Fused deposition modeling (FDM)-3D printing has been the favored technology to build functional components in various industries. The present study investigates infill percentage and infill pattern effects on the printed parts’ impact properties through the 3D printing technique using coconut wood-filled PLA composites. Mathematical models are also proposed in the present study with the aim for future property prediction. According to the ASTM standard, fifteen specimens with different parameter combinations were printed using a low-cost FDM 3D printer to evaluate their impact properties. Statistical analysis was performed using MINITAB to validate the experimental data and model development. The experimental outcomes reveal the honeycomb pattern with 75% infill density achieves the highest energy absorption (0.837 J) and impact energy (5.1894 kJ/m^{2}). The p-value from statistical analysis clearly shows that all the impact properties are less than the alpha value of 0.05, suggesting all the properties are vital to determine the impact properties. The validation process affirms that the generated mathematical model for the energy absorbed and the impact energy is reliable at an acceptable level to predict their respective properties. The errors between the experimental value and the predicted value are 3.98% for the energy absorbed and 4.06% for impact energy. The findings are expected to provide insights on the impact behavior of the coconut wood-filled PLA composites prepared by FDM-3D printing and a mathematical model to predict the impact properties.

At present, the application of 3D printers has gained researchers’ interest worldwide. 3D printing technology facilitates the fabrication of products under the control of 3D CAD data. 3D printing technology is considered one of the techniques in the additive manufacturing process. Additive manufacturing (AM) processes promising privileges include accuracy, precision in the final products, and optimized use of base materials, which reduces the wastage and enhances the cost efficiency of the procedure. Cost efficiency and time consumption are the primary benefits of AM technologies. Fused deposition modeling (FDM) is used extensively to process thermoplastic materials-based components production among all AM processes. FDM technology works base on the injection of melted filament onto the building platform through the nozzle. The printing process is followed by reinforcement of each layer onto the subsequent layer.

FDM technology proved efficient in producing highly accurate products from high-efficiency processes. However, the printing process parameters were shown to affect the mechanical properties of the final products significantly. There are several advantages of using an FDM printer, such as cost efficiency, convenience, and high material usage efficiency [

The present study investigates infill percentage and infill pattern effects on the printed parts’ impact properties through the 3D printing technique using coconut wood-filled PLA composites. A statistical model is developed based on the acquired experimental data accompanied by response surface methodology. This is the first study investigating the effects of infill percentage and infill patterns on the impact properties of the FDM 3D printing technique using response surface methodology to the best of the authors’ knowledge. The developed coconut wood-filled PLA composite as an integrated filament with a variation of infill percentage and infill patterns is considered for this analysis. The response surface methodology is adopted to develop the mathematical models related to impact properties, focusing on varying infill percentages and infill patterns. Impact test results reflect that infill percentage, infill pattern, the interaction effect of infill pattern with infill percentage, the second-order term of infill pattern, and the second-order term of infill percentage significantly impact both impact properties which is the energy absorbed and impact energy. Honeycomb infill pattern with 75% infill percentage gives the highest energy absorbed and impact energy in the experiment. The validation process affirms that the generated mathematical model for the energy absorbed and the impact energy is reliable at an acceptable level to predict their respective properties. The errors occur between the experimental value, and the predicted value is 3.98% for the energy absorbed and 4.06% for impact energy. The optimum impact properties are honeycomb infill pattern and 75% infill percentage. Thus, the effect of infill percentage is essential in the impact properties. A high infill percentage develops highly packed composition of product that can withstand a higher level of energy.

The material used in this research is coconut wood-filled PLA in the form of filament. The filament is then melted in the 3D printer nozzle and injected into the building platform to form a 3D shape object. The filament is composed of 40% coconut wood particles and 60% PLA. Coconut wood is selected due to its good thermal stability and high corrosion resistance. In this research project, the standard filament diameter of 1.75 mm (±0.05 mm) is used. It is from FormFutura, which is located in the Netherlands. The EasyWood coconut wood filaments characteristics are: the color of the filaments is in wood color, and the processing temperature is between 200°C until 240°C. An additional notable characteristic of EasyWood coconut wood filament is it can be printed without a heating board. This is due to its lower shrinkage, distortion-free, and hard to be warped off upon the printed parts are cooled.

In this research, the selected 3D printer is WANHAO Duplicator i3 desktop 3D printer. This 3D printer has fulfilled all the minimum requirement that needed to print the specimen for mechanical testing. Firstly, the maximum extruder temperature is 240°C (coconut wood filament requires 200°C) to melt the filament. The printing speed of the 3D printer is in the range of 10 to 60 mm/min. The next consideration is the bed temperature. Although the coconut wood filament does not require a heated bed to print, it can also be printed on the heated bed, giving better quality.

