Particleboard Panels Made from Sugarcane Bagasse: Characterization for Use in the Furniture Industry

The objective of this study was to evaluate the physical and mechanical properties of commercial panels produced with sugarcane bagasse to investigate the possibility of us for the production of furniture. We evaluated industrial MDP (Medium density particleboard) panels made of Eucalyptus and Pinus by two Brazilian companies and an industrial MDP panel made from sugarcane bagasse, produced in China. We tested the physical characteristics of water absorption and thickness swelling of the panels after 2 and 24 h of immersion in water. To estimate the moisture content and density, we followed the procedure detailed in NBR 14810-31. The mechanical properties were characterized by performing bending (moduli of elasticity and rupture), compression (moduli of elasticity rupture), internal bonding, screw pullout and Janka hardness tests. We found that panels made from sugarcane bagasse showed comparable with or superior physical and mechanical properties to those made from Eucalyptus and Pinus.


Introduction
The furniture industry is characterized by the combination of several production processes that involve different raw materials and a broad variety of final products.This industry can be classified mainly according to the materials from which the furniture is made (for instance, wood and metal) or how the furniture is used (for instance, home and office furniture).The seven largest economies of the furniture industry (United States, Italy, Germany, Japan, France, Canada, and United Kingdom) produce approximately US$131 billion in furniture².
Silva³ states that solid wood was the first raw material used to produce furniture, but the use of this material has become increasingly rare due to the reduction in the availability of timber.In the opinion of this author, reconstituted wood panels have taken the place of large timber as the main raw material in the furniture industry, as they allow more efficient use of wood.Between 80% and 90% of the particleboard panels produced are intended for furniture production.
With the growth of the wood panel industry due to the demands in the furniture industry, the search for new raw materials has also intensified.In this context, in addition to the increase in areas planted with species of Eucalyptus and Pinus, there is a search for new fast-growing species that can be used for furniture production and facilitate the development of new products.
Caraschi et al. 4 report that the agroindustry produces many residues that can be used for making particleboard panels such as lignocellulosic materials.Among agricultural residues, sugarcane bagasse has been highlighted due to its large-scale production.According to the data of the Food and Agriculture Organization of the United Nations 5 , the amount of sugarcane produced in the world in 2010 was approximately 1711 trillion tons.This amount represents the generation of ~479 trillion tons of sugarcane bagasse, as according to the Sugarcane Research Center 6 each processed sugarcane ton generates 280 kg of residues.
Mendes et al. 7 , Freire et al. 8 , Tabarsa 9 , Mendes et al. 10 , Battistelle et al. 11 , Barros Filho 12 , and Widyorini et al. 13 have already assessed the use of sugarcane bagasse in the production of particleboard panels and their quality in comparison with panels made from other materials, and they found that boards made from sugarcane bagasse are very promising for furniture-based applications.
In addition to adding value to the residue, the production of particleboard panels with sugarcane bagasse may be able to satisfy the growing demand in the wood panel industry for raw materials.Furthermore, the production of particleboard panels with sugarcane bagasse may allow the industry to expand, reduce the use of timber (and thereby, reduce the pressure on forests), and reduce the costs of panel production; these properties have the potential to make such boards very competitive 14 .
With the high consumption of reconstituted wood panels by the furniture industry, the use of panels made from residues in the furniture production process becomes highly economically feasible.Hence, the objective of the present study was to assess the physical and mechanical properties of commercial panels made from sugarcane bagasse in order to investigate the possibility of using these panels for producing furniture.

Materials and obtainment of samples
We obtained five industrial MDP panels made from Eucalyptus and Pinus produced by two Brazilian companies and five industrial MDP panels made from sugarcane bagasse, which were produced in China.
These panels were sent to the Experimental Unit for Wood Panel Production (Unidade Experimental de Produção de Painéis de Madeira) located in the Federal University of Lavras, state of Minas Gerais, where the samples were removed with a sliding table circular saw.Next, the samples were placed in an acclimatization room with controlled conditions of temperature and humidity (20±2 o C and 65±5% relative humidity) until reaching a constant mass.
To assess the quality of the sugarcane bagasse panels by comparing them with Eucalyptus and Pinus panels already used on a commercial scale for furniture production, the three panel types were characterized in terms of their physical and mechanical properties.

