MODELING THE RELATIONSHIPS BETWEEN WOOD PROPERTIES

AND QUALITY OF VENEER AND PLYWOOD OF CHINESE

PLANTATION POPLARS
Fucheng Bao
Professor
Research Institute of Wood Industry
Chinese Academy of Forestry
Wan Shou Shan
Beijing, 100091, China

and

Shengquan Liu
Associate Professor
Anhui Agricultural University
Hefei 230036, China
(Received February 2000)

ABSTRACT

Following our earlier studies on the relationships between wood properties and the quality of veneer
and plywood of Chinese plantation poplars, this study attempted to model their relationships with
stepwise and multiple linear regression. The results show that the selected three indices of wood quality
of veneer and plywood of plantations poplars grown on the shore of the Yangtse River in China,
which are the variance of veneer thickness (VVT), the ratio of lathe check in veneer (ratio of the
depth of check to the thickness of veneer) (RLC), and the glue-bond strength of plywood (GS), can
be satisfactorily predicted with the key wood properties using the individual regression equations. The
variance of veneer thickness (VVT) is a function of fiber width (FW), proportion of vessel (VP), and
fiber (FP): VVT = -3.414 + +
0.063 FW 0.049 FP - 0.033 VI? The ratio of lathe check in veneer
(RLC) is a function of wood hardness on tangential surface (TH), modulus of elasticity in bending
(MOE), air-dry wood density (WD), and total volumetric shrinkage of wood (SV): RLC = 9.472 t
0.005 TH + 0.003 MOE + 8.366 WD - 2.302 SV. The glue-bond strength of plywood (GS) is a
function of pH values (PH), fiber length (FL), fiber width (FW), vessel length (VL), and proportion
of vessel (VP) and ray (RP): GS = 3.326 - 0.268 PH - 0.002 FL + 0.191 FW - 0.004 VL - 0.033
VP + 0.095 RP The correlation coefficients (r) of the above regression equations are 0.734.93. The
determination coefficients (r2) for the regression equations are 0.54-0.88. The regression equations are
highly statistically significant at the 1% level. The differences between the model predicted and ex-
perimentally measured values are not statistically significant.
Keywords: Plantation poplar, wood properties, quality of veneer and plywood, regression equations

INTRODUCTION growing interest in combining studies on these

two aspects: wood properties and wood prod-
It is well known that wood processing and uct quality. Moreover, since wood quality is
utilization are directly affected by wood prop- directly affected by growth conditions, some
erties (Panshin and De Zeeuw 1980; Haygreen authors emphasize the importance of connect-
and Bowyer 1982). For a long time, wood sci- ing silviculture and wood quality through
entists placed emphasis on determining wood modeling approaches and simulation proce-
properties, while wood technologists mainly dures (Goudie 1999; Lindeberg et al. 1999;
studied methods of wood processing and uti- Nepveu and Malan 1999; Seifert 1999).
lization. In recent years, there has been a Fast-growing plantation poplars have been
Wood 'md Fiber Sclmce. 33(2). 2001. pp. 264-274
0 2001 by the Society of Wood Sc~enceand Technology
 Bao und Liu-RELATIONSHIP OF WOOD PROPERTIES TO QUALITY OF VENEER AND PLYWOOD 265

