Variation in density and shrinkage of birch (Betula pendula Roth) timber from plantations and naturally regenerated forests.
Large variations in wood density and in the behavior of sawn timber during sawmilling and further processing are typical for the birch raw material used in the wood industry. In addition to the differences in wood properties between site, geographic, and within-tree related sources, seasonal variation in the behavior and usability of sawn birch timber are also known to occur. In this study, basic density, weight density (at a moisture content of 5%), and shrinkage during drying were determined from full-size sawn timber pieces. Seasonal variation of wood density and shrinkage in sawn timber was found. Both of them were the highest in winter. Physiological changes in birch trees between seasons were presumed to be the basis for the variation. The basic density of plantation-grown birch timber was significantly lower than that of timber from naturally regenerated stands. Also, the relationship between the weight density of dried timber and the basic density of timber varied between the origins and between the seasons. These findings indicated differences in the drying behavior of birch wood.
Density is an important characteristic of wood used as raw material for wood products because it is easy to measure and it quite accurately predicts the strength and wear properties of wood (e.g., Hoadley 1995). For example, a high positive correlation exists between basic density and both bending strength and stiffness (Herajarvi 2002) and Brinell hardness of birch wood (Herajarvi 2004). Density also plays an important role in the behavior of wood during drying; the higher the density gradient within sawn timber, the stronger the shrinkage and warping (e.g., Keey et al. 2000). Drying of birch timber by artificial methods is relatively difficult. Uneven shrinkage, drying stresses, deformations and uneven discoloration are typical drying defects (Paukkonen et al. 1999, Luostarinen and Verkasalo 2000, Luostarinen et al. 2002).
The increasing supply of birch sawlogs from plantations will increase the variation of raw material properties of birch in the future. Due to the breeding of plant material and the selection of high-fertility sites for plantations, the growth rate of birch trees in plantations has been considerably higher than that of naturally regenerated birch trees, at least in the early stages of the stand development (Saksa 1998). In general, the high growth rate of plantation trees is connected with the high proportion of juvenile wood in them (cf. Zobel and Sprague 1998). Only preliminary studies on wood properties have been made for plantation-grown birch trees (Velling 1979, Verkasalo 1998); although, with many other tree species, both the physical properties and drying behavior of wood (Zhang 1995, Deresse et al. 2002), as well as the wood machining properties (Hernandez et al. 2001), are known to differ between trees from plantations and trees from naturally regenerated stands.
In addition to the variation caused by the stand and growth rate of the trees, season-dependent variation, caused by the physiological changes in trees, also occurs in wood properties of birch in northern latitudes. For example, there are clear seasonal changes in the total amount of extractives (Perila and Toivonen 1958), storage compounds (Piispanen and Saranpaa 2001), and some phenolics (Luostarinen and Mottonen 2004) of birch wood. However, the effects of the variation of extractive content on the physical and mechanical properties of wood are considered to be so small that they are usually ignored, although extractive content is known to have an effect on the wood density (Sjostrom 1981, Ona et al. 1997, Singleton et al. 2003). According to the empirical knowledge of woodworkers, many properties of birch wood (e.g., machining properties, color, and durability) are best in winter-felled wood; it is recommended that birch wood used in the wood-working industry is felled and sawn in winter (Keinanen and Tahvanainen 1995). Thus, it can be assumed that the season-dependent physiological changes in trees growing in boreal region and the freezing of wood may also change the physical properties and behavior of wood during drying.
[FIGURE 1 OMITTED]
The objective of this study was to investigate the variation of basic density of sawn birch (Betula pendula Roth) timber. Origin of timber, felling season, within-tree location of wood, and drying method were studied as sources of variation. In addition, our aim was to study the behavior of sawn timber during drying and the dependence of weight density of dried wood on basic density.
