Process optimization of compressive property and dimensional error on wood polylactic acid gyroid-structured polymer composite

Abstract

In the fused deposition modeling (FDM) process, various thermoplastic filaments may be used as a feedstock material for the component fabrication. The present study involves incorporating a gyroid structure in wood/polylactic acid (PLA) polymer composite. The strength of the sample may reduce while incorporating lattice structure in PLA polymeric samples. There are certain difficulties due to the wide availability of process parameters that may vary the quality and strength during the fabrication of the sample through the FDM process. Determining the influence of certain important process parameters such as raster angle, layer thickness, and wall thickness is carried out in this research to attain higher mechanical strength and less dimensional error in the geometry of the fabricated sample. Taguchi L₉ orthogonal array is used in these experiments to optimize process parameters in the gyroid incorporated samples. The output responses in the present study are compressive strength and dimensional error. Both the output responses were predicted using the ANOVA technique. Both the output responses are greatly influenced by the raster angle with the value of 60.78% in compressive strength and 90.43% in the dimensional error, and Wall thickness is the least influenced process parameter with the value of 7.17% in compressive strength and 0.93% in dimensional error. The sequential order of influencing process parameters in both the compressive strength and dimensional error were raster angle > layer thickness > wall thickness. 31.019 MPa of compressive strength was observed in the confirmational compression test. The prepared composite can be used as a structural material in the replacement of balusters and handrails.

Keywords

taguchi ANOVA structure compression dimensional error

Introduction

Additive manufacturing (AM) is the process of manufacturing components by layer. The process integrates several layers accordingly to the design of the required component compared with the conventional manufacturing process. A limited amount of material is wasted during the process of additive manufacturing. Manufacturing of components through AM is also very simple compared to other conventional manufacturing methods. In recent years, industries have widely used additive manufacturing due to its cost and time efficiency, and it is a well-known technology for simply fabricating complex geometries.¹ Other than the AM machine, the process just requires the CAD model and the slicer. Direct digital manufacturing, layer manufacturing, additive fabrication, rapid prototyping are the various terms that represent additive manufacturing. Additive manufacturing plays a keen role in fabricating small-sized prototypes. At the same time, the fabrication of small-sized components through other conventional methods should not be cost-efficient compared to additive manufacturing.

There is no post-cleaning process in additive manufacturing, and the complicated structures can easily be built through this method. Stereolithography (SLA), inkjet printing (IP), selective laser sintering (SLS), laminated object manufacturing (LOM), and fused deposition modeling (FDM) are the commercially available methods of additive manufacturing.² Fused deposition modeling (FDM) is a well-known method for fabricating polymeric samples, and composite materials can also be manufactured by the FDM process.³ The feedstock material used in the printing process is thermoplastic filaments such as polycarbonate (PC), polyethylene, PLA, ABS, acrylic, and nylon.⁴ So, there may be a possibility of poor bonding between the layers, warpage, difference in the width of the walls, and some other issues. To overcome all such things and to improve the quality and compression strength of the samples, the process parameters must have to be optimized. Among the various mechanical properties, the structure applied may fail easily on various loading conditions (compression, tensile). So, there is a necessity to analyze the mechanical property of newly developed polymer material and structure incorporated materials. Various results are available in optimization of the process parameters and their respective effect on the mechanical property, and some of the examples are tabulated in Table 1.

Table 1.

Various research studies under the topic of optimization.

Sl. No	Reference	Machine/Equipment	Material	Methods/Tools	Process parameters	Part characteristics
1	Elkholy et al⁵	Ultimaker 2	PLA	Energy equation	Layer height, raster width	Thermal conductivity
2	Liu et al⁶	MakerBot Replicator2	PLA	Taguchi method, S/N	Build orientation, layer thickness,	Tensile, flexural and impact strength
				Ratio, ANOVA gray relational	Raster orientation
				Analysis	Raster width, air gap
3	Torres et al⁷	MakerBot Replicator2	PLA	Taguchi method,	Layer thickness, infill density and post-processing	Ultimate shear strength, 0.2% yield
				Taguchi method,	Layer thickness, infill density and post-processing	strength
				Regression	Heat-treatment	Proportional limi
				ANOVA	Time	Shear modulus, and fracture strain
4	Es-Said et al⁸	FDM 1650 machine	ABS	—	Raster orientation	Ultimate tensile strength, yield strength, flexural strength, and impact strength
5	Lee et al⁹	FDM3000 machine	ABS	Taguchi method (L9), S/N ratio, ANOVA	Layer thickness, raster orientation, raster width, air gap	Elastic performance
6	Ziemian et al¹⁰	Stratasys Vantage-i machine	ABS	ANOVA	Raster orientation	Tensile strength and fatigue performance

The process parameters such as extruding temperature, infill density, infill pattern, and layer thickness are varied in the research carried out by Qattawi.¹¹ The results depict that the optimum process parameter for obtaining high tensile strength is the highest extruding temperature (210°C), highest infill density (100%), hexagonal pattern, and lowest layer thickness (0.2 mm). Wang et al.¹² highlight that the thickness of the layer is the main influencing parameter in the fused deposition modeling (FDM) process for PLA filaments. Dong et al.¹³ found that the build orientation, raster angle, infill density, and layer thickness are the main influencing parameter while fabricating PLA samples through the FDM process. Strength-wise, additively manufactured components are lesser than the molded samples,¹⁴ but there is no material wastage, and any complex structures can be fabricated via AM process. In recent years, AM has been employed in various fields because of such advantages. Hikmat et al.¹⁵ investigated the tensile property of the 3D printed PLA parts through the Taguchi optimization method. The influential parameter in the tensile strength was determined by developing an L18 orthogonal array using seven parameters and three levels. They concluded that the build orientation is the most influencing process parameter in attaining the greater tensile strength among six other process parameters. ANOVA results depict that the build orientation, nozzle diameter, and infill density are the three significant parameters in terms of high tensile strength. The sample has a tensile strength of 58.05 MPa, which was fabricated at the optimal combinations.

