Properties of multi-laminated plywood produced with Hovenia dulcis Thunb . and Pinus elliottii wood under different pressing pressures

This study aimed at evaluating the effect of different pressing pressures and the influence of the wood features on the properties of plywood produced with sapwood and heartwood Hovenia dulcis combined with Pinus elliottii. To support the discussion, the anatomical and physical characterizations of the wood were carried out. The panels were produced by applying three different press pressures (0.88, 1.18 and 1.47 MPa) and with six combinations of wood veneers. Phenol-formaldehyde resin was employed, 160 g m in a simple line and 35% solid content. The anatomical analysis revealed that the sapwood is more permeable than the H. dulcis heartwood. The H. dulcis wood basic density it was higher than that of the P. elliottii. Increased press pressure raised the values of the apparent density, thickness swelling and thickness swelling plus recovery of the plywood and water absorption reduction. The panels produced with H. dulcis veneers presented higher apparent density, MOR, MOE and bonding line resistance, as well as lower water absorption and moisture content, than those produced with P. elliottii veneers. No difference was noticed regarding the plywood properties when the main effects were evaluated in relation to the use of H. dulcis heartwood and sapwood veneers.


Introduction
The physical mechanical properties of multilaminated plywood can be better than those of wood particle panels.When compared to sawn wood, they present better applicability due to the dimensions in which they are produced.These characteristics are important for the destination sectors of these panels, mainly building industry and furniture production in Brazil.Plywood might be used inside buildings when the glue urea-formaldehyde is applied and externally or 'water proof' when employing a phenol-formaldehyde based glue (Iwakiri, 2005).
In Brazil, multi-laminated plywood is produced with fast growing wood species such as slash pine, eucalyptus and more recently the paricá, a tropical species with large planted area in the north of the country (Pinto & Iwakiri, 2013).
Hovenia dulcis is a fast growing pioneer species, and studies have been developed on its plantation to implement reforestation and characterize this wood.The wood is moderately heavy, with basic density values ranging between 0.58 and 0.62 g cm -3 (Napoli, Sanches, Iwakiri, & Hillig, 2013).The authors confirmed that the timber is suitable for producing particleboards when mixed with eucalyptus wood, providing panels with good dimensional stability.
The wood presents few pores, with small and low frequency rays, short fibers with walls varying from thin to thick.Its basic density of 0.577 g cm -³ is considered medium, 7.68% tangential, 4.36% radial, and 11.76% volumetric shrinkage and 1.77 tangential to radial shrinkage ratio.(Motta, Oliveira, Braz, Duarte, & Alves, 2014).
When evaluating the natural durability of tree species of Eucalyptus sp. and H. dulcis woods submitted to a deterioration test in two environments, field and forest, H. dulcis wood presented lesser deterioration probability in both environments (Carvalho et. al., 2016).However, its wood was considered not resistant to fungal attack from Gloeophyllum trabeum and Trametes versicolor species (Carvalho, Santini, Gouveia, & Rocha, 2015).
The specie H. dulcis is considered invasive, due to its easy dispersion through several means and high adaptability (Lima, Dechoum, & Castellani, 2015).It is found widespread in Brazilian subtropical forests, however, no differences were found between plant communities invaded and non-invaded by H. dulcis at three different succession stages according to Dechoum et al. (2015).
The region of Irati, in the state of Paraná, is one of the places where there is great dispersion of this species.Considering that the region presents large part of its wood sector comprising plywood industries, this sector might become an important consumer of this wood.
Regarding plywood production, Iwakiri (2005) explains that the hot pressing aims to finish the interaction adhesive-wood, through physical and chemical reactions and its variables are pressure, temperature and time.The pressure aims at transferring the glue from one veneer to the other and keeping suitable contact between the veneers.For panels made of several species veneers, the pressure must be adjusted taking into consideration the lowest density wood, since high pressure in light woods might result in reduction in the thickness of the panels and volumetric loss.
When considering species, the differences observed in the glue quality might be explained by the intrinsic qualities of the plywood structure material, highlighting those related to the wood anatomical and physical constitution (Guimarães Júnior, Mendes, Mendes, & Guimarães, 2012).The wood anatomical characteristics influence adhesion, taking as an example the difference in porosity observed in early wood and late wood, heartwood and sapwood, and juvenile and adult wood (Albuquerque & Latorraca, 2005).
The heartwood might present higher density, higher concentration of 'strange' materials such as oil, fat and phenolic compounds when compared to the sapwood, which affect permeability and consequently the adhesive movement, and can also interfere in the chemical reactions of the adhesive curing process (Iwakiri, 2005).The author also recommends the use of lower pressure when the plywood is produced with high gluing rates in the process, aiming to prevent the transfer of adhesive to the external veneer surface and leakage through the panel sides.
According to Vital, Maciel and Della Lucia (2006), the adhesive differentiated behavior on each wood species is possibly due to the variability in density and permeability of each kind of wood.
Research has been developed to evaluate the viability of using alternative species in the production of plywood with phenol-formaldehyde resin.The production of multi-laminated plywood using Sequoia sempervirens wood veneers was studied by Iwakiri et al. (2013) and the quality of plywood made of Schizolobium amazonicum Huber ex veneers was evaluated by Iwakiri et al. (2011).Hevea brasiliensis veneers were used by Palma, Escobar, Ballarin and Leonello (2012).The yield in lamination and quality of plywood made of Criptomeria japonica veneers were evaluated by Pinto and Iwakiri (2013).Veneers of the species Toona ciliate were used by Albino, Sá, Bufalino, Mendes and Almeida (2011).
This study aimed at evaluating the effect of wood characteristics and the use of different press pressures on the properties of multi-laminated plywood, produced with H. dulcis sapwood and heartwood veneers in combination with Pinus elliottii veneers.