Materials and methods
A total of 108 silver birch (Betula pendula Roth) trees were felled from two naturally regenerated stands (48 trees) and from two plantations (60 trees) in summer (July), autumn (September), and winter (November/December). During each season, 36 trees were felled. Both the plantations and the naturally regenerated stands were located in North Carelia (in eastern Finland). The fellings were carried out in 1998 (naturally regenerated stands) and in 1999 (plantations). The naturally regenerated stands, which were 70 to 80 years old, were representative of a low fertile Vaccinium site and a medium fertile Myrtillus site, later abbreviated as VT and MT, respectively (Kuusipalo 1996). The plantations, both of which were 33 years old, were representative of a high fertility Oxalis-Myrtillus site, later abbreviated as OMT and an abandoned agricultural land. The 2.5-m butt logs were cut from the trees and sawn into boards with dimensions of 35 mm by 80 mm by 2500 mm. From each log, 2 to 3 boards were selected for this study and each board was cut further into 2 shorter boards 1200 mm in length. The short boards were then grouped by the radial (Fig. 1) and longitudinal locations in the log. The boards located at a length of 0 to 1.2 m in the butt log were vacuum dried, and the boards located at a length of 1.2 to 2.4 m were kiln-dried conventionally to a moisture content (MC) of 5 percent. The drying schedules were similar to those used earlier by Luostarinen et al. (2002); dry-bulb temperatures 37[degrees] to 65[degrees]C and 65[degrees] to 82[degrees]C were used for conventional and vacuum drying, respectively. The conventional drying and the vacuum drying were both carried out with small, computer-controlled laboratory kilns (Brunner Trockentechnik). The boards for conventional drying were stored overnight covered at +6[degrees]C before drying, during which period the melting of frozen winter-sawn boards began. In the vacuum drying, the melting period was included in the drying schedule.
Usually only sound wood free from knots and other defects is used for density measurements. In this study, the wood of the sample boards included knots, but the sample boards with other visible defects (e.g., decayed wood or false-heart wood) were not included. The dimensions of green and dried boards were determined using an electronic calliper by measuring the width and thickness of boards at eight longitudinal positions, two on each side, which were located at subsequent distances of approximately 30 cm from the board ends. The length of the green boards was not measured because they were sawn precisely to the length of 1200 mm with a circular cutting saw. The boards were weighed before and after drying. In addition, to determine the exact dry weight of the boards, the residual MC of the dried boards was determined gravimetrically by using small (10-g to 20-g) subsamples. The average MC of conventionally kiln-dried and vacuum-dried boards was 5.5 and 6.2 percent, respectively.
Basic density of the wood in boards was calculated as follows (Panshin and de Zeeuw 1980):
[D.sub.b] = [M.sub.0]/[V.sub.g] 
where [D.sub.b] = basic density (g [cm.sup.-3]) of the wood; [M.sub.0] = dry weight of the wood (g); [V.sub.g] = green volume of the wood ([cm.sup.3]).
Weight density of the wood of dried boards (at the target MC of 5%) was calculated as follows (Panshin and de Zeeuw 1980):
[D.sub.d] = [M.sub.d]/[V.sub.d] 
where [D.sub.d] = the weight density of dried wood (g [cm.sup.-3]); [M.sub.d] = the weight of the wood after drying (g); [V.sub.d] = the volume of the wood after drying ([cm.sup.3]). Both the basic density and weight density were transformed into kg [m.sup.-3] for the further calculations.
Shrinkage of the wood during drying (from the green state to the target MC of 5%) was determined in the tangential and radial direction on the basis of the dimension measurements of the boards. In most cases, the width of the boards was representative of the tangential direction in wood (Fig. 1). Only in a few cases, the growth rings were oriented in boards in the opposite direction. Determination of the volumetric shrinkage was based on shrinkage in the main directions. The variation in longitudinal shrinkage during drying was assumed to be small in relation to radial and tangential shrinkage. Hence, the longitudinal shrinkage of 0.6 percent (Wagenfuhr 1996) was used for the calculation of volume of dried (MC of 5%) boards.
Analysis of variance (Tukey's test) was used to compare means between the groups according to the following classification variables: origin of raw material (plantation, natural regeneration), felling season (summer, autumn, winter), within-tree location (radial, longitudinal), and drying method (conventional kiln-drying, vacuum drying). Dummy variable regression, which tests the equality of the regression coefficients between groups, was used to analyze the variation in drying behavior of wood by comparing the dependence of weight density on basic density between the groups according to the classification variables.