widely planted all over the country in China growth ring width, density, shrinkage, hartl-
since the 1960s. It is estimated that the area ness, modulus of elasticity, and modulus of
of plantation poplar currently totals 6,670,000 rupture. The quality criteria of veneer are th~e
ha (Wang 1995). Studies on the properties of variance of veneer thickness, the ratio of lath~e
wood from fast-growing forest plantations checks in veneer; the degree of surface rough-
were listed in the Eighth Five-Year (1999- ness of veneer, and the transverse modulus of
1995) and Ninth Five-Year (1996-2000) Plan rupture. Among these four indices, the first
of National Key Technologies R&D Programs and the second are primary; the larger and th~e
in China; and preliminary study of wood prop- deeper the lathe checks, the lower the modulus
erties of fast-growing plantation poplar in of rupture and the rougher the veneer surface.
combination with pulping and plywood pro- The quality of veneer depends mainly on th~e
cessing has been carried out (Bao and Jiang variance of veneer thickness and the ratio of
1998), but this field needs to be further de- lathe checks of veneer (the ratio of the depth
veloped. of check to the thickness of veneer) (Cahalin
In our previous research (Bao and Liu 1987; Northeast Forestry University 1989).
1999), the relationships between wood prop- Therefore, the variance of veneer thickness
erties and the quality of veneer and plywood and the ratio of lathe checks of veneer were
of plantation poplars were studied. It was ob- selected as the criteria of veneer quality. Th~e
served that the quality of veneer and plywood physical and mechanical properties of ply-
is closely related to some of the wood anatom- wood mainly include moisture content, den-
ical, chemical, and physical-mechanical prop- sity, glue-bond strength, heat conduction, and
erties. In the present study, an attempt was ratio of strength to weight. Among these, glue-
thus made to model their relationships with bond strength is the primary index that reflects
stepwise and multiple linear regression in or- the mechanical strength of plywood (Cahalin
der to explore the possibility for predicting the 1987). It is also the property that must be test-
quality of veneer and plywood through the key ed according to the Chinese National Standard
wood properties measured. (Chinese National Technique Monitoring Bu-
reau 1981). In our studies, the glue-bond
BACKGROUND
strength of plywood is determined and anii-
lyzed as the index of plywood quality.
In our previous study (Bao and Liu 1999), The relationships between wood properties
the indices of wood properties that were listed and quality of veneer and plywood in Chinese
in the IAWA list of microscopic features for plantation poplars were developed in our pre-
hardwood identification (IAWA Committee vious study (Bao and Liu 1999) as follows; it
1989) and the Chinese National Standards was found that the variance of veneer thick-
(Chinese National Technique Monitoring Bu- ness (VVT) was highly statistically signifi-
reau 1981, 1991) were followed as much as cantly related with the fiber width (FW), the
possible. The anatomical properties include: fiber proportion (FP), and the vessel propor-
fiber dimensions (fiber length, fiber width, tion (VP), and statistically significantly related
thickness of cell wall, diameter of lumen, ratio to the fiber length (FL), the thickness of cell
of cell wall to lumen, ratio of lumen to cell wall (TCW), the ratio of cell wall to lumen
diameter), vessel dimensions (vessel length, (TCWLD), the ratio of lumen to cell diameter
tangential vessel diameter, number of vessels (LDID), the vessel length (VL), and the ray
per mm2), microfibril angle, vessel proportion, proportion (RP). The ratio of lathe checks of
fiber proportion, ray proportion, and percent- veneer (RLC) was highly statistically signifi-
age of cell wall. The chemical properties in- cantly related with the lumen diameter (LDl),
clude: pH value, acid buffer, and alkali buffer. the modulus of elasticity (MOE), and statisti-
The physical-mechanical properties include: cally significantly related with the air-dry
266 WOOD AND FIBER SCIENCE, APRIL 2001, V. 33(2)