Zharylkassyn et al.¹⁶ analyzed the effects of process parameters in attaining better dimensional accuracy of the additively manufactured parts through FDM. It was concluded that the minimal layer thickness (0.1 mm and 0.2 mm) shows better dimensional accuracy in the PLA or AMS resins. In terms of nylon and ASA, a larger layer thickness value (0.3 mm) was preferred in that literature.

Polymer composites are reinforced with particle reinforcement to improve the strength and ductility of the composite. Some researchers have attempted to use various types of reinforcement such as wood,¹⁷ ceramic¹⁸ as a reinforcement in polymeric samples. Wood reinforced polymer composite, namely wood PLA composite (WPC) filaments, are used as a feedstock material for fabricating compression samples in this research. The polymeric material is deposited in the build plate as a layer during the AM process. Vishal et al.¹⁹ experimented using composite filament composed of silicon and PLA in the FDM process. Vinoth et al.²⁰ investigated the influence of slicing parameters on mechanical properties and surface quality of the additively manufactured carbon fiber/PLA material. They concluded that the slicing parameters such as infill density, layer thickness, and infill pattern greatly influenced the mechanical properties of the 3D printed CF/PLA.

Kottasamy et al.²¹ predicted and investigated the mechanical properties of additively manufactured samples fabricated from copper-reinforced PLA. Mechanical properties such as tensile, flexural, and compression were carried out in their research by varying weight percentage (wt.%) of copper and by varying the infill patterns. Their results depict that the greater tensile and flexural strength were observed in the sample fabricated at concentric infill pattern with 25 wt.% of Cu. A greater compressive strength was observed in the sample prepared at 25 wt.% of Cu with a grid infill pattern.

Lattice structures are generated by the sequential arrangement of a unit cell and are hugely applied in aerospace, biomedical, and automobile fields.²² Nowadays, incorporating lattice structure in the design of the fabricating samples for reducing the material usages and obtaining comparatively a similar mechanical property of a complete solid sample is an active area of research. Several lattice structures were incorporated in the design of the polymeric samples depending on the required field, and it is tabulated in Table 2.

Table 2.

Application of lattice structure in additively manufactured components.

S. No	Author	Process	Parts	Lattice type	Industrial	Year
1	Mahshid et al²³	SLM	Cooling channel in injection molding tools	Modified FCC-z	Tooling	2016
2	Gorguluarslan et al²⁴	SLM	Bone scaffold	BCC	Biomedical	2017
3	Dumas et al²⁵	SLM	Femoral stem	Diamond	Biomedical	2017
4	Jette et al²⁶	SLM	Femoral stem	Diamond	Biomedical	2018
5	Magerramova et al²⁷	SLM	Gas turbine engine, fan blades	Rhombic dodecahedron, Dode medium	Aircraft	2018
6	Rahmani et al²⁸	SLM	Implant	Rectangular, gyroid, Primitive	Biomedical	2019
7	Soro et al²⁹	SLM	Bone implant	Schwartz	Biomedical	2019

Simultaneously, while incorporating lattice structures, the weight of the component is also reduced. A structure that comes under triply periodic minimal surface (TPMS), namely gyroid³⁰ is incorporated in the design of the sample. The surface area of the gyroid structure is maximum among the other TPMS and lattice structures. At the same time, the strength of the sample is reduced while incorporating a gyroid structure. Composite filaments are used for the sample preparation to improve the sample’s strength. Abueidda et al.³¹ compared some TPMS structures on the polymer material through AM process. The results concluded that the gyroid structure’s mechanical (compressive strength) properties are much better than all the compared TPMS structures. Abidin et al.³² optimized the printing process parameter in the FDM process for improving the porosity accuracy in the scaffolds of PLA. Taguchi method was used in this research for optimizing the porosity. Their results depict that the nozzle temperature is the most influencing parameter in attaining less porosity in the PLA samples, with a percentage contribution of 42.22%. Layer thickness contributes 10.81% next to nozzle temperature.

TPMS types of structures are difficult to fabricate and limited researches were available with the incorporation of gyroid structures in the polymeric composite samples for experimental optimization for attaining better mechanical properties. The present work concentrates on the process optimization for fabricating TPMS structure incorporated WPC composites. The gyroid structure was incorporated in the sample’s design to reduce weight and material usage. The samples are prepared from the WPC filaments, which comprise 70% Poly Lactic Acid (PLA) and 30% Wood flour and the samples are fabricated through the FDM process. Compressive strength and the dimensional error in the samples are selected as the output responses for the optimization process. Taguchi optimization technique determines each process parameter’s contribution to the respective output responses. The influence of each parameter and the contribution of every parameter is analyzed using the ANOVA technique.

