INFLUENCE OF REINFORCEMENT ON WOOD TENSILE STRENGTH SUBMITTED TO WEATHERING

Wood is an important building material used in roof structures, bridges, viaducts, and pedestrian bridges. In order to improve its use, as well as rehabilitate structural parts deteriorated by weathering, reinforcement of beams and columns with composite material is performed. Given the climate action on structures, it is important to know the infl uence and behavior of wood and composite material when subjected to weathering. This study aimed to evaluate the infl uence of the reinforcement of CCB-treated wood with carbon fi ber reinforced polymer (CFRP) and subjected to artifi cial weathering on its tensile strength parallel to the fi bers. The results showed a 30% reduction in strength due to weathering and a 25% increase in tensile strength, showing the positive infl uence of reinforcement even under artifi cial weathering conditions


INTRODUCTION
Wood is an important material for many applications, which is used in construction, furniture, sport equipment and musical instruments. It has a good ratio between mechanical strength and density, besides being natural, renewable and abundant in Brazil (Adamopoulos and Passialis, 2010;Beech, 2017;Viluma, 2017;Rodrigues and Christoforo, 2019).
Considering the use of wood in civil construction, it can be employed in roof elements, anchors, structures of reinforced and prestressed concrete, sleepers (railways), works of art as in bridge structures, viaducts and walkways, storage structures (silos), among others (O'Born, 2018;Ramage, 2016;. Reinforcing wooden beams and columns with durable and economical materials, such as glass and carbon fi bers, can optimize this versatile material and increase its load carrying capacity, making possible and easier, for instance, the restoration of historic buildings (Johns and Lacroix, 2000;Alam, 2009;García, 2016). There are several studies in the literature about reinforcing wooden structural members with composite materials (Ramage, 2016;Cestari, 2013;Burawska, 2015;Dewey, 2018).
Some works have characterized Erisma uncinatum species, though the studies didn't consider applying preservative treatment or any reinforcement on the wood surface. As a result, the studies found out values of tensile strength parallel to the grain [f t0 ] of 58 MPa and 45 MPa (Dias and Lahr, 2004;Lahr et al., 2016).
The application of polymeric resin reinforced with carbon nanotubes in old wooden structures was tested. Samples of some non-tropical species were extracted from an ancient structure, reinforced with the application and tested in laboratory. The results pointed to an increase of up to 35% in bending strength of the samples (Cestari et al., 2013).
Local Reinforcement of CFRP covering 5% of the entire beam length was used to analyze its eff ectiveness on wood strength through the static bending test. As a result, local reinforcement led to a signifi cant increment in strength and stiff ness of the beams (Burawska et al., 2015).
The possibility of reinforcing wood components of a degraded bridge in Australia using carbon fi bers (CFRP) was verifi ed. Some specimens were removed from a bridge structure, so that their mechanical properties could be assessed in laboratory. It was concluded that the reinforcement increased their ductility, ultimate load capacity and bending stiff ness compared to non-reinforced samples (Dewey et al., 2018).
The use of unidirectional and bidirectional carbon fi ber reinforced polymer (CFRP) could be an alternative to increase bending strength of round timber girders. Unidirectional fi bers increased the ultimate load capacity, stiff ness and ductility by 25%, 20% and 30%, respectively, compared to the nonreinforced girders (Globa et al., 2018).
Taking into account that wood structures are subjected to the action of weather, such as humidity, insulation, abrasion, among others, it is important to know how wood reinforced with carbon composite material behaves and how the reinforcement infl uences the properties of strength and stiff ness of wood after aging process (Dewey et al., 2018;Dietsch and Winter, 2018). It was verifi ed that there is no study of reinforcement on wood (CFRP) and its infl uence on tensile strength parallel to the grain using wood treated with preservative, and subjected to weathering.
A study investigating the eff ects of weathering on macroscopical behavior of wood, and wood color characteristics in connection with changes in its molecular and anatomical structure was realized with some untreated tropical wood species.
The results pointed to changes on the top surfaces of all species, such as: creation of visible longitudinal macrocracks, darkening in exterior exposure mainly due to pollutants, this last, except for ipé (Tebebuia serratifolia Nichols.; T. spp.), and vice versa lightening in Xenotest, as well as greening and blueing in both modes of exposure, faster decrease of guaiacyl than syringyl lignin, absolute decrease of conjugated and unconjugated carbonyl groups in the newly formed lignin-polysaccharide-extractive substrate in the photo-oxidized and washed-out cell walls, and decrease of cellulose crystallinity, damaging of cell-walls by microcracks, and their degradation by thinning (Reinprechtet al., 2017).
Two softwood species with surfaces modifi ed with melamine formaldehyde resins were studied aiming to investigate the eff ects of artifi cial weathering. The increase in hardness due to melamine treatment was well preserved after the wood being exposed to long weathering process. The treated samples also showed advantages compared to untreated referential samples regarding discoloration and crack formation. Low molar mass and low degree of methylolation of the melamine resin used was found out to be favorable for a successful treatment (Hansmann et al., 2008).
Objectifying a better understanding of weathering infl uence on tensile strength of CCB-treated wood reinforced with carbon fi ber reinforced polymer (CFRP), this study aimed to analyse the eff ect of carbon fi ber composite on the tensile strength of wooden beams exposed to artifi cial weathering.

