To compute vapor diffusivity, both finite difference and finite element techniques [15] were applied to the 3-D digital 100 pixel x 100 pixel x 100 pixel binary images. The finite element techniques consider corner and edge connections as well as pixel face connections and thus result in higher computed values for relative diffusivity. Both techniques were executed because of the extreme fineness of the slit-like pores in the clinker brick images, as seen in Fig. 1. For pores this small, the overall connectivity of the pore system can be largely influenced by diagonal pixel connections in the 3-D image. To convert the relative diffusivities to absolute values, a value of 0.0922 m2/h was used as the diffusivity of water vapor in air [16]. For the lime silica brick, relative diffusivities of 0.007 and 0.015 were assigned to the solid phases, corresponding to the values determined for the 16 % porosity phase using the finite difference and finite element techniques, respectively. Thus, we are assuming that the lime silica brick has a self-similar microstructure such that the relative diffusivity computed for the coarse pores can be used to provide the values needed for the fine-scale porosity present in the " solids" in the 3-D µCT images.
To compute permeability, Stokes equation for slow incompressible flow was solved using a finite difference scheme along with non-centered difference equations [4,5]. These calculations were performed on both 100 pixel x 100 pixel x 100 pixel and 200 pixel x 200 pixel x 200 pixel central portions of the 3-D microstructure to examine the effects of sample size (with respect to obtaining a representative elementary volume) on permeability. For the lime silica brick, no attempt was made to account for the contribution of the fine-scale porosity to the overall permeability, as it was considered that the measured permeability would be dominated by the porosity present at the coarser scale imaged by the X-ray tomography technique. In all cases, transfer coefficients were computed for each of the three principal directions.