Materials and method

WPC filaments are purchased from HartSmart Products, US, and such filaments are used for the sample preparation. Wood reinforced PLA filaments are dark brown due to the addition of wood filler in the PLA matrix. The samples are fabricated at the FDM-based 3D printer, namely, PRATHAM 3.0, which can print composite filaments, and this printer is capable of printing samples with a maximum dimension of 300*300*300 mm.

Design modeling of gyroid structure in wood PLA composite samples

The gyroid-structured compression sample was designed by the CAD software, namely CATIA V5 R20. The dimension of the sample is 15 mm*15 mm*15 mm,³³ which is nothing but a cube of side 15 mm. The radius of the curves present in the gyroid structure is 1.2 mm. In most cases, gyroid structures generate lightweight internal structures and higher strength with minimum manufacturing time. 2D and 3D images of gyroid structures are shown in Figure 1(a) and (b).

Preparation of triply periodic minimal surface incorporated wood PLA composite samples

After designing the part, the file should be saved in the stereolithography (.Stl) format. Then, the (.Stl) formatted file has to be imported into any slicing software. All the process parameters have to be mentioned in the slicing section. In this research, the software used to slice the model is Ultimaker Cura, an open-source software. Table 3 shows the parameters which are kept constant during the sample fabrication.

Table 3.

Process parameters for 3D printing the TPMS structure.

Sl. No	Printing parameters	Unit	Value
1	Bed temperature	°C	60
2	Extrusion temperature	°C	210
3	Printing speed	mm/s	20
4	Nozzle diameter	mm	0.4
5	Filament diameter	mm	1.75
6	Infill pattern	—	Line
8	Build adhesion	—	Skirt

The sample will be sliced into several layers based on the parameters defined in the slicing stage. For example, the layer thickness is the only parameter determining the number of layers in the sample. G-code is generated for the sliced sections of the samples to be printed. G-code is the geometric code, where all the instructions for the fabrication are saved. After that, the G-code has to be fed into the FDM machine for the fabrication of compression samples. Pictorial representations of the 3D model, sliced model, and fabricated model are displayed in Figure 2.

Selection of process parameters

The Taguchi method is used to minimize the experimental counts without compromising the influence of each parameter. By varying raster angle, layer height, and wall thickness, the L₉ orthogonal array is developed. The array is developed by three parameters and three levels, and the process parameter and their respective levels are tabulated in Table 4.

Table 4.

Levels and parameters for optimization process of TPMS structure.

Levels	Parameters
	A	B	C
	Raster angle (°)	Layer thickness (mm)	Wall thickness (mm)
1	0	0.1	0.4
2	45	0.2	0.6
3	90	0.3	0.8

Gyroid incorporated WPC samples are fabricated in all the nine sets of combinational printing process parameters such as raster angle (0°, 45°, 90°), layer thickness (0.1 mm, 0.2 mm, 0.3 mm), and wall thickness (0.4 mm, 0.6 mm, 0.8 mm) for finding out the dimensional error occurred in the samples and compressive behavior. Five samples are prepared for each set of printing process parameters, and the average value of the respective combination is used as the experimental results.

Table 5 shows the output responses (compressive strength and dimensional error) along with the L₉ orthogonal array of the WPC 3D printed samples. The average of five samples determines the compressive strength and the dimensional error of the samples. Five samples were prepared for all nine sets of parameters, and then the compression tests were conducted on all the fabricated samples. These experimental values are used for the optimization process.

Table 5.

L₉ orthogonal array for the gyroid incorporated WPC composite.

Sl. No	Raster angle (°)	Layer thickness (mm)	Wall thickness (mm)	Compressive strength (MPa)	Dimensional error (mm)
1	0	0.1	0.4	30.44	0.02
2	0	0.2	0.6	27.65	0.05
3	0	0.3	0.8	28.34	0.08
4	45	0.1	0.6	34.25	0.19
5	45	0.2	0.8	32.24	0.21
6	45	0.3	0.4	29.31	0.24
7	90	0.1	0.8	29.11	0.13
8	90	0.2	0.4	24.78	0.18
9	90	0.3	0.6	26.44	0.15

Mechanical testing of triply periodic minimal surface incorporated wood PLA composite samples

A Universal Testing Machine (UTM) (Make: Bluestar, India) is used for conducting the compression test on the gyroid incorporated compression samples. The UTM machine used in this study can provide a maximum load of 50 kN. The UTM machine consists of two flat surfaces, one flat surface at the bottom is for placing the sample at the center, and the load will be applied to the sample through the upper flat surface. The compression test was carried out on the TPMS structure incorporated WPC samples at a feed rate of 5 mm/min. Samples are tested in the UTM, which are all fabricated at different process parameters. Figure 3(a) shows the loaded WPC samples for the compression experiment, and Figure 3(b) shows the samples which are fabricated at the various combination of process parameters before the compression tests.

Measurement of dimensional error in Samples

The width and wall of the structure incorporated TPMS WPC sample is measured for finding out the variation in the dimension of the fabricated samples with its design. The wall and width of all the samples are measured by Video Measuring System (VMS 2010). The dimensions of the sample in its design are a wall of 2.5 mm and a width of 15 mm. After fabrication, the pre-mentioned two parameters are measured in all the samples for figuring out the optimum process parameter for attaining lesser dimensional error in the design of the sample. The dimension of the wall and width of the gyroid incorporated WPC sample is shown in Figure 4.