Material
Two wood species were used, the H. dulcis Thunb.and the P. elliottii (slash pine), with three trees per species.H. dulcis trees have an invading origin and are situated at the Campus of the Midwestern State University, Irati, Paraná State, coordinates 25º 27' 56" S/ 50º 37' 51" O, and the age of the trees is estimated to be between 14 and 18 years old.The P. elliottii trees, which were 14 years old, were extracted from commercial plantations belonging to a company in the region.

Wood characterization
The H. dulcis and the P. elliottii sapwood and heartwood basic density was determined according to the standard NBR 11941/2003 (Associação Brasileira de Normas Técnicas [ABNT], 2003).
The qualitative anatomic analysis of the H. dulcis heartwood and sapwood was carried out in permanent micro sections produced according to the methodology described by Burger and Richter (1991).Blocks measuring 1 x 1 x 1 cm were used, extracted from the heartwood and sapwood from four discs of each tree.Photographs of the permanent micro sections were taken by employing an electronic microscope with coupled digital camera with 50 x resolution.

Producing veneers and plywoods
The logs were turned into veneers in a logs turning lathe with a fuse traction system, branded Thoms Benato, with the following adjustment: knife sharpening angle 21°, knife angle 90°30', horizontal opening 1.35 mm and vertical opening 0.75 m.The veneers produced were 1.70 wide x 1.5 mm thick.The sheet was cut into a 1.70 x 1.35 m veneer.
The H. dulcis wood veneers were separated according to their position in the log, and comprised 38.32% sapwood veneers, 32.83% heartwood veneers and 28.75% mixed veneers.For the panel production, the sapwood and heartwood veneers were used separately.
The veneers were dried in a roll drier system, branded Benecke, at 130°C and 21m min. - speed, resulting in a 5% final moisture content.
Phenol-formaldehyde (PF) resin was used to produce the panels, with 54% solid content.The gluing was carried out with 67.3% adhesive, 16.35% wheat flour and 16.35% water, so that the mixture reached a 35% solid content and 40 to 60 second viscosity.The gluing was applied manually at a rate of 160 g m -² (simple line), using a spatula across the surface of the veneer, and 30 minutes assembling time.
The relevant variables under analysis were the hot press specific pressure and the combination of wood veneers, totaling 18 treatments with three replications each, according to the experimental design presented in Table 1.
The panels produced measured 60 X 60 cm and had five-veneer layers, 1.5 mm thick each, obeying to the cross lamination assembling principle, pressed at 140°C for 7.5 minutes.
After being pressed the panels were cut and placed in the air-conditioned chamber at 20 ± 2ºC and 65 ± 5% relative moisture up to stabilization.The physical and mechanical properties were determined according to the NBR standards (Associação Brasileira de Normas Técnicas [ABNT], 2011a;2011b;2011c;2011d;2012;2006a;2006b).The panel properties were analyzed after the variance homogeneity and normal distribution prerogatives were satisfied.The factorial ANOVA was applied, with 5% error probability for the main factors and their interactions.When difference was found between the factors, the Tukey test was carried out to compare the averages.

Log basic density
The H. dulcis sapwood and heartwood presented 0.50 g cm -3 and 0.57 g cm -3 basic density, respectively, which was higher than the one presented by the P. elliottii, 0.42 g cm -3 on average.In general, the timber up to 0.50 g cm -3 is preferred for the plywood production, with the sapwood and heartwood presenting the proposed limit and above it, respectively.These results are similar to those found by other authors.H. dulcis basic density values of 0.58 g cm -3 were found by Motta et al. (2014).Basic density values of 0.41 g cm -3 for 13-year old P. elliottii wood were found by Santini, Haselein and Gatto (2000).