Density of wood
The basic density of birch timber from plantations was significantly lower (p = 0.000) than that from naturally regenerated stands (Table 1). Within a log, the basic density increased from the pith toward the surface. The radial increment of density was larger for plantation stands (8.3%) than for naturally regenerated stands (3.2%). The difference in basic density between the wood near the pith and the wood near the surface was significant (plantations: p = 0.000, naturally regenerated stands: p = 0.009). Longitudinally, within a 2.4-m butt log, the basic density of plantation-grown timber was significantly higher (p = 0.009) in the lower half (0 to 1.2 m) than in the upper half (1.2 to 2.4 m) of the log. For the timber from naturally regenerated stands, the difference between longitudinal locations was insignificant (p = 0.092).
Regarding the felling season, the basic density of timber was the lowest for autumn-felled timber and the highest for winter-felled timber. The difference was significant between winter and the two other seasons (p = 0.000 for both plantations and naturally regenerated stands) (Table 1). The basic density was 5.9 and 5.7 percent higher in winter than in autumn for plantation-grown birch trees and for naturally regenerated birch trees, respectively.
The two-tailed Pearson correlation between weight density of dried sawn timber and basic density of timber was 0.989 (p = 0.000, n = 433) for the whole material (Fig. 2a). Dummy variable regressions predicting the weight density of dried timber on the basis of the basic density, presented in Table 2, indicated a significant difference (p < 0.05) in the regression coefficient between the timber origins and felling seasons, but it showed an insignificant difference between drying methods. The weight density of timber from naturally regenerated birch trees increased more than that of timber from plantation-grown birch trees when basic density increased (Fig. 2b). Similarly, the weight density of timber from winter-felled trees increased more with the basic density than that of the timber from summer-felled or autumn-felled trees (Fig. 2c). According to the test, the relationship between the basic density and weight density of dried sawn timber was similar in both drying methods (Fig. 2d).
Shrinkage of sawn timber
The average volumetric shrinkage of all the sawn timber specimens during drying (from the green condition to the target MC of 5%) was 12.8 percent. In radial and tangential directions, shrinkage was 5.6 and 7.4 percent, respectively. The sawn timber from naturally regenerated stands shrank significantly more (p = 0.000) than that from plantation stands (Table 1). No differences were observed in the volumetric shrinkage between radial locations.
When the drying methods were compared, the shrinkage of sawn timber was larger in conventional kiln-drying (at the longitudinal location 1.2 to 2.4 m) than in vacuum drying (at the longitudinal location 0 to 1.2 m) (Table 1). Volumetric shrinkage was the greatest in sawn timber from winter-felled trees, and it differed significantly from that of summer- and autumn-felled trees for timber from the naturally regenerated stands (p = 0.000), but not for timber from plantations (p = 0.088).
Both the volumetric and tangential shrinkage of sawn timber correlated significantly with basic density (p = 0.000) and weight density of dried timber (p = 0.000) (Table 3). In addition, a negative correlation was observed between tangential and radial shrinkage (p = 0.000).
Density characteristics of wood are typically determined by the water-immersion method in which the increase in the weight of a container of water is measured when a small wood sample or disc is submerged in the water. In this study, however, the aim was to measure all the variables from specimens that are of normal size in sawn timber production. Therefore, it was most practical to determine the volumes of sample boards needed for density characteristics of specimens based on their space geometry. In addition, the same measurements could be used for analysis of shrinkage of the sawn timber. In general, the method based on space geometry is used for volume measurement, when the specimens have a regular shape (Karkkainen 2003).
[FIGURE 2 OMITTED]
In this study, the density of knots was included in density measures of the sawn timber, which was not the case with the referenced studies. Knots increase the density of sawn timber because the wood of the knot is usually higher in density than the surrounding wood (Hoadley 1995). Differences in basic density of 19 kg [m.sup.-3] (Bjorklund and Ferm 1982) and 50 kg [m.sup.-3] (Karkkainen 1976) for small-sized and mature birch, respectively, have been found between wood of branches and stem wood. However, as the proportion of knots from the volume of stem wood in hardwood species varies typically between 1.3 to 1.5 percent (Karkkainen 2003), the influence of knots on density of sawn timber is quite small.