Taguchi analysis for the optimization process

Taguchi is a statistical method developed for improving the quality of manufactured samples.³⁴ Taguchi method is utilized for determining the maximum compressive strength and minimal dimensional error on the WPC composite samples, which are additively manufactured through the FDM process. For determining the optimum process parameter for the fabrication of gyroid (TPMS) incorporated WPC polymer sample, analysis of variance is performed. A mathematical equation is progressed based on the process parameter and the experimental results. The regression equation is utilized here for establishing the percentage contribution of each and every process parameter. Compressive strength and error in the dimension of the fabricated samples are considered as the resilience parameter for attaining greater results in it without compromising the quality of the sample. The compressive strength from the experimental results is converted into a signal to noise (S/N) ratio for obtaining the optimal process parameter.

The output response parameters in this study are compressive strength and dimensional error. The process parameters are optimized to incorporate the gyroid structure in WPC polymer material to attain high compressive strength and less dimensional error. In terms of acquiring high compressive strength, the condition followed here is “larger the best.” Expressions for the compressive strength response are shown in equation (1).

\frac{S}{N} = - 10 * l o g_{10} (\frac{1}{Y^{2}})

(1)

where Y represents the compressive strength of the gyroid incorporated WPC samples

In the case of dimensional error, the “Smaller the best” condition is followed, and the expression for the dimensional error response is shown in equation (2).

\frac{S}{N} = - 10 * l o g 10 (s u m \frac{Y^{2}}{N})

(2)

where Y represents the dimensional error in the samples which are fabricated through the FDM process.

Analysis of variance technique is followed in this research for determining the most influenced parameter and the contribution of each process parameter of the fabricated WPC samples. Compressive strength and the error in the dimension of the samples are the output response. The experiment was conducted at a 95% confidence level, and 5% significant level for manufacturing the gyroid incorporated samples.

The experimental results are used for establishing the regression equation by diversing the process parameter such as raster angle, layer thickness, and wall thickness concerning the compressive strength and dimensional error. Minitab is the statistical software used in this study for generating the mathematical model for the output response (compressive strength and dimensional error) of the additively manufactured WPC samples.

‘Minitab 2020’ software is used for conducting the statistical analysis for the process optimization. The Analysis of variance technique is utilized to determine the most influencing parameter among other various process parameters. The confidence and significance levels are set at 95% and 5% in the current FDM process for finding out the interaction of each parameter during the incorporation of the TPMS structure in the WPC composite. Amidst several process parameters such as raster angle, layer height, and wall thickness, the most influencing process parameter will be selected from the experimental results, best fit model and the regression equation were predicted with regards to the raster angle, layer height, and wall thickness as an independent variable and compressive strength and the dimensional error as a dependent variable.

Results and discussion

Signal to Noise ratio for compressive strength

In order to achieve greater compressive strength and less dimensional error in the gyroid incorporated WPC samples fabricated via the FDM process, the larger the best condition is followed in the S/N ratio plot of the compressive strength, and smaller the best condition is followed in the S/N ratio plot of the dimensional error. Signal to Noise (S/N) ratio plot of the compressive strength on the gyroid incorporated polymer samples with different process parameters such as raster angle, layer thickness, wall thickness is shown in Figure 5.

Figure 5(a) represents the parameter raster angle with three different angles, notably 0°, 45°, and 90° for the compressive strength response. Figure 6(a) shows the various raster angles for printing the gyroid structure, in which the compressive load can transfer effectively in the prepared sample. The maximum S/N ratio value was obtained on the gyroid incorporated samples prepared at a 45° raster angle. This shows that the samples prepared at 45° exhibit higher compressive strength than the sample fabricated at the other two raster angles. The compressive strength is maximum at 45° inclined samples due to the inclined stacking of WPC layers, and these samples absorb more energy before breaking. Samples fabricated at 0° (horizontal) and 90° (vertical) attain the lowest signal to noise ratio value compared to a sample prepared at a 45° raster angle. A similar observation was observed by Saty and Rajeev.³⁵ and the results concluded that the sample prepared at 45° possessed higher compressive strength, and that result coincided with the experimental results of this research.

Figure 5(b) represents the parameter layer thickness with three levels, notably 0.1 mm, 0.2 mm, and 0.3 mm for the compressive strength response. Figure 6(b) shows various layer thicknesses in the 3D printing process. The samples prepared at 0.1 mm layer thickness possess the maximum S/N ratio value. This shows that the samples prepared at 0.1 mm layer thickness possess the highest compressive strength compared to other layer thickness values. The compressive strength is maximum at 0.1 mm layer thickness sample because the gap between the two layers is minimized in such cases. The gap between the layers increases as the layer thickness of the sample increases. As the thickness of the layer decreases, the number of layers increases. Therefore, the compressive load will act uniformly and transfer to each layer equally in the printed composite samples. Simultaneously, the production time of the sample will be increased in terms of lower layer thickness. The ductility of the samples increases with the decrease in gaps. So, as the ductility increases, compressive strength also increases. Samples fabricated at the layer thickness of 0.2 mm and 0.3 mm possess a lower signal to noise ratio value than the gyroid incorporated WPC sample prepared at 0.1 mm layer thickness. Heidari et al.³⁶ investigated the effect of process parameters for optimizing the mechanical property in the tested samples. The results show that the samples fabricated at minimum layer thickness obtain maximum mechanical property. These results are in line with Heidari’s experiment.

