65
transformed stiffness matrix, a new approximate orthotropic material was obtained. The
core strands had correct axial moduli, but the 2D calculations cannot account for tension-
shear coupling because that is a 3D effect.
The modeling of OSB with VTC strands followed the methods described earlier
for normal strands.
In this case, the geometry (length and width) of the normal strands
was the same as before but the length of the VTC strands were 125 mm (25 mm shorter
than the normal strands). Real VTC strands were thicker but here we had to use same
thickness to align with the grid. To account for thickness, strand length and weight for the
20% weight VTC, there were two VTC strand layers (one on each surface) out of 21
strands total (19 normal strand layers were randomly oriented).
This structure was as
close as possible to 20% by weight we could achieve as the grid and it gave 19.7% VTC
by weight. For 40% weight VTC strands, the closest possible structure has two VTC
strands on each surface out of 18 total strands (14 normal strands randomly for core
layer). The actual MPM simulation of VTC weight was 40.0% for the 40% VTC case.
To get the base line for comparison between VTC and normal strands, we also did
simulations for the control cases for 20% and 40% of weight
of normal strands on the
surface. In the 20% control case, there were total of 4 normal strands on the surfaces (2
for each surface) out of 20 total strands (16 normal strands randomly for core layer). In
the 40% control case, there were 8 normal strands on the surfaces (4 for each surface) out
of 20 total strands (12 normal strands randomly for core layers). Both these cases exactly
matched to target weight fraction.
There was at least a 7 % increase in MOE for 20% weight VTC compared to the
control. This result is shown in Figure 3.16. This trend is the same for the case of 40%
weight VTC on the surface layers, but there was a 28% increase in MOE for 40% weight
VTC compared to no VTC control (see Figure 3.17). The values of MOE decreased as
1/D
t
increased and the difference got slightly smaller at high 1/D
t
.
66
Figure 3.16. Comparison of MPM calculation of modulus of OSB panel with 20% by
weight of VTC on the surface to the control at 20% compaction.
Table 3.2 has a summary of simulated moduli results of Figure 3.16 and Figure
3.17 for different values of adhesive coverage compared to experimental results (Rathi
2009). The experimental results were in bending but the MPM simulations were in
tension. Bending MOE is expected to be much higher in MOE than tension MOE (as it
is) but relative changes provide some indication of the validity of the model results. In
studies by Rathi (2009), the 20% control is 11.6GPa +/- 1.57GPa for MOE and for 20%
VTC, MOE is 12.4 +/-1.57 for a 6.9% increase. In the case of 40% control, MOE is 12.9
+/-1.72 GPa and for 40% VTC, MOE is 16.1+/-2.47 GPa for a 24.8% increase. The
results in our studies vary with the interfacial stiffness. In our simulation of 20% VTC,
there is 7.9% increase in MOE for 1% adhesive coverage and 9.1%
increase in MOE for
25% adhesive coverage levels and 12.9% increase in MOE for 100% glue. For 40%
VTC, there is 23 % increase in MOE for 1% adhesive coverage, 27% increase in MOE
for 25% adhesive coverage levels, and 31% increase in MOE for 100% adhesive
coverage. In average there is 9% increase in MOE for 20% VTC and 26%
increase for
40% VTC which is close to experiment values of Rathi (2009).
3400
3600
3800
4000
4200
4400
4600
4800
5000
5200
0.00
0.01
0.02
0.03
0.04
0.05
Do'stlaringiz bilan baham: