Selected properties of hybrid poplar clear wood and composite panels. (Composites and Manufactured Products).
D.A. Bender *
M.P. Wolcott *
Eleven hybrid poplar clones grown in the inland and coastal regions of the Pacific Northwest were studied. Small, clear specimens were tested to determine specific gravity, shrinkage, and bending properties. Oriented strandboard (OSB) panels were fabricated and tested to determine flexural properties, internal bond, density, water absorption, and thickness swell. Clear wood properties of the hybrid poplar clones were lower than those of native aspen and cottonwood species. Differences were observed among the clear wood mechanical and shrinkage properties for the clones, indicating potential for selective breeding to genetically engineer trees with targeted structural end uses. OSB produced from individual clones performed beyond industry specifications for mechanical properties and moisture resistance. Future research is needed to evaluate the use of hybrid poplar in composite lumber products.
Intensively managed hybrid poplar is a promising source of wood fiber given the high yields and relatively short growing cycles (6 to 10 years). Forest products companies are interested in finding new industrial uses for hybrid poplar; however, little is known about the engineering properties of these hybrids. This information is essential to drive genetic research and to assess utilization options.
The first large-scale project in which poplars were hybridized in the United States began in 1925 (24, 25). Since that time, considerable effort and resources have been expended in researching areas such as genetics, biomass yield, silviculture treatments, and disease and insect resistance, to facilitate the pulping industry's need for improved woody biomass yield. In fact, the pulping industry's interest is what has driven the evolution of these hybrids. Genetic selection for the suitability of trees for solid wood end use has received relatively little attention (29). With relatively low pulp and paper prices, the forest products industry could benefit from research identifying product opportunities for hybrid poplar. Additionally, the heritability of certain wood traits provides a unique opportunity to genetically engineer trees with mechanical and physical properties favorable to specific structural end uses.
The objectives of the study were to:
1. Characterize physical and mechanical properties of solid wood for 11 clones of hybrid poplar, including: specific gravity, longitudinal, radial, and tangential shrinkage, and flexural properties;
2. Characterize mechanical and physical properties of oriented strandboard (OSB) panels produced from the clones, including flexural properties, internal bond, panel density, water absorption, and thickness swell;
3. Critically evaluate the relative performance of the hybrid poplar clones with regard to clear wood and OSB properties;
4. Assess the ability of mixed clone, hybrid poplar panels to meet commercial OSB panel specifications.
Hybrid poplars (Populus spp.) are produced by cross breeding different species within the genus Populus. Hybrid poplars have been grown most efficiently as an agricultural crop using short rotation, intensive culture techniques. Within the Populus genus are two general groups, the cottonwoods and the aspens. The parents of the hybrids grown in the Pacific Northwest come from two cottonwood species, eastern cottonwood (P. deltoides) and black cottonwood (P. trichocarpa). Other species that have been used in creating hybrids include P. maximowiczii from Asia and P. nigra from Europe.
The trees from intensively managed, short rotation plantations contain high percentages of juvenile wood. It is well known that when compared to mature wood, juvenile wood typically exhibits lower specific gravity (SG), lower transverse shrinkage, higher longitudinal shrinkage, lower strength, higher moisture content (MC), and larger microfibril angle in the S2 cell wall (3,28). Lower specific gravity is generally considered characteristic of hybrid poplars as compared to native or slower grown trees (22). Bendtsen et al. (5) and Beaudoin et al. (2) observed lower specific gravity in the clone P. x euramericana (P. deltoides x P. nigra) than native cottonwood species. Lower average densities have also been reported in P. x euramericana compared to eastern cottonwood (P. deltoides) and trembling aspen (P. tremuloides) (13). Additionally, significant differences were found to exist between the densities of 10 P. x euramericana clones (2) and a study indicates that plant spacing and age have no effect on specif ic gravity in three different hybrid crosses (14,20).
Koubaa et al. (15) reported transverse and volumetric shrinkage values in 10 clones from the hybrid P. x euramericana. The study indicated that growth rate had little effect on radial, tangential, and volumetric shrinkage characteristics, but differences were found among the 10 hybrids. The same study indicated slightly higher transverse shrinkages among hybrid poplars than native poplars. No published research is available on longitudinal shrinkage in hybrid poplars.
Research has been conducted that indicates some hybrid poplar clones have mechanical properties similar to native cottonwood, but slightly lower that those of aspen (13). Matyas and Peszlen (19) report the average modulus of elasticity (MOE) and modulus of rupture (MOR) of 10- and 15-year old P. euramericana hybrids to increase from the pith to the bark by 30 percent. Researchers have also shown the bending properties of select hybrid poplar clones to be at or below those of eastern cottonwood (5,16).
Native aspen is regarded as an exceptional furnish source for strand-type exterior panels (12). There are several reasons for its widespread use, including low cost, low density, wide availability, and ease of flaking (5). Perhaps most important to its physical performance, is its low density. Aspen and woods of similar specific gravity are easily compressed into medium density panels, which are characterized by good flake interaction and good internal bonding (7). It is reasonable to expect similar results with hybrid poplars due to their relatively low specific gravities.
The research available on the performance of hybrid poplar in OSB panels is limited. Hybrid poplar, 4 to 6 years in age, has been found to perform well in OSB panels with regard to bending properties (11). In their study of five different Populus clones, Geimer and Crist (11) found that with alignment of face flakes, the average MOR was 47.6 MPa while the average MOE was 6,650 MPa. The internal bond (13) ranged from 448 kPa to 586 kPa. The swelling properties were typically above 30 percent. Additionally, the study indicated differences in strength properties between the different clones tested. Maloney et al. (18) reported hybrid poplar to perform above Canadian and American OSB grading standards and indicated its high potential as an OSB furnish once it becomes commercially available. Research has been conducted that indicates the properties of MOE, MOR, IB, thickness swell (TS), and linear expansion (LE), in intensely cultured poplar, are superior when compared to intensely cultured tamarack and pine and compared favorably to native aspen (10). The study also indicated TS in excess of 30 percent.
