Thursday, 20 October 2016

La relation entre les propriétés des panneaux de fibres de densité moyenne et les caractéristiques du bois - Collection des mémoires et thèses à l'Université Laval

Chapter I Literature Review
The relationships between fiber characteristics and paper properties were well studied. Dinwoodie (1965) has given a good review of the principal properties of paper in relation to various anatomical and chemical characteristics of pulpwood fiber. According to the review, fiber density, fiber length and fiber strength were concluded as the three major factors controlling paper strength properties. The conclusion may apply, to some extent, to fiberboard panel products.
The strength of an individual fiber has little influence on strength properties of low density fiberboard panels; it becomes significant in medium density fiberboard; and it might be the predominant factor for hardboard (900 kg/m3) and higher (Jones 1960). In low density fiberboards, the failures usually occur in the bond between fibers. If the density of the panel is high enough, contact between fibers is more intimate, thus, causing failure in the fiber itself (Suchsland and Woodson 1990). Therefore, at a low density range, panel strength can be enhanced by improving bond quality, either through adhesive system or increasing fiber to fiber contact area. In high density fiberboards, the strength of fiber itself appears to be more important.
However, the strength of an individual fiber is difficult to measure due to the limitation of the current available facilities. In the last few years, the USDA Southern Research Station Forest Service Research and Development, the Advanced Engineered Wood Composites (AEWC) Center of University of Maine and the University of Southwestern Louisiana have worked cooperatively on developing methods for characterization of the mechanical and physical properties of individual wood fibers (Groom et al. 1998).
Alternatively, mechanical properties of an individual fiber can be known by microfibrils orientation on cell wall S2 layer. Numerous studies demonstrated that fiber strength is highly related to microfibril angle (MFA) (Preston 1934; Bailey and Vestal 1937; Barber and Meylan 1964; Harris and Meylan 1965; Cave 1966; Manwiller 1966; Meylan 1968; 1969; Page 1969; Boyd 1977; Leney 1981; Megraw 1985; Senft and Bendtsen 1985; Cave and Walker 1994; Megraw 1998). MFA is considered to be the major factor in governing fiber strength (Panshin and DeZeeuw 1980).
A tracheid or fiber typical cell wall consists of two walls, the thin primary wall (pr) and the thick secondary wall (S). The secondary wall can be divided into three layers: a thin outer layer with a nearly horizontal helix of microfibrils (S1); a thick central layer with nearly parallel to the cell axis microfibrils (S2); and an inner thin layer with microfibrils almost parallel to those in the outer layer (S3) (Panshin and DeZeeuw 1980). The S2 layer is the thickest in a tracheid or fiber cell wall. Microfibril angle can be defined as the angle between tracheid or fiber axis and microfibril orientation in the S2 layer. Microfibril angle in wood cell walls can be measured using X-ray diffraction (Cave 1966; Boyd 1977), polarized light microscopy (Preston 1934; Manwiller 1966; Page 1969; Leney 1981), and other observations (Bailey and Vestal 1937; Cockrell 1974; Senft and Bendtsen 1985).
Microfibril angle is a critical micro-structural characteristic that influences wood (shrinkage and MOE) and fiber properties (fiber strength and fiber length) significantly (Barber and Meylan 1964; Harris and Meylan 1965; Meylan 1968; 1969; Cave and Walker 1994; Megraw 1985; 1998). The curvilinear relationships of longitudinal and tangential shrinkage to microfibril angle are shown in Figure 1-2.
The relationship between the longitudinal Young’s modulus of a single fiber and its mean microfibril angle is also non-linear (Navi 1988). Figure 1-3 shows that when the value of microfibril angle is small it has a dominant influence on the longitudinal modulus of a single fiber. This is to say that a small variation in microfibril angle causes a large variation in Young’s Modulus (Mark and Gillis 1973). With the increase of microfibril angle value, the degree of modulus variation reduces in terms of microfibril angle variation. A study by Groom et al. (1998) shows that microfibril angle was related to the modulus of elasticity of MDF panels that were made from 51-year-old plantation loblolly pine ( Pinus taeda L.). Figure 1-4 shows the panel modulus of elasticity as a function of the longitudinal modulus of elasticity of the individual fibers. The study also shows an inverse correlation between panel stiffness/strength and the individual fiber mechanical properties in the longitudinal direction.  Note: The unit of modulus of elasticity in Groom et al.’s (1998) study was psi. It was converted to MPa by the author. 
