“…In general, there are three classes of stochastic pore space generation methods that are commonly used. The first prescribes particular attributes, such as porosity or grain distribution, and randomly places geometric objects, such as cubes or spheres, into a domain until target values of those attributes are obtained [5,12,28,37,43]. The second-generation method relies on statistical quantities obtained from images, e.g., two-point correlation functions or linear path functionals, obtained from images of real pore spaces to reconstruct samples with similar characteristics [4,8,24,27,29,31,46].…”