wood density (WD), the total volumetric: mechanical properties. The fiber length was
shrinkage (SV), and the tangential hardness measured under a projector with the macerated
(TH). The glue-bond strength (GS) was highly fibers. The microfibril angle was determined
statistically significantly related to the pH val- on 15-20km thickness tangential sections
ues (PH), the fiber width (FW), the ray pro- with the help of a polarized microscope at
portion (RP), the vessel proportion (VP), the 10OX magnification. The other cell dimen-
vessel length (VL), the fiber length (FL), the sions were measured using an image analysis
lumen diameter (LD), the thickness of cell system (Quantimet-570). The chemical and
wall (TCW), the ratio of cell wall to lumen physical-mechanical properties of wood were
(TCWLD), the microfibril angle (MFA), and determined according to the Chinese National
statistically significantly related with fiber pro- Standards (Chinese National Technique Mon-
portion (FP). itoring Bureau 1981, 1991).
Obviously, the quality of veneer and ply- The procedures of rotary cutting veneer and
wood is closely related to the anatomical, making plywood are as follows: log + rotary
chemical, and physical-mechanical properties cutting + drying + gluing + matching +
of the wood from which it is made. Thus, it solidifying -+ heat pressuring + trimming -+
is possible to model their relationships with sanding.
stepwise and multiple linear regression for Rotary cutting.-The wood of fast-growing
prediction of the quality of veneer and ply- poplars is soft as well as high in moisture con-
wood through the key wood properties. tent, so the green wood was rotary cut directly.
The parameters for rotary cutting follow: the
MATERIALS AND METHODS
grinding angle of the rotary cutting knife was
Testing materiuls 19-21", the clearance angle was 4-lo, the ve-
Three plantation poplar clones, poplar 72 neer compression rate by the nose-bar was
(Populus X euramericana cv.1-72/58), poplar 8%. In order to compare the effect of different
63 (P. deltoides cv.1-6315I), and poplar 69 (P. veneer thickness on veneer quality under the
deltoides cv.1-69/55), grown on three different same rotary cutting condition, three lunds of
shores of the Yangtse River in China (Islet veneer thickness, namely 1.01 mm, 2.02 mm,
beaches of Xinzhou in Anhui provinces, rive:r 3.03 mm were adopted for each of the three
beaches of Huangzhou in Hubei province:;, poplars. In total, 18 logs were cut; 9 logs were
and lake beaches of Yuanjinag in Hunan prov- cut for 1.Ol-mm veneer, 6 logs for 2.02-mm
inces), and at three different planting densities veneer, and 3 logs for 3.03-mm veneer. All of
( 3 m X 4 m , 4 m X 5 m , 5 m X6m)were the veneer was peeled with a rotary lathe
selected as the testing materials. In total, 18 (3VKKTL665 made in Finland) in the testing
plantation poplar trees were cut down for this factory of Research Institute of Wood Industry
study. The growth conditions of Chinese plan- of Chinese Academy of Forestry.
tation poplars are given in Table 1 . Cutting samples for testing quality of ve-
neer.-The veneer strips for testing were cut
Test methods from the rotary cut veneer pith outward. In the
From each poplar tree, one disc about 5 cm 1.01-mm veneer, they were cut every ten ro-
in thickness at the height of 1.3 m was re- tations, and in the 2.02-mm and 3.03-mm ve-
moved for measuring anatomical and chemic(a1 neer every 5 rotations. Every time a veneer
properties. One log (about 90 cm in length (at strip with width 50 mm (W) and length 1000
the height of 0.3 m to 1.2 m) was used for mm (L) was cut for testing, the ratio of lathe
rotary cutting of veneer, and the other log check in veneer and roughness of veneer sur-
(about 2 m in length at the height of 1.3 m to face were measured. Four veneer strips 80 mm
3.3 m) was used for measuring wood physical- (W) and 1,000 mm (L) were taken for testing
Bao and Liu-RELATIONSHIP OF WOOD PROPERTIES TO QUALITY OF VENEER AND PLYWOOD
268 WOOD AND FIBER SCIENCE, APRIL 2001, V. 33(2)

the variance of veneer thickness and differ- the 1 .Ol-mm veneer, 420 g/m2 in the 2.02-mm
ence in shrinkage. veneer, and 440 g/m3 in the 3.03-mm veneer.
The veneer sheets were cut into veneers A solidifying agent, 1.5 g NH,CL, and 15 g
with width 900 mm (W) and length 1,000 mm flour as filling agent were added to every 300
(L), which were numbered from pith outward g urea-formaldehyde resin glue. About one
in order to readily distinguish the heartwood hour after being glued, the veneer was pressed
and sapwood in the course of making ply- into plywood. Testing of pressing was done at
wood. the Research Institute. In this test, the loading
Testing the quality of veneer.-Variation in pressure was 10 kg/cm2, and the pressure was
veneer thickness was measured by micrometer sustained 2.5 min for three layers of 1.01-mm
calipers on four veneer strips (80 mm X 1,000 veneer, and 5 min and 7.5 min for 2.02-mm
mm) at each of the two ends and the middle and 3.03-mm veneer, respectively. The tem-
point of the strip. The thickness of veneer was perature was controlled at 115OC. The ply-
measured at the same points before and after wood so produced was numbered according to
being dried in order to determine the ratio of the original veneer number (i.e., number of ro-
radial shrinkage of veneer. The ratio of lathe tation) in order to examine the glue-bond
check in veneer was tested on the veneer strip strength variation of plywood from the pith
of 50 mm (W) X 1,000 mm (L). The method outward.
of testing was as follows: two bands of blue Testing of the samples of glue-bond strength
ink were coated on the two ends of veneer of plywood.-The tests of glue-bond strength
strip and they were cut out after drying. If were made according to standard GB 9846.9-
there was a check in veneer strip coated with 88 (Chinese National Technique Monitoring
ink, the check was blue and could be seen eas- Bureau 1988). The size of samples was 100
ily. The number and depth of lathe check in mm (L) X 25 mm (W). Twelve samples were
veneer were observed with a microscope cut from a piece of plywood (300 X 300 mm).
(lox). Five pieces of plywood were made for each
Drying.-All the veneer drying work was log from pith outward. Therefore, 18 trees pro-
done in a Mode RD-6 Roll dryer, made in duced 18 X 5 X 12 = 1,080 test samples. The
MARUNAKA TEKKOSHO INC, in the ply- glue-bond strength of plywood was deter-
wood testing factory of the Research Institute. mined according to standard GB 9486. 12-88
The machine uses heated oil as a heating me- (Chinese National Technique Monitoring Bu-
dium. The 1.01-mm veneer needed drying reau, 1988). The wood properties and quality
only one time; however, 2.02-mm and 3.03- of veneer and plywood have been tested ac-
mm veneer had to be dried twice to satisfy the cording to the Chinese National Standards
requirement for final moisture content of ve- (Chinese National Technique Monitoring Bu-
neer (surface board 10-12%, heart board 8- reau 1981, 1988, 1991).
10%).
Cutting, gluing, and heating pressure.-
Statistical methods
The veneer sheets of 900 mm (W) X 1,000
mm (L) were further cut into small veneer of Based on the relationships obtained between
300 mm (W) X 300 mm (L) after being dried; wood properties and the quality of veneer and
three pieces of small veneer were selected ev- plywood in our paper mentioned above (Bao
ery 4-8 rotations from pith outward to glue and Liu 1999), we used stepwise regression
and press into plywood. Urea-formaldehyde and multiple linear regression to model their
resin glue was used in the study (named MN- relationships in two steps. At first, we used
2, degree of stick 1.35 Pa.5, content of resin stepwise regression to be sure which of the
55.72%, time to solidify 25 s). The quantity wood properties were the key in affecting the
of glue used (double faces) was 280 g/m2 in quality of veneer and plywood (Lang and Tang
 Bao cznd Liu-RELATIONSHIP OF WOOD PROPERTIES TO QUALITY OF VENEER AND PLYWOOD 269