Physical characterization
To test water absorption (WA) and thickness swelling (TS) after 2 and 24 h of immersion and springback, we followed the ASTM D-1037 standard 15 .To estimate moisture content and bulk density, we followed the NBR 14810-3 standard 1 .

Mechanical characterization
For the mechanical characterization of panels, we performed static bending modulus of elasticity (MOE) and modulus of rupture (MOR), compression (MOE and MOR), internal bond, top and surface screw pullout, and Janka hardness tests, following the NBR 14810-3 standard 1 .
The static bending tests were performed with a universal testing machine (Time Group Inc.) with a capacity of two tons, which was equipped with a computerized system for the control of test variables and data collection.

Statistical analysis
For the data analysis, we used the Sisvar software.For the physical and mechanical characterization, we conducted an analysis of variance at 5% significance, with a post hoc Tukey test at 5% significance.

Physical characterization 3.1.1. Bulk density
The average values of bulk density of the industrial MDP panels made from Eucalyptus, Pinus, and sugarcane bagasse are presented in Figure 1.The average values of bulk density measured for Pinus, Eucalyptus, and sugarcane bagasse panels were 0.612, 0.623, and 0.616 g/cm 3 , respectively.The panels were classified as medium density panels, which refers to panels with bulk density between 0.60 and 0.80 g/cm 3 16 .We observed no difference in density among the industrial panels.
The average values of bulk density for industrial panels found by Mendes et al. 17 were 0.683 g/cm 3 for Pinus, 0.655 g/cm 3 for Eucalyptus, and 0.669 g/cm 3 for sugarcane bagasse.The authors observed no significant difference among panels.
Despite the lack of a significant difference, sugarcane bagasse had a lower density compared with Pinus and Eucalyptus.When Mendes et al. 7 assessed the use of sugarcane bagasse for the production of particleboard panels, they found an average standard density value of 0.098 g/cm 3 ; whereas Barros Filho 18 obtained average values of 0.096 g/cm 3 for bagasse from a cane sugar mill and 0.099 g/cm 3 for bagasse from a cachaça distillery (locally known as alambique).
This ratio between panel density and raw material density, defined as the compression ratio, may directly affect the physicomechanical properties of particleboard panels 19,20 .

Panel moisture content
The average values of moisture for the panels made from Pinus, Eucalyptus, and sugarcane bagasse are presented in Figure 2.
According to the results obtained, all treatments differed from each other.The Pinus panels showed the highest moisture value, followed by the Eucalyptus panels, and the sugarcane bagasse panels.
The sugarcane bagasse may have presented the lowest average value of moisture for two reasons.The first is related to the type of additives used for panel production; for example, the chemical formula of the urea-formaldehyde adhesive used in the production of particleboard panels made from sugarcane bagasse.The second is related to the chemical composition of the raw material, as the sugarcane bagasse in general presents higher values of extractives than wood, and this chemical compound may have decreased the moisture of panels 20,21 .
Barros Filho 18 assessed the chemical composition of the sugarcane bagasse from a cachaça distillery and cane sugar mill for the production of particleboard panels and obtained average values of extractives of 11.1% and 9.1%, respectively.These values are above those commonly found for softwood (coniferous wood; 5% ± 3%) and hardwood (locally known as madeira de folhosas; 3% ± 2%) 22 .
Melo et al. 23 assessed the physicomechanical properties of particleboard panels produced with different proportions of wood (Eucalyptus grandis) and rice husk.The values found for moisture varied from 8.3% to 8.6% among treatments.
The EN 312 trading standard 24 defines a suitable panel moisture value as that ranging from 5% to 13%.Hence, all treatments were within this range.

