Soil internal erosion stands as a pivotal contributor to the loose soil landslides and the failure of landslide-induced dams. The hydrodynamic effect leads to the detachment, migration, and eventual washout of fine particles (fines) from pores, while conversely, the dynamic state of these fines influences fluid pressure drop. To elucidate the intricate interplay between fluid pressure variations and particle migration during internal erosion, we employed numerical simulations utilizing computational fluid dynamics coupled with discrete element methods (CFDEM). The simulation was validated against classical laboratory experiments. A systematic investigation was then conducted into the effects of coarse-to-fines size ratio, fines content, and hydraulic gradient. The findings reveal that as fines are lost, there is a gradual increase in permeability coefficients and the growth rate of permeability coefficient increases with the increase of hydraulic gradient and fines content. Specifically, when coarse-to-fines size ratio is equal to 10, a natural logarithmic relationship emerges between the growth rate of permeability coefficient and the product of hydraulic gradient and fines content; conversely, at coarse-to-fines size ratio is equal to 8, this relationship manifests as linear. Furthermore, an inverse correlation exists between erosion rates and permeability coefficients. Additionally, our analysis uncovers that the probability distribution governing quiet times associated with fluid pressure drops during internal erosion adheres to an exponential distribution pattern. These insights significantly enhance our comprehension of internal erosion mechanisms while facilitating more accurate assessments regarding various stages of soil erosion.