Basic density of clear and sound silver birch wood varies typically between 460 and 800 kg [m.sup.-3] and weight density (at MC of 12% to 15%) between 510 and 830 kg [m.sup.-3] (Wagenfuhr 1996). The average basic density of timber from naturally regenerated stands of this study, 507.4 kg [m.sup.-3], was comparable with the values of 497 kg [m.sup.-3] (Hakkila 1966) and 512 kg [m.sup.-3] (Herajarvi 2004) presented earlier for mature, naturally regenerated silver birch wood. On the other hand, the average basic density of timber from plantation stands, 454.5 kg [m.sup.-3], was markedly lower than that of birch trees from naturally regenerated stands. The difference between timber origins can to some extent be explained by the age of the trees, which was earlier found to be the most important variable explaining the variation in basic density of birch wood (Hakkila 1966). Bhat (1980) showed that the cambium of birch produces denser wood the higher the age of the cambium. Low basic density of wood of young birch trees is also reported by Hakkila (1979), Velling (1979), and Verkasalo (1998).
Within the log, the basic density of timber was the highest in the lower part of the log near the surface, which is in accordance with the results of Hakkila (1966) and Herajarvi (2004). In the radial direction, the increase of basic density from pith to bark was clearly larger in plantation-grown birch trees than in naturally regenerated birch trees. This may be due to the larger proportion of juvenile wood in plantation-grown trees. Obviously, the average basic density of wood of plantation-grown birch trees will increase, as trees grow older. However, the results indicate that the high growth rate of birch trees in plantations may lead to decreased density of birch raw material, despite the fact that in diffuse porous trees, like birch, the effect of growth rate on the basic density (Zhang 1995) and the differences in wood properties between juvenile and mature wood (Zobel and Sprague 1998) are considered to be small.
The seasonal variation in basic density and in weight density of dried timber found in this study is in accordance with the earlier findings of Peterson and Winquist (1960) and Burmester (1983). Burmester (1983) presumed that the seasonal changes in these properties are a consequence of seasonal changes in reserve carbohydrates. Based on the results of this study, the seasonal variation in weight density of dried timber was even greater than the variation in basic density. It is, therefore, evident that the season-dependent physiological changes had an effect on the drying behavior of wood as well. In addition to chemical changes in wood with seasons, changes in wood-water relations caused by the subzero temperatures may also have influenced the drying behavior of wood. At subzero temperatures, water tends to move from the cell wall tissue to the ice crystallized in cell cavities due to the lower vapor pressure of ice compared to that of free water (Skaar 1988), which can be seen as changes in the dimensions of wood (coldness shrinkage) in living trees (Winget and Kozlowski 1964, Leikola 1969, Kubler 1988) and in wood blocks (Kubler 1962). The freezing of water in wood has been shown to have an effect on the drying behavior of wood when artificial freezing of wood has been tested as a pretreatment method for drying of sawn timber (Erickson 1968, Ilic 1999). The main advantages of this method were reductions in shrinkage and drying degrade, and, for some species, reduction of drying time in the early phase of the kiln-drying period. With this material, the rate of drying, especially below the fiber saturation point in the late phase of kiln-drying, was observed to be lower during winter than during other seasons (unpublished result from Mottonen and Luostarinen), which may be an indicator of changes in the drying behavior of wood with seasons.
The seasonal difference in drying behavior was seen as a larger shrinkage of wood during winter than during other seasons. This result was logical because the shrinkage of wood tends to decrease with an increased rate of drying (Stevens 1963). On the other hand, shrinkage is, in mature wood, proportional to the basic density of wood (Keey et al. 2000), which explains the difference in shrinkage in this study between naturally regenerated and plantation wood. Also, Burmester (1983) found seasonal differences in shrinkage of birch wood at different relative humidity (RH) ranges; however, the total shrinkage between seasons at the RH range of 100 to 65 percent was the same.
