EVALUATION OF THE USE POTENTIAL OF NINE SPECIES OF GENUS Eucalyptus FOR PRODUCTION OF VENEERS AND PLYWOOD PANELS

The objective of this study was to evaluate the use potential of nine species of Eucalyptus for production of veneer sheets and multilaminated plywood panels. Veneers were cut using a pilot laminating lathe to a nominal thickness of 2.0 mm. Analysis included finding values of overall yield and yield according to three quality classes for the nine relevant species. Plywood panels were manufactured in a laboratory, consisting of five 2.0 mm veneer sheets which were bonded together with phenol-formaldehyde resin at a weight of 360 g/m2 (double line). The panels were compressed using a specific pressure of 10 kgf/cm2, a temperature of 140°C and a pressing time of 10 minutes. Results indicated that, with the exception of E. phaeotricha and E. pellita, all other Eucalyptus species had above 50% average veneer yield after lamination. Results of glue line shear testing and static bending parallel and perpendicular demonstrated that species Eucalyptus grandis, Eucalyptus saligna, Eucalyptus dunnii, Eucalyptus globulus, Eucalyptus viminalis, Eucalyptus robusta and Eucalyptus pellita have great potential within the parameters of this study for use in the production of veneer sheets and plywood panels intended for outdoor use.


INTRODUCTION
In Brazil, the timber used for production of plywood panels is both sourced from nonnative forests, whose most common species are Pinus taeda and Pinus elliottii, and from native forests consisting of tropical Amazonian species.
The economic, social and environmental contribution made by the nonnative forest sector is very expressive in Brazil.According to 2008 data, Brazil boasts 6.5 million hectares of nonnative forests, out of which 4.26 million are Eucalyptus and 1.87 million are Pinus forests.This sector contributes around R$9 billion in tax collections and creates 2.2 million direct and indirect jobs.And environmentally speaking, that means seven to ten tons of CO 2 being sequestered per hectare/year (SOCIEDADE BRASILEIRA DE SILVICULTURA -SBS, 2008).
Figures provided by a study known as Estudo Setorial da Associação Brasileira de Madeira Processada Mecanicamente -ABIMCI (2009) demonstrated that Brazil produced around 2.6 million m³ of plywood in 2008, out Iwakiri, S. et al.
of which 2.0 million m³ were produced with species from nonnative forests (Pinus spp.), against 0.60 million m³ produced with species from native forests (tropical woods).Figures demonstrated that around 70% of these plywood panels are produced with Pinus sourced from forest stands.Extensive exploitation of traditional Pinus species (Pinus taeda and Pinus elliottii) has created problems relating to supply of timber from large-diameter trees with which to derive good quality veneers for production of plywood panels.
Several studies have been conducted in search for alternative species, looking to increase the supply of quality timber for production of veneers and plywood panels.Some studies conducted by Brazilian researchers and involving Eucalyptus species for veneer and plywood production are reported below.Jankowski (1979) was a pioneer in the study of veneer cutting and plywood production in Brazil.This author used wood from Eucalyptus grandis, E. saligna and E. urophylla and found good results with E. grandis for veneering and plywood production.Jankowski and Aguiar (1983) evaluated the species Eucalyptus saligna, E. grandis, E. triantha, E. microcorys and E. pellita at age 10 years for veneer cutting and plywood production, and concluded that only E. triantha and E. saligna offered acceptable quality veneers which were suitable for manufacture of plywood panels.
Keinert Júnior and Interanmense (1994) studied the behavior of six species of genus Eucalyptus in production of veneers and plywood panels and found good results in their qualitative material assessments.Similarly, Interanmense (1998) and Pio (1996) studied the behavior of four Eucalyptus species for production of plywood panels intended for exterior use and found positive results regarding their mechanical properties.
Bortoletto Júnior (2003) evaluated the behavior of 11 species of genus Eucalyptus for production of plywood panels and concluded that the following species are suitable for production of plywood for exterior use: Eucalyptus pilularis, E. propingua, E. microcorys, E. maculata, E. pyrocarpa, E. tereticornis, E. urophylla, E. pellita, E. torelliana and E. saligna. Iwakiri et al. (2007) studied the properties of structural panels from Eucalyptus grandis and Eucalyptus dunnii as manufactured with different formulations and concluded that both species comply with minimum requirements set in standard EN 314 (EUROPEAN COMMITTEE FOR STANDARDIZATION -EN, 1993) concerning glue line strength in panels intended for exterior use.These authors also found that variations in the amount of glue being used did not significantly influence the mechanical properties of resulting plywood panels.
As far as plywood production is concerned, technological factors should be analyzed carefully, in particular those concerning the physicochemical process involved in bonding veneer sheets (MARRA, 1992).The choice of resin to be used in veneer bonding should consider suitability for the end environment, more specifically urea-formaldehyde should be used for interior applications while phenol-formaldehyde should be used for exterior applications.Parameters such as formulation and quantity of adhesive will directly influence not only the bonding quality but also the production cost of plywood panels (BALDWIN, 1993;SELLERS, 1993).These authors argue that wood density is a critical factor in defining these parameters, given the interactions that occur between wood porosity and glue absorption to finally form an adhesive bond between the veneer sheets.According to Marra (1992), low-density woods absorb larger quantities of adhesive, on account of their high porosity.In such cases, adhesive viscosity should be increased in order to avoid excessive glue line absorption due to porosity.Other factors such as pH and extractives content are also critical as they may affect adhesive curing during the hot-press process.Baldwin (1993) argues that some extractives present in veneers may obstruct the vaporization process and migration from one glue line to another, and then to the edges of the panel.This process being very slow will result in increased vapor pressure within, resulting in a burst at the time the press is opened and consequently in a delamination of the panel.Therefore, heterogeneity and variability in the anatomical, physical and chemical properties of wood, whether among species or among different parts of a single tree, indeed may affect the bonding conditions of veneer sheets during the manufacturing process of plywood.
This study aimed to evaluate the use potential of nine species of genus Eucalyptus for production of veneers and plywood panels intended for exterior use.