Item/Properties | Specifications |

Body material | Powder Coated Steel |

Extruder | MK10 Single-Extruder |

Print technology | FDM |

Build volume | 8 × 8 × 7 Inches (200 mm × 200 mm × 180 mm) |

Printing speed | 10–60 mm/min |

Extruder temperature | 180–240°C |

Heating plate temperature | 70–120°C |

AC input | 2A/110 V; 1A/220 V, 50–60 hz, 250W |

Net weight | 10 kg |

Overall dimension | 40 cm × 41 cm × 40 cm |

Firmware | Marlin |

Price | RM 1300.00 |

Throughout the printing process, all the parameters were kept constant except the selected varying parameters. This is to ensure that the experiment can be conducted successfully with a satisfying outcome, which means the properties will only be affected by the varying parameters. The Repetier–Host is used as slicing software in this work compatible with most firmware. The printing parameters are manipulated or keep constant with the help of this software. All the parameters that were kept constant in this research are summarized in

Parameters | Values (Kept constant) |

Height of first layer | 0.3 mm |

Layer height | 0.3 mm |

Horizontal Shell: solid layer | Top: 1 layer, Bottom: 1 layer |

Diameter of nozzle | 0.4 mm |

Diameter of filament | 1.75 mm (± 0.05 mm) |

Extruder temperature | 200°C (± 2°C) |

Print bed temperature | 60°C (± 2°C) |

Printing speed | 30 mm/s |

The impact test is a method to determine the amount of energy that a material can absorb before failure. Generally, there are three impact test types: Charpy impact test, Izod impact test, and Keyhole impact test. The only difference between the three tests is the direction of the specimen placed. In this research study, the Izod impact test is used. For the Izod impact test, the printed specimen will be placed vertically, whereas the printed specimen is placed horizontally for the Charpy impact test. The ASTM standard used for this Izod impact test is ASTM D256. CEAST 9050 impact pendulum is used to conduct the Izod impact test. The pendulum has an energy range between 0.5 to 50 J.

Charpy impact test | Izod impact test | |

Type of notch | U or V-notch | V-notch |

Specimen’s position | Horizontal and notch facing away from the pendulum | Vertically, a notch is facing towards the pendulum |

Point of strike | Centre of the test specimen | The upper part of the test specimen |

Dimensions (W × D × L) | 10 × 10 × 55 (mm) | 12.7 × 3.2 × 64 (mm) |

The impact properties, comprising energy absorbed and impact energy, are investigated comprehensively in this research study. Energy absorbed is the amount of energy that the specimen gains after being hit in the testing. The value of energy absorbed can be obtained from the time the load starts to rise until the first occurrence of 0 loads after the maximum point. The obtained value can be used to represent the toughness of a material. Thus, the higher the value obtained means the tougher the material. The impact of energy is a measurement of the work performed to break down a specimen. When the striker is released to hit the sample, the sample will immediately absorb the energy from the striker, causing the specimen to yield. At this moment, the specimen starts to undergo plastic deformation at the notch. The sample continues to absorb the striker’s energy and work-hardens at the plastic zone at the notch. The specimen’s fractures upon it reached the maximum energy absorption. In this experiment, the energy absorbed is obtained from the machine software, whereas the impact of energy is calculated manually. Therefore, it does not require any stress-strain curve in this experiment. The average for all the impact properties is listed in

Infill pattern | Infill percentage (%) | Energy absorbed (J) | Impact energy (kJ/m^{2}) |

Octagram Spiral | 25 | 0.352 | 2.1824 |

Grid | 25 | 0.459 | 2.8458 |

Rectilinear | 25 | 0.474 | 2.9388 |

Concentric | 25 | 0.552 | 3.3170 |

Honeycomb | 25 | 0.597 | 3.3418 |

Octagram Spiral | 50 | 0.434 | 2.6908 |

Grid | 50 | 0.505 | 3.1310 |

Rectilinear | 50 | 0.557 | 3.4534 |

Concentric | 50 | 0.576 | 3.5712 |

Honeycomb | 50 | 0.588 | 3.6456 |

Octagram Spiral | 75 | 0.453 | 2.8086 |

Grid | 75 | 0.623 | 3.8626 |

Rectilinear | 75 | 0.731 | 4.5322 |

Concentric | 75 | 0.789 | 4.8918 |

Honeycomb | 75 | 0.837 | 5.1894 |

Based on the summarized results in

Based on the summarized results in ^{2}. Based on the acquired results, it is clearly seen that the impact energy increased with respect to the infill pattern while the infill percentage increases simultaneously. Meanwhile, these results prove that the 75% infill percentage has the highest impact energy, followed by the 50% and 25% infill percentages. Overall, the specimen with the honeycomb pattern has the highest impact energy, followed by the concentric, rectilinear, grid, and octagram spiral pattern. The acquired results are attributed to more contact points/sheared points per unit area perpendicular to the impactor [