Fiber morphology such as cell wall thickness, length and width is of importance in determining paper sheet properties (Dinwoodie 1965; Rudie 1998). These fiber morphological characteristics may influence fiberboard properties.
a. Fiber Density
Fiber density, which is closely related to wood density, is usually defined as cell wall thickness or sometimes defined as the ratio of cell wall thickness to lumen diameter (Panshin and DeZeeuw 1980). Dix et al. (1999) studied the properties of wood-based panels manufactured from poplar and eucalyptus wood, and indicated that cell wall thickness is an important factor to the manufacture of both fiberboard and particleboard. Generally, the thin-walled fiber can readily collapse and be compressed. Fiber-to-fiber contact is more intimate for those thin-walled fibers. Therefore, good bonded panels can be easily made with thin cell-walled fibers. Groom et al. (1999) found that MDF stiffness and strength properties increased with increasing the percentage of thin-walled loblolly pine (Pinus taeda L.) juvenile wood fiber into furnish.
b. Fiber Length
Longer fiber may have the tendency to yield higher bulk density (Suchsland and Woodson 1990). In general, low bulk density is associated with good panel strength properties (Nelson 1973). Fiber length was found to be positively related to linear stability of fiberboards (Nelson 1973). Fiber orientation in-plane of a board can be affected by the length of the fibers. Shorter fibers are more likely to develop a vertical component in a board (Suchsland and Woodson 1990). The vertical component can weaken the fiber to fiber bond since the contact area between the fibers is diminished.
However, a study by McMillin (1969) shows that most properties of wet-formed hardboard were improved by using fiber refined from wood having short, slender tracheids with thin walls. The hardboards were made from loblolly pine ( Pinus taeda L.) groundwood. The fiber morphological properties measured were single cell-wall thickness, radial lumen diameter, and radial tracheid width and tracheid length for both earlywood and latewood. It was concluded that short tracheids were more desirable than long tracheids because of a greater number of fibers crossing per unit weight in the mat. Using multiple regression analysis, equations were developed to estimate fiberboard properties. The equations showed that most properties of fiberboard panels were relevant to tracheid length. All equations were of type: y = b0 + b1x1 + b2x2 + …, where y is a dependent variable representing fiberboard properties; bi is a regression coefficient; xi is an independent variable representing weighted tracheid morphological characteristics.
c. Fiber Width
Fiber width or lumen diameter is also important since fiber width combined with fiber length determine fiber surface area. Generally, with a specific resin content, the good bond can be achieved with fiber materials having larger surface area. 
Wood pH and buffering capacity are considered as the important chemical characteristics on many occasions when gluing and coating with adhesives (Johns and Niazi 1980; Johns et al. 1985). The pH of wood reflects the acidity of the wood. Buffering capacity is the amount of acid or base that a buffer can accept without a change in pH. The greater the amount of acid or base needed to change pH, the more resistant they are to change in pH. Buffered solutions or buffers are solutions which resist a change in pH when small amounts of acid or base are added. Buffers contain an acidic species to neutralize OH- ions and a basic species to neutralize H+ ions. However, the two species must be able to co-exist in a solution without completely neutralizing each other.
For a weak acid HX, the following equilibrium occurs:
HX ⇔ H+ + X-
If an amount of H+ ions is added, the reaction will shift to the left. The reaction will cause the H+ to decrease to what it was before, and thus the pH will stay fairly constant. If OH- ions are added, they will react with H+ ions to form water, thus increasing the pH. The equilibrium reaction will shift to the right as H+ are removed. The most effective buffering solutions are those that have similar concentrations of HX and X- so as to have the capacity to absorb both acid and base with the same effectiveness in either direction.
The acid buffering capacity equals to the number of ml of NaOH solution required to raise the starting pH of the wood (or wood fiber) extract to a pH of 7 multiplied by the normality of the base solution. The base buffering capacity can be calculated using the number of ml of H2SO4 solution required to reduce the starting pH of the wood (or wood fiber) extract to a pH of 3 multiplied by the normality of the acid solution. Total buffering capacity is the sum of the acid and base buffering capacity (Lambuth 1967).
As we know, different wood species have different acidity and buffering capacity. In general, wood pH ranges from 3.0 to 5.5 (Stamm 1964; Subramanian et al. 1983). In panel making industry, pH and buffering capacity are used to measure the acidity of wood or fiber and its capability to change the field of acidity to a slower or less reactive field. UF resins are more sensitive than phenol-formaldehyde (PF) resins since UF resins are acid-catalyzed. Buffering capacity can affect the cure rate of resins. If a UF resin was used during composite panel process, higher pH value and lower buffering capacity of wood or fiber are always desirable (Maloney 1993; Hsu 1997; Myers 1977; Myers 1978; Johns and Niazi 1980; Albert et al. 2002). Nelson (1973) found that the pH of fiber influenced all strength properties and linear stability of MDF panels, and higher pH of the pulp caused stronger and more stable (in length) panels. 