The difference in shrinkage between the drying methods was clear in this study. Similar results on shrinkage between the same drying methods have been obtained for sawn red oak timber (Harris and Taras 1984). The average difference of 0.6 percent units in final MC may have a role in this difference, but the main reason is apparently the difference in the drying rate and the drying mechanism between the drying methods. During vacuum drying, a steep moisture gradient occurs in the surface layer of a wood specimen (Avramidis et al. 1994) and the drying proceeds as a boiling front from the surface to the center (Chen and Lamb 2001). However, the difference in shrinkage obtained between the drying methods in this study had no significant effect on the weight density of dried timber. One reason for this can be the sampling of the boards, which resulted in the lower and upper part of the log being treated by vacuum drying and conventional drying, respectively. The basic density of the boards that were vacuum-dried was higher than that of the boards that were dried conventionally; the difference in basic density covered the difference in shrinkage.
In conclusion, the low density of plantation-grown birch wood compared with wood from naturally regenerated stands is important in considering the use of birch wood for those products where the strength and wearing properties of wood are of great importance (e.g., Mottonen et al. 2004). To maintain the adequate quality of consumer products, attention should be paid to sawing patterns to maximize the utilization of the densest wood near the log surface, particularly with the plantation-grown wood material. At present, this is not sufficiently taken into account by modern sawmills that use the double-cut method with small-sized sawlogs for its speed and the ability to automate the sawing. The seasonal variation in the density measures and drying behavior of wood found in this study is so large that it has an influence on the properties of the products made of birch wood.
Avramidis, S., M. Liu, and B.J. Neilson. 1994. Radio frequency/vacuum drying of softwoods: Drying of thick western red cedar with constant electrode voltage. Forest Prod. J. 44(1):41-47.
Bhat, K.M. 1980. Variation in structure and selected properties of Finnish birch wood. I. Interrelationships of some structural features, basic density and shrinkage. Silva Fenn. 14:384-396.
Bjorklund, T. and A. Ferm. 1982. Pienikokoisen koivun ja harmaalepan biomassa ja tekniset ominaisuudet. (Biomass and technical properties of small-sized birch and grey alder). Folia For. 500:1-37.
Burmester, A. 1983. Veranderung von Holzfeuchtigkeit, dichte und Schwindung bei Laubholzern durch jahrezeitlich bedingte Einflusse. (Seasonal changes in moisture content, density and shrinkage of wood in deciduous trees). Holz Roh-Werkstoff. 41:493-498.
Chen, Z. and F.M. Lamb. 2001. Investigation of boiling front during vacuum drying of wood. Wood and Fiber Sci. 33(4):639-647.
Deresse, T., R.K. Shepard, and R.W. Rice. 2002. Longitudinal shrinkage, kiln-drying defects, and lumber grade recovery of red pine (Pinus resinosa Ait.) from a 125-year-old natural stand and a 57-year-old plantation. Forest Prod. J. 52(5):88-93.
Erickson, R.W. 1968. Drying of prefrozen redwood--Fundamental and applied considerations. Forest Prod. J. 18(6):49-56.
Hakkila, P. 1966. Investigations on the basic density of Finnish pine, spruce and birch wood. Commun. Inst. Forestalis Fenn. 61(5):1-98.
__________. 1979. Wood density survey and dry weight tables for pine, spruce and birch stems in Finland. Commun. Inst. Forestalis Fenn. 96(3):1-59.
Harris, R.A. and M.A. Taras. 1984. Comparison of moisture content distribution, stress distribution, and shrinkage of red oak lumber dried by a radio frequency/vacuum drying process and a conventional kiln. Forest Prod. J. 34(1):44-54.
Herajarvi, H. 2002. Properties of birch (Betula pendula, B. pubescens) for sawmilling and further processing in Finland. Res. Papers 871. Finnish Forest Res. Inst., Helsinki, Finland. 52 pp.
__________. 2004. Variation of basic density and Brinell hardness within mature Finnish Betula pendula and B. pubescens stems: Wood and Fiber Sci. 36(2):216-227.