Term | Coef | SE Coef | |

Constant | 0.5608 | 0.0169 | 0.000 |

Infill Pattern | 0.1175 | 0.0114 | 0.000 |

Infill Percentage | 0.10740 | 0.00984 | 0.000 |

Infill Pattern × Infill Pattern | −0.0576 | 0.0192 | 0.015 |

Infill Percentage × Infill Percentage | 0.0472 | 0.0170 | 0.022 |

Infill Pattern × Infill Percentage | 0.0484 | 0.0139 | 0.007 |

According to ^{2} is equal to 96.59%, indicating that the higher the value of R2, the better the model fits the data. The S value equal to 0.0311246 suggests that the models’ lower S value predicts the response better. Meanwhile, adjusted ^{2} equal to 94.69% is responsible for the number of predictors in the model that describes the relationship’s significance. The higher adjusted ^{2} indicates that the proposed mathematical model presents an impressive elaboration of the relationship between the properties and the response. On the other hand, the model’s predictive ability level, known as predicted ^{2}, is 85.90%. The p-value obtained for the model is 0.000, which is less than the alpha value of 0.05. These results suggest that the model considered is statistically significant. The model produced from this analysis is as shown in

Source | DF | Contribution | Adj SS | Adj MS | F-value | P-value |

Model | 5 | 96.59% | 0.246807 | 0.049361 | 50.95 | 0.000 |

Linear | 2 | 85.69% | 0.218953 | 0.109477 | 113.01 | 0.000 |

Infill pattern | 1 | 40.55% | 0.103606 | 0.103606 | 106.95 | 0.000 |

Infill percentage | 1 | 45.14% | 0.115348 | 0.115348 | 119.07 | 0.000 |

Square | 2 | 6.32% | 0.016141 | 0.008071 | 8.33 | 0.009 |

Infill pattern × Infill pattern | 1 | 3.41% | 0.008715 | 0.008715 | 9.00 | 0.015 |

Infill percentage × Infill percentage | 1 | 2.91% | 0.007426 | 0.007426 | 7.67 | 0.022 |

2-Way interaction | 1 | 4.58% | 0.011713 | 0.011713 | 12.09 | 0.007 |

Infill pattern × Infill percentage | 1 | 4.58% | 0.011713 | 0.011713 | 12.09 | 0.007 |

Error | 9 | 3.41% | 0.008719 | 0.000969 | ||

Total^{2 }=^{ }96.59%^{2} – adjusted = 94.69%^{2} – predicted = 85.90% |
14 | 100.00% | 0.255526 |

A comparison is made between the energy absorbed from the experimental procedure and the theoretical results from the model generated, as shown in

Infill pattern | Infill percentage (%) | Average experimental energy absorbed (J) | Predicted energy absorbed (J) | Error (%) |

Octagram spiral | 25 | 0.352 | 0.374 | 5.93 |

Grid | 25 | 0.459 | 0.452 | 1.55 |

Rectilinear | 25 | 0.474 | 0.501 | 5.39 |

Concentric | 25 | 0.535 | 0.521 | 2.65 |

Honeycomb | 25 | 0.539 | 0.513 | 5.15 |

Octagram spiral | 50 | 0.434 | 0.387 | 12.17 |

Grid | 50 | 0.505 | 0.489 | 3.29 |

Rectilinear | 50 | 0.557 | 0.562 | 0.91 |

Concentric | 50 | 0.576 | 0.607 | 5.03 |

Honeycomb | 50 | 0.588 | 0.622 | 5.48 |

Octagram spiral | 75 | 0.453 | 0.495 | 8.41 |

Grid | 75 | 0.623 | 0.621 | 0.35 |

Rectilinear | 75 | 0.731 | 0.718 | 1.78 |

Concentric | 75 | 0.789 | 0.787 | 0.28 |

Honeycomb | 75 | 0.837 | 0.827 | 1.26 |

Another analysis was conducted is to find out the maximum energy absorbed through the response optimization method. This response optimization helps identify the preferable value of the printing parameters used in the given constraint or a specific range to produce the highest energy absorbed. According to the analysis, it is clearly shown that the maximum energy absorbed, which could be reached, is 0.824 J, and it can be achieved by the parameter combination of honeycomb pattern and 75% infill percentage.

In this research, the primary purpose is to evaluate the impact properties of coconut wood-filled 3D printed components by using fused deposition modeling. The experimental and statistical evaluation performed elaborates on the effect of the infill pattern and infill percentage on the impact properties of printed coconut wood. In conclusion, the impact test reflects that infill percentage, infill pattern, the interaction effect of infill pattern with infill percentage, the second-order term of infill pattern, and second-order term of infill percentage give a significant impact on both impact properties, which is the energy absorbed and impact energy. Honeycomb infill pattern with 75% infill percentage gives the highest energy absorbed and impact energy in the experiment. The p-value clearly shows that all the impact properties are less than the alpha value of 0.05, which means that all the properties are essential to determine the impact properties. The validation process affirms that the generated mathematical model for the energy absorbed and the impact energy is reliable at an acceptable level to predict their respective properties. The errors occur between the experimental value, and the predicted value is 3.98% for the energy absorbed and 4.06% for impact energy. The optimum impact properties are honeycomb infill pattern and 75% infill percentage. Thus, the effect of infill percentage is essential in the impact properties due to high infill percentage will give the highly packed composition of product which can withstand a higher level of energy.