The trees were harvested from two separate plantation sites: a "coastal" site located in western Washington, and an "inland" site located in northeastern Oregon. Each grower identified the promising clones from their respective plantations. The selection criteria were that the clone must exhibit exceptional growth and maintain resistance to diseases and insects. The trees ranged from 6 to 8 years old at the time of harvesting. The clones chosen included a hybrid cross between native black cottonwood (Populus trichocarpa) (T) (western Washington) and eastern cottonwood (P. deltoides) (D) (Texas, Mississippi, Oklahoma, and Illinois), which served to capture the rapid growth characteristics of both parents. Also included in this study were the more recent crosses between P. trichocarpa and both P. nigra (N) from Europe and P maximowiczii (M) from Japan that have been likewise selected for exceptional growth. Table 1 contains a listing of the 11 hybrids tested.
The clones were harvested from the plantation as 137-cm-long bolts. Two bolts were taken from each of three trees for a total of six bolts representing each clone. The bolts were transported to Washington State University in the green moisture condition, at which point they were sawn into 4-sided cants. Three radial, tangential, and longitudinal clear wood shrinkage specimens were cut from each of the cants for a total of 18 specimens per clone for each shrinkage property. A pair of flexural clear wood specimens was cut from two locations in each cant. These matched specimens consisted of one taken from the sapwood and one from the heartwood, near the pith, for a total of 24 flexural specimens per clone. These specimens were cut slightly oversized before being conditioned in a chamber maintained at 22[degrees]C and 65 percent relative humidity to an equilibrium moisture content (EMC) of 12 percent based on ovendry weight, at which point they were cut to their final test dimensions.
CLEAR WOOD TESTING
The flexural specimens were tested for failure in center-point loading, and the MOE and MOR were determined. Initial length and weight measurements were made on each of the green shrinkage specimens. They were then placed in a conditioning room until constant mass was obtained. Upon removal from the conditioning room, the shrinkage specimens were once again weighed, measured, and placed in an oven until a constant mass was obtained. Final weight and length measurements were made and recorded. The longitudinal shrinkage tests were carried out on specimens of nominal 2.5- by 2.5- by 10-cm dimensions, which is consistent with previous studies (27). Ovendry specific gravity was determined from a 2.5- by 2.5- by 5-cm section cut from each of the flexural specimens. All other clear wood tests were carried out in accordance with ASTM D143-94, Standard Methods of Testing Small Clear Specimens of Timber (1). All dimensional measurements needed for obtaining structural properties were made using digital calipers accur ate to [+ or -] 0.01 mm. Weight measurements were made using a digital balance accurate to [+ or -] 0.01 grams. Specimen conditioning for the "air-dry" test condition was achieved in a room at 21[degrees]C and 70 percent relative humidity. These conditions yield an EMC of approximately 12 percent. All ovendrying was done at 103 [+ or -] 2[degrees]C until a constant mass was obtained.
ORIENTED STRANDBOARD MANUFACTURE
The raw material for OSB specimens was sprinkled to maintain the green moisture condition. This material consisted mainly of sections of the original cants with relatively large dimensions but also included some of the larger trimmings from the solid test specimens. The effort to use only the largest material resulted in more uniform strand geometry, which is important in maintaining consistent board quality (17). Strands were prepared from all of the clones except for the coastal TM hybrid 272-102, which was excluded for lack of material.
The strands were produced using a PRZ Hombak drum flaker. Flake dimensions were a nominal 64 mm long by 6 to 38 mm wide and 0.5 mm thick. They were cut so that the grain of the wood ran along the 64-mm length. The strands were dried to an MC of 3 percent (ovendry), at which point they were stored in polyethylene bags until they could be screened. The strands were screened using a Black Clawson rotary shaker table to remove overs and fines. The shaker table was equipped with two screens: a top screen with 4.5-cm-square openings and a bottom No. 4 screen (1.6 divisions/cm). Material retained on the bottom screen was returned to the sealed polyethylene bags for use in the OSB. Material passing the bottom screen accounted for a relatively small portion of total material weight, ranging between 10 and 14 percent.
A long retention, 1.2-in-diameter rotary drum blender was used for applying 5 percent by weight, powdered PF resin. When specified, wax was added at either 1 or 2 percent by solids weight. Sufficient furnish to manufacture two panels was blended for 5 minutes. The mats were hand-formed using a laboratory mechanical former. The former consisted of an open 56- by 56-cm box fitted with 19 baffles on 2.5-cm spacing. The former was suspended above a steel caul plate that oscillated over a range of 4 cm in the direction perpendicular to the baffles. A former-to-mat distance of 5 cm was maintained constant during forming. The panel was composed of two face layers and one core layer oriented at 0-90-0 degrees to the panel direction. Each layer comprised one-third of the total panel weight.
Panels were pressed to a target density of 640 kg/[m.sup.3] at 1.27 cm target panel thickness. A computer-controlled Siempelkamp press was set at a target platen temperature of 193[degrees]C. A position controlled press cycle of 630 seconds was used, which included 30-second closing and degassing periods. Once cool, the panels were trimmed to 50 by 50 cm and cut into test specimens using a standard pattern. Additional experimental details are given in a thesis by Peters (23).
ORIENTED STRANDBOARD TESTING
All testing of the OSB was carried out in accordance with ASTM D 1037-93 Standard Test Methods for Evaluating Properties of Wood-Base Fiber and Particle Panel Materials (1). Dimensions were measured using digital calipers accurate to 0.01 mm. Weight measurements were made using a balance accurate to 0.01 grams except for determination of panel density, in which case, a balance accurate to 0.00 1 grams was used. Ovendrying was done at 103[degrees]C until a constant mass was obtained.
