Study on the application of 3D printing to wooden furniture connectors

With the rapid development of high technology in the 21st century, the manufacturing industry cannot be exempted from it. In addition, 3D printing is a multi-layer manufacturing technology that has emerged in recent years, and it is gradually integrated into the manufacturing process of components in various industries, such as aerospace, medical, military, transportation and construction industries. However, the use of 3D printing to make wooden furniture connectors has gradually replaced the traditional way of connecting wooden furniture. Like other 3D printed products, the size of the connectors of 3D printed wooden furniture is also affected by its molding accuracy. In order to obtain the best precision 3D printed wooden furniture connectors, the experimental parameters of this study are designed by Taguchi experiment method, and the measurement is carried out by 3D scanner equipment. The research results show that when the experimental parameters are the nozzle temperature of 200 °C, the layer thickness of 0.1 mm, the filling degree of 20 %, and the extrusion speed of 50 mm/s outside and 100 mm/s inside, the printed 3D wooden furniture connectors have the best accuracy.


Introduction
Due to the continuous improvement of high technology in recent years, the progress of the manufacturing industry is astonishing.However, 3D printing is a new and unique manufacturing technology, and the direction of 3D printing has become more complex and high-precision workpiece processing to make up for the inability of cutting to perform this complicated and important work.3D printing is a rapid prototyping technology.Compared with traditional subtractive manufacturing technology, 3D printing technology is additive manufacturing (AM).Its principle is to first convert the 3D design CAD file in the computer into a digital mockup, and then cut the 3D solid model into 2D slices, and then build the parts by layer-by-layer material accumulation and printing.The materials used in 3D printing are bondable materials such as powdered metal or filamentary plastic, which have irreplaceable advantages, especially for the manufacture of hollow, recessed and interlocked construction parts, and can be automated, fast and direct.Therefore, 3D printing has been gradually applied to medical, military, aerospace and construction industries.Nowadays wooden furniture has become an option for more and more families.Especially, for Chinese families, wooden furniture, which has a vivid and special wood grain, as well as a unique appearance, has high plasticity and can have different styles and presentations according to everyone's likes and dislikes.Private custom furniture has gradually become the first choice of every family.However, privately-made wooden furniture is not perfect, and it mainly has several defects: (1) The buying crowd is vague.Due to various reasons, the number of people who buy wooden furniture is uncertain, resulting in an inaccurate positioning of specific wooden furniture among the crowd.(2) The price is high.The price of wood consumed by wooden furniture itself varies, and the price of some wood is even more expensive, plus the cost of private customization, the price is even more expensive.(3) Customized wooden furniture cannot avoid the waste of many auxiliary materials and leftovers.Gellert [1] uses polyethylene material to 3D print the connectors of different types of panel furniture, so that future furniture can be combined in any way without restriction.Compared with the traditional nails, screws and glue to fix furniture wood, this 3D printed wooden furniture connector is more beautiful in appearance, and can be designed into different shapes to suit personal customization.The diversity of wooden furniture is shown in Figure 1 [2].Taguchi experimental design has been widely used in academia, industry and research fields [3][4].Kumar and Singh [5] performed rotary ultrasonic drilling of BK-7 glass by the Grey-Taguchi method, and found that the feed rate was the main factor affecting the material removal and surface roughness of the workpiece.Hu et al. [6] used grey relational analysis and found that ultrasonic vibration assistance was positively correlated with material removal rate and increased surface roughness.However, higher ultrasonic power was beneficial to suppress the increase of surface roughness.The quality of the workpieces printed by 3D largely depends on the technology and printing parameters of the 3D printer [7][8][9][10].The technology of the 3D printer is naturally the more expensive the 3D printer uses the more advanced technology, but the relative cost is also high, and the connector of the wooden furniture itself is a cheap object, if the machine purchase cost is too high, then the gains outweigh the losses.How to obtain the best printing parameters to improve the quality of wooden furniture connectors is a more reasonable reason and means.

Influence of experimental parameters on 3D printing
There are many types of 3D printing machines today, and the more common 3D printing methods are: Stereolithography (SLA), selected laser sintering (SLS), selective laser melting (SLM) and fused deposition modeling (FDM), and the material properties and advantages and disadvantages of different 3D printing methods are compared, as shown in Table 1.The technology of FDM, however, is the most widely used.

