Optimizing Layer Thickness and Width for Fused Filament Fabrication of Polyvinyl Alcohol in Three-Dimensional Printing and Support Structures

Abstract

Polyvinyl Alcohol (PVA) is frequently applied as a support material in 3D printing, especially in the crafting of intricate designs and projecting elements. It functions as a water-soluble filament, often paired with materials like ABS or PLA. PVA serves as a momentary scaffold, supporting the jutting segments of a 3D model throughout the printing process. Subsequent to printing, the primary component can be effortlessly isolated by dissolving the PVA support using water. PVA, being a pliable and eco-friendly polymer, is susceptible to moisture. Its aqueous solubility renders it a prime selection for bolstering 3D print structures. In this investigation, equivalent-sized samples were 3D printed utilizing an Ultimaker 3D printer to assess the potency of PVA-generated specimens. Tensile examinations were executed on each sample employing a testing apparatus. The durability of the specimens was notably impacted by the input parameters, specifically the stratum width and stratum thickness. Strength dwindled as stratum width increased, whereas it rose with augmented stratum thickness. A few specimens with heightened stratum width and compromised quality displayed subpar performance during the tensile assessment. The findings unveiled a peak tensile strength of 17.515 MPa and a maximum load of 1600 N. Attaining an optimal degree of material utilization led to a decrease in filament consumption by 8.87 g, all the while upholding a MTS (maximum tensile strength) of 10.078 MPa.

1. Introduction

Three-dimensional printing, also known as additive manufacturing (AM), involves the construction of three-dimensional objects from digital models by depositing, connecting, or solidifying material under computer control [1,2,3]. Unlike traditional manufacturing methods that involve material removal, 3D printing builds objects by layering materials [4,5,6,7]. The printing process involved dual extrusion and fused filament fabrication techniques. The quality and precision of the printed object were found to be influenced by factors such as temperature management, material viscosity, layer adhesion, and the mechanical, thermal, and optical properties of the material [8,9,10]. Three-dimensional printing has revolutionized manufacturing by enabling rapid prototyping, shorter lead times, and the production of customized products. It has found applications in various industries, including healthcare, aerospace, defense, automotive, and education, facilitating innovation, efficiency, and experiential learning opportunities [11,12,13,14].

Sikidar et al. [15] investigated the effects of layer thickness on the mechanical characteristics of 3D-printed ABS polymer samples. They utilized a fused deposition modelling (FDM) 3D printing device to produce samples with varying layer thicknesses and compared them to a sample made through conventional injection molding. The results indicated that the samples produced via injection molding exhibited the highest values of tensile strength, impact strength, and hardness. Chaudhery et al. [16] comprehensively explored the 3D printing process, including methodology, materials, feed, technology, and applications. They introduced 3D/4D printing technologies and highlighted the use of PVA as a feedstock. The authors emphasized that traditional methods lacked the ability to produce complex structures, whereas 3D printing technologies, such as light-, droplet-, and extrusion-based systems, enabled their creation. The study also investigated the temperature response of PVA. Terranova et al. [17] investigated the synthesis of PVA-based nanocomposites using diamond nanograins as fillers and their application in 3D printing through additive manufacturing. They explored the use of PVA and detonation nanodiamond dispersions as innovative inks for layer-by-layer additive manufacturing of variously shaped objects. The researchers developed 3D printing technology and methods to shape hybrid materials while fabricating nanocomposites, utilizing PVA-DND inks as a test system. They emphasized the importance of aligning the chemical and physical properties of the materials with the 3D printer to enhance the quality of the final printed products. The study demonstrated that well-defined and shaped structures of PVA-DND nanocomposites can be successfully printed, offering potential applications across various fields. Dwiyati et al. [18] conducted research on the axial and lateral tensile characteristics of Acrylonitrile Butadiene Styrene (ABS) material for 3D printing. Specimens were printed with varying layer thicknesses of 0.1, 0.2, and 0.3 mm, following ASTM D 632-02 standards [19]. Tensile tests were performed using the Zwick Roell Series Z 021 machine, and SEM testing was used to analyze the fracture surface. The study concluded that the axial direction of 3D-printed specimens exhibited higher maximum force and tensile strength compared to the lateral direction, with thicker layers more likely to exhibit greater maximum force and tensile strength. Yu et al. [20] investigated the mechanical behavior of 3D preforms and their composites produced via additive manufacturing, focusing on the influence of printing direction. The compressive behavior of 3D braid preforms, and composites was analyzed for three different printing orientations (0°, 45°, and Z-direction). Pores induced by the fabrication process were observed in sections printed at 0° and 45° orientations. Solid cube specimens were then created and injected with a silicone matrix. Performs printed in the 45° direction exhibited improved inter-yarn adhesion, leading to enhanced initial modulus. However, Z-direction specimens displayed greater structural ductility due to inter-yarn slippage. Paul et al. [21] explored the application of novel biomaterials and advanced 3D printing techniques in bioprinting. The study highlights recent advancements in 3D printing technology and new materials, emphasizing their superiority over traditional methods, particularly in the field of biomedicine. Key considerations discussed include printing speed, feasibility of cell growth, and the ability to achieve complex shapes in bioprinting applications. Several studies have been conducted to investigate the mechanical properties of 3D-printed samples made from polyvinyl alcohol (PVA) and establish a correlation between printer parameters and mechanical characteristics [22,23,24,25,26,27]. These research efforts aim to understand how variations in printer settings, such as layer thickness, infill density, and printing speed, affect the resulting mechanical properties of PVA-based 3D-printed objects [28,29,30,31,32]. By identifying these relationships, researchers can optimize printer parameters to achieve desired mechanical properties in PVA 3D-printed samples [33,34,35,36]. AM, particularly Material Extrusion (MEX), refers to a swift and convenient manufacturing technique that employs raw materials in filament structure. It holds the potential to effectively utilize recycled plastics and fibers sourced from industrial waste and household recycling, making it a viable approach for sustainability [37,38].