MATERIAL AND METHODS
The following species were used in this research: Eucalyptus grandis, Eucalyptus saligna, Eucalyptus dunnii, Eucalyptus globulus, Eucalyptus viminalis, Eucalyptus robusta, Eucalyptus phaeotricha, Eucalyptus deanei and Eucalyptus pellita.At the time of sample Evaluation of the use potential of nine species ... collection, the first two above species were 21 years and the other species were 18 to 19 years.Two trees of each species were sampled from forest stands located in the region of Piên -PR and Jaraguá do Sul -SC.For bonding the veneer sheets a phenol-formaldehyde resin (FF) was used with a solids content of 49%, a pH of 11.5 and Brookfield viscosity of 420 cP.
Each tree was bucked into three logs 0.60 m in length each, to a total of six logs per species prior to veneer cutting.Veneers were obtained using a pilot lathe to a thickness of 2.0 mm and divided using a pneumatic guillotine so as to obtain 600 mm x 600 mm pieces.Samples 20 mm x 300 mm in size were randomly taken from veneers of each species for determination of apparent specific mass, using a total of 20 test pieces per species.
Veneers were graded as A, B and C, according to requirements specified in standard ABNT 31:000.05-001/1(ASSOCIAÇÃO BRASILEIRA DE NORMAS TÉCNICAS -ABNT, 1986).Also determined were the proportions of veneers in each of the three quality classes contributing to the overall yield per species.Veneer yield per species was determined by the ratio of produced veneer volume to log volume without bark.Determinations were made of total yield and yield per quality class for each species being studied, as well as of veneer surface evenness after drying.
Prior to being assembled, the veneer sheets were oven-dried to an average moisture content of 8% and finally cut to a square shape 500 mm x 500 mm in size.Each panel was made with five veneers using phenolformaldehyde resin with the following formulation, in parts by weight: resin = 100 p/w, wheat flour = 15 p/w and water = 15 p/w.Adhesive was applied manually at a rate of 360 g/m² (double line), using a spatula across the surface of the veneer.The panels were compressed using a specific pressure of 10 kgf/cm², a temperature of 140°C and a pressing time of 10 minutes.Two panels were produced per species, to a total of 18 experimental panels.
After keeping the panels in a temperaturecontrolled chamber set at a temperature of 20 + 2ºC and relative humidity of 65 + 5%, test pieces were taken and submitted to the following laboratory tests: glue line strength to shear stress (dry test, boil/dry cycle and 72-h boil) and static bending for determining the modulus of elasticity (MOE) and modulus of rupture (MOR) parallel and perpendicular to panel plane.Tests were performed according to procedures described in European Standards EN 314 and EN 320 respectively.Test results were examined using analysis of variance and Tukey test at the 95% probability level, outlining a completely randomized design.

Apparent specific mass of veneers
Mean values of apparent specific mass for the veneers of each species are illustrated in Table 1.As can be noted, mean values of apparent specific mass ranged from 0.436 g/cm³ for veneers of Eucalyptus grandis to 0.687 g/cm³ for Eucalyptus saligna.These values fall within the medium to low range of reference values reported in literature and thus can be considered suitable for manufacture of plywood.

Yields in the lamination process / veneer quality
Yield results in the lamination process, both overall and by quality class, are given in Table 2.