Limited size
Generally speaking, the main printing parameters that affect the forming accuracy of 3D printing are: nozzle temperature, platform temperature and layer thickness [11][12].(1) Nozzle temperature: The unsuitable temperature of the nozzle will cause the workpiece to warp.As the temperature of the nozzle increases, the warpage of the workpiece will gradually increase.The main reason is that the temperature of the nozzle determines the molten state of the printing filament.When the heating temperature of the nozzle is low, the filament is not completely melted, and the extruded filament is not uniform, which will cause printing problems such as insufficient filling and insufficient filament section width, resulting in a gap between the filament and the bottom plate or the filament.The bonding force is poor, so it is easy to cause warping and deformation of the workpiece.On the contrary, when the temperature of the nozzle is too high, the temperature difference between the extruded filament and the bottom plate or the printed workpiece will be large, and the cooling time of the molten filament will be longer, which will lead to an increase in the internal stress of the workpiece.It will leads to an increase in the deformation of the workpiece and a decrease in the accuracy.The high heating temperature of the nozzle will also lead to the degradation of the wood powder in the composite material, causing the surface of the workpiece to be scorched and causing defects.The temperature of the nozzle is not easy to be too high or too low.For polylactic acid (PLA), the reasonable nozzle temperature is 190-220 o C, within this range the warpage of the printed workpiece is small and the printing quality is good.(2) Platform temperature: As the platform temperature increases, the deformation of the workpiece gradually decreases.Because when the filament is extruded from the nozzle to the platform, if the platform is not heated, there will be a large temperature difference when the workpiece is cooled, so the internal stress of the workpiece will increase, which will easily lead to deformation of the workpiece.If the platform is continuously heated to keep the temperature constant during the printing process, the temperature difference between the nozzle and the platform can be reduced, the internal stress and the amount of deformation can be reduced.If the heating temperature of the platform is too high, the printed filament will not be easily cooled and solidified on the platform, resulting in sticky filament and unsuccessful printing.Therefore, when the temperature of the platform is 50-80 o C, it will not be deformed much in 3D printing.(3) Layer thickness: With the increase of layer thickness, the degree of deformation of the material first decreases and then increases.The reason is that in the 3D printing of FDM technology, the filament is heated by the nozzle to become a molten state and printed on the platform, and the layers are stacked on the platform to obtain the final workpiece.If the layer thickness is set smaller than the diameter of the nozzle, it will cause the nozzle to exert a certain squeezing effect on the printed matter.After the filament is printed on the platform, the extrusion force from the nozzle disappears, and the internal stress of the material is restored, and the workpiece will deform.The smaller the layer thickness, the greater the extrusion effect.When the layer thickness is too large, there will be no extrusion between layers during printing, which will reduce the dimensional accuracy of the workpiece and increase the amount of deformation.Therefore, the layer thickness of 3D printing is between 0.2mm and 0.3mm is more reasonable.

