An abstract of the thesis of

Time(msec)
E(M
Pa
)

---

56 
Work on compression of foam, however, shows that quasi-static results can be 
obtained provided the loading rate is lower than about 3% of the wave speed of the 
material (Bardenhagen et al 2005). The compaction of OSB was first run using 2m/sec 
(0.18% of transverse wave speed) but it took too long to finish (one month per 
calculation). Accordingly,
the transverse compression in the radial direction of OSB 
specimen was then tried at 4m/s (0.36% of transverse wave speed), 8m/s (0.73% of 
transverse wave speed), 16m/s (1.45% of transverse wave speed), 32 m/s (2.9% of 
transverse wave speed) and 64m/s (5.81% of transverse wave speed) to compression ratio 
of 0.6. All these speeds were less than 1% of longitudinal wave speed and gave similar 
results. The compaction speed used to compact OSB panel was thus set to 32 m/sec (see 
chapter 6 for details on how compaction rate affects density profile) to give faster 
simulations with minimal artifacts due to loading rate. 
1900
2400
2900
3400
3900
4400
4900
5400
5900
0.00
0.05
0.10
0.15
0.20
0.25
1/D
t
(mm/MPa)
MO
E
 (
MP
a
)
1% Glue
25% Glue
PF
20%
10%
0%
30%
100% Glue
Figure 3.8. Simulation results of MOE versus 1/D
t
at different levels of compaction with 
commercial OSB.

---

57 
Figure 3.9. Sample simulation of OSB that loaded in tension. 
Next, structures at selected compactions were input into new MPM simulations 
and loaded in tension. The loading rates were set to 5 m/sec which is 0.094% of 
longitudinal wave speed. Figure 3.9 represents tension test that was fixed on one end 
(left) and pulled on the other end (right end). A series of simulations were run with 
different values of D
t
and MOE was found by the energy methods described before. 
Figure 3.8 and 3.10 shows the results of axial MOE versus 1/
D
t
as different levels of 
compaction for commercial OSB. Figure 3.10 is the zoom-in of Figure 3.8 for 0% to 30% 
compaction. As the level of compaction increased, MOE increased. Using experimental 
results for 
D
t
(in previous study, chapter 2) for PF resins, if the strands are covered 1% 
(by area) with glue, the stiffness of the OSB panel would be about 14% lower than could 
be expected for a panel with 100% coverage and about 10% lower than it could be for 
25% rather 100% coverage. In Figure 3.8, all the data in this plot are obtained from 
average of at least five runs. The error bars on one curve show the standard deviation of 
the results. There was little change in standard deviation for different values of 1/D
t
. 

---

58 
2900
3400
3900
4400
4900
5400
5900
0.00
0.01
0.02
0.03
0.04
0.05