This study delves into the complexities of gravel formation, attributing its characteristics to the intricate interplay of material properties and particle arrangement orientations. The variability and diversity inherent in experimental outcomes underscore the challenges faced in sampling processes, which are exacerbated by the logistical and financial implications, leading to a scarcity of comprehensive testing. Conventional approaches, such as large-scale triaxial and in situ direct shear tests, although reliable, fail to encompass all factors influencing gravel's strength properties, such as particle size distribution, packing, and interlocking mechanisms. This gap highlights the difficulty of capturing representative properties of gravel formations. To bridge this gap, the design of experiments (DOEs) method and discrete element particle flow software were utilized in this study to explore the influence of microproperties on the mesoscale strength properties of gravel. A statistical analysis of variance (ANOVA) was used to establish regression equations, delineating the range of Mohr-Coulomb failure criterion strength parameters of the gravel formation through in situ direct shear testing. The Taoyuan gravel layer in Taiwan was used as an example, applying these equations, microscale parameters of a rock specimen with mesoscale cohesion (c) values ranging from 0.011 to 2.552 MPa and friction angle (phi) values ranging from 12.66 degrees to 40.75 degrees can be determined through a few trial steps to demonstrate the application and potential of the proposed methodology in addressing the aforementioned challenges.