Fiber composition refers to the percentages of fine fibers, fiber bundles, whole fibers and broken fibers in the pulp (Myers 1987). It may be true that less fiber bundles and fines are more desirable since fiber bundles destroy uniformity of fiber distribution and fines result in more resin consumption (Groom et al. 1999; Moss and Retulainen 1995). Barnes (2002) found that increasing fine fiber content in evenly distributed aspen OSB panels reduces the panel bending properties in a linear relationship between 100 percent strands and the 100 percent fines products. Additional information can be found in Groom et al. (1999) showing that the stiffness, strength, and IB strength of MDF made from loblolly pine ( Pinus taeda ) decrease with inclusion of fines in the furnish.
To determine fiber composition, there are several approaches in practice. One (Myers 1987) is to prepare numerous microscope slides using refined fibers, and count the numbers of fine fibers, fiber bundles, whole fibers and broken fibers on each slide, thereafter, calculate the percentage of each composition. This procedure is tedious and time consuming. Fiber Quality Analyzer (FQA) can be an alternative that gives a detailed profile of fiber coarseness, fiber mean length and width, and fine fiber percentage quickly and accurately. FQA is frequently used in pulp and papermaking industry or research to evaluate the pulp quality. However, this method is limited when the pulp contains many large particles because these large particles cannot pass through the analyzer unless they are screened. Via Bauer McNett Classifier, the fiber size distributions can be determined using four screens with different mesh sizes. Fiber size distributions can be known by calculating the percentages of weight of the fibers retained on the screens with mesh size of 3.240 mm2, 0.828 mm2, 0.281 mm2, 0.017 mm2, and passed through the screen with mesh size of 0.017 mm2.
A high level of refining energy spent on wood chips, flakes, or shavings can result in reduction in fiber length and production of debris (Clark 1978; Kure and Dahlqvist 1998; Kure et al. 1999). Many variables influence the fiber composition or fiber size distribution during the refining process. These variables include moisture content, size, and shape of wood chips, flakes, and shavings fed into a refiner, the steaming time and pressure, disc diameter, plate pattern, loading rate, retention time, and gap between the two grinding plates (Maloney 1993). High refining energy can also cause damage on fiber surface, while fiber bundles are likely to be produced by low refining energy level (Myers 1983; Kure and Dahlqvist 1998; Kure et al. 1999). Thus, fibers with preferable morphological properties can be generated through adjusting refining parameters. Therefore, to handle a new wood species or type, the refining parameters should be taken into account primarily in order to produce high performance fibers. 
In general, compared to mature wood, juvenile wood in conifers has thinner tracheid cell wall, shorter tracheid length, larger lumen diameter, higher wood moisture content and lower wood specific gravity (Bendtsen and Senft 1986, Yang et al. 1986, Fahey and Laundrie 1968, Wheeler et al. 1966). MFA of S2 layer of tracheid cell wall is greater in juvenile wood than it is in mature wood in conifers, resulting in lower stiffness and strength and larger shrinkage in the longitudinal direction (Haygreen and Bowyer 1982; Megraw 1985; Bendtsen and senft 1986; Meylan 1968; Cave and Walker 1994; Barber and Meylan 1964; Harris and Meylan 1965; Meylan 1968; Megraw 1985; Megraw et al. 1998). The gradual change in properties from juvenile to mature wood is presented in Figure 1-5 (Bendtsen 1978). The chemical composition of juvenile wood differs from that of mature wood. In loblolly pine, juvenile wood contains more extractives than it does in mature wood (Shupe et al. 1997). Juvenile wood is sometimes called “weak wood”. However, it is more adequate to call it “different wood” because some of the characteristics of juvenile wood are even desirable whenever juvenile wood is used as raw materials for some products such as pulp and paper (Dinwodie 1965; Maloney 1986). In most solid wood applications, juvenile wood is considered to be an undesirable feature. For example, the presence of juvenile wood causes structural lumber to be weaker; it also leads lumber to warp more at drying.  In Dahurian larch ( Larix gmelinii ), the characteristics of juvenile wood from young thinnings are even lower than those of juvenile wood from mature trees (Chow and Lu 1980). 