The flexural properties for the OSB were determined on specimens of nominal 76- by 356- by 13-mm dimensions. Two specimens were taken from each panel with their long dimension parallel to the outer layer orientations and two with their long dimension perpendicular to the outer layer orientations for a total of six specimens per clone for each orientation direction. Cross sectional dimensions were measured using the digital calipers and each specimen was weighed on the electronic balance. The specimens were loaded at their center point on a span of 30 cm in order to maintain a minimum span of 24 times the nominal thickness. The load was applied throughout the test at a rate of 0.1 mm/s. Upon specimen failure, the load deflection curve was recorded and a 76-by 13- by 25-mm section was cut from near the failure in order to determine the MC of each specimen as well as panel density. The specific gravity of each flexural specimen was determined by using the determined MC to calculate the ovendry weight of the ent ire flexural specimen, and dividing it by the volume of the specimen at the time of testing.
Four internal bond specimens of a nominal 50 by 50 by 13 mm were cut from each panel for a total of 12 specimens per clone. The dimensions of each specimen were measured and then recorded. A tensile load was applied at a rate of 1 mm/min. and the stress was determined using the load at which the specimen failed.
Two nail head pull-through specimens of nominal 76 by 152 by 13 mm were cut from each panel for a total of six specimens per clone. Two 6d common wire nails were driven through each specimen at right angles until their heads were flush with the surface of the board. The pointed end of each nail was fitted with a tension grip and a uniform load was applied at a rate of 1.5 mm/min. The maximum load required to pull the head of the nail through the board was determined.
A nominal 152- by 152- by 13-mm. specimen was cut from the center of each panel for use in determining the water absorption and TS after 24-hour submersion. The specimens were immediately weighed and their length and width were measured. The average thickness was determined by taking several measurements at specific locations on the specimen. The specimens were then submerged in a horizontal position under 2.5 cm of distilled water maintained at room temperature (20[degrees]C). After 24 hours of submersion, the specimens were drip dried for 10 minutes, wiped of any surface water, weighed, and the thickness determined as before.
DATA ANALYSIS AND STATISTICAL METHODS
Data analysis was conducted using the statistical software Statistical Analysis System (SAS). When the data set was balanced, analysis of variance (ANOVA) procedures were employed to compare the mean properties for the different clones. In certain instances, shrinkage specimens were difficult to measure because of severe checking or warping. These specimens were not included, resulting in unbalanced data sets between the clones. When data sets were not balanced because of removed data points, the more robust general linear model (GLM) procedure was used to compare the property means.
In the case of the OSB comparisons, it was deemed appropriate to normalize the different properties with respect to panel density. This is because processing parameters can have a profound effect on mechanical properties of interest. Thus, the clonal variation effects, which were the object of study, could be masked. The analysis of covariance procedure in SAS was used for this purpose.
RESULTS AND DISCUSSION
The results of the ANOVA procedure indicated the clonal differences to be significant at an [alpha]-level of 0.01 for all of the solid wood properties measured. Tables 2 and 3 present the average and coefficient of variation (COV) for the specific gravity, shrinkage properties, and sapwood and heartwood MOE and MOR. Individual property comparisons were conducted using Duncan's multiple range tests at an [alpha]-level of 0.05, which enabled comparisons and subsequent conclusions about the relative performance of the different clones with respect to the different properties. These results are contained in Table 4.
The average values of MOE and MOR for all of the clones combined are 5,680 MPa and 52.3 MPa, respectively. The ranges of the average flexural properties for the individual clones are from 4,540 MPa to 6,310 MPa for MOE, and from 44.5 MPa to 57.2 MPa for MOR. These results are higher than the average values of 5,380 MPa and 33.6 MPa reported by Bendtsen et al. (5) for P. deltoides x P. nigra. The average MOE is considerably less than the values of 8,140 MPa and 9,860 MPa given in the Wood Handbook (8) for quaking and bigtooth aspen, respectively. The average is also lower than those given for the cottonwoods: balsam poplar, black, and eastern, which have MOEs of 7,580 MPa, 8,760 MPa, and 9,450 MPa, respectively. The average hybrid MOR was lower, albeit to a lesser extent, than quaking and bigtooth aspen, which are 57.9 MPa and 62.7 MPa, respectively. The MORs were closer to those published for the cottonwoods: 46.9 MPa, 58.6 MPa, and 58.6 MPa for balsam poplar, black, and eastern cottonwoods, respectively. In short, the average flexural properties generally tested lower than those of any commercially available species listed in the Wood Handbook (8).
The average values of heartwood MOE and MOR for all the clones combined are 4,930 MPa and 48.8 MPa, respectively. The ranges of the average flexural properties for the individual clones are from 4,080 MPa to 5,970 MPa for MOE, and from 40.1 MPa to 59.4 MPa for MOR. Individually paired observations of heartwood and sapwood showed that MOE and MOR were about 12.2 and 5.7 percent lower, respectively, in the heartwood. This result is consistent with those published by several researchers, indicating increases in mechanical properties from pith to bark (4,19,26).
Because of an oven malfunction, a portion of the ovendry information was lost. It was necessary to use ASTM. D 2395-93 Standard Test Methods for Specific Gravity of Wood and Wood-Base Materials (1) to extrapolate the ovendry data from the ambient data. The average sapwood specific gravity, based on ovendry volume for all of the clones combined, was 0.322. The range was from 0.284 to 0.361. The specific gravity numbers are lower than the average ovendry specific gravity of three clones of 0.45 published by Murphey et al. (20). The results are also lower than those published by Beaudoin et al. (2) who found the average specific gravity on a green volume basis of 10 clones of P. deltoides x P. nigra to be 0.350 and the range to be from 0.284 to 0.407. The average specific gravity is lower than the values given by the Wood Handbook for quaking and bigtooth aspen: 0.38 and 0.39, respectively. It is also lower than the values given by the Wood Handbook (8) for the cottonwoods: balsam poplar, black and eastern, whi ch are 0.34, 0.35, and 0.40, respectively.