Hernandez, R.E., C. Bustos, Y. Fortin, and J. Beaulieu. 2001. Wood machining propertics of white spruce from plantation forests. Forest Prod. J. 51(6):82-88.
Hoadley, R.B. 1995. Understanding Wood. A Craftsman's Guide to Wood Technology. 10th printing. The Taunton Press Inc., Newtown, CT. 256 pp.
Ilic, J. 1999. Influence of prefreezing on shrinkage related degrade in Eucalyptus regnans F. Muell. Holz Roh- Werkstoff. 57:241-245.
Keey, R.B., T.A.G. Langrish, and J.C.F. Walker. 2000. Kiln-drying of Lumber. Springer-Verlag, Berlin Heidelberg, Germany. 326 pp.
Keinanen, E. and V. Tahvanainen. 1995. Pohjolan jalot puut. Erikoispuiden mitta-, laatu- ja kasittelyopas. (The Noble Tree Species of the North. Requirements for Dimensions, Quality, and Processing). Kuopion kasi- ja taideteollisuusakaternia (Academy of Crafts and Arts of Kuopio), Kuopio, Finland. 160 pp. (in Finnish).
Kubler, H. 1962. Schwinden und quellen des holzes durch kalte. (Shrinkage and swelling of wood by coldness). Holz Roh- Werkstoff 20:364-368.
__________. 1988. Frost cracks in stems of trees. Arboric. J. 12:163-175.
Kuusipalo, J. 1996. Suomen metsatyypit. (The Forest Site Types of Finland). Kirjayhtyma, Helsinki, Finland. 144 pp. (in Finnish).
Karkkainen, M. 1976. Puun ja kuoren tiheys ja kostcus scka kuoren o-suus koivun, kuusen ja mannyn oksissa. (Density and moisture content of wood and bark, and bark percentage in the branches of birch. Norway spruce, and Scots pine). Silva Fenn. 10(3):212-236.
__________. 2003. Puutieteen perusteet. (Basics of Wood Science). Metsalehti Kustannus, Hameenlinna, Finland. 451 pp. (in Finnish).
Leikola, M. 1969. The influence of environmental factors on the diameter growth of forest trees. Auxanometric study. Acta For. Fenn. 92:1-144.
Luostarinen, K. and V. Mottonen. 2004. Effect of growing site, sampling date, wood location in trunk and drying method on concentration of soluble proanthocyanidins in Betula pendula. Scand. J. Forest Res. 19(3):234-240.
__________ and E. Verkasalo. 2000. Birch as sawn timber and in mechanical further processing in Finland. A literature study. Silva Fenn. Monographs 1. The Finnish Society of Forest Science, Finnish Forest Research Institute, Vammala, Finland. 40 pp.
__________, V. Mottonen, A. Asikainen, and J. Luostarinen. 2002. Birch (Betula pendula) wood discolouration during drying. Effect of environmental factors and wood location in the trunk. Holzforschung 56:348-354.
Mottonen, V., H. Herajarvi, H. Koivunen, and J. Lindblad. 2004. Influence of felling season, drying method and within-tree location on the Brinell hardness and equilibrium moisture content of wood from 27 to 35-year old Betula pendula. Scand. J. Forest Res. 19:241-249.
Ona, T., T. Sonoda, K. Ito, and M. Shibata. 1997. Relationship between various extracted basic densities and wood chemical components in Eucalyptus camaldulensis. Wood Sci. and Tech. 31:205-216.
Panshin, A.J. and C. de Zeeuw. 1980. Textbook of Wood Technology. 4th ed. McGraw-Hill Book Company, New York. 722 pp.
Paukkonen, K., J. Luostarinen, J. Asp, and A. Asikainen. 1999. Koivusahatavaran kayttaytyminen kuivauksessa. Folia Forestalia. (Behavior of birch sawn timber during drying). Metsatieteen aikakauskirja 2:227-238. (in Finnish).
Perila, O and A. Toivonen. 1958. Investigations concerning the seasonal fluctuation in the composition of the diethyl-ether extract of birch (Betula verrucosa). Paperi ja Puu. 40(4a):207-213.
