<p>Understanding and accurately identifying the morphological characteristics of granite residual soil (GRS) particles are crucial for conducting in-depth research on the simulation of soil structures. In this research, the GRS particle samples are directly prepared and subjected to computer tomography scanning to quantify and simulate their morphological characteristics. Additionally, the simulation based on the coupled computational fluid dynamics-discrete element method evaluates the influence of different mesh face counts on the morphological changes of particles during their descent in water. The findings of this study indicate that direct scanning of the GRS particle samples improves scanning efficiency and accurately captures the morphological characteristics. Moreover, three-dimensional analysis shows that the particles present a rough surface, predominantly characterized by prolate and compact shapes. In numerical simulations, decreasing the number of mesh faces on the surface of the particles has a certain impact on the parameters used to evaluate morphological characteristics. Nevertheless, the evaluation parameters based on the geometric positions of the starting and ending points reveal no significant changes. Furthermore, simulation results for particles falling in water simplified to 50 mesh faces retain properties closely aligned with those of the original mesh count. This investigation provides a technical foundation for implementing simplified particle models in numerical simulations, as well as enriching the understanding of the microstructure of GRS.</p>