Water absorption
The average values of water absorption 2 (WA2h) and 24 h after immersion (WA24h) are presented in Figure 3.
For both WA times (2 and 24 h of immersion), the average values of all treatments differed from each other.In both cases (WA2h and WA24h), the Pinus panels showed the lowest average absorption values, whereas the sugarcane bagasse panels showed the highest average values.
The reason the sugarcane bagasse presents the highest average value of WA could be the low density of the material compared with Pinus and Eucalyptus.Therefore, a higher amount of particles per area is needed to form a panel mattress of predetermined density, which increases the availability of sorption sites and consequently the WA values.
Pedreschi 25 assessed the properties of industrial panels made from sugarcane bagasse and found average values of 10.7% for WA2h and 40.6% for WA24h.For the industrial panels made from Eucalyptus, the author found average values of 12.3% for WA2h and 35.5% for WA24h.
Souza et al. 26 assessed the physicomechanical properties of industrial and experimental panels made from Pinus and obtained average values of 14.5% for WA2h and 49.6% for WA24h for industrial panels.
Mendes et al. 7 studied the effect of the addition of sugarcane bagasse on the physicomechanical properties of particleboard panels made from Pinus spp.They found higher WA2h and WA24h values as the percentage of sugarcane bagasse was increased, with average values ranging from 32.1% to 54.1% for WA2h and from 49.8% to 64.2% for WA24h.
In general, the data obtained for industrial panels made from Pinus, Eucalyptus, and sugarcane bagasse are consistent with those reported in the literature, and in some cases, we obtained lower average values.

Thickness swelling and springback
The average values of thickness swelling 2 (TS2h) and 24 h (TS24h) after immersion and the average values of springback are presented in Figure 4.
The industrial panels made from Pinus differed from other panels in TS2h, presenting the lowest average value.There was no significant difference in TS between panels made from Eucalyptus and sugarcane bagasse.
The industrial panels made from sugarcane bagasse did not differ significantly from the panels made from Pinus and Eucalyptus in TS24h, but the panels made from Pinus and Eucalyptus differed significantly from each other in this property.The lowest average value was obtained for Pinus panels.
All treatments differed significantly from each other in springback.The Pinus panels showed the lowest average value, whereas the Eucalyptus panels showed the highest average value and a direct relationship with the property TS24h.
The high TS24h and springback values obtained for the Eucalyptus panels may be explained by a lower compaction ratio between particles, resulting in larger empty spaces and therefore better penetration of water in the panel structure.Figures 5 and 6 present the amount of empty spaces in Eucalyptus panels.
When comparing the quality of particleboard panels produced with sugarcane bagasse and Eucalyptus, Pedreschi 25 found average values of 3.1% for TS2h, 12.5% for TS4h, and 8.0% for springback for sugarcane bagasse panels.For the particleboard panels produced with Eucalyptus, Pedreschi 25 obtained average values of 5.8% for TS2h, 17.4% for TS4h, and 17.4% for springback.
Barros Filho et al. 12 assessed the quality of particleboard panels with sugarcane bagasse mixed with Eucalyptus using urea-formaldehyde and melamine-formaldehyde as adhesives and obtained average values between 7.0% and 26.5% for TS2h, 16.3% and 36.2% for TS24h, and 11.1% and 27.7% for springback.
Tabarsa et al. 9 assessed the potential use of bagasse as an alternative raw material for the production of panels and obtained values ranging between 15.1% and 18.1% for TS2h, and 22.6% and 24.5% for TS24h.
Mendes et al. 17 compared industrial panels and found average values for TS2h of 5.9% for Pinus, 1.8% for Eucalyptus, and 4.5% for sugarcane bagasse, and average values for TS24h of 19.7% for Pinus, 12.8% for Eucalyptus, and 12.0% for sugarcane bagasse.
The average values for the three types of panels studied here are consistent with the literature, and in some cases, we obtained lower average values.However, the average values obtained for TS24h in all treatments did not meet the EN 312 standard 24 , which defines a maximum value of 8%.All the panels met the Commercial Standard -236-66 27 , which dictates a maximum value of TS 24 h after immersion of 35%.
In general, we observed that in terms of the assessed physical properties, the particleboard panels made from sugarcane bagasse are inferior to the panels produced with Pinus.However, sugarcane bagasse panels are equal to or even superior to Eucalyptus panels in some properties.Hence, we may say that in terms of quality, sugarcane bagasse panels present properties in accordance with the panels commonly used for furniture production, and are, in terms of their physical properties, suitable for this use.

