernier Caliper
	5	3.33	0.13	Vernier Caliper
	6	3.36	0.16	Vernier Caliper
	7	3.17	0.03	Vernier Caliper
	8	3.19	0.01	Vernier Caliper
	9	3.2	0	Vernier Caliper

. The difference between the measured length and the design dimension of 20 mm ranges from 0.18 mm to 0.64 mm. Similarly, for the thickness (design dimension of 3.2 mm), the difference ranges from 0.01 mm to 0.16 mm. These measurements were obtained using Vernier calipers, ensuring precise evaluation of dimensional variations.

The results demonstrate that the length of the printed parts deviated by a percentage ranging from 0.9% to 3.2%, while the thickness exhibited a higher variation, ranging from 0% to 5%. These findings highlight that dimensional accuracy for length is relatively consistent, whereas thickness is more affected by variations. Overall, the results demonstrate that the 3D printer maintains acceptable performance for length and thickness dimensions, although further optimization may be required to minimize deviations, particularly for thickness.

The default flow rate settings of inexpensive 3D printers may not be suitable for all geometries and sizes. Road width was identified as another contributing factor to part distortion, as increasing it raises residual stress, though this factor was not controlled in the study. Despite these limitations, the Creality Ender 3 V2 demonstrated promising performance. The majority of printed parts and features stayed within a 5% difference from the defined dimensions, in comparison with the performance of other commercial desktop 3D printers evaluated in prior research [19].

The main effects plots for length and thickness, presented in Figure 1 and Figure 2, highlight how process parameters affect dimensional accuracy. Figure 5 demonstrates that the error in length is minimized at a layer thickness of 0.12 mm, a printing speed of 60 mm/s, and a temperature of 190 °C. Among the parameters, temperature has the most significant effect on length accuracy, with the mean length increasing consistently as temperature rises.

Similarly, Figure 6 shows that the error in thickness is at its lowest at a layer thickness of 0.20 mm, a printing speed of 40 mm/s, and a temperature of 210 °C. In this case, layer thickness exerts the greatest influence, with a noticeable peak in accuracy at 0.20 mm before declining at lower values.

These main effects plots offer important insights into how process parameters influence dimensional accuracy, helping to pinpoint conditions that minimize deviations.

4. Conclusions

This paper studied the roughness of surfaces and the dimensional precision of PLA parts produced by Fused Deposition Modeling (FDM) technology. The main conclusions are as follows:

• Layer thickness significantly impacts surface roughness (Ra). Surfaces are smoother with a smaller layer thickness (0.12 mm) compared to thicker ones (0.2 mm).
• Printing speed and temperature also influence RA and dimensional accuracy. A lower speed (40 mm/s) gives smoother surfaces, and optimal temperatures (200 °C) improve filament bonding without increasing surface roughness.
• Dimensional Accuracy: It was confirmed that the dimensional deviations of parts using optimum parameters remain within a 5% threshold.
• Inexpensive FDM potential: FDM printers produce high-quality products by fine-tuning parameters such as layer thickness, printing speed, and temperature.

This study contributes to improving additive manufacturing by focusing on the potential of inexpensive 3D printed parts to achieve accurate and smooth products for industrial and commercial applications.