The characteristics of juvenile wood are different from those of mature wood. How does juvenile wood affect the properties of composite panels? In fact, it is not well known (Maloney 1986). Species may be the first important factor that should be considered when juvenile wood is used as raw material for composite panel manufacturing (Maloney 1986). In other words, juvenile wood from some species may perform better than others. Populus juvenile wood is a good raw material for both particleboard and fiberboard manufacture (Dix et al. 1999). On the other hand, the type of wood composites matters. Generally, fiberboard may be less affected by juvenile wood than oriented strand board (OSB) and particleboard (Maloney 1986). Studies by Pugel et al. (1989; 1990) show that properties of fiberboard made from loblolly pine ( Pinus taeda L.) were least affected by the differences between juvenile and mature wood furnishes compared to flakeboard and particleboard.
Tension, bending strength, and dimensional stability of fiberboards made from loblolly pine ( Pinus taeda L.) core wood (juvenile wood) were found superior to those of panels made from outer wood (mature wood) (McMillin 1968).
Pugel et al. (1989; 1990)’s studies show that MOE, MOR, and IB of the composite panels made from southern pine ( Pinus taeda L.) juvenile wood were comparable or superior to those of panels made from mature wood. In these studies, flakeboard, particleboard and fiberboard panels manufactured from four different sources of southern pine juvenile wood: 1) fast-grown trees; 2) the inner core of older trees; 3) branches and 4) tops. The durability of these panels was assessed by subjecting specimens to an ovendry-vacuum-pressure-soak (ODVPS) treatment, and then evaluated for MOE, MOR and IB. Results indicate that composite panels made from juvenile wood had properties equivalent to or better than mature wood panels. But dimensional stability of the panels made from juvenile wood, which was evaluated by measuring thickness swell (TS) and linear expansion (LE) of specimens under ODVPS condition and specimens exposed to a single cycle of 30 to 90 % relative humidity, appeared to be greater when compared to composite panels made from mature wood, particularly in LE.
Strength properties and dimensional stability of structural composite panels made with strands from Douglas fir ( Pseudotsugamenziesii var. menziesii [Mirb.] Franco) five age classes: 0-7, 8-14, 15-21, 22-28, and 29-bark were investigated by Wasniewski (1989). The finding is that panel MOR, MOE, IB, TS, and LE were reduced with increasing the age of the furnish.
Red pine ( Pinus resinosa ) thinnings as a raw material for waferboards bonded with 2.5 % powdered phenolic resin were studied by Li et al. (1991). Variables studied were red pine content (red pine/aspen ratio from 0/100 to 100/0), red pine wafer thickness and panel density. The temperature used for panel consolidation was 207 oC, time to stops 0.75 min, time at stops 6.0 min, and decompression 0.75 min. Results indicate that static bending properties were not influenced by the red pine/aspen ratio, internal bond and thickness swell increased as red pine content increased. Static bending and thickness stability were improved by increasing the content of thin wafer in the 100 percent of red pine boards. Compared with aspen that is an excellent raw material for waferboard and OSB, red pine plantation thinnings can be considered to be a comparable raw material for waferboard panels.
Shupe et al. (1999) found a minimal difference in MOR, MOE, IB, thickness swell, and water absorption between MDF panels made from loblolly pine ( Pinus taeda L.) inner wood (juvenile wood) and outer wood (mature wood). 
6-year-old Populus clones were studies as raw material for structural flakeboard panels making (Geimer and Crist 1980). These clones were 5261 coming from P. deltoides and P. balsamifera hybrid family, 5260 from P. tristis and P. balsamifera family, 5273 from P. deltoides , 5339 from P. alba and P. grandidentata , and 5351 from P. balsamifera family. Panels made with clone 5351 showed lowest bending properties, but the highest internal bond strength; water absorption and thickness swell of such panels were lower than those of panels made from the other clones.
Another study done by Peter et al. (2002) shows that OSB panels fabricated from some hybrid poplar ( Populus spp.) clones performed better than the panels made from the other clones. The clones selected for OSB manufacture were 11-11, 15-29, 49-177. 50-194, 50-197, 184-411 from a coastal P. trichocarpa and P. deltoides hybrid family, 310-85 coming from P. trichocarpa and P. nigra hybrid family, 272-102 from P. trichocarpa and P. maximowiczii hybrid family, 50-194i and 184-411i from inland P. trichocarpaand P. deltoides family, and 272-102i coming from P. trichocarpa and P. maximowiczii family. The best performing clones in MOE were the hybrids 50-194, 15-29, 50-194i, and 49-177. Clones 50-194, 15-29, 50-197, and 50-194i were among the best performers in regard to panel MOR. The hybrids 184-411i and 272-102i yielded the highest IB.

For further details log on website :
http://archimede.bibl.ulaval.ca/archimede/fichiers/23592/ch02.html

No comments:

Post a Comment