In green onion harvesting, the problems of easy dumping and low rate of clean digging can be encountered. In this paper, a kind of harvesting device for digging and pulling green onions, referred to simply as the device, was designed. The device mainly consists of a digging shovel, screen bars, clamping conveyor belt, etc. This paper focuses on the analysis of the model forces of green onions and soil in the two states of the onion digging process without dumping and clamping. The key factors affecting the model state of onions and soil were identified as: screen bar length l(2), screen bar inclination angle beta, and pulling point position x. Based on the discrete element simulation technology of EDEM, the mechanism-crop-soil model was established, and a single-factor simulation test was conducted to determine the range of values for each factor. Taking the advantages of field test and three-factor five-level orthogonal experimental design, the parameter combinations of green onion harvesting operation evaluation indices were optimized, including a pulling point position of 166 mm, screen bar length of 242 mm, and screen bar inclination angle of 14 degrees. As the results of the field test show, the harvester operation was stable without congestion or damage, the harvesting effect of green onions was improved, and the clean digging rate reached 100%, which meets the agronomic requirements for onion harvesting and the expectations of users.

Given the significant damage rate observed during the transportation of current garlic combine harvesters in China, this study aims to design a new garlic combine harvester capable of achieving minimal harvest losses. The designed machine can simultaneously complete operations for garlic digging, clamping transport, seedling-bulb separation, soil cleaning, and fruit collection across two rows. Through the use of theoretical analysis and calculation of garlic harvesting operations, the key parameters of soil-breaking device, clamping transport device, length-limiting cutting device, and soil cleaning conveyor were determined. The BoxBehnken test technique was utilized within Design-Expert software, and orthogonal experiments were conducted with the unit's forward speed, digging depth, and soil-breaking angle as test factors, and the stem cutting rate and bulb damage rate as test indices. The test results showed that when the unit's forward speed, digging depth, and soil-breaking angle were 0.49 m/s, 100 mm, and 20 degrees, respectively, the working parameter combination was the best, and the rate of stem cutting and damage were 95.71% and 3.10%, respectively. The findings from the field experiment and optimization aligned closely. This study can provide reference for the development of mechanized garlic harvesting.