Mechanical characterization 3.2.1. Internal bond
The internal bond average values are presented in Figure 7.The industrial panels made from Pinus and Eucalyptus did not show significant differences from each other, but differed from the panels made from sugarcane bagasse, which showed the lowest average values.The reason for the lowest average value of internal bond for the panels made from sugarcane bagasse may be related to the lower amount of adhesive available per particle.To obtain a particleboard panel with the same density using this residue, a larger number of particles is required, mainly due to the lower density of the material.
Mendes et al. 17 found average values of internal bond of 0.5 MPa for industrial panels made from Pinus, 0.6 MPa for panels made from Eucalyptus, and 0.2 MPa for panels made from sugarcane bagasse.Barros Filho et al. 12 assessed the quality of particleboard panels made from sugarcane bagasse mixed with Eucalyptus and by using urea-formaldehyde and melamine-formaldehyde as adhesives, and they obtained average values of internal bond ranging between 0.20 and 0.63 MPa.
In general, the characteristics of panels produced with Pinus, Eucalyptus, and sugarcane bagasse assessed in the present study are consistent with data from the literature.Only the treatments with Pinus and Eucalyptus met the minimum values established by the A 208.1 28 and NBR 14810-2 standards 29 , which is 0.40 MPa, as well as the EN 312 standard 24 , which sets a minimum value of 0.30 MPa.

Modulus of elasticity and modulus of rupture to static bending
The average values of the MOE and MOR are presented in Figures 8 and 9.
The industrial panels made from Pinus and sugarcane bagasse did not differ from each other in MOE and MOR for static bending.However, the Eucalyptus panels differed significantly from the other two treatments and showed the highest average value.
The highest MOE and MOR values obtained for the panels produced with Eucalyptus may be explained by the physical and anatomical characteristics of this material as well as by associations between these factors and variables of panel production.
The lower average value observed for panels produced with sugarcane bagasse compared with panels produced with Eucalyptus may result from the presence of medulla in sugarcane bagasse 30 , which is a material with low mechanical resistance, and from the lower internal bond between particles, as observed in Figure 7.
Widyorini et al. 13 assessed the quality of particleboard panels produced with sugarcane bagasse with and without medulla in association with some processing variables and obtained MOE values between 400 and 1600 MPa and MOR values between 2 and 11 MPa for static bending.
Mendes et al. 7 assessed the effect of the percentage of addition of sugarcane bagasse to Eucalyptus (25, 50, and 75%) with different adhesive types (urea-formaldehyde and phenol-formaldehyde) and contents (6, 9, and 12%) on the production of particleboard panels.They obtained MOE values between 915.3 and 1064.7 MPa and MOR values between 9.5 and 11.5 MPa for static bending.
Mendes et al. 31 assessed different types of lignocellulosic materials produced in Brazil for the production of particleboard panels with urea-formaldehyde adhesive.For static bending, they obtained average MOE and MOR values of 1643.2 and 20.9 MPa for panels produced with sugarcane bagasse, 1622.6 and 19.9 MPa for panels produced with Pinus, and 1654.8 and 21.8 MPa for panels produced with Eucalyptus, respectively.For static bending, the average values of MOE and MOR for the three types of panels assessed in the present study are consistent with the literature.All the panels met the ANSI A208-1 trading standard 28 , which defines minimum values of 1943.8MPa for MOE and 12.8 MPa for MOR, for standard panels (M-S).All the panels also met the EN    312 standard 24 , which defines minimum values of 1800 MPa for MOE and 13 MPa for MOR for particleboard panels intended for internal use.The three types of panel differed significantly in both properties.The Eucaliptus panels had the highest average values, followed by Pinus panels and sugarcane bagasse panels, which had the smallest MOR and MOE values in the compression test.This may be explained by the lower density of sugarcane bagasse, which requires a larger volume of particles to produce a panel of predetermined density, which reduces the amount of adhesive per particle, and thus decreases the bond among particles (Figure 7) and consequently the parallel compression.