Figure 5(c) represents the parameter wall thickness with three different values, notably 0.4 mm, 0.6 mm, 0.8 mm for the compressive strength response. Figure 6(c) shows the various wall thicknesses in the FDM process. Certain voids were observed in the wall of the sample prepared at 0.4 mm layer thickness. Simultaneously, voids generation in the samples were decreased while increasing the wall thickness. Samples prepared at 0.8 mm wall thickness show the maximum S/N ratio among the three wall thickness levels. This shows that the samples prepared at 0.8 mm exhibit higher compressive strength than the sample prepared at other wall thicknesses. The compressive strength is maximum at the 0.8 mm wall thickened sample. It is due to the less number of walls present in the sample. As the number of walls decreases, the fusion of the polymer layer increases, and the possibility of the failure of the whole wall decrease.

Similarly, the developed voids were also reduced while increasing the wall thickness of the sample. Samples fabricated at 0.6 mm wall thickness possess the value of 29.5 in signal to noise ratio, and the samples prepared at 0.4 mm wall thickness attain the least signal to noise ratio with the value of 28.2 on comparing the other two-layer heights. Rodríguez et al.³⁷ studied the mechanical properties of the FDM samples and concluded that the mechanical properties of the samples are better when the samples are prepared at greater wall thickness. Results of wall thickness in this research are in line with that research.

Signal to Noise ratio for dimensional error

Signal to noise (S/N) ratio plot of the dimensional error on the gyroid incorporated polymer samples with different process parameters such as raster angle, layer thickness, wall thickness is shown in Figure 7.

Figure 7(a) represents that the raster angle parameter greatly influences the dimensional error in the fabricated samples. Samples fabricated at 0° show a less dimensional error with a value of 27.4 in signal to noise ratio while comparing the other two raster angles. The samples prepared at 45° have a dimensional error with a value of 13 in signal to noise ratio in the walls and width of the sample. The samples fabricated at 90° possess the value of 16.3 in signal to noise ratio. It is expected that the 45° inclined sample will have a high dimensional error due to the inclined stacking of layers. A similar result was obtained from research carried out by Alafaghani and Qattawi.¹¹

Figure 7(b) shows that while increasing the layer thickness, the error in the dimension of the fabricated samples increases. Sample prepared at the layer thickness of 0.1 mm shows a minimal dimensional error with a value of 22 in the signal to noise ratio, the layer thickness of 0.2 mm possesses 18, and layer thickness of 0.3 mm possesses 17 in the signal to noise ratio. While increasing the layer thickness of the samples, the void between the layers increases, and hence the dimensional error also increases.

Figure 7(c) depicts that the error in the dimension of the samples is decreased when the wall thickness of the samples is decreased. While observing the wall thickness of the samples, the 0.4 mm wall thickness samples possess less dimensional error with an SN ratio of 20.5. The samples fabricated at 0.6 mm and 0.8 mm possesses 19 and 17.5 in the signal to noise ratio value. The reason for such observation is as the count of layers in the wall of the sample increases, the dimensional error of that sample decreases. A similar result is available for wall thickness as well as the layer thickness in the research carried out by Tasdemir et al. .³⁸

The optimum process parameter for acquiring higher compressive strength in the gyroid incorporated WPC samples is determined from the experimental results. The optimum process parameters are a 45° raster angle, a minimum layer thickness of 0.1 mm, a maximum wall thickness of 0.8 mm. For attaining the least dimensional error, the optimum process parameter is 0° raster angle, 0.1 mm layer thickness, and 0.4 mm wall thickness. Delta values of the parameters such as raster angle, layer thickness, and wall thickness are extracted from the difference of maximum S/N ratio and minimum S/N ratio. The response table, along with the delta values for both the properties, is tabulated in Table 6. Based on the delta values, the ranking sequence is determined.

Table 6.

Response table for compressive strength and dimensional error.

Compressive strength
Level	Raster angle (°)	Layer thickness (mm)	Wall thickness (mm)
1	29.18	29.88	28.96
2	30.07	28.96	29.32
3	28.54	28.94	29.5
Delta	1.53	0.94	0.54
Rank	1	2	3
Dimensional area error
Level	Raster angle (°)	Layer thickness (mm)	Wall thickness (mm)
1	27.31	22.04	20.42
2	13.46	18.16	18.97
3	16.36	16.94	17.74
Delta	13.85	5.10	2.68
Rank	1	2	3

From Table 6, the sequential order of influencing parameter for attaining the high compressive strength is raster angle > layer thickness > wall thickness. The sequential order of influencing parameter for obtaining the minimal dimensional error is raster angle > layer thickness > wall thickness. The results are calculated based on the delta value from the maximum and minimum values obtained from the S/N ratio value. The maximum delta value of 1.53 is obtained for the raster angle with respect to compressive strength response and 13.85 for the dimensional area error response. From that, it is concluded that the raster angle has a major contribution and most influential parameter for printing WPC composite.

Analysis of variance for the prepared gyroid structure

ANOVA results of the compressive strength and the dimensional error are mentioned in Table 7. The influence of every process parameter and its contribution is determined from the output responses. The percentage contribution of process parameters in compressive strength and dimensional error is shown in Figure 8. The most influencing process parameter for attaining better compressive strength is the raster angle of 60.78%, layer thickness contributes 29.71%, and the wall thickness contributes 7.17%. In the case of dimensional error, the contribution of raster angle, layer thickness, and wall thickness is 90.43%, 6.82%, 0.94%, respectively.

Table 7.

