A truncated Gaussian random filed method was employed to generate three-dimensional reconstructed models of polyolefin catalysts based on the porosity evaluated from absorption method and the correlation function from small angle X-ray scattering (SAXS) method. It has been found that the reconstructed models have the same porosity and the correlation function as the original catalysts. Moreover

the way to obtain the correlation function of three-dimensional form directly from SAXS method is easier than the process that the correlation function extracted from two-dimensional images is expanded into three-dimensional form. And the correlation function from SAXS method is more abundant and accurate than that from two-dimensional images

because the nanoscale resolution is hardly achieved in some ordinary images but can be characterized by the SAXS method which provides more statistics. The three-dimensional models of catalysts establish the basis for calculation on the effective heat and mass transport coefficients such as diffusivity

permeability

and heat conductivity

and for prediction of the catalyst fragmentation.