This study explores various methods employed in 3D printing technologies, specifically utilizing a dual extrusion print head and fused filament fabrication. The objective is to investigate the influence of input parameters, including layer thickness and wall thickness, on the strength of specimens. To conduct the experiment, shape samples were produced using PVA as the material on an Ultimaker 3D printer, while adhering to the specified input parameters. Subsequently, measurements were taken from the output samples and subjected to a tensile test. Results were obtained through graphical analysis, providing insights into material strength and force analysis. Furthermore, optimization and design of experiments were applied, utilizing the response surface method to optimize the weight and maximum tensile strength of the 3D-printed samples. The goal is to enhance the weight efficiency and mechanical characteristics of PVA samples subsequent to the 3D printing procedure.

2. Experimental Work

The experiment aimed to use an Ultimaker 5 3D printer to construct samples and investigate the impact of layer thickness and width thickness on their mechanical properties. The process involved creating a 3D design using SolidWorks software (version 11) and converting it into a series of cross-sectional layers. Fused filament fabrication (FFF) technology was employed, where a molten material was layered to build the items. Thermoplastic polymers like ABS and Nylon were used, and a dual extrusion print head with an auto nozzle lifting system was utilized. PVA material was prepared by being feed into the printer’s filament spool and then into the print head, and the parameters were defined based on the response surface method levels (

Table 1. Independent process parameters with design levels.

Variable	Notation	Unit	−2	−1	0	1	2
Layer thickness	LT	[mm]	0.1	0.15	0.2	0.25	0.3
Width thickness	WT	[mm]	0.4	0.6	0.8	1	1.2

). In this study, the selected infill pattern was the honeycomb design. The specimens were fabricated following the dimensions specified for the ISO 527-2 tensile test samples [39].

Table 2. Chemical properties of PVA.

Formula	(C2H40) x
Melting Point	200 °C
Density	1.19 g/cm3
Boiling Point	228 °C
Solubility in	Water
EPA	DTXSID431930
Log P	0.26

provides the chemical properties of PVA. The printing process involved extruding PVA layer by layer to form the shape of the object. In some cases, support structures were used and later removed. Once printing was completed, the product was removed from the print bed and any remaining support structures were taken out. The samples were then subjected to tensile testing to analyze their mechanical properties.

After the completion of 3D printing, additional post-processing steps can be performed to improve the surface quality of the object, including sanding or polishing. The removal of PVA support material is achieved by submerging the printed object in water, causing the PVA to dissolve and leaving behind the primary 3D-printed component. Once dried, the part may require further post-processing, such as sanding or finishing, to achieve the desired outcome. The result is a 3D-printed object constructed by PVA materials. This procedure was repeated to produce a total of eleven samples, and the final products can be observed in Figure 1. Geometrical parameters, including weight, length, and thickness, were measured for each sample (

Table 3. Experimental layout and multi-performance results.