TABLE
2. Vuriubies entered or rejected through stepwise regression analysis
VVT RLC GS
-
Var~ablcs Variables Variable, Variables Variables Variables
Model entered rejected entered rejected entered rejected

WD
SV
MOE
FL TH
TCW
TCWILD
LDID LD
VL TCW
RP TCWLD
MFA
FP

1987); then the relationships between the key Yi(VVT, RLC, GL)
wood properties and the quality of veneer and
plywood were further modeled by multiple = Po + Pilxl~+ Pi2~i2 + . . . + Pirnxtm + E

linear regression (Hu and Feng 1996). (1)

Stepwise regression.-The variance of ve-
where i = 1 to 3, Y,, instead of VVT, RL,C
neer thickness (VVT), the ratio of lathe check and GL; xi,, xi,, . . . , xi,, instead of the key
of veneer (RLC), and the glue-bond strength
wood properties; E, represents the errors in tlhe
(GS) of plywood were taken as dependent var- model. Then, we used the obtained poplar dalta
iables. The anatomical, chemical, and physi- to perform the regression equations and at last
cal-mechanical properties mentioned above, to test the regression equations by comparing
which were statistically significantly related to the measured data with the predicted data.
VVT, TLC, and GS respectively, were regard-
ed as independent variables. The independent RESULTS AND DISCUSSION
variables were examined one by one through
stepwise regression of statistical software Results of relationships between the key
package (Lang and Tang 1987) to find whether wood properties and the quality of veneer
the sum of squares of partial regression of the and plywood of Chinese plantation poplars
independent variable exceeded the selected In order to quantitatively determine the k'ey
critical F-values. Generally, the selected criti- wood properties affecting the quality of ke-
cal F value is F,,, (1, n - ql2) or F,,, (1, n neer, the wood properties that were high~ly
- q/2), where q is the number of independent statistically significant or significantly correlat-
variables, n is the number of observations. If ed to the variance of veneer thickness (VV'T),
the sum of squares exceeds the selected criti- the ratio of lathe check of veneer (RLC), and
cal F-value, it means that the independent var- the glue-bond strength (GS) reported in alur
iable affected the dependent variable statisti- previous study (Bao and Liu 1999), were ta.k-
cally significantly, and is chosen to enter the en as the independent variables, and the vari-
regression equation; otherwise, the indepen- ance of veneer thickness, the ratio of lathe
dent variable is rejected. check of veneer, and the glue-bond strength
Models of multiple linear regression.-Af- were taken as the dependent variables. Then
ter the key wood properties were chosen, we we used the statistical software package (Lang
used the following equations to express the re- and Tang 1987) to undertake the stepwise re-
lationships between key wood properties and gression analysis. Table 2 shows the indepen-
quality of venzer and plywood: dent variables chosen to enter the regression
27 0 WOOD AND FIBER SCIENCE, APRIL 2001, V. 33(2)