Results from the ANOVA analysis.

Compressive strength (MPa)
Source	DF	Adj SS	Adj MS	Percentage contribution (%)
Raster angle	2	40.481	20.2404	60.78
Layer thickness	2	19.775	9.8876	29.71
Wall thickness	2	4.774	2.3869	7.17
Error	2	1.562	0.7808	0.78
Total	8	66.592	—	100
Dimensional error (%)
Source	DF	Adj SS	Adj MS	Percentage contribution (%)
Raster angle	2	0.040956	0.020478	90.43
Layer thickness	2	0.003089	0.001544	6.82
Wall thickness	2	0.000422	0.000211	0.94
Error	2	0.000822	0.000411	1.81
Total	8	0.045289	—	100

Regression analysis of the TPMS structure incorporated WPC samples

The regression equation is used for predicting the compressive strength of the WPC samples, and the mathematical model is shown in equation (3)

C o m p r e s s i v e S t r e n g t h (M P a) = 30.85 - 0.0226 R a s t e r a n g l e - 16.2 L a y e r t h i c k n e s s + 4.30 W a l l t h i c k n e s s

(3)

The regression equation used to predict the dimensional error of the WPC samples is shown in equation (4).

D i m e n s i o n a l e r r o r (%) = 0.054 - 0.001148 R a s t e r a n g l e + 0.217 L a y e r t h i c k n e s s - 0.017 W a l l t h i c k n e s s

(4)

Confirmational experiment on the 3D printing process parameter

The optimum process parameter is determined from the experimental results for attaining higher compressive strength. Images from video measuring system (VMS) depict the optimum parameter for fabricating the gyroid incorporated WPC samples. Experimental and predicted values of the output responses along with the percentage error are tabulated in Table 8.

Table 8.

Confirmational test results of optimal combination along with the percentage error.

Response parameter	Optimal combination	Predicted value	Experimental value	Percentage error
Compressive strength (MPa)	Raster angle–45°	31.653	31.019	2.78
	Layer thickness–0.1 mm
	Wall thickness–0.8 mm
Dimensional error (%)	Raster angle–0°	0.0859	0.0913	6.24
	Layer thickness–0.1 mm
	Wall thickness–0.6 mm

The error percentage for the optimal combination was calculated based on the experimental and predicted results of the gyroid lattice structure incorporated on the WPC material. The results show that the predicted and experimental compressive strength for the optimized condition was 31.653 MPa and 31.019 MPa. Based on the observed results, the calculated error percentage is 2.78% which is within the acceptable limit. Figure 9 shows the stress–strain plot of the sample, which was fabricated at the optimal conditions. The sample withstands a greater load and possesses the greatest compressive strength with the value of 31.019 MPa. After completing the test, it is observed that the sample was compressed to a great extent, and the final height is measured approximately as 4 mm. The sample before and after the testing was also attached, along with the stress versus strain graph. The compressed sample has not undergone any buckling. This shows the sample can absorb the load uniformly, which is evident in lesser void content. The results are in line with the predicted sample results. Dhinakaran et al.³⁹ detected a similar observation for the compressive properties of the Almond shell/PLA composite. The experimental results highlight that the optimal combination of the compressive sample experience an error percentage of 3.70%.

Figure 10 shows the VMS image of the gyroid incorporated WPC sample, which was prepared at the optimum conditions. The gyroid sample was designed with a wall thickness of 2.5 mm. The image from the VMS depicts that the sample fabricated at optimal process parameters has a variation of 0.09 mm in its wall dimension after fabrication. This variation is due to the localized melting of polymer and shrinkage that appeared during the printing of polymer samples. Most importantly, surface features reveal that the walls of the printed section have uniform printing layers, and the gyroid structure is free from structural damages. These structures are continuous throughout the thickness of the sample. This could be the reason for attaining the higher compressive strength of the composite. Sabarinathan et al.⁴⁰ experienced a similar observation on the optimal combination of tensile results for the hexagonal-shaped lattice structure on the PLA sample. The predicted and experimental results of the dimensional error response parameter are 0.0859 and 0.0913. The percentage error for the dimensional error response is 6.24%, and it is within the nominal range of the acceptable limit. A similar kind of result is available in the research carried out by Sabarinathan et al. on the sol-gel recovery process.⁴¹

The compression test is conducted on all the samples fabricated by incorporating gyroid structures in WPC polymer composites to determine the optimum process parameter. Figure 11 shows the tested gyroid incorporated WPC samples at the end of the compression before unloading. Observing the image, the samples fabricated at the eighth and ninth sets of the orthogonal array are separated into two pieces. The reason is both sets of samples are fabricated at a 90° raster angle, and the crack in the samples is initiated at the edges of the wall. Most importantly, both samples possess lower compressive strength than the other samples. Therefore, it is clearly seen that the sample, which is fabricated at a raster angle of 90° resists lesser load and fails more quickly than the other raster angles. Samples fabricated at the seventh set of L₉ orthogonal array are at 90° raster angle. Still, it resists better load than the eighth and ninth conditioned samples and has undergone local buckling due to the minimum layer thickness and maximum wall thickness. From this observation, the influence of the other optimal process parameters such as layer thickness and wall thickness in WPC samples were clearly observed. Samples prepared at the second set of parameters also undergo minimal deformation than other samples.

