Abstract: The cone penetration test, considered one of the most crucial in-situ testing methods in geotechnical engineering, is unable to directly measure soil parameters. Generally, the empirical relationship between the penetration resistance and the soil parameters is established through tests or numerical simulations. In this research, a coupled numerical approach of the discrete element method (DEM) and the finite difference method (FDM) is employed to simulate the entire cone penetration test process within a calibration chamber and unveil the mechanisms that influence the penetration resistance. Firstly, the microscopic parameters of the sand are calibrated based on its macroscopic behaviors. Additionally, the effects of calibration chamber size, sand density and confining stress on the penetration resistance are thoroughly analyzed. Finally, a relationship is established between the normalized penetration resistance Q and the peak internal friction angle φ_peak. The findings indicate that the size effects in the simulation become negligible when the continuous-discrete model size ratio R_df reaches 0.67 and the normalized radial length R_d of the calibration chamber is set at 20. Furthermore, the simulated penetration resistance closely matches the results obtained by the cavity expansion methods, confirming the reliability of the numerical simulation. An exponential relationship is observed between Q and φ_peak, which closely aligns with the in-situ test results. This further validates the accuracy of the coupled simulation method. These outcomes offer valuable insights for establishing empirical relationships between the cone penetration resistance in sand and the soil parameters in geotechnical engineering.