The longitudinal shrinkage ranged from 0.132 to 0.521 percent with COVs as high as 0.683. Radial and tangential shrinkage ranged from 3.17 to 4.30 percent and from 6.64 to 9.11 percent, respectively. The average values for longitudinal, radial, and tangential shrinkage were 0.257, 3.89, and 7.49 percent, respectively. Limited information is available with which to compare longitudinal shrinkage. Foulger's (9) work indicated longitudinal shrinkage in eastern white pine of 0.25 percent, which is comparable to that observed in this study. More important is the relative comparison of longitudinal shrinkage between the clones, which will be discussed later in this section. Average radial shrinkage is comparable to eastern cottonwood, which is 3.9 percent, but slightly higher than bigtooth and quaking aspen, which have radial shrinkage values of 3.3 and 3.5 percent, respectively. The balsam poplar cottonwood and black cottonwood have slightly lower radial shrinkage of 3.0 and 3.6 percent, respectively. The average tangential shrinkage of the hybrids is lower than the bigtooth aspen value of 7.9 percent, and the black and eastern cottonwood values of 8.6 and 9.2 percent, respectively. The average tangential shrinkage is slightly higher than the quaking aspen value of 6.7 percent and the balsam poplar cottonwood tangential shrinkage value of 7.1 percent. Flexural and shrinkage properties seemed to be inversely related, particularly with regard to longitudinal shrinkage, which is most likely due to their strong connection through the microfibril angle. Generally, the higher the microfibril angle, the higher the longitudinal shrinkage and the lower the flexural properties.
Table 4 presents results of Duncan's multiple range tests for sapwood specific gravity, and shrinkage and flexural properties. The higher strength clones include 15-29, 50-194, 50- 194i, 49-177, 310-85, and 11-11. As indicated, the MORs of clones 50-197, 272-102, 184-411, and 184-411i are statistically lower than all but 272-102i, which represents the transition between the lower and higher strength clones. The best performing clones with regard to MOE were the TD hybrids 50-194i, 15-29, 50-194, and 49-177. The results of the Duncan's grouping indicate clone 184-411 has a statistically higher specific gravity than the remaining clones. The clones 11-11 and 50-197 have statistically lower specific gravities than the remaining clones. Little difference was found between the remaining clones, which fell between the high and low Duncan groups. For this grouping, air-dry weight was divided by air-dry volume, which was done to enable calculation of specific gravity using only measured data and no data extrapolated after the oven malfunction. With regard to shrinkage, once again the TD hybrids 50-194i and 15-29 are included among the best performers possessing the lowest shrinkage values. Additional standouts included the TM hybrid 272-102 in the tangential direction, and the TD hybrids 50-197 and 184-411 in the radial direction. The hybrids generally performed better with regard to dimensional stability than they did in bending, as evidenced by the native property comparisons.
Although statistical differences were found between identical clones from different growing regimes, for practical purposes, the effect of the site and treatment was minimal. This is particularly true of the property MOE, in which no statistical differences were found between identical clones harvested from different sites. The TD hybrids consistently outperformed the other clones tested. Furthermore, clones 15-29, 50194, and 49-177 constituted the top performers in both flexural properties.
PROPERTIES OF ORIENTED STRANDBOARD PRODUCED FROM INDIVIDUAL CLONES
Panels were manufactured from strands of individual clones to assess the potential differences that may exist among the different hybrid poplars. These panels were produced without wax to assess the potential for swelling within the specific clone. Mechanical properties were assessed relative to industrial standards to determine commercial viability of using these hybrid poplars in production.
The average and COV for the density and flexural properties of the OSB from the different clones are presented in Table 5. The average flexural properties for the different clones are presented as bar charts in Figures 1 and 2. The values in Table 5 and Figures 1 and 2 are presented as measured, and have not been adjusted for panel density. The average value of MOE parallel to the orientation is 7,170 MPa and the average value perpendicular is 2,960 MPa. The average value of MOR parallel to the orientation is 49.4 MPa and the average value perpendicular is 25.0 MPa.
Table 6 contains the average and COV for the internal bond, fastener, and swelling properties. Figure 3 is a bar chart comparing the internal bond performance of the different clones. The values in Table 6 and Figure 3 are presented as measured and have not been adjusted for panel density. For all of the individual clones assessed, the panel mechanical properties were significantly greater than specifications set by the CSA O-2 grade standard.
Panel density is influenced by furnish and manufacturing parameters (e.g., flake geometry, dimensions, alignment, and pressing variables). Panel density, in turn, has been shown to have considerable effect on final panel properties (21). To assess OSB property comparisons among the individual clones, the confounding effect of panel density was isolated and removed using an analysis of covariance.
An analysis of covariance fits regression lines between the dependent property (e.g., MOE) and the covariate (panel density) and then tests the hypothesis that the regression intercepts are equal. Subsequent adjustments are then made on the property means to bring them to a common density. The validity of the covariance analysis was examined by checking the heterogeneity of slopes for each of the properties. The heterogeneity of slopes analysis indicated the regression slopes were equal for all mechanical properties. It indicated that no significant differences existed between the perpendicular MORs. It also indicated a simple ANOVA could be used for comparing the nail head pull-through, as density had little effect on this particular property. The subsequent analysis of covariance indicated significant differences between the clones for all the mechanical properties measured with the exception of perpendicular MOR. No significant differences were found between the physical properties of water absorption and TS.
Individual property comparisons were conducted using Duncan's multiple range tests at an [alpha]-level of 0.05. Once again, this enabled comparisons and subsequent conclusions about the relative performance of the different clones in an OSB panel. The Duncan's grouping for the OSB flexural properties are contained in Table 7. The average properties in Table 7 have been adjusted to the panel density of hybrid 272-102i. The results were similar to those of the solid wood flexural properties. Among the best performing clones in MOE were the TD hybrids 50-194, 15-29, 50-194i, and 49-177. The same clones were the top four performers with regard to solid wood MOE. Clones 50-194, 15-29, 50197, and 50-194i were among the best performers with regard to panel MOR. Similar results were also observed in the solid wood MORs. Among the poorest performers for both flexural properties were the TM hybrid 272-102i, the TD hybrid 134-411, and the TN hybrid 310-85. Similar results had previously been observed in the solid wood.
