Abstract
Due to the lack of an exact simulation model for spinach seeds, existing seed-metering devices exhibit poor seeding performance and struggle to achieve precise seeding of two seeds per seed-metering hole. This study proposes a method for calibrating discrete element parameters of spinach seeds and develops a seed-metering device with rectangular seed-metering holes tailored to the two-seeds-per-hole requirement. Firstly, a spinach seed discrete element model was constructed. Based on the Box-Behnken Design experimental scheme, simulations of angle of repose and fluidity, combined with physical of angle of repose (38.27°) and physical mass flow rate (18.60 g/s), were used to calibrate contact parameters: (1) between spinach seed models (coefficient of restitution: 0.385; coefficient of static friction: 0.481; coefficient of rolling friction: 0.042); and (2) between seed models and PVC plate (coefficient of restitution: 0.339; coefficient of static friction: 0.600; coefficient of rolling friction: 0.408). Subsequently, simulated and physical bulk density tests were conducted to verify the validity of the established spinach seed model. Guided by the agronomic requirements for spinach seed planting, the range of dimensional parameters for the seed-metering holes was defined. Using the Box-Behnken Design, simulated seeding tests were performed to optimize the device’s structural parameters (hole depth, hole length, hole width, and seed-filling angle), resulting in optimal values of 2.52 mm, 3.67 mm, 5.15 mm, and 37.99°, respectively. Finally, physical seeding tests were conducted, achieving a qualified seeding rate of 92.23% at a seeding speed of 10 r/min. These results confirm the design accuracy and high operational efficiency of the device. This study lays a foundation for the overall design of future spinach planters.