Peterson, O. and T. Winquist. 1960. Vikt- och fuktighetsvariationer hos bjork under olika arstider. Skogshogskolan. (Variations of weight and moisture in birch with seasons). Research notes / Royal School of Forestry, Dept. of Forest Products 28:1-20. (in Swedish)
Piispanen, R. and P. Saranpaa. 2001. Variation of non-structural carbohydrates in silver birch (Betula pendula Roth) wood. Trees 15:444-451.
Saksa, T. 1998. Rauduskoivun uudistaminen--luontaiscsti vai viljellen. (Comparing the regeneration methods of silver birch--natural regeneration or planting). In: Rauduskoivu tanaan--ja tulevaisuudessa. P. Niemisto and T. Vaara, eds. Res. Pap. 668. Finnish Forest Res. Inst. pp. 5-10. (in Finnish).
Singleton, R., D.D. DeBell, and B.L. Gartner. 2003. Effect of extraction on wood density of western hemlock (Tsuga heterophylla (Raf.) Sarg.). Wood and Fiber Sci. 35(3):363-369.
Sjostrom, E. 1981. Wood Chemistry, Fundamentals and Applications. Academic Press, Inc., Burlington, MA. 223 pp.
Skaar, C. 1988. Wood-Water Relations. Springer-Verlag, Berlin-Heidelberg-New York. 283 pp.
Stevens, W.C. 1963. The transverse shrinkage of wood. Forest Prod. J. 13(9):386-389.
Velling, P. 1979. Wood density in two Betula pendula Roth progeny trials. Folia For. 416. 24 pp. (in Finnish with English summary).
Verkasalo, E. 1998. Raudus- ja hieskoivun laatu puuaineen tiheyden perusteella arvioituna. In: Rauduskoivu tanaan--ja tulevaisuudessa. (Quality of silver birch and white birch based on the wood density). P. Niemisto and T. Vaara, eds. Res. Pap. 668. Finnish Forest Res. Inst., Res. Pap. 668. pp 127-140. (in Finnish).
Wagenfuhr, R. 1996. Holzatlas. (Wood Atlas). 4th ed. VEB Fachbuchverlag, Leipzig, Germany. 688 pp. (in German).
Winget, C.H. and T.T. Kozlowski. 1964. Winter shrinkage in stems of forest trees. J. of Forestry 62(5):335-337.
Zhang, S.Y. 1995. Effect of growth rate on wood specific gravity and selected mechanical properties in individual species from distinct wood categories. Wood Sei. and Tech. 29:(6)451-465.
Zobel, B.J. and J.R. Sprague. 1998. Juvenile Wood in Forest Trees. Springer Verlag, Berlin, Germany. 300 pp.
The authors are, respectively, Research Scientist and Senior Assistant, Univ. of Joensuu, Faculty of Forestry, Joensuu, Finland (firstname.lastname@example.org; email@example.com). This paper was received for publication in July 2004. Article No. 9903.