The Duncan's groupings for IB and nail head pull-through are presented in Table 7. Groupings for the swelling properties have been excluded because as stated earlier, statistical differences were not found among them. The best performers with regard to IB were the hybrids 184-411i and 272-102i. The best performers in IB were generally the lowest in flexural properties. This result was expected because the less dense, more flexible flakes often provide better mat consolidation. This consolidation serves to enhance resin efficiency and bond formation (21). The differences among clones with regard to nail head pull-through were very subtle, as evidenced by Table 7. The best performers were the hybrids 50-197, 11-11, and 15-29, although they were not statistically superior. The poorest performers were the hybrids 310-85 and 184-411i, with the latter being lower statistically than every other clone with the exception of the former.
ORIENTED STRANDBOARD PRODUCED FROM MIXED CLONES
Any commercial producer of OSB utilizing hybrid poplars is unlikely to obtain material strictly from a single clone. Furnish was produced by mixing equal proportions of the various clones to assess the viability of utilizing hybrid poplars as a raw material source in a commercial operation. As with single clone panels, these materials were produced with 5 percent powdered PF resin; however either 1 or 2 percent wax was added to mimic commercial formulations. The average mechanical and physical properties of these panels are presented in Table 8.
At 1 and 2 percent by solids wax levels, the TS averaged 15.1 and 9.8 percent, respectively. These values are compared to the 15 percent standard set for CSA O-2 grade panels. As in the single clone panels, the mechanical properties substantially exceed the GSA standard. In fact, the properties of MOE, MOR, and IB increase with the addition of wax, presumably from the improved resin distribution from the added moisture of the liquid wax.
SUMMARY AND CONCLUSIONS
The clear wood properties of the hybrid poplar clones were lower than those published for native aspen as well as other commercially available species. This result was true of all of the clones tested. There were statistically significant differences among the clones for the mechanical and shrinkage properties. The best performers with regard to clear wood flexural properties were the TD hybrids: 50-194i, 50-194, 15-29, and 49-177. The most dimensionally stable clones once again included TD hybrids 50-194i and 15-29, this time, accompanied in the tangential direction by TM hybrid 272-102, and in the radial direction by TD hybrids 50-197 and 184-411. The differences among the clones would seem to indicate the unique possibility of selective breeding to genetically engineer trees with structural end uses in mind.
OSB produced from individual clones performed beyond the specifications set for GSA 0-2 panels for all mechanical properties. In addition, mixed clone panels produced with 1 and 2 percent wax exceeded both mechanical and moisture standards set by the GSA O-2 specifications. Combined, these results indicate that commercial use of these clones for OSB production is viable.
Among the best performing clones in MOE were the TD hybrids 50-194, 50194i, 15-29, and 49-177. Clones 50-194, 50-194i, 15-29, and 50-197 were among the best performers with respect to MOR. These results agree very closely with the results of the clear wood tests. The best performers with regard to IB were the hybrids 184-41li and 272-102. Generally, the best performers in the flexural test were lowest in IB.
The clear wood properties of the hybrid poplar were lower than reported values for native aspen and cottonwood species. Many utilization options are available for low-strength woods, including low grade lumber, pallets, crates, millwork, formwork, and paneling. The hybrids showed great promise for use in structural panel products based on superior flexural and IB properties. The TS was comparable to native aspen panels manufactured without wax. Upon addition of wax to the blending process, the panels exhibited TS properties that met both the Canadian and American grading agency standards. This study indicates potential for substituting hybrid poplar for native aspen in OSB sheathing products. Future research is needed to evaluate the use of hybrid poplar in composite lumber products.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