Modulus of elasticity and modulus of rupture for parallel compression
Iwakiri et al. 32 assessed particleboard panels of Pinus spp.produced with 8% of urea-formaldehyde adhesive, with densities between 0.60 and 0.90 g/cm³, and found average values of MOR for compression between 6.6 and 15.9 MPa, respectively.
Mendes & Mendes 33 produced particleboard panels made from sugarcane bagasse from industry and a cachaça distillery, with a nominal density of 0.70g/cm 3 and by using a content of 9% urea-formaldehyde adhesive.The average value of MOR for compression was 6.2 MPa for the panel made from industrial sugarcane bagasse and 5.2 MPa for the panel made from sugarcane bagasse from the cachaça distillery.
Pedreschi 25 found average values of MOR for compression of 7.3 MPa for particleboard panels made from sugarcane bagasse.For the industrial particleboard panels made from Eucaliptus, the average value was 9.1 MPa.
In general, the average values of the three panels studied here are consistent with the literature.

Janka hardness
The average values for Janka hardness are presented in Figure 12.There was no significant difference between treatments.However, we obtained higher average values for Eucalyptus and sugarcane bagasse panels.The higher average value for Eucalyptus panels may be explained by anatomical factors such as higher wall fraction, whereas the higher average value for sugarcane bagasse panels may be explained by a higher compaction ratio of the panels, as reported by Mendoza 34 .
Bianche et al. 35 assessed the quality of particleboard panels produced with Eucalyptus urophylla and Schizolobium amazonicum woods mixed with broom fibers (Sida spp.) and 6% or 8% of urea-formaldehyde adhesive, and they obtained Janka hardness values between 34.1 and 50.5 MPa.
Milagres et al. 36 determined the properties of panels produced with 8% urea-formaldehyde adhesive with a mixture of particles of Eucalyptus grandis, high-and low-density polyethylene, and polypropylene.In the panels composed of wood only, the average value of Janka hardness was 49.6 MPa.
The trading standard ANSI A280-1 28 determines a minimum value of 22.7 MPa for Janka hardness of particleboard panels.
The average values obtained in the present study for the three types of panels were below the values reported in the literature but met the trading standard.

Screw pullout -top and surface
The average values of the top and surface screw pullout tests are presented in Figures 13 and 14, respectively.
The panels made from Pinus and sugarcane bagasse did not differ from each other in resistance to screw pullout, whereas the panel made from Eucalyptus differed significantly from the other two, as it had a higher average value.
In general, the average values obtained for screw pullout on the surface of panels were higher than the average values obtained for screw pullout on the top.This difference may be explained by a lower densification of the particles in the core of the panel by the smaller size of particles on the surfaces, which promotes their higher compaction, and by the density gradient created in the pressing of the panels.
Iwakiri et al. 37 assessed the quality of homogeneous and multilayer particleboard panels produced with the wood of  Melia azedarach and Pinus taeda with 6%, 8%, and 10% of urea-formaldehyde adhesive.In multilayer panels, the average values ranged between 103.2 and 182.6 Kgf for screw pullout on the surface and between 75.4 and 174.2 Kgf for screw pullout on the top.
Weber 38 assessed the potential use of residues originated in the production of plywood panels, MDF (Medium density fiberboard), and MDP to produce particleboard panels with 6% and 10% of urea-formaldehyde adhesive and recorded average values for screw pullout on the surface between 73.2 and 117.7 Kgf and for screw pullout on the top between 58.0 and 132.0 Kgf.Silva et al. 39 assessed the screw pullout of particleboard panels made from sugarcane bagasse mixed with Eucalyptus (25%, 50%, and 75%) and with 6%, 9%, and 12% of urea-formaldehyde adhesive.They found average values of screw pullout on the surface between 32 and 110 Kgf and for screw pullout on the top between 25 and 78 Kgf.
The ANSI A208.1 trading standard 28 defines a minimum value of 92 Kgf, whereas the NBR 14810-2 standard 29 defines a value of 81.6 Kgf.
The average values obtained in the present study for the three types of panels were close to or below those reported in the literature.No treatment met the value defined by the ANSI A208.1 standard 28 , and only the Eucalyptus panels met the NBR 14810-2 standard 29 .

