The physical sampling to grasp the distribution of deep-sea polymetallic nodules is essential for efficient mining. A hydraulic sampling scheme is proposed that balances low environmental disturbance, high sampling precision, broad coverage, and statistical data viability to address the limitations of existing methods. This paper utilizes a coupled approach of Computational Fluid Dynamics (CFD) and Discrete Element Method (DEM) to explore innovative solutions for precise and continuous hydraulic sampling of deep-sea polymetallic nodules. Focusing on the optimization of the double-row nozzle sampling mechanism design, simulations are conducted to analyze the dynamic characteristics of solid-liquid two-phase flow during the hydraulic sampling process under various nozzle configurations and operational water pressures of the pump. Laboratory experiments with the hydraulic sampling mechanism validate the accuracy of the simulation and the effectiveness of the optimization suggestions. The configuration is finalized with a front nozzle diameter of 10 × φ9 mm, a rear nozzle diameter of 10 × φ7 mm, and an initial working water pressure for the pump of 80 kPa. Subsequently, integrated driving-sampling experiments are conducted using a self-propelled tracked vehicle, validating the feasibility of the proposed configuration.