MATERIAL AND METHODS
The wood species evaluated in this research were Couratari spp. (Tauari) and Erisma uncinatum (Cambará). These wood species are suitable for structural purposes, such as, for beams. (Lahr et al., 2016;Coimbra et al., 2018). Forty-eight (48) test specimens were prepared to evaluate tensile strength parallel to grain, according to Annex B of Brazilian Standard ABNT NBR 7190 (1997). For each species, 24 test samples were prepared and treated with CCB preservative.
CCB preservative is composed of the mixture of chromium, copper and boron oxide that aims to protect wood from biological attacks. The process of wood impregnation with CCB was carried out under pressure (full cell impregnation), with retention of 7.5 kg/m³, which is higher than the lowest value allowed of 6.5 kg/m³ (ABNT P-EB-474 1973a; Ramos et al., 2006) and also respects the requirements of the standard ABNT NBR 7190 (1997). In Brazil, the preservatives are performed considering chemical and biological attack on wood, which is treated with two possible preservatives for structures: CCA or CCB (Magalhães et al., 2012;Rodrigues et al., 2012). After treatment with CCB, the fi ber-carbon composite was applied to 24 specimens: being 12 of each species (Couratari spp. and Erisma uncinatum). The specimens were sanded and cleaned for subsequently application of the castor-based bicomponent resin to glue the CFRP composite. The laminated composite was glued and the resin healing last three days. (Figure 1(a)) shows the specimens after the preservative treatment and application of carbon fi ber reinforced polymer. Proof tests were made according to the Brazilian Standard NBR 7190 (1997), and the side view of specimens is exposed in (Figure 1(b)).
The Atlas Weather-Ometer Equipment model XW65-WR1 was selected to execute the aging process of the samples, operating with a 6500 W xenon lamp. The specimens were subjected to extreme conditions of temperature and humidity, following the recommendations of the standard ASTM G155 (1999), in which the samples underwent 16 aging cycles in 8 days, being 2 cycles per day, and each cycle duration of 12 hours, totalizing 192 hours, which is equivalent to 2 months under natural weathering.
Twenty-four specimens were used in the artifi cial aging process, in which 12 specimens (6 per species) without the carbon fi ber reinforced polymer, and 12 specimens with the reinforcement (6 per species) being glued before the aging process. The glue line behavior between CFRP and wood, and CFRP infl uence on ancient structures rehabilitation were analyzed (Santos et al., 2013;Chang, 2015 For the validation of ANOVA, the normality of the residual distribution was evaluated using Anderson-Darling normality test, also checked at 5% level of signifi cance, so that the test conditions were: P-value higher and lower than the level of signifi cance [0

.05] implies that the residual distribution is normal [H ] and not normal [H ], respectively.
It is worth mentioning that six specimens were prepared, tested and evaluated for each of the four experimental treatments on wood, totalizing 48 determinations of tensile strength parallel to the grain.

