Safflower is one of the most important oil crops worldwide. To improve the efficiency and quality of harvesting soft safflower filaments, minimize filament damage by the blade, and ensure the integrity of the filaments, an end-effector with a rotary push-cut-type negative-pressure airflow was designed. The design of the rotary blade edge curve was based on Archimedean and logarithmic spirals. The effect of the Archimedean spiral on the pushing of the scattered filaments was investigated. The rotary-cutting force and speed of the logarithmic helix on the soft filaments were analyzed. The sliding cut angle and feed speed of the blade are the key factors governing filament removal and damage. A simulation model of the cutting chamber flow field was established for the change in airflow with the position and speed rotation of the blade, allowing analysis of the parameters of the negative-pressure airflow. Single-factor and Box-Behnken tests were performed with the sliding cut angle, blade feed speed, and negative-pressure airflow speed as influencing factors and filament removal, damage, and drop rate as indices. The response surface clarified the effects of various factors on these indices. Optimal parameters were a sliding cut angle of 32.20°, a blade feeding speed of 0.031 m/s, and a negative-pressure airflow speed of 4.57 m/s. The corresponding filament removal, damage, and drop rates were 93.47%, 6.94%, and 4.33%, respectively. Optimization results showed that filament removal, damage, and drop rates were 93.50%, 7.02%, and 4.43%, respectively. The harvesting process met the requirements of high efficiency and low damage.

Understanding the biomechanical properties of safflowers is essential for appropriately designing harvesting machinery and optimizing the harvesting process. Safflower is a flexible crop that lacks a basis for relevant simulation parameters, which causes difficulties in designing harvesting machinery. In this study, a calibration method for safflowers was proposed. First, a discrete element model was established by measuring the intrinsic parameters of a safflower, such as its geometric parameters, density, Poisson’s ratio, and modulus of elasticity. Second, the contact and bonding parameters were calibrated using a combination of physical and simulation tests. In the contact parameter tests, the Hertz-Mindlin (no-slip) model was implemented for the stacking angle tests conducted regarding the safflower filament. A regular two-level factorial design was used to determine the important factors and perform the steepest climb test. Moreover, the Box-Behnken design was adopted to obtain the optimal contact parameters. In the bonding parameter tests, the Hertz-Mindlin model with bonding contact was applied for the safflower shear simulation tests; moreover, the optimum bonding parameters were obtained through the central composite design test. The results demonstrated that the relative errors between the simulated and measured stacking angles and maximum shear were 3.19% and 5.29%, respectively. As a result, the safflower simulation parameters were accurately calibrated, providing a reference for appropriately setting the simulation parameters and designing key mechanical components.