Experiment No.	Input Variables		Responses
	LT [mm]	WT [mm]	MTS (MPA)	Weight (g)
1	0.2	1.2	8.892	8.34
2	0.25	0.6	13.878	9.26
3	0.2	0.4	17.515	9.40
4	0.2	0.8	12.708	9.58
5	0.2	0.8	14.175	9.68
6	0.1	0.8	13.431	9.46
7	0.3	0.8	14.507	9.47
8	0.15	0.6	17.925	9.68
9	0.15	1	10.458	8.83
10	0.2	0.8	14.587	9.40
11	0.25	1	4.417	7.95

). Subsequently, all the 3D-printed samples underwent tensile strength testing using an Instron 5567 Universal Testing Machine (UTM). The nozzle temperature was set within the range of 180 to 280 degrees Celsius with a nozzle diameter of 0.6 mm.

3. Result and Discussion

3.1. Maximum Tensile Strength (MTS)

The ANOVA

Table 4. Analysis of variance (ANOVA) for MTS.

Source	Sum of Squares	df	Mean Square	F Value	p-Value Prob > F
Model	2.930 × 105	2	1.465 × 105	13.39	0.0028
A-LT	9437.48	1	9437.48	0.86	0.3802
B-WT	2.835 × 105	1	2.835 × 105	25.91	0.0009
Residual	87530.92	8	10941.36
Lack of Fit	80375.07	6	13395.84	3.74	0.2258
Pure Error	7155.85	2	3577.93
Cor Total	3.805 × 105	10

presents the designed model for the MTS of the samples after 3D printing. In this model, the effects of input parameters on the MTS are analyzed. Based on the p-values and the significant model experiments, the effects of the parameters are deemed significant. Referring to the ANOVA Table 4 and the coefficients Equation (1), it can be concluded that the effects of the WT parameter on the MTS are greater compared to the effects of the LT parameter in the printed samples.

(MTS)^2.26 = +1090.15231 − 560.87673 × LT − 768.57399 × WT

(1)

In Figure 2, the output graphs generated by the “Design Expert” software for the printed sample’s MTS are displayed. Figure 2a represents the normal plot of residuals for the MTS, showing that the residuals align closely with the trend line, indicating the conformity of the DOE for this output. Figure 2b, which is the perturbation plot for the MTS, examines the simultaneous effects of two parameters, WT and LT, on the MTS of the samples. Figure 2c presents the surface plot for the MTS of the samples. Based on this plot and the 2D contour plot of the MTS (Figure 2d), it can be observed that the MTS increases with a decrease in the WT of the printer. This is because reducing the WT results in thinner printed layers, leading to better filament consolidation in the samples. Increased consolidation of the printed layers requires more force to separate the samples, ultimately resulting in an increase in the MTS. Furthermore, considering the red regions in Figure 2d, the MTS of the samples is maximized when the WT parameter is set to a lower value for the printer. This is because decreasing the WT increases the print density and filament consumption, which leads to an increase in the MTS of the samples.

The tensile test results were obtained using a computer-controlled tensile testing machine. The results were analyzed and presented in the form of graphs. In Figure 3a, the load applied to sample 1 is plotted against the corresponding extension. It is evident that the sample exhibits elastic behavior up to an extension of 1.75 mm with a force of 510 N. Beyond this point, the material deviates from its elastic state. The maximum load observed is 760 N, achieved at a stretch of 3 mm. Subsequently, as the sample is stretched further, a decrease in force is observed until fracture occurs at a 7 mm extension. Similarly, in Figure 3b, the load applied to sample 2 is plotted against the extension. The graph shows that the sample retains its elastic properties up to an extension of approximately 1.00 mm, with a force of around 500 N. Upon exceeding this limit, the material loses its elastic behavior. The maximum load recorded is 1050 N at an extension of 2.5 mm. As the sample is stretched beyond this point, a decrease in force is observed until the sample fractures at an extension of over 7.25 mm. Comparing the results of sample 1 and sample 2, it is noteworthy that sample 2, which was printed with a layer thickness of 0.25 mm and width thickness of 0.6 mm, exhibited a 36 percent increase in maximum load and a 56 percent increase in the MTS of Elasticity compared to sample 1. Furthermore, the width thickness of sample 2 was reduced to half of that in sample 1.

3.2. Weight

In ANOVA,

Table 5. Analysis of variance (ANOVA) for Weight.

