1/D
t
(mm/MPa)
S
p
eci
m
en
k (
N
/m
m
)
∆
Measured D
t
Expt k
∆
k
(1/D
t
)
23
2.3 Results and Discussion
Figure 2.4 shows the mean modulus E in longitudinal direction of over 1000
specimens for both normal strands and VTC strands. The modulus for normal strands was
close to solid wood for poplar wood which is 10GPa (Wood Handbook 2007). The
Modulus E of modified VTC strands was much higher than the unmodified normal
strands. Kamke and Kutnar (2007) demonstrated that the mechanical properties of
modified VTC strands increases linearly as the level of density increases. This is the
major advantage of modified VTC strands because the VTC process can utilize small
logs and fast growing species like poplar and still obtain high MOE through
densification.
0
5
10
15
20
25
30
E
L
(
GP
a
)
Normal Strand
VTC strand
Figure 2.4. Modulus of Elasticity E
L
for normal strand and VTC strand.
Over 300 DLS specimens were prepared and loaded in tension. The stiffness k
was measured from the displacement and load. The stiffness and input geometries, elastic
modulus and shear modulus were input into the shear-lag model to extract the interfacial
properties. Three variations of adhesive coverage of PF resin and PVA were prepared,
1%, 25% and 100%. The 100% level is an extreme case of a lot more PF or PVA applied
to glue each strand together for making DLS specimens. Table 2.1 lists the average
values of interfacial properties and 95 percent confidence intervals for PF resin and PVA
24
wood glue at different adhesive coverages for normal and VTC strands. Although over
500 DLS specimens were prepared and tested, Table 2.1 only includes the last set of data
that was tested using an Instron machine (model 5445 with clip gage). At 100% coverage
both VTC and normal strands gave a negative
D
t
for PF resin but not for PVA glue with
normal strands. VTC gave a better confidence interval than normal strands. PVA wood
glue gave a smaller relative error than PF resin.
For a better representation of results in Table 2.1, Figure 2.5 shows the PVA
adhesive compliance (1/
D
t
) for specimens with 1%, 25% and 100% coverage in DLS
specimens with unmodified hybrid polar strands. No VTC experiments were done for the
PVA resin. The adhesive stiffness property increased (compliance decreased) as the
amount of glue coverage increased. Figure 2.6 plots the results for interfacial properties
1/
D
t
of Table 2.1 as the function of adhesive coverage for PF resin. VTC strands had a
larger
D
t
than normal strands. As the level of adhesive coverage increased 1/
D
t
approaches zero or even become negative for both VTC and normal strands at 100%
coverage. This demonstrates that as the adhesive goes from discrete droplets to a
continuous line, the bonding stiffness of adhesive and wood is greater than the stiffness
of the wood itself.
The surfaces of VTC strands were much smoother than normal strands (Kutnar
and Kamke 2007) and densification was likely to limit resin penetration. This was
confirmed by microscopy and SEM images of these bond lines (see appendix and Kutnar
et al 2007). Nevertheless, the interfacial properties (
D
t
) of VTC were higher than for
normal strands. Thus VTC strands should work well in OSB and may lead to better stress
transfer between strands. Furthermore, the fact that higher
D
t
was achieved with less
penetration suggest that penetration may not be important for good bonding.
25
Do'stlaringiz bilan baham: |