Anatomical characteristics of the H. dulcis wood.
Figure 1 shows photographs of permanent anatomical micro sections of H. dulcis wood sample in transversal, radial longitudinal and tangential longitudinal cuts.
The Hovenia dulcis heartwood presented higher amount of extractives, apparently smaller pores and rays with fewer cell rows, when compared to the sapwood.Also, differences were found in pore distribution, since the sapwood presented multiple distributions up to three and the heartwood up to two.These differences revealed that the Hovenia dulcis sapwood is more permeable than the heartwood.

Plywood apparent density, thickness and moisture content
In Table 2, the apparent density average values factorial analysis, thickness and moisture content for each level of factor under study can be observed.The press pressure influenced the apparent density and thickness of the panels.Thus, the use of 1.47 MPa pressure was not advantageous to the species under study, since it produced heavier plywood and presented higher plywood thickness reduction when compared to the application of 0.88 MPa.The use of 0.88 or 1.18 MPa pressure provided panels with lower apparent density than the use of 1.47 MPa.The application of high pressure on low density wood results in reduction in the plywood thickness (Iwakiri, 2005).
Regarding plywood composition, the compositions with H. dulcis veneers presented higher apparent density values when compared to the mixed and pine compositions.The plywood apparent density variation is related to the basic density of the species used in this study.
The pine plywood was verified to reach higher moisture content, followed by the mixed plywood (H.dulcis and pine) and the panels made of H. dulcis veneers only.This fact is related to the physical and chemical characteristics of each wood species.

Dimensional stability
Table 3 shows the average values of dimensional stability properties for each level of factor under study.Regarding water absorption, statistical difference was observed in the average values obtained from different pressures and compositions used in the plywood, which revealed factor interactions.
Plywood thickness swelling (TS) and thickness swelling plus recovery (TSR) only presented statistical difference for the factor pressure.
There was reduction in water absorption by the panels produced at 1.18 MPa pressure when compared to those produced at 0.88 MPa.This fact was ascribed to the reduction of empty spaces in the wood due to compression, an effect that did not occur when the pressure was increased from 1.18 to 1.47 MPa.
The factor composition revealed that the panels produced with H. dulcis heartwood veneers presented the lowest water absorption, followed by the mixed (sapwood and heartwood), sapwood and pine.The flow of liquids through the woody structure is related to its porosity; this fact can be explained by the anatomical characteristic differences between the H. dulcis sapwood and heartwood, as well as by the wood basic density difference between pine and H. dulcis.
Table 4 presents average values of each treatment resulting from the combination of factors such as plywood composition and press pressure on water absorption.
It was seen that the increase in pressure from 0.88 to 1.18 MPa was efficient to reduce water absorption in all panels that contained pine veneers and in those only made of H. dulcis heartwood.For panels only produced with H. dulcis sapwood veneers, the water absorption reduction only occurred with 1.47 Mpa pressure, while in the mixture heartwood with sapwood there was no water absorption reduction with the pressure increase.
Regarding thickness swelling and thickness swelling plus recovery, results revealed that averages were higher when the pressure 1.47 MPa was applied to the panel production process, since with pore and fiber compaction there was increase in the panel apparent density and in the internal compression tensions.

Static bending
Table 5 shows the modulus of rupture (MOR) and modulus of elasticity (MOE) average values in static bending, parallel and perpendicular to the cover veneer, for each level of the factors under study.The perpendicular and parallel MOR results analyzed revealed statistical difference in average values obtained for the different panel compositions.Regarding perpendicular and parallel MOE, the panel composition and different pressures applied influenced the results, and interaction between composition and pressure factors was also observed in the perpendicular MOE results.
The 0.88 MPa pressure might not have been enough to transfer the adhesive from one veneer to the other, and at 1.47 MPa excessive absorption of the adhesive by the veneer might have occurred, both influencing the panel rigidity.Thus, the plywood with the highest elasticity module was produced at 1.18 MPa press pressure.
Panels produced with H. dulcis venners presented greater parallel MOR and MOE than panels produced with pinus.Panels only made of H. dulcis veneers that presented higher apparent density (Table 2) and higher perpendicular MOR and MOE when compared to the others.Better MOR and MOE static bending results in panels with higher apparent density were also observed by Albino et al. (2011).This is an important factor for the panel structural uses since they require the increased strength and rigidity.
The interaction of plywood composition and press pressure in the perpendicular MOE results can be seen in Table 6.
Panels containing H. dulcis heartwood veneers obtained increase in the elasticity module (MOE) when pressure above 1.18 MPa was applied, since they present more obstructed pores and need higher pressure for adhesive transfer.In the composition containing only H. dulcis sapwood veneers and in the mixture of sapwood and pine, the 1.47 MPa press pressure caused reduction in MOE values, probably due to the excessive adhesive absorption by the panel veneers.In mixed panels made of H. dulcis heartwood and pine veneers, the use of 1.18 MPa in the press process was enough to distribute the adhesive, and the application of 1.47 MPa might have resulted in uneven distribution between veneers, due to the higher impermeability of the H. dulcis heartwood when compared to pine, which caused higher adhesive absorption by the latter.
In the panels only made of pine wood, the application of 0.88 MPa in the press process was enough and provided efficient gluing, agreeing with Iwakiri (2005) which recommends the use of lower pressure when the plywood is produced with high gluing rates.