It is clear that the samples prepared for the fourth and fifth experimental conditions observe more energy, withstand a greater load, and deforms better than the other samples. The similarity between these two samples is that both the samples are fabricated at a 45° raster angle. A failure occurred earlier in the samples, fabricated at seventh, eighth, and ninth experimental conditions. Samples made for the eighth and ninth experiments are broken into two pieces. The correlation between samples 7, 8, and nine is that these samples are fabricated at a 90° raster angle. This shows that the raster angle would be an influencing parameter in attaining a better compressive strength.

Conclusions

Concise research was carried out in this present work for finding the optimal printing process parameter for the WPC samples fabricated through the FDM process. The results observed in this present work are mentioned below.

(1) From the observation from the S/N ratio plot of compressive strength results, it is clear that the samples fabricated at a 45° raster angle, 0.1 mm layer thickness, and 0.8 mm wall thickness will absorb more energy, possess maximum compressive strength.

(2) From the S/N ratio plot of dimensional error, the optimum process parameter is clearly identified as 0° raster angle, 0.1 mm layer thickness, and 0.4 mm wall thickness

(3) The results show that the percentage of error in the compressive strength and dimensional error is 2.34% and 6.24%, which are in the acceptable range.

(4) The most influencing parameter for attaining high compressive strength and dimensional error is raster angle. It contributes 60.78% in compressive strength and 90.43% in dimensional error. Other parameters have shown minimum contribution when compared to raster angle.

(5) Samples fabricated at 45° are absorb more energy and are compressed greatly, but the dimensional error in that sample is higher when comparing other raster angles. The impact of wall thickness in both compressive strength and dimensional error is very less, and it contributes 7.17% in compressive strength, and it contributes 0.93% only in the dimensional error

(6) Finally, from all such results, raster angle at 45°, layer thickness at 0.1 mm, and wall thickness at 0.8 mm are concluded as the optimized parameter for attaining high compressive strength. For acquiring less dimensional error, the optimum process parameter will be 0° raster angle, 0.1 mm layer thickness, and 0.4 mm wall thickness in the gyroid incorporated WPC polymer samples.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Chennai institute of technology (CIT/CAR/2021/003).

ORCID iDs

Dhinakaran Veeman

Sabarinathan Palaniyappan

References

Wong

Hernandez

. A review of additive manufacturing. Int Sch Res Notices 2012; 2012.

Gibson

Rosen

Stucker

, et al. Addit. Manuf. Tech. Cham, Switzerland: Springer, 2021.

Carneiro

Silva

Gomes

. Fused deposition modeling with polypropylene. Mater Des 2015; 83: 768–776.

Sabarinathan

Dhinakaran

Rajkumar

, et al. Photovoltaic industrial waste as substitutional reinforcement in the preparation of additively manufactured acrylonitrile butadiene styrene composite. Arab J Sci Eng 2022, doi:10.1007/s13369-022-06806-5, In press.

Elkholy

Kempers

. Investigation into the Influence of Fused Deposition Modeling (FDM) Process Parameters on the Thermal Properties of 3D-Printed Parts. In: Proceedings of the 2018 Canadian Society for Mechanical Engineering (CSME) International Congress. Toronto, ON: Canada; 2018.

Liu

Zhang

, et al. Mechanical property parametric appraisal of fused deposition modeling parts based on the gray Taguchi method. The Int Addit. Manuf. Tech 2017; 89(5): 2387–2397.

Torres

Cotelo

Karl

, et al. Mechanical property optimization of FDM PLA in shear with multiple objectives. Jom 2015; 67(5): 1183–1193.

Ang

Leong

Chua

, et al. Investigation of the mechanical properties and porosity relationships in fused deposition modelling‐fabricated porous structures. Rapid Prototyp J 2006; 12(2): 100–105.

Lee

Abdullah

Khan

. Optimization of rapid prototyping parameters for production of flexible ABS object. J Mater Process Tech 2005; 169(1): 54–61.

10.

Ziemian

Okwara

Ziemian

. Tensile and fatigue behavior of layered acrylonitrile butadiene styrene. Rapid Prototyp J 2015; 21(3): 270–278.

11.

Qattawi

. Investigating the effect of fused deposition modeling processing parameters using Taguchi design of experiment method. J Manuf Process 2018; 36: 164–174.

12.

Wang

Lin

. Optimizing the rapid prototyping process by integrating the Taguchi method with the Gray relational analysis. Rapid Prototyp J 2007; 13(5): 304–315.

13.

Dong

Wijaya

Tang

, et al. Optimizing process parameters of fused deposition modeling by Taguchi method for the fabrication of lattice structures. Additive Manufacturing 2018; 19: 62–72.

14.

Huynh

Smilo

Blourchian

, et al. Property-map of epoxy-treated and as-printed polymeric additively manufactured materials. Int J Mech Sci 2020; 181: 105767.

15.

Hikmat

Rostam

Ahmed

. Investigation of tensile property-based Taguchi method of PLA parts fabricated by FDM 3D printing technology. Results Eng 2021; 11: 100264.

16.

Zharylkassyn

Perveen

Talamona

. Effect of process parameters and materials on the dimensional accuracy of FDM parts. Mater Today Proc 2021; 44: 1307–1311.

17.

Sabarinathan

Rajkumar

Annamalai

, et al. Static and dynamic behavior of micrometric agro Prunus amygdalus particulate distributed interpolymer layer‐kenaf composite. Polym Composites 2020; 41(8): 3309–3321.

18.