RESULTS
( Figure 2(a; b)) shows mean values, confi dence intervals [Confi dence level at 95%] and coeffi cients of variation [CV] of the tensile strength parallel to the grain for both wood species. The coeffi cients of variation indicate a considerable variation between treatments and species, especially in treatment 4 for Couratari spp. despite it was the opposite for Erisma uncinatum. Moreover, treatment 2 was the lowest value for Couratari spp., but the contrary for Erisma uncinatum. (Figure 2(c)) shows a specimen reinforced with carbon fi ber reinforced polymer after rupture, which the surface darkened and cracked longitudinally. Results of the normality tests of residuals for ANOVA results are shown in (Figure 3). From (Figure 3), it can be seen that the distribution of residuals of f t0 for both wood species were normal, validating the ANOVA model. (Table 2)

DISCUSSION
Considering the data presented in (Figure 2), it was verifi ed that the artifi cial weathering causes  considerable reductions in tensile strength [f t0 ] of both species, which could be inferred from the surface reduction and its darkening, which collaborates to the conclusions in (Reinprechtet al., 2017;Hansmann et al., 2008).
Regarding the carbon fi ber reinforced polymer, it was observed an increase in tensile strength parallel to grain in both species. It can be noted in the literature that the combination of wood with CCB preservative treatment, artifi cial weathering process and reinforcement of composite material (CFRP) for subsequent determination of tensile strength (f t0 ) has not been considered in any studies.
Analyzing the treatments 1 and 2 in (Figure 2(a) In conclusion, the results confi rmed the negative infl uence of artifi cial weathering on wood, and the reason for the higher reduction on tensile strength for Couratari spp. could be attributed to anatomical characteristics, such as, axial parenchyma, which is thicker into this species (IPT, 2013).

The comparison between values of treatment 3 [Reinforced without aging] and 4 [Reinforced with
aging] showed in (Figure 2(a)) revealed a reduction of 30% on tensile strength of Couratari spp. [Tr 3: 100 MPa; Tr 4: 69.7 MPa], likewise the reduction of 16.4% that can be inferred in the (Figure 2(b)) for In percentage terms, the reductions on tensile strength remained close to the rates obtained considering the analysis of treatments 1 and 2. Observing (Figure 2(c)), the failure in the weathered specimens strengthened with CFRP composite sheet occurred in the wood lamination, indicating that the interface wood/adhesive/CFRP was not aff ected by the aging process, which demands more studies to be concluded.
Substituting the four pairs of values [Ref, Nc] associated with the four treatments presented in the  1,8], the use of reinforcement contributed to increase the ft0 by 17%, highlighting the positive infl uence of reinforcement insertion even in condition of exposure to artifi cial weathering.

CONCLUSION
The results of the present research revealed that there was a considerable increase in tensile strength due to the reinforcement of carbon fi ber composite added to the samples of Couratari spp. and Erisma uncinatum species. In addition, it was verifi ed that the weathering action applied to the wood specimens (Couratari spp. and Erisma uncinatum) considerably reduced their tensile strength. The failure of aged wood reinforced with CFRP induced by tensile tests indicates that the aging process did not aff ect the interface wood/CFRP, demonstrating the good adherence of the adhesive between wood and CFRP, demanding more studies to be concluded.
Considering the experimental treatments applied to wood and comparing non-reinforced with reinforced specimens submitted to the action of manmade aging, there was an increase of 30% or higher in tensile strength parallel to grain, showing the infl uence of CFRP as reinforcement.
For more comprehensive understanding of the reinforcement (carbon fi ber Reinforced Polymer) behavior and its infl uence on tensile strength parallel to grain, it is necessary to develop more studies with a larger number of species, weathering cycles, specimens as well as more experimental treatments.

AUTHOR CONTRIBUTIONS
Giovana Gobatto Balanco contributed to carrying out the experimental tests;

Edson
Fernando Castanheira Rodrigues contributed to the writing of the text in addition to the bibliographical research.
Vinicius Borges de Moura Aquino assisted in the translation and writing of the text, in addition to bibliographical research; Túlio Hallak Panzera assisted in the treatment and analysis of the results; Francisco Antonio Rocco Lahr helped in the preparation of the text and also collaborated in the feasibility of the experimental tests; André Luis Christoforo assisted in the analysis, interpretation of results and also in writing Chapters 3 to 5.