TABLE 1 Hybrid poplar clones studied. Plantation Parentage Clone # Coastal P. trichocarpa x deltoides (TD) 11-11 15-29 49-177 50-194 50-197 184-411 P. trichocarpa x nigra (TN) 310-85 P. trichocarpa x maximowiczii (TM) 272-102 Inland P. trichocarpa x deltoides (TD) 50-194i 184-411i P. trichocarpa x maximowiczii (TM) 272-102i TABLE 2. Clear wood specific gravity and bending properties. Spawood (n = 12) Hybrid Clone MOR MOE Air-dry SG (MPa) Coastal TD 11-11 Avg. 54.0 6,010 0.282 COV 0.077 0.057 0.043 15-29 Avg. 57.1 6,180 0.321 COV 0.054 0.063 0.055 49-177 Avg. 55.7 6,130 0.322 COV 0.045 0.058 0.086 50-194 Avg. 57.0 6,170 0.312 COV 0.06 0.079 0.040 50-197 Avg. 49.9 5,620 0.277 COV 0.075 0.116 0.086 184-411 Avg. 44.7 5,100 0.349 COV 0.156 0.166 0.074 Coastal TN 310-85 Avg. 54.8 5,790 0.306 COV 0.119 0.14 0.031 Coastal TM 272-102 Avg. 48.8 5,060 0.319 COV 0.056 0.1 0.043 50-194i Avg. 56.3 6,310 0.314 COV 0.051 0.079 0.033 Inland TD 184-411i Avg. 44.5 4,540 0.311 COV 0.063 0.141 0.062 Inland TM 272-102i Avg. 52.3 5,590 0.324 COV 0.155 0.147 0.049 Spawood (n = Heartwood (n = 12) 12) Hybrid Ovendry SG MOR MOE Air-dry SG Ovendry SG (MPa) Coastal TD 0.289 51.0 5,100 0.323 0.334 0.044 0.068 0.111 0.051 0.053 0.331 48.9 4,990 0.313 0.322 0.056 0.234 0.174 0.081 0.084 0.332 45.7 4,380 0.320 0.331 0.089 0.124 0.173 0.065 0.067 0.321 54.4 5,630 0.326 0.337 0.041 0.143 0.147 0.063 0.065 0.284 40.1 4,080 0.304 0.313 0.089 0.206 0.171 0.061 0.063 0.361 41.6 4,670 0.277 0.285 0.077 0.118 0,136 0.058 0.060 Coastal TN 0.315 46.1 4,760 0.338 0.349 0.032 0.156 0.137 0.062 0.064 Coastal TM 0.329 47.6 4,660 0.327 0.337 0.042 0.127 0.122 0.101 0.104 0.323 59.4 5,970 0.341 0.352 0.034 0.077 0.096 0.044 0.046 Inland TD 0.320 46.9 4,440 0.320 0.330 0.064 0.112 0.173 0.044 0.045 Inland TM 0.334 55.2 5,580 0.344 0.356 0.050 0.137 0.11 0.060 0.062 TABLE 3. Clear wood shrinkage properties. Longitudinal (n = 18) Hybrid Clone Air-dry Ovendry Coastal TD 11-11 Avg. 0.125 0.270 COV 0.674 0.458 15-29 Avg. 0.073 0.173 COV 0.646 0.359 49-177 Avg. 0.105 0.246 COV 0.452 0.304 50-194 Avg. 0.055 0.131 COV 0.908 0.534 50-197 Avg. 0.066 0.188 COV 1.304 0.683 184-411 Avg. 0.237 0.521 COV 0.828 0.653 Coastal TN 310-85 Avg. 0.141 0.189 COV 1.389 0.531 Coastal TM 272-102 Avg. 0.199 0.330 COV 0.709 0.649 50-194i Avg. 0.053 0.169 COV 0.932 0.366 Inland TD 184-411i Avg. 0.122 0.277 COV 0.810 0.607 Inland TM 272-102i Avg. 0.141 0.287 COV 0.572 0.420 Radial (n= 18) Tangential (n = 18) Hybrid Air-day Ovendry Air-day Ovendry Coastal TD 1.84 3.67 4.66 7.28 0.331 0.116 0.110 0.063 1.92 3.51 4.25 6.88 0.125 0.102 0.107 0.097 2.25 4.30 5.14 8.21 0.248 0.176 0.199 0.163 1.84 3.77 4.91 8.01 0.156 0.150 0.093 0.066 1.97 3.35 4.57 7.07 0.257 0.183 0.070 0.061 1.41 3.17 4.71 7.37 0.221 0.177 0.095 0.066 Coastal TN 2.38 3.87 6.24 9.11 0.142 0.092 0.073 0.059 Coastal TM 2.11 3.72 3.98 6.64 0.128 0.113 0.075 0.077 1.90 3.38 4.19 6.90 0.168 0.189 0.112 0.106 Inland TD 2.27 3.76 5.08 7.55 0.134 0.099 0.147 0.106 Inland TM 2.10 3.75 4.74 7.37 0.108 0.089 0.120 0.095 TABLE 4. Duncan's grouping for clear wood properties. Specific gravity Clone 184-411 272-102i 49-177 15-29 0.386 0.358 0.357 0.355 Longitudinal Shrinkage (%) Clone 184-411 272-102 272-102i 184-411i 0.521 0.330 0.287 0.277 Radial Shrinkage (%) Clone 49-177 310-85 50-194 184-411i 4.30 3.87 3.77 3.76 Tangential Shrinkage (%) Clone 310-85 49-177 50-194 184-411i 9.11 8.21 8.01 7.55 MOE(MPa) Clone 50-194i 15-29 50-194 49-177 6,310 6,180 6,170 6,130 MOR(MPa) Clone 15-29 50-194 50-194i 49-177 57.2 57.0 56.3 55.7 Clone 272-102 50-194i 50-194 184-411i 0.353 0.347 0.345 0.344 Longitudinal Shrinkage (%) Clone 11-11 49-177 310-85 50-197 0.270 0.246 0.232 0.188 Radial Shrinkage (%) Clone 272-102 270-102 11-11 15-29 3.75 3.72 3.67 3.51 Tangential Shrinkage (%) Clone 184-411 272-102i 11-11 50-197 7.37 7.37 7.28 7.07 MOE(MPa) Clone 11-11 310-85 50-194 272-102i 6,010 5,790 5,620 5,580 MOR(MPa) Clone 310-85 11-11 272-102i 50-197 54.7 54.1 52.3 49.8 Clone 310-85 11-11 50-197 0.339 0.313 0.306 Longitudinal Shrinkage (%) Clone 15-29 50-194i 50-194 0.173 0.169 0.132 Radial Shrinkage (%) Clone 50-194i 50-197 184-411 3.38 3.35 3.117 Tangential Shrinkage (%) Clone 50-194i 15-29 270-102 6.90 6.88 6.64 MOE(MPa) Clone 184-411 272-102 184-411i 5,100 5,060 4,540 MOR(MPa) Clone 272-102 184-411 184-411i 48.7 44.7 44.5 