Experimental procedure
Based on the consideration of the cost of the machine and tools, this study uses the KINGSSEL K-2327 3D printing machine (KINGTEC technical Co.) to make furniture connectors, as shown in Figure 2. As for the filaments commonly used in FDM printers at present, there are mainly two kinds of acrylonitrile butadiene styrene (ABS) and PLA materials, and the cost and price of these two materials are similar, but the ABS material has a pungent exudate, and is slightly toxic and susceptible to thermal deformation.In contrast, PLA materials are more environmentally friendly and safer.Therefore, this study selected PLA as the 3D printing filament used in this study.The shapes of the wooden furniture connectors made in this study are shown in Figure 3.The dimensions of the wooden furniture connector are about 60 mm×40 mm×20 mm.First use SolidWorks software to draw its 3D model, then export it into an STL file, and load it into the 3D printing control software for slicing processing, and finally become a Geode code that can be recognized by the 3D printer.Among the 3D printing parameters, the important parameters affecting the accuracy of the workpiece mainly include the nozzle temperature and layer thickness.The range of printing parameters used in this study is determined by the test before the experiment.When the print head temperature is 200-220 o C, the error degree of the printed workpiece is small, so this study selects 200, 210 and 220 o C as the three levels of the print head temperature.The layer thickness greater than 0.1-0.4mm, the error degree of the workpiece obtained by printing is also small, so this study selects 0.1, 0.2 and 0.3 mm as the three levels of layer thickness.In addition to the above two experimental parameters, when using a 3D printer to try out samples, it was found by trial and error that the filling degree and extrusion speed also have a great influence on the accuracy of the 3D printed product.After production, it is found that when the filling degree is 15-25 %, and the extrusion speed is 30 mm/s on the outer side and 80 mm/s on the inner side, 50 mm/s on the outer side and 100 mm/s on the inner side, the workpiece error is small.This study selects 15, 20 and 25 % as the three levels of filling degree, while the outer 30 mm/s inner 80 mm/s, outer 40 mm/s inner 90 mm/s and outer 50 mm/s inner 100 mm/s 3 levels as extrusion speed.The experimental printing parameters selected in this study are shown in Table 2. Taguchi method is an effective method to improve the optimal design and control the product cost.
Parameter design is the core of Taguchi method.It uses experimental optimization method to obtain the best combination of parameters of the system, so that the sensitivity of the product to environmental conditions and other noise factors is reduced, and the final effect is not to increase product cost, even on the basis of reducing costs, the loss of product quality is minimized.In this experiment, in order to explore the influence of nozzle temperature, layer thickness, filling degree and extrusion speed on the deformation of wooden furniture connectors after printing, the Taguchi method of L 9 was used to optimize the printing parameters.The printing parameters are then verified by experiments to obtain the optimal printing parameters that really affect the quality of the finished product after printing.The sample is prepared with the data of sample 1 in the L 9 orthogonal experiment table, as shown in Figure 4.

Measurement and analysis
Because the wooden furniture connectors obtained by 3D printing are relatively complex and delicate, if the traditional Vernier calipers are used to measure, there will be uncontrollable factors on the dimensions of the wooden furniture connectors due to uneven force, resulting in a larger measurement result error.In this study, a more advanced ATOS Compact Scan (ROAD AHEAD TECHNOLOGIE) was selected to perform 3D scanning measurement on nine selected sample points on wooden furniture connectors, as shown in Figure 5.
The point data of the sample obtained after scanning is the grid data.Then use the GOM software for comparison, import the point data (STL file) and the CAD model (STP file) into the GOM software, align the reference points of the two image files, and then perform the actual mesh on the surface.By comparison, the size error of the 3D printed sample can be clearly seen, as shown in Figure 6.It can be known from the image in Figure 6 that the image color distribution of the two files reflects the difference in the amount of deformation of the two print samples.From the color bar on the right of Figure 6, it can be seen that the greenish part represents no deformation, and its value is close to zero.The reddish area upwards indicates that the size of the printed sample is larger than that of the CAD model.On the other hand, the blue part means that the size of the printed sample is smaller than that of the CAD model.As can be seen from Figure 6, the top edge of the wooden furniture connector is more seriously recessed, while the middle part of the wooden furniture connector tends to be slightly larger in size.In order to make a better comparison of the error degree of each sample, this study selects three fixed reference points on the CAD model, as shown in Figure 7. Compare the errors of the three reference point data obtained for each sample.The information about the three fixed reference points obtained for each sample is shown in Table 3.The signal-to-noise (S/N) ratio measured for each sample, it is shown in Table 4.  Analysis of variance (ANOVA) is used to analyze the degree of influence of experimental parameters on the quality of the sample, and to judge the significance of the parameters.On the basis of Taguchi experiment method, this research analyzes the influence of nozzle temperature, layer thickness, filling degree and extrusion speed on the size error of wooden furniture connectors through ANOVA.The ANOVA results of the size error of wooden furniture connectors are shown in Table 5.The larger the contribution rate, the greater the influence of the experimental parameters on the error amount.
Through the contribution rate results in Table 5, it can be known that the printing parameters used in this experiment have an influence on the size error of the wooden furniture connectors.The extrusion speed is the best.It can be seen from the results in Table 6 that the S/N ratio difference of the extrusion speed is the largest, and its value is 0.8277 dB.It can be seen that the extrusion speed has the greatest influence on the size error of printed wooden furniture connectors, and the higher the extrusion speed, the better the printing quality.When the material is extruded from the nozzle faster, the plastic is less likely to exudate out, and the problem of 3D printing filament drawing can be solved.The second influencing factor is the temperature of the nozzle, the difference in the S/N ratio is 0.7232 dB.When the temperature of the nozzle is higher, the temperature difference between the extruded material and the room temperature is larger.The material stacked on the sample will lead to aggravated thermal shrinkage reaction, resulting in increased size error of the sample.The third influencing factor is the printing layer thickness, and the difference of the S/N ratio is 0.4863 dB.From the point of view of the S/N ratio of the printing layer thickness, when the printing layer thickness is 0.1mm, the size error to the sample is the smallest.As the thickness of the printed layer increases, the size error of the sample also increases, and this phenomenon can also be explained by the concept of heat-affected zone.Among the four printing parameters selected in this experiment, the filling degree has the smallest influence on the size of the sample, and the difference of the S/N ratio is 0.11 dB.From the results in Table 6, it can be known that a filling degree of 20 % can obtain the smallest size error of the sample.
Because the size error of the samples collected in this experiment is expected to be small, it means that the larger the S/N ratio value of the printing parameter that affects the size error of the sample is better.
According to the results in Table 6, the optimal parameter combination for the size error of the desired sample is A 1 B 1 C 2 D 3 , that is, when the nozzle temperature is 200 o C, the printing layer thickness is 0.1 mm, the filling degree is 20 %, and the extrusion speed is outside 50 mm/s, inner side 100 mm/s.In order to confirm whether the optimal parameter combination obtained by the Taguchi method can obtain the expected size error on the wooden furniture connector, it is necessary to use the optimal printing parameters to print two more models to verify the optimal value.accuracy of the program.When the printing temperature is 200 o C, the printing layer thickness is 0.1 mm, the filling degree is 20 %, and the extrusion speed is 50 mm/s for the outer edge and 100 mm/s for the inner edge, and then measured by a 3D scanner after that, the size error results of the optimized printed product obtained are shown in Table 7. From the results in Table 7, it can be seen that using the optimal printing parameters to print 2 finished products, the measured size error on the wooden furniture connector is 3.7 %, which fully proves the optimization obtained by using the Taguchi method are indeed valid and feasible.

