Next: Intrinsic properties for
Up: Exact solutions for
Previous: Field fluctuation result
A result that can be very useful in testing elastic programs is the equal shear modulus result, true for any microstructure [21,22]. If there are only two phases present, with equal shear moduli G but different bulk moduli K1 and K2, then the effective shear modulus of the entire system is just G, and the effective bulk modulus K of the system is, in 2-D,


We choose the same cubic image as was used for the small contrast in conductivity case to test the equal shear moduli result. A 22 x 22 x 22 pixel cube (phase 2) is centered in a 30 x 30 x 30 unit cell, so that the volume fractions are c1 = 0.60563 and c2 = 0.39437. The two shear moduli are taken equal to unity, G1 = G2 = 1, and the two bulk moduli are K1 = 1, and K2 = 20. The exact answer, according to eq. (53), is K = 2.263250, while the numerical answer is K = 2.269684, a difference of only 0.3%.
In 2-D, a test of the equal shear moduli result can be combined with a test of the effect of digital resolution by considering the checkerboard problem. The shear moduli are both equal to 2, and there is a ratio of 10 between the bulk moduli (K1 = 1, K2 = 10) (see Fig. 6 for a picture of the microstructure). The effective bulk modulus is computed as a function of system size L x L, where each "check" is L/2 x L/2. The exact value of K is 2.8 using eq. (60). Table 9 displays the data from this test. For L > 16, the error in the effective bulk modulus is about 1% or less.
| System size |
K | % diff. |
| 2 | 5.500 | 96.4 |
| 4 | 3.271 | 16.8 |
| 8 | 2.937 | 4.9 |
| 16 | 2.841 | 1.5 |
| 32 | 2.812 | 0.4 |
| 64 | 2.804 | 0.1 |
| 128 | 2.8012 | 0.04 |
| 256 | 2.8003 | 0.01 |
| 512 | 2.8001 | 0.004 |
Table 9: Effective bulk modulus for checkerboard--equal shear moduli case. |
|---|