Abstract
FIB‐nanotomography (FIB‐nt) is applied to record high‐resolution 3D pore networks from cementitious materials. Based on these data, it is examined as to why the pore size distribution (PSD), which is obtained from traditional analysis by mercury intrusion porosimetry (MIP), principally deviates from the findings that are achieved by common back‐scattered electron image analysis. The paper does not reflect the vulnerability of the physical model assumptions, but merely focuses on the fundamental issues of the geometrical definition of a PSD. A computationally fast approach for the PSD assessment from 3D data as well as for the simulation of MIP is presented and the varying concepts for the definition of a PSD are compared with each other.