equation or rejected through stepwise regres-

sion analysis. It was revealed that the key
wood properties affecting the variance of ve-
neer thickness (VVT) were the fiber width
(FW), fiber proportion (FP), and vessel pro-
portion (VP). However, the independent vari-
ables of fiber length (FL), thickness of cell
wall (TCW), ratio of cell wall to lumen (TCWI
LD), ratio of lumen to cell diameter (LDD),
vessel length (VL), and ray proportion (RP)
were rejected. The key wood properties af-
fecting the ratio of lathe check of veneer
(RLC) were air-dry wood density (WD), total
volumetric shrinkage (SV), modulus of elas-
ticity (MOE), and tangential hardness (TH);
but the independent variable of lumen diam-
eter (LD) was rejected. The key wood prop-
erties affecting the glue-bond strength (GS)
were pH values (PH), fiber width (FW), ray
proportion (RP), vessel proportion (VP), ves-
sel length (VL), fiber length (FL); but the in-
dependent variables of lumen diameter (LD),
thickness of cell wall (TCW), ratio of cell wall
to lumen (TCWLD), microfibril angle (MFA),
fiber proportion (FP) were rejected. These re-
sults were partly in accordance with the other
reports on poplar (Wang et al. 1995) and pine
(Li and Wang 1996).

Models of relationships between the key

wood properties and the quality of veneer of
Chinese plantatiox poplars
Based on the above analysis, the multiple
linear regression was used to further analyze
the key wood properties, the variance of ve-
neer thickness, and the ratio of lathe checks,
respectively. Table 3 shows the results. It in-
dicated that the variance of veneer thickness
(VVT) is a function of fiber width (FW), the
vessel proportion (VP) and the fiber propor-
tion (FP):
VVT = 3.414 + 0.063 FW + 0.049 FP

The correlation coefficient (r) of this re-

gression equation is 0.9076. The determina-
tion coefficient (r2) is 0.8237, which indicates
TABLE4. The comparison between the predicted values and the measured values of the variance of veneer thickness, the ratio of lathe check of veneer, and
glue-bond strength of plywood of Chinese plantation poplars.
The vanance of veneer
thicknesr The ratlo of lathe check of veneer Glue-bond strength
Measured Predicted Relative Measured Predicted Relative Measured Predicted Relative
Sample values values errors values values errors values values errors
site No. of tree (mm) (mm) (%I (%) (%I (Mpa) (MP~) (%I
River 7 2X3X4 0.15 0.144 3.953 25.5 26.1 15 2.270 2.21 2.118 3.928
Beach 6 9X3X4 0.13 0.085 3 1.965 37.7 37.991 0.737 1.97 1.925 2.295
6 3X3X4 0.07 0.076 9.009 37.7 33.462 11.227 1.90 1.916 0.642
7 2X4X5 0.1 1 0.102 5.296 31.3 31.071 0.818 1.77 1.904 7.416
6 9X5X6 0.22 0.187 13.130 26.3 27.038 2.804 2.09 2.129 1.880
6 3X5X6 0.07 0.090 28.571 25.3 31.836 25.996 1.85 1.698 8.015
Islet 6301 0.36 0.343 4.649 34.2 26.983 21.196 2.03 1.928 5.027
Beach 6303 0.05 0.045 9.873 25.7 26.365 2.636 1.33 1.340 0.774
6304 - - 29.0 30.852 6.349 - - -

6905 0.04 0.060 33.333 32.4 28.530 12.067 1.49 1.536 3.349
6906 - - - 28.4 22.673 20.194 - - -

7209 - - - 35.7 30.067 15.801 - - -

Lake 7201 0.21 0.178 13.861 18.4 23.330 26.929 1.85 1.823 1.478
Beach 7202 - - - 28.3 28.498 0.843 - - -