Sabarinathan

Annamalai

Xavier Kennedy

. On the use of grains recovered from spent vitrified wheels in resinoid applications. J Mater Cycles Waste Manag 2020; 22(1): 197–206.

19.

Vishal

Rajkumar

Sabarinathan

, et al. Mechanical and Wear Characteristics Investigation on 3D Printed Silicon Filled Poly (Lactic Acid) Biopolymer Composite Fabricated by Fused Deposition Modeling. Silicon 2022; 1–3, doi:10.1007/s12633-022-01712-9, In press.

20.

Vinoth Babu

Venkateshwaran

Rajini

, et al. Influence of slicing parameters on surface quality and mechanical properties of 3D-printed CF/PLA composites fabricated by FDM technique. Mater Tech 2021; 1–18, doi:10.1080/10667857.2021.1915056, In press.

21.

Kottasamy

Samykano

Kadirgama

, et al. Experimental investigation and prediction model for mechanical properties of copper-reinforced polylactic acid composites (Cu-PLA) using FDM-based 3D printing technique. Int J Adv Manuf Syst 2022; 119: 5211–5232.

22.

Nagesha

Dhinakaran

Varsha Shree

, et al. Review on characterization and impacts of the lattice structure in additive manufacturing. Mater Today Proc 2020; 21: 916–919.

23.

Mahshid

Hansen

Højbjerre

. Strength analysis and modeling of cellular lattice structures manufactured using selective laser melting for tooling applications. Mater Des 2016; 104: 276–283.

24.

Gorguluarslan

Choi

S-K

Saldana

. Uncertainty quantification and validation of 3D lattice scaffolds for computer-aided biomedical applications. J Mech Behav Biomed Mater 2017; 71: 428–440.

25.

Dumas

Terriault

Brailovski

. Modelling and characterization of a porosity graded lattice structure for additively manufactured biomaterials. Mater Des 2017; 121: 383–392.

26.

Jetté

Brailovski

Dumas

, et al. Femoral stem incorporating a diamond cubic lattice structure: Design, manufacture, and testing. J Mech Behav Biomed Mater 2018; 77: 58–72.

27.

Magerramova

Volkov

Afonin

, et al. Application of light lattice structures for gas turbine engine fan blades. In: Proccedings of the 31st Congress of the International Council of the Aeronautical Sciences, ICAS 2018.

28.

Rahmani

Antonov

Kollo

, et al. Mechanical behavior of Ti6Al4V scaffolds filled with CaSiO3 for implant applications. Appl Sci 2019; 9(18): 3844.

29.

Soro

Attar

Brodie

, et al. Evaluation of the mechanical compatibility of additively manufactured porous Ti-25Ta alloy for load-bearing implant applications. J Mech Behav Biomed Mater 2019; 97: 149–158.

30.

Khaderi

Deshpande

Fleck

. The stiffness and strength of the gyroid lattice. Int J Sol Struct 2014; 51(23–24): 3866–3877.

31.

Abueidda

Elhebeary

Shiang

C-S

, et al. Mechanical properties of 3D printed polymeric Gyroid cellular structures: Experimental and finite element study. Mater Des 2019; 165: 107597.

32.

Abidin

Zainal

YanisKadir

M.M. Zahri

Akbar Teguh

, et al. Optimization of FDM 3D printing process parameter for improving porosity accuracy of PLA Scaffold. In: 4th Forum in Research, Science, and Technology. Atlantis Press; 2021, pp. 155–161. (FIRST-T1-T2-2020).

33.

Tao

Pan

Liu

, et al. A case study: Mechanical modeling optimization of cellular structure fabricated using wood flour-filled polylactic acid composites with fused deposition modeling. Compos Structures 2019; 216: 360–365.

34.

Karna

Sahai

. An overview on Taguchi method. Int J Eng Math 2012; 1(1): 1–7.

35.

Dev

Srivastava

. Experimental investigation and optimization of FDM process parameters for material and mechanical strength. Mater Today Proc 2020; 26: 1995–1999.

36.

Heidari-Rarani

Ezati

Sadeghi

, et al. Optimization of FDM process parameters for tensile properties of polylactic acid specimens using Taguchi design of experiment method. J Thermoplast Compos Mater 2020, doi:10.1177/0892705720964560, In press.

37.

Rodríguez

Zapico

Robles

, et al. Evaluation of mechanical properties of FDM components reinforced with fibre. In: IOP Conferences Series: Materials Science and Engineering, 1193. IOP Publishing, 2021, p. 012069. No. 1.

38.

Taşdemir

. Investigation of dimensional integrity and surface quality of different thin-walled geometric parts produced via fused deposition modeling 3d printing. J Mater Eng Perform 2021; 30(5): 3381–3387.

39.

Gnanavelbabu

Saravanan

Rajkumar

, et al. Mechanical strengthening effect by various forms and orientation of glass fibre reinforced isopthalic polyester polymer composite. Mater. Today: Proc 2018; 5(13): 26850–26859.

40.

Palaniyappan

Veeman

Narain Kumar

, et al. Effect of printing characteristics for the incorporation of hexagonal-shaped lattice structure on the PLA polymeric material. J Thermoplast Compos Mater 2022, doi:10.1177/08927057221089832, In press.

41.

Sabarinathan

Annamalai

. Removal of aluminosilicate bond and process optimization on recovery of sol gel alumina abrasive grain from abrasive industry waste. Silicon 2021; 13(2): 495–505.