Next: Discussion and Conclusions
Up: Main
Previous: Tests of elastic
A series of five realizations of microstructures were created using the algorithm discussed in
Sec. II, for a range of concentrations c1
and c2. The ratio of E1/E2 was held fixed at 10,
and a
range of Poisson ratios
1 =
2
was investigated. Fig. 2 shows the 3-D results for the
composite Poisson's ratio, averaged over the 5 realizations, as a function of c1. The variation of
the effective moduli between configurations was about 5%, as discussed above, with the
Poisson's ratios sometimes varying up to 10% over the five realizations. The system size was
643, and w = 5 in the Gaussian kernel of
eq. (3). The solid lines are the EMT
predictions of
eqs. (4)-(7). The EMT results describe the numerical results well. The numerical results are in
agreement with the EMT predictions for the behavior of the effective Poisson ratio, so that when
1 =
2 = 1/5,
= 1/5 as well for all volume fractions. Also, when
1 =
2 > 1/5,
1/5 <
<
1 =
2,
and when
1 =
2 < 1/5,
1 =
2 <
< 1/5.
This result is also predicted by a "differential method" EMT [21]. The
EMT also accurately fits the Young's modulus results, which are not shown.
Figure 2: Showing the 3-D effective Poisson ratios vs. phase fraction for a stiffness ratio of 10. The lines are the graphs of the EMT equations (4)-(7), and the symbols are numerical results for the Gaussian kernel-based microstructure of Fig. 1.
We have carried out the equivalent simulations in 2-D, for values of Poisson's ratio that are now accessible to our previous algorithm, and have confirmed the EMT results found before [1]. A 2-D version of the microstructure algorithm in eqs. (1)-(3) was used, with w = 5 and a 128 x 128 unit cell. Results are displayed in Fig. 3, with the agreement between the simulation data and the 2-D EMT [3] slightly worse than in the 3-D case shown in Fig. 2.
Figure 3: Showing the 2-D effective Poisson ratios vs. phase fraction for a stiffness ratio of 10. The lines are the graphs of the 2-D EMT equations [3], and the symbols are numerical results.