6301 - - - 24.9 27.168 9.107 - - -

6302 0.02 0.029 42.708 32.0 33.365 4.364 1.62 1.660 2.673
6901 0.09 0.120 38.462 30.3 27.874 7.995 1.90 2.025 6.460
6902 - - - 23.7 28.825 21.497 - - -
272 WOOD AND FIBER SCIENCE, APRIL 2001, V. 33(2)

TABLE
5 . Dzfferences between the predicted and measured values of VVT, RLC, and GS in Chinese plantation poplars.
Qualily ~ n d ~ c eofr Average measured Average predicted Alpha values Significant
veneer and plywood values values (a) difference*

VVT (mm) 0.126 0.123 0.893 No

RLC (%) 29.269 29.002 0.861 No
GS (MPa) 1.834 1.834 0.995 NO
* There was no significant difference between the predicted and measured values if the alpha value >O 05
stat~st~cally

that 82.37% of the variance of veneer thick- The correlation coefficient (r) of the regres-
ness can be accounted for by the fiber width, sion equation (GS) between the glue-bond
the fiber proportion, and the vessel proportion. strength of plywood and the key wood prop-
The regression equation is highly statistically erties is 0.9399, the determination coefficient
significant at the 1% level. The ratio of lathe is 0.8834, and the regression equation is high-
check (RLC) is a function of wood hardness ly statistically significant at the 1 % level.
on the tangential surface (TH), modulus of Based on the coefficient of determination, it is
elasticity in bending (MOE), air-dry wood shown that 88.34% of the variation of the
density (WD), and total volumetric shrinkage glue-bond strength can be explained through
of wood (SV): the pH values, fiber length (FL), fiber width
(FW), vessel length (VL), vessel proportion
RLC = 9.472 + 0.005 TH + 0.003 MOE (VP), and ray proportion (RP).
+ 8.366 WD - 2.302 SV. (3)
The correlation coefficient (r) of this regres- Test of the regression equations
sion equation is 0.7375. The determination co- Before these equations can be accepted,
efficient (r2) is 0.5407, which indicates that they must be tested. The regression equations
54.1% of the ratio of lathe check of veneer presented in Table 3 were evaluated by com-
can be explained by the tangential hardness, paring model-predicted values with experi-
the modulus of elasticity, the air-dry wood mentally measured values. As shown in Table
density, and the total volumetric shrinkage. 4, the predicted values are generally close to
The regression equation is highly statistically the measured values, and the difference be-
significant at the 1 % level. tween the predicted values and the measured
values are not statistically significant (Table
The model of the relationships between wood 5). This verifies that the quality of veneer and
properties and glue-bond strength of plywood can be predicted satisfactorily with
plywood of Chinese plantation poplars. the selected key wood properties through the
Table 3 shows that the regression equation regression equations obtained. However, it is
of the glue-bond strength (GS) changes with also found from Table 4 that there exist dif-
the key wood properties of Chinese plantation ferent relative errors between the predicted
poplar. It is observed that the glue-bond values and measured values which shows that
strength of plywood (GS) is a function of pH further researches are recommended to consid-
values (PH), fiber length (FL), fiber width er a greater number of sample trees.
(FW), vessel length (VL), and proportion of
vessel (VP) and ray (RP): CONCLUSIONS

GS = 3.326 - 0.268 PH - 0.002 FL Results from this study led to the following
+ 0.191 FW - 0.004 VL conclusions:

- 0.033 VP + 0.095 RP. (4) The variance of veneer thickness (VVT) is a

 Bao and Liu-RELATIONSHIP OF WOOD PROPERTIES TO QUALITY OF VENEER AND PLYWOOD 273

function of fiber width (FW), the proportions and Simulation Softwares," IUFRO WP
of vessels (VP), and proportion of fibers (FP): S5.01-04, Biological Improvement of Wood
Properties in La Londe-Les-Maures, France,
VVT = -3.41 4 + 0.063 FW + 0.049 FP 1999.

REFERENCES
The ratio of lathe check (RLC) in veneer is a BAO, F., AND Z. JIANG.1998. Wood properties of main
function of wood hardness on the tangential tree species from plantation in China. China Forestry
surface (TH), modulus of elasticity (MOE), Publishing House. Beijing, China, 558 pp. (in Clii-
nese).
air-dry wood density (WD), and oven-dry vol- , AND S. LIU. 1999. Studies on the relationship be-
umetric shrinkage (SV): tween wood properties and quality of veneer and plly-
wood of plantation poplars. Australian Forestry 62(1):
RLC = 9.472 + 0.05 TH + 0.003 MOE 21-26.
CAHALIN, H. B. 1987. Manual of plywood processing,
Chinese Forestry Publishing House, Beijing, China.
CHINESENATIONAL TECHNIQUE MONITORING BUREAU.
The glue-bond strength of plywood is a func- 1981. National Standard GB 2677-1-8 1. Chemical com-
tion of the pH values (PH), fiber length (FL), position testing method for material of paper industry.
fiber width (FW), vessel length (VL), and the China Standard Press. Beijing, China (in Chinese).
. 1988. National Standard GB 9846.1-9846.12-88
proportions of vessels (VP) and rays (RP): Plywood. Chinese Standard Press, Beijing, China (in
Chinese).
. 1991. National Standard GB 1932-1942-91. Test-
ing methods for physical and mechanical properties of
wood. China Standard Press. Beijing, China (in Chi-
nese).
GOUDIE,J. W. 1999. Modeling the impact of silviculture
activities on the wood characteristics of coastal western
The correlation coefficients (r) of the above
hemlock in British Columbia. Proc. IUFRO WP S5.01-
regression equations are 0.73 - 0.93. The de- 04, Biological Improvement of Wood Properties. Third
termination coefficients (r2) for the regression workshop. La Londe-Les-Maures, France.
equations are 0.54 - 0.88. The regression HAYGREEN, J. G., AND J. L. BOWYER. 1982. Forest products
equations are highly statistically significant at and wood science-An introduction. The Iowa St.ate
University Press, Ames, IA.
the 1% level. The difference between the pre- Hu, J., AND T. FENG.1996. Applied statistics. Beijing Uni-
dicted and measured values is not statistically versity Press. Beijing, China (in Chinese). IAWA Com-
significant. It is possible to satisfactorily pre- mittee. 1989. IAWA list of microscopic features For
dict the quality of veneer and plywood of Chi- hardwood identification.
IAWA COMMITTEE. 1989. IAWA List of microscopic fea-
nese plantation poplars with the key wood
tures for hardwood identification. RijksherbariumIHor-
properties selected. tus Botanicus, Leiden, Netherlands.
LANG,K., AND Z. TANG.1987. Statistical Software Pack-
age Applied to IBM-PC. China Forestry Publishi~ng
ACKNOWLEDGMENTS
House. Beijing, China (in Chinese).
This research was part of the Ninth Five- LI, C., AND J. WANG.1996. Study on relationship among
loblolly pine log peeling, veneer quality and wood
Year (1996-2000) Plan of National Key Tech- properties. China Wood Industry lO(2): 10-1 6 (in Chi-
nologies R & D Program supported by the nese).
State Science and Technology Commission, I? LINDEBERG, J., U. BERGSTEN, S. CLAESSON, K. SAHLEN, AND
R. China, and granted by the National Nature G. EGNELL.1999. Connection between silviculture and
Science Foundation, I? R. China. This com- wood quality. Proc. IUFRO WP S5.01-04, Biological
Improvement of Wood Properties. Third Workshop. La
munication was presented to the Third Work- Londe-Les-Maures, France.
shop "Connection Between Silviculture and NEPVEU,G., AND E MALAN.1999. Connection between
Wood Quality Through Modeling Approaches silviculture and wood quality through modeling ap-
274 WOOD AND FIBER SCIENCE, APRIL 2001, V. 33(2)

proaches and simulation software. Proc. IUFRO WP spruce depending on silvicultural treatment. Proc. IUF-
S5.01-04, Biological Improvement of Wood Properties. RO WP S5.01-04, Biological Improvement of Wood
Third workshop. La Londe-Les-Maures, France. Properties. Third Workshop. La Londe-les-Maures,
NORTHEAST FORESTRY UNIVERSITY. 1989. Plywood pro- France.
cessing. China Forestry Publishing House. Beijing, Chi- WANG,J., C. LI, AND C. Lu. 1995. Study on relationship
na. (in Chinese). among poplar log peeling, veneer quality and wood
PANSHIN, A. J., AND C. DE ZEEUW.1980. Textbook of properties. China Wood Industry 9(5):1-6 (in Chi-
wood technology, 4th Ed. McGraw-Hill, New York, nese).
NY. 722 pp. WANG,S. 1995. Progress in poplar research. China For-
SEIFERT,T. 1999. Modeling wood quality of Norway estry Publishing House. Beijing, China (in Chinese).