Table 1. -- Basic density of wood and volumetric shrinkage by the timber origin for the whole material, and by radial location, longitudinal location, and felling season for plantations and forests of natural regeneration. Basic Volumetric density shrinkage n Mean (a) SD n Mean (a) SD (kg [m.sup.3]) (%) Whole material Plantation 240 454.5 A 34.6 240 12.5 A 1.0 Natural forest 235 507.4 B 28.4 241 13.1 B 1.3 Plantations Radial location Near to the pith 82 433.9 A 27.7 82 12.6 A 1.0 In the middle 76 459.8 B 34.2 76 12.4 A 1.0 Near to the surface 82 470.0 B 31.3 82 12.6 A 1.0 Longitudinal location (b) 0 to 1.2 m 120 460.2 A 36.1 120 12.1 A 0.9 1.2 to 2.4 m 120 448.7 B 32.2 120 13.0 B 0.9 Felling season Summer 80 449.5 A 34.0 80 12.4 A 1.0 Autumn 80 443.8 A 34.9 80 12.5 A 0.9 Winter 80 470.0 B 29.3 80 12.7 A 1.1 Natural forests Radial location Near to the pith 52 497.5 A 29.2 55 12.9 A 1.3 In the middle 118 508.4 AB 27.0 121 13.2 A 1.3 Near to the surface 65 513.4 B 28.7 65 13.2 A 1.3 Longitudinal location (b) 0 to 1.2 m 116 510.5 A 28.9 119 12.8 A 1.2 1.2 to 2.4 m 119 504.3 A 27.8 122 13.4 B 1.4 Felling season Summer 80 504.8 A 26.8 80 12.8 A 1.1 Autumn 76 494.4 B 28.7 81 12.8 A 1.4 Winter 79 522.5 C 22.4 80 13.7 B 1.1 (a) Between timber origins, radial and longitudinal locations, and felling seasons in each group, means followed by the same capital letter did not differ significantly ([alpha] = 0.05). (b) The boards from 0 to 1.2 m and 1.2 to 2.4 m were vacuum dried and conventionally dried, respectively. Table 2. -- Linear regression tables for dummy-variable regression models predicting the weight density of dried timber on the basis of basic density. Linear regression model: [D.sub.d] = a + [B.sub.1][D.sub.b] + [B.sub.2][X.sub.1] + [B.sub.3][X.sub.2]. (a) Unstandardized coefficients (B) Model and standard errors (SE) (dummy variable) Variable B SE T p-value Timber origin: Plantation = 0 Natural = 1 Intercept (a) -3.436 6.247 -0.550 0.583 RMSE: 7.3102 [D.sub.b] 1.153 0.014 84.155 0.000 [r.sup.2]: 0.979 [X.sub.1] 0.073 0.023 3.197 0.001 Sig. 0.000 [X.sub.2] -33.020 11.167 -2.957 0.003 n: 433 Felling season: Summer/Autumn = 0 Winter = 1 Intercept (a) -8.438 5.024 -1.679 0.094 RMSE: 7.2270 [D.sub.b] 1.165 0.011 108.992 0.000 [r.sup.2]: 0.980 [X.sub.1] 0.076 0.019 4.021 0.000 Sig. 0.000 [X.sub.2] -34.263 9.214 -3.718 0.000 n: 433 Drying method: Conventional drying = 0 Vacuum drying = 1 Intercept (a) -19.067 5.621 -3.392 0.001 RMSE: 7.1681 [D.sub.b] 1.194 0.012 101.120 0.000 [r.sup.2]: 0.980 [X.sub.1] 0.020 0.017 1.220 0.223 Sig. 0.000 [X.sub.2] -13.771 7.984 -1.725 0.085 n: 433 (a) [D.sub.d] = weight density, kg [m.sup.-3]; [D.sub.b] = basic density, kg [m.sup.-3]; [X.sub.1] = [D.sub.b] x [X.sub.2]; [X.sub.2] = dummy variable; T = t-value for the t-test. Table 3. -- Correlation coefficients between the density and shrinkage variables for the whole material. [D.sub.b] = basic density; [D.sub.d] = weight density; [B.sub.t] = tangential shrinkage; [B.sub.r] = radial shrinkage; [B.sub.v] = volumetric shrinkage. [D.sub.b] [D.sub.d] [B.sub.t] [B.sub.r] [D.sub.d] 0.989 (**a) -- -- -- (433) [B.sub.t] 0.321** 0.403** -- -- (473) (432) [B.sub.r] -0.007 0.060 -0.256** -- (473) (432) (480) [B.sub.v] 0.256** 0.385** 0.615** 0.598** (474) (433) (480) (480) (a**) = significant at the 1 percent level.
|Printer friendly Cite/link Email Feedback|
|Author:||Mottonen, Veikko; Luostarinen, Katri|
|Publication:||Forest Products Journal|
|Date:||Jan 1, 2006|
|Previous Article:||Stress-wave velocity of wood-based panels: effect of moisture, product type, and material direction.|
|Next Article:||Effect of methylene bisthiocyanate on propugales and established mycelium of two sapstain fungi.|