Glue line strength to shear stress testing
Results of glue line strength to shear testing on panels are illustrated in Table 3.  CV: coefficient of variation; FM: percentage of wood defects; Means followed by the same letter in a column are statistically similar at the 95% probability level.
In the dry test, the mean values of shear stress ranged from 1.37 MPa for E. deanei to 2.78 MPa for E. viminalis.The mean values found for panels made from E. viminalis were statistically similar to values found for panels made from E. grandis and E. robusta but higher in comparison with panels made from the remaining Eucalyptus species.
In the boil cycle test, the mean values of shear stress ranged from 0.81 MPa for E. robusta to 1.82 MPa for E. saligna.Panels made from E. saligna had statistically similar mean values to panels made from E. viminalis but higher values in comparison with the remaining Eucalyptus species.
As for the 72-hour boil test, the mean values of shear stress ranged from 0.86 MPa for E. dunnii to 1.86 Evaluation of the use potential of nine species ... MPa for E. viminalis.Panels made from E. viminalis had statistically similar mean values to panels made from E. grandis and E. saligna but higher values in comparison with panels made from the remaining Eucalyptus species.
Except for species E. robusta, E. dunnii and E. deanei, the mean values of shear stress found for panels from all other species, both in the boil cycle test and in the 72-hour boil test, were found to be above the minimum value of 1.0 MPa defined by standard EN 314-2 (EN, 1993) as being a requirement for panels intended for exterior use.
Results of glue line strength to shear stress found in this study for all nine Eucalyptus species can be considered satisfactory in relation to reference values reported in literature.Iwakiri et al. (2007) studied phenolic plywood panels produced with veneers from E. grandis and E. dunnii and found mean values of shear stress of 2.69 MPa and 2.43 MPa in the dry test, against 1.57MPa and 1.20 MPa in the boil cycle test, respectively for each of the two species.Bortoletto Júnior ( 2003) studied 11 species of Eucalyptus and found mean values of shear stress in the range of 1.9 MPa to 2.4 MPa in the boil cycle test.

Static bending parallel
Test results of static bending in the parallel direction are illustrated in Table 4.
The range of variation in mean values of MOE parallel was 9,378 MPa for panels made from E. deanei and 18,494 MPa for panels made from E. dunnii.Regarding MOR parallel, variation ranged from 72.23 MPa for panels made from E. deanei and 115.68 MPa for E. viminalis.
The mean values of MOE parallel found for panels from E. dunnii, E. viminalis and E. saligna were statistically higher in comparison with panels made from other species of Eucalyptus.As for MOR parallel, the mean values found for panels from E. viminalis were statistically similar in relation to panels from E. dunnii, E. saligna, E. phaeotricha and E. globules but higher in relation to panels from other Eucalyptus species.
With reference to results reported in literature, Iwakiri et al. (2007)   Therefore, it can be said that results found in this study in static bending parallel testing for all nine species of Eucalyptus are considered satisfactory.
The mean values of MOE and MOR in the parallel direction found in this study for all nine Eucalyptus species fully comply with the minimum requirements established in standard DIN 68792 (DEUTSHE INSTITUT FÜR NORMUNG -DIN, 1979) for plywood panels intended for exterior use -concrete moulds, whose values are 5,000 MPa for MOE and 45 MPa for MOR.

Static bending perpendicular
Results of static bending tests in the perpendicular direction are illustrated in Table 5.
The range of variation in mean values of MOE perpendicular was 2,738 MPa for panels made from E. deanei and 4,627 MPa for panels from E. viminalis.For MOR perpendicular, variation ranged from 34.46 MPa for panels made from E. deanei to 53.43 MPa for E. viminalis.As was expected, values of MOE and MOR in the perpendicular direction were much lower in relation to values found for the parallel direction.

Table 2 -
Almeida et al. (2004)veneer quality.Eucalyptus species, ranged from 41.37% for E. phaeotricha to 67.22% for E. saligna.With the exception of E. phaeotricha and E. pellita, all other species had mean lamination yields above 50%.As regards veneer quality, only three species, namely E. grandis, E. viminalis and E. dunnii, derived superior quality veneers (A), despite the percentages below 3%.Higher percentages of class B veneers were found for E. viminalis (45.22%) and E. robusta (40.34%).As for class C, higher percentages of class C veneers were found for E. saligna and E. dunni, with 65.83% and 47.29% respectively.Results found in this study are satisfactory in comparison with values of lamination yield reported in literature for Eucalyptus species.While studying clones I and II of Eucalyptus grandis x Eucalyptus urophylla,Almeida et al. (2004)found mean yields of 51.74% and 56.81% respectively.Pio (1996)found mean yields of 36.47% and 44.00% for species E. scabra and E. robusta respectively.This author cites, as sources for comparison, mean yields for the following species: RT: overall yield; QS: quality of veneer surface (after dry); SO: no undulation; PO: few undulations; MO: many undulations.Mean values of overall lamination yield, for each of the nine

Table 3 -
Glue line strength to shear stress testing.
found 11,058 MPa and 12,699 MPa as mean values of MOE parallel, against 83.70 MPa and 77.50 MPa as mean values of MOR parallel, for plywood panels made from E. grandis and E. dunnii respectively.Bortoletto Júnior (2003), in a study with 11 species of Eucalyptus, found mean values of MOE parallel in the range of 12,336 MPa to 19,331 MPa, against mean values of MOR parallel in the range of 84 MPa to 130 MPa.
MOE: modulus of elasticity; MOR: modulus of rupture; CV: coefficient of variation; Means followed by the same letter in a column are statistically similar at the 95% probability level.
MOE: modulus of elasticity; MOR: modulus of rupture; CV: coefficient of variation; Means followed by the same letter in a column are statistically similar at the 95% probability level.