Conclusion
This research uses the 3D printer of FDM technology to make wooden furniture connectors, and uses 3D scanning technology to obtain its external dimension data, and finally uses Taguchi experimental method to analyze it, and the following two conclusions are obtained.
(1) Through the ANOVA in the Taguchi experiment, among the four FDM printing parameters (nozzle temperature, layer thickness, filling degree and extrusion speed), the extrusion speed and nozzle temperature are connected to the PLA wooden furniture produced by FDM printing.The size error of the samples has a greater impact, while the layer thickness is second, and the filling degree has the least impact.
(2) According to the results of ANOVA, the optimal printing condition is A 1 B 1 C 2 D 3 (named nozzle temperature of 200 o C, layer thickness of 0.1 mm, filling degree of 20 %, and the extrusion speed is 50 mm/s outside and 100 mm/s inside).The size error of the printed wooden furniture connectors by FDM is the smallest, and the size accuracy can be improved by 3.7 % compared with the original printing condition (A 1 B 1 C 1 D 1 ).

Figure 2 .
Figure 2. Experimental equipment used in this study.Figure 3. 3D printed models of wooden furniture connectors.

Figure 3 .
Figure 2. Experimental equipment used in this study.Figure 3. 3D printed models of wooden furniture connectors.

Figure 4 .
Figure 4. Sample prepared from the data of sample 1 in the L 9 orthogonal experiment table.

Figure 5 .
Figure 5. 3D scanner measures nine selected sample points on wooden furniture connectors.

Figure 6 .
Figure 6.Comparison of the size error for the 3D printed sample.

7 Figure 7 .
Figure 7. Three fixed reference points were selected on the CAD model in this study.

Table 1 .
Comparison of material properties and advantages and disadvantages of different 3D printing

Table 2 .
Experimental printing parameters selected in this study.

Table 3 .
Data of three fixed reference points obtained for each sample.No. Reference point 1 (mm) Reference point 2 (mm) Reference point 3 (mm)

Table 4 .
Signal-to-noise (S/N) ratio measured for each sample.

Table 5 .
ANOVA of size errors on wooden furniture connectors.

Table 6 .
S/N ratio response table of printing parameters for wooden furniture connectors.

Table 7 .
Optimal printed product error amount and S/N ratio results.