Source	Sum of Squares	Df	Mean Square	F Value	p-Value Prob > F
Model	1.250 × 105	3	41,653.09	5.12	0.0348
A-LT	7185.70	1	7185.70	0.88	0.3787
B-WT	85,196.52	1	85,196.52	10.47	0.0143
B2	32,577.04	1	32,577.04	4.00	0.0855
Residual	56,966.61	7	8138.09
Lack of Fit	53,971.72	5	10,794.34	7.21	0.1263
Pure Error	2994.90	2	1497.45
Cor Total	1.819 × 105	10

, the designed model for the weight of the samples after 3D printing is presented. In this model, the effects of the input parameters on the weight of the samples, as well as the second-order effects of WT, are analyzed. Based on the p-values, the effects of the parameters and the experimental model are found to be significant. Considering ANOVA, Table 5, and the coefficients Equation (2), it can be concluded that the effects of the WT parameter on the weight of the printed samples are greater compared to the effects of the LT parameter.

(Weight)³ = +655.92146 − 489.41132 × LT + 1087.08498 × WT − 942.73992 × WT²

(2)

In Figure 4, the output graphs of the Design Expert software (version 11)for the weight of the printed samples are shown. In Figure 4a, the normal plot of residuals for weight is displayed, indicating the residuals align closely to the diagonal line, indicating the adequacy of the DOE for this output. In Figure 4b, the surface plot for weight is presented. Based on this plot and the 2D contour plot of weight (Figure 4c), it can be observed that increasing the weight of the samples is directly related to an increase in LT. This is because an increase in LT leads to thicker printed layers, resulting in higher filament consumption and ultimately increasing the weight of the samples. Additionally, considering the red regions in Figure 4c, the weight of the samples is minimized when the WT parameter is low and the LT parameter is set on high for the printer, as reducing WT increases print density and filament consumption, leading to an increase in sample weight.

4. Optimization

In this section, the optimization of 3D-printed PVA samples using the FFF method is addressed. In this optimization, two input parameters of the printer, namely WT and LT, were optimized based on the optimization

Table 6. Constraints and criteria of input parameters and responses.

Name	Goal	Lower Limit	Upper Limit	Lower Weight	Upper Weight	Importance
A: LT	is in range	0.1	0.3	1	1	3
B: WT	is in range	0.4	1.2	1	1	3
MTS	maximize	4.41779	17.925	1	1	3
Weight	minimize	7.95	9.68	1	1	3

, where the parameter ranges for both samples were evaluated, considering the upper and lower limits. Since the responses for this study are MTS and Weight, a goal was set for each in the optimization analysis, aiming to increase the MTS and decrease the Weight of the printed PVA samples. Furthermore, according to

Table 7. Predicted optimum results and experimental validation.

No.	LT	WT	Desirability		MTS (MPa)	Weight (g)
#1	0.3	0.85	1	Predicted	13.540	9.15
				Actual	11.985	9.05
				Error %	12.97	1.10
#2	0.25	0.85	0.953	Predicted	11.603	8.58
				Actual	12.091	8.99
				Error %	4.03	4.56
#3	0.25	0.95	0.947	Predicted	10.815	9.07
				Actual	10.078	8.87
				Error %	7.31	2.25

, three optimized samples were shown for this experiment, and the predicted values by the Design Expert software and the actual data from the tests were reported for these three optimized samples. As evident, the error values for the samples are below 15%, which is highly suitable for predicting the optimization of the samples based on previous studies, and conducted research. Additionally, considering the overlay plot in Figure 5, it demonstrates the optimized regions for the input parameters of the printer, where selecting a parameter within the yellow regions will result in optimal responses.

5. Conclusions

The constraints and future potential of PVA 3D printing encompass its susceptibility to moisture, causing degradation, print issues, and extruder clogs, necessitating proper storage. Primarily used as support material in dual-extrusion 3D printing, PVAs limited application excludes standalone objects due to properties and cost. Slower print speeds increase project duration, deterring time-sensitive tasks. Higher cost hampers larger projects and cost-effectiveness goals. Its biocompatibility is promising for temporary medical implants. PVA finds utility in education and rapid prototyping due to easy support removal. Anticipated technological advancements could alleviate limitations and expand prospects as researchers strive to optimize PVA-based 3D printing for diverse applications. Below are notable accomplishments from this research:

• Mechanical properties of printed samples were significantly influenced by the width thickness parameter.
• Low layer thickness and high width thickness combo reduces tensile strength due to poor layer adhesion in width direction.
• Using 0.25 mm layer thickness and 1 mm width thickness requires less material but increases voids and lowers tensile strength.
• Sample 8: 0.15 mm layer thickness, 0.6 mm width thickness, MTS of 17.925 MPa.
• Sample 3: 0.2 mm layer thickness, 0.4 mm width thickness, MTS of 17.515 MPa, maximum load 1600 N. Sample 3 maintains good MTS with a reduced filament consumption of 8.87 g (10.078 MPa).