While researching the literature in order to complete my review article
of mixture theories,
I found it was almost impossible to find data on elastic constants for
composites that
would lie properly between the Hashin-Shtrikman bounds.
In attempting to come to some resolution of this quandary,
I discovered in most cases that the discrepancies were of the same
direction and type -- suggesting that the presence of porosity
might account for the lack of agreement with the theory.
This paper was the result of that study, and shows that
significant errors can arise from assuming only two solid constituents
are present when in fact a third constituent (void space)
is also present.