This study investigates a velocity PDF-based model for predicting particle dispersion in porous media flow. Models in this area operate along a spectrum: at one end, fully resolved simulations capture fine-scale physics with high computational cost; at the other, reduced-order models sacrifice detail in favor of macroscopic parameterizations and computational efficiency. Our objective is to assess a model that occupies a position closer to the middle of this spectrum, leaning toward the high-detail end, and to identify its operating limits in terms of numerical and physical parameters. We applied the model to a dataset spanning a range of Peclet numbers, different flow regimes, and varying numerical parameters such as time step size and number of bins. The results show that the model accurately predicts the mean tracer position, while consistently underestimating the standard deviation by approximately 40% across different conditions. This error is attributed to a misrepresentation of the relative contributions of dispersion and diffusion in the relevant Peclet number range and suggests potential for further improvement through parameter adjustments. Nevertheless, the model offers significant computational savings, particularly in turbulent flow regimes. These findings suggest that velocity PDF-based models can serve as efficient tools for studying dispersion in complex flow systems, with potential applications in environmental and industrial processes.     «

This study investigates a velocity PDF-based model for predicting particle dispersion in porous media flow. Models in this area operate along a spectrum: at one end, fully resolved simulations capture fine-scale physics with high computational cost; at the other, reduced-order models sacrifice detail in favor of macroscopic parameterizations and computational efficiency. Our objective is to assess a model that occupies a position closer to the middle of this spectrum, leaning toward the high-de...     »