TABLE 5. Unadjusted OSB density and flexural properties. Parallel (n=6) Hybrid Clone MOR MOE Density (MPa) (kg/[m.sup.3]) Coastal TD 11-11 Avg. 48.4 7,320 646 COV 0.116 0.092 0.056 15-29 Avg. 58.9 8,200 647 COV 0.165 0.091 0.038 49-177 Avg. 50.2 7,520 634 COV 0.225 0.129 0.046 50-194 Avg. 55.8 7,780 618 COV 0.094 0.064 0.038 50-197 Avg. 49.4 6,670 612 COV 0.088 0.031 0.026 184-411 Avg. 39.2 5,810 607 COV 0.344 0.399 0.056 Coastal TN 310-85 Avg. 42.0 6,280 618 COV 0.124 0.082 0.036 Inland TD 50-194i Avg. 53.5 8,170 654 COV 0.123 0.099 0.06 184-411i Avg. 50.7 7,100 641 COV 0.146 0.042 0.024 Inland TM 272-102i Avg. 45.6 6,560 657 COV 0.207 0.12 0.061 Perpendicular(n=6) Hybrid MOR MOE Density (MPa) (kg/[m.sup.3]) Coastal TD 24.9 2,610 670 0.129 0.3 0.064 24.6 2,850 641 0.061 0.107 0.061 26.0 3,130 647 0.232 0.154 0,082 24.8 2,900 618 0.102 0.072 0.045 23.0 3,040 610 0.295 0.168 0.072 25.6 3,350 622 0.145 0.049 0.066 Coastal TN 21.9 2,610 666 0.182 0.179 0.026 Inland TD 28.2 3,330 673 0.121 0.117 0.028 25.0 2,730 655 0.114 0.068 0.041 Inland TM 25.6 3,090 622 0.308 0.17 0.063 TABLE 6. Unadjusted OSB properties: IB, fastener, thickness swell, and water absorption. Thickenss swell (n = 3) Internal bond Nail pull Hybrid Clone (n=12) (n=6) 2-hr. (kPa) (N) (%) Coastal TD 11-11 Avg. 593 2,010 11.3 COV 0.076 0.076 0.197 15-29 Avg. 600 1,970 11.7 COV 0.118 0.118 0.205 49-177 Avg. 621 1,930 9.4 COV 0.124 0.124 0.127 50-194 Avg. 579 1,920 9.6 COV 0.169 0.169 0.213 50-197 Avg. 600 2,110 10.2 COV 0.178 0.178 0.231 184-411 Avg. 572 1,730 11.9 COV 0.219 0.219 0.455 Coastal TN 310-85 Avg. 607 1,870 12.1 COV 0.257 0.257 0.396 Inland TD 50-194i Avg. 621 1,870 9.8 COV 0.085 0.085 0.113 184-411i Avg. 710 1,560 7.9 COV 0.131 0.131 0.293 Inland TM 272-102i Avg. 696 1,870 9.6 COV 0.237 0.237 0.285 Thickenss swell Water absorption as a (n = 3) % of cond. wt. (n=3) Hybrid 24-hr. 2-hr. 24-hr. (%) Coastal TD 27.1 29.3 63.9 0.108 0.352 0.245 25.5 31.2 70.8 0.004 0.244 0.195 26.8 25.0 62.1 0.046 0.099 0.117 26.8 27.9 68.6 0.089 0.175 0.138 25.5 27.2 72.2 0.026 0.194 0.116 26.6 31.9 68.4 0.257 0.329 0.177 Coastal TN 29.4 31.5 75.6 0.236 0.416 0.339 Inland TD 27.3 23.0 59.4 0.077 0.128 0.121 22.7 23.4 58.4 0.197 0.294 0.238 Inland TM 26.0 26.6 65.3 0.329 0.304 0.336 TABLE 7. Duncan's grouping for OSB properties. MOE - parallel (MPa) Clone 50-194 15-29 50-194i 49-177 8,000 8,000 7,860 7,520 MOR - parallel (MPa) Clone 50-194 15-29 50-197 50-194i 57.5 57.4 51.8 51.4 MOE - perpendicular (MPa) Clone 50-197 310-85 50-194i 49-177 3,290 3,170 3,140 3,120 Internal bond (kPa) Clone 184-411i 272-102i 310-85 49-177 710 696 641 621 Nail head pull-through (N) Clone 50-197 11-11 15-29 49-177 2,110 2,010 1,970 1,930 MOE - parallel (MPa) Clone 11-11 50-197 184-411i 310-85 7,170 6,960 6,960 6,960 MOR - parallel (MPa) Clone 49-177 184-411i 11-11 310-85 50.1 49.8 47.0 46.4 MOE - perpendicular (MPa) Clone 50-194 272-102i 15-29 184-411 3,090 2,980 2,890 2,780 Internal bond (kPa) Clone 50-194i 184-411 15-29 50-197 621 607 600 600 Nail head pull-through (N) Clone 50-194 50-194i 272-102i 184-411 1,920 1,870 1,870 1,870 MOE - parallel (MPa) Clone 184-411 272-102i 6,690 6,210 MOR - parallel (MPa) Clone 184-411 272-102 45.1 43.1 MOE - perpendicular (MPa) Clone 184-411i 11-11 2,670 2,440 Internal bond (kPa) Clone 11-11 50-194 593 579 Nail head pull-through (N) Clone 310-85 184-411i 1,720 1,560 TABLE 8. Results of follow-up study to control thickness swell. Parallel (n = 6) Perpendicular (n = 6) MOE MOR MOE (MPa) No wax 6,620 36.5 2,240 1% wax 6,920 43.8 2,520 2% wax 6,750 46.2 2,900 Perpendicu 24 hr. water soak (n = 3) lar (n = 6) MOR IB (n = 12) SG (n = 6) Water abs. (MPa) (kPa) (%) No wax 14.4 200 0.66 115.0 1% wax 18.8 324 0.64 36.7 2% wax 22.6 476 0.62 22.9 24 hr. water soak (n = 3) TS (%) No wax 56.5 1% wax 15.1 2% wax 9.8
This paper was received for publication in January 2001. Reprint No. 9258.
(1.) American Society for Testing and Materials. 2000. Annual Book of ASTM
Standards, Section 4: Construction, Volume 04.10: Wood. ASTM D-143-94, ASTM D-1037-93. ASTM, Philadelphia, PA.
(2.) Beaudoin, M., R.E, Hernandez, A. Koubaa, and J. Poliquin. 1992. Interclonal, intraclonal and within-tree variation in wood density of poplar hybrid clones. Wood and Fiber Sci. 24(2):147-153.
(3.) Bendtsen, B.A. 1978. Properties of wood from improved and intensively managed trees. Forest Prod. J. 28(l0):61-72.