Conclusion
We observed that the panels made from sugarcane bagasse were inferior to the panels made from Pinus in terms of their physical properties.However, they were equal or even superior to the panels made from Eucalyptus in some properties.The three panels met the CS 236-66 trading standard 27 .
In a general analysis of all mechanical tests, we observed that the particleboard panels made from sugarcane bagasse were only inferior in terms of average values to those of Pinus and Eucalyptus for internal bond and MOE and MOR for parallel compression.For other properties, the performance of these panels was similar to those made from Pinus and Eucalyptus.The sugarcane bagasse panels did not meet the requirements of the standards for internal bond and screw pullout.
In general, we observed that the panels made from sugarcane bagasse showed physical and mechanical properties close or even superior to those measured for Pinus and Eucalyptus panels, which are commonly used for furniture production.Hence, we conclude that sugarcane bagasse panels have great potential for use in the furniture industry.

Figure 1 :
Figure 1: Average values of bulk density.Average values followed by the same letter did not differ from each other in the Tukey test at 5% significance level.

Figure 2 :
Figure 2: Average values of moisture.Average values followed by the same letter did not differ from each other in the Tukey test at 5% significance level.

Figure 4 :
Figure 4: Average values of thickness swelling two and twenty-four hours after immersion and springback.Average values followed by the same letter did not differ from each other in the Tukey test at 5% significance level.

Figure 3 :
Figure 3: Average values of water absorption two and twenty-four hours after immersion.Average values followed by the same letter did not differ from each other in the Tukey test at 5% significance level.

Figure 7 :
Figure 7: Average values of internal bond.Average values followed by the same letter did not differ from each other in the Tukey test at 5% of significance level.

Figure 8 :
Figure 8: Average values of the modulus of elasticity (MOE) to static bending in MPa.Average values followed by the same letter did not differ from each other in the Tukey test at 5% significance level.

Figure 9 :
Figure 9: Average values of the modulus of rupture (MOR) to static bending in MPa.Average values followed by the same letter did not differ from each other in the Tukey test at 5% significance level.

Figure 10 :
Figure 10: Average values of the modulus of elasticity (MOE) to compression in MPa.Average values followed by the same letter did not differ from each other in the Tukey test at 5% significance level.

Figures 10 and 11
Figures 10 and 11 indicate average values for the MOE and MOR for compression.The three types of panel differed significantly in both properties.The Eucaliptus panels had the highest average values, followed by Pinus panels and sugarcane bagasse panels, which had the smallest MOR and MOE values in the compression test.This may be explained by the lower

Figure 11 :
Figure 11: Average values of the modulus of rupture (MOR) to compression in MPa.Average values followed by the same letter did not differ from each other in the Tukey test at 5% significance level.

Figure 12 :
Figure 12: Average values of Janka Hardness in MPa.Average values followed by the same letter did not differ from each other in the Tukey test at 5% significance level.

Figure 13 :
Figure 13: Average values of screw pullout -top in Kgf.Average values followed by the same letter did not differ from each other in the Tukey test at 5% significance level.

Figure 14 :
Figure 14: Average values of screw pullout -surface in Kgf.Average values followed by the same letter did not differ from each other in the Tukey test at 5% significance level.