Abstract: Disk-shaped particles are widely found in nature and industrial environments. Existing literature lacks a comprehensive study on drag models for disk-shaped particles applicable to industrial application scenarios. To expand the applicability of the existing model, this paper investigates the drag coefficients and flow characteristics of disk-shaped particles in a flow field based on particle-resolved direct numerical simulation (PR-DNS).The OpenFOAM-body-fitted mesh method is adopted to calculate the drag coefficient of disk-shaped particles with low aspect ratios (0 < Ar < 1), Reynolds numbers (Re≤1000), and incidence angles (0°≤θ≤90°). A new drag coefficient correlation is proposed using genetic algorithm (GA) optimization. Results show that the flow characteristics around disk-shaped particles are jointly affected by aspect ratio, Reynolds number, and incidence angle. The increase of Reynolds number significantly enhances the flow instability, while incidence angle and aspect ratio alter the particle's projected area, thereby changing its drag coefficient and flow characteristics. Numerical verification indicates that the new correlation has high accuracy, with an RMSE of 0.2993 and a mean relative error of 4.16%. It shows good consistency with existing drag coefficient correlations for disk-shaped particles. It improves the existing models and provides reliable theoretical support for computational fluid dynamics (CFD Eulerian-Lagrangian) numerical simulation and engineering applications of disk-shaped particles.