(4.) _____ and J. Senft. 1984. Mechanical and anatomical properties in individual growth rings of plantation-grown eastern cottonwood and loblolly pine. Wood and Fiber Sci. 18(l):23-38.
(5.) ______, R.R. Maeglin, and D. Frederick. 1981. Comparison of mechanical and anatomical properties of eastern cottonwood and populus hybrid NE-237. Wood Sci. 14(1):1-14.
(6.) Canadian Standards Association. 1992. Standards on OSB and waferboard. CAN/CSA-0437-92. CSA, Rexdale, Ontario, Canada.
(7.) Dickerhoof, H.E., J.A. Younquist, and C.G. Call. 1982. U.S. wood-based panel industry: Production trends and changing markets. Forest Prod. J. 32(6):14-23.
(8.) Forest Products Laboratory. 1999. Wood Handbook: Wood as an Engineering Material. Forest Prod. Soc., Madison, WI.
(9.) Foulger, A.F. 1966. Longitudinal shrinkage pattern in eastern white pine stems. Forest Prod. J. 16(12):45-47.
(10.) Geimer, R.L. 1986. Properties of structural flakeboard manufactured from 7-year-old intensively cultured poplar, tamarack, and pine. Forest Prod. J. 36(4):42-46.
(11.) _____ and J.B. Crist. 1980. Structural flakeboard from short-rotation, intensively cultured hybrid populus clones. Forest Prod. J. 30(6):42-48.
(12.) Gunn, J.M. 1963. Wafer dimension control: number I design criteria for plant producing particleboard for building construction uses. Forest Prod. J. 13(4):163-167.
(13.) Hernandez, R.E., A. Koubaa, M. Beaudion, and Y. Fortin 1997. Selected mechanical properties of fast-growing poplar hybrid clones. Wood and Fiber Sci. 30(2):138-147.
(14.) Holt, D.H. and W.K. Murphey. 1978. Properties of hybrid poplar juvenile wood affected by silviculture treatments. Wood Sci. 10(4): 199-203.
(15.) Koubaa, A., R.E. Hernandez, and M. Beaudoin. 1998. Shrinkage of fast-growing hybrid poplar clones. Forest Prod. J. 48(4): 82-87.
(16.) Kretschmann, D.E., J.G. Isebrands, G. Stanosz, J.R. Dramm, A. Olstad, D. Cole, and J. Samsel. 1998. Structural lumber properties of hybrid poplar. Res. Paper FPL-573. USDA Forest Serv., Forest Prod. Lab., Madison, WI.
(17.) Maloney, T.M, 1993. Modem Particleboard and Dry-Process Fiberboard Manufacturing. Miller Freeman Publications, San Francisco, CA.
(18.) _____, M.T. Lentz, and G. Cambron. 1995. Oriented strandboard from Douglas fir, hemlock, spruce, cedar, and poplar. Res. Rept. 3808-2996. Wood Materials and Engineering Laboratory, Washington State Univ., Pullman, WA.
(19.) Matyas, C. and I. Peszlen. 1997. Effect of age on selected wood quality traits of poplar clones. Silvae Genetica 46(2-3):64-72.
(20.) Murphey, W.K., T.W. Bowersox, and P.R. Blankenhorn. 1979. Selected wood properties of young populus hybrids. Wood Sci. 11(4):263-267.
(21.) Olson, B.D. 1996. Developing wood composites using small diameter timber resources from dense, stagnant stands. M.S. thesis. Washington State Univ., Pullman, WA.
(22.) Paul, B.H. 1956. Specific gravity of Populus species and hybrids. Res. Rept. FPL-2060. USDA Forest Serv., Forest Prod. Lab., Madison, WI.
(23.) Peters, J.J. 1999. Selected engineering properties of hybrid poplar clear wood and composite panels. M.S. thesis. Dept. of Civil and Environ. Engineering, Washington State Univ., Pullman, WA.
(24.) Stout, A.B. and E.J. Shreiner, 1933. Results of project in hybridizing poplars. J. of Heredity 24:2 16-229.
(25.)_____ R.H. McKee, and EJ. Schreiner 1927. The breeding of forest trees for pulp wood. J. of New York Botanical Gardens 28:49-63.
(26.) Wasniewski, J.L. 1991. Evaluation of juvenile wood and its effects on Douglas-fir structural panels. M.S. thesis. Washington State Univ., Pullman, WA.
(27.) Ying, L., D.E. Kretschmann, and B.A. Bendtsen. 1994. Longitudinal shrinkage in fast-grown loblolly pine plantation wood. Forest Prod. J. 44(l):58-62.
(28.) Zobel, B.J. 1984. The Changing Quality of the World Wood Supply. Springer-Verlag. Berlin, Germany.
(29.) _____ and J.B. Jett. 1995. Genetics of Wood Pproduction. Springer-Verlag. Berlin, Germany.
The authors are, respectively, Engineer/Quality Control Director, Sentinel Structures Inc., Peshtigo, WI; Director and Professor, Wood Materials and Engineering Laboratory, Washington State University (WSU), P0 Box 641806, Pullman, WA 99164-1806; Associate Professor, WSU; and Associate Scientist, Puyallup Research and Extension Center, WSU, Pullyallup, WA. This research was supported by the USDA Wood Utilization Research Program. Materials were provided by the Potlatch Corporation and the WSU Puyallup Research & Extension Center.
* Forest Products Society Member.
[C] Forest Products Society 2002.
Forest Prod. J. 52(5):45-54.
|Printer friendly Cite/link Email Feedback|
|Author:||Peters, J.J.; Bender, D.A.; Wolcott, M.P.; Johnson, J.D.|
|Publication:||Forest Products Journal|
|Article Type:||Statistical Data Included|
|Date:||May 1, 2002|
|Previous Article:||Green marketing of softwood lumber in western North America and Nordic Europe. (Management).|
|Next Article:||A model of the effect of fines content on the strength properties of oriented strand wood composites. (Composites and Manufactured Products).|