Bonding quality
Table 7 shows the bonding line resistance average values, after the sample treatments using the humid and boiling methods, for each level of factor under study.According to the NBR ISO 12466/2 (Associação Brasileira de Normas Técnicas [ABNT], (2006b)bonding quality requirements, for BLR equal to or over 1.0 MPa there is no need to determine the percentage of wood failure in the bonding quality test.Since the average values are around this minimum or more, the evaluation of bonding quality was carried out according to the bonding line resistance values only.
The pressing pressure factor did not influence the bonding line resistance (BLR) results.Statistical difference was observed in the BLR after humid treatment average values, obtained for the different panel compositions.For BLR after boiling, there was statistical difference regarding the factor composition as well as interaction between the factors.
The panels produced with H. dulcis heartwood veneers presented better BLR results when compared to the panels that contained pine veneers in their composition.The anatomical characteristics between the species influenced gluing, since the H. dulcis wood has short fibers, small and scarce rays, with smaller pores in the heartwood, absorbing less adhesive than the pine.
According to results obtained by Albino, Mori and Mendes (2012), when pore diameter dimension, the width and the fiber length and ray width are too high, they might result in a thick glue line since more adhesive is absorbed.
The interaction between plywood composition and press pressure in the BLR after boiling results is presented in Table 8.For panels produced with H. dulcis sapwood veneers, BLR after boiling was reduced with the application of 1.47 MPa, while for panels produced with mixture of the sapwood and heartwood veneers the inverse occurred.For other compositions there was no statistical difference between the pressures used.
Since it was found that the sapwood is more permeable than the heartwood, the use of less pressure to this type of veneer prevents high absorption of adhesive.The mixture veneer composition of heartwood and sapwood for the use of 0.88 MPa did not provide a good adhesive absorption for the heartwood veneers, damaging the bonding.

Conclusion
The plywood composition with H. dulcis veneers resulted in heavier panels than those containing pine.The 1.47 MPa press pressure led to increase in the plywood density and thickness reduction, and the 1.18 MPa pressure was the most suitable, since it was enough to reduce water absorption and maintaining the resistance and rigidity.
The plywood properties were higher for H. dulcis veneers, except for thickness properties, and there was no difference of heartwood and sapwood veneers.
The H. dulcis wood presents potential for use in structural multi-laminated plywood panels, mainly in uses that require greater resistance to static bending.

1
The equation used to calculate thickness swelling was modified to match the concepts supplied by the standard NBR.Acta Scientiarum.Technology Maringá, v. 39, n. 2, p. 185-192, Apr.June, 2017

Table 2 .
Influence of press pressure and veneer composition on the plywood apparent density, thickness and moisture content.

Table 3 .
Influence of press pressure and veneer composition on the panel water absorption, thickness swelling and thickness swelling plus recovery.

Table 4 .
Interaction of press pressure with plywood composition in the panel water absorption property (%).

Table 5 .
Influence of factors on the MOR and MOE static bending, both parallel and perpendicular.

Table 6 .
Interaction of press pressure with plywood composition in perpendicular MOE (MPa).

Table 7 .
Influence of press pressure and plywood composition on the bonding line resistance (BLR).

Table 8 .
Interaction of pressing pressure and the plywood composition in the bonding line resistance (BLR) after boiling.
Note: Averages followed by the same letter were not statistically different (capital letter in the lines and small letter in the columns), Tukey test at 5% error probability; P: P. elliottii veneers; S-Hd: H. dulcis sapwood veneers; H-Hd: H. dulcis heartwood veneers.