Currently, there is a lack of theoretical basis for setting operational parameters such as cleaning and conveying Jerusalem artichoke tubers during mechanical harvesting. A bonding model that reflected the root flexibility and tuber detachment mechanical characteristics was established by using the discrete element method, and its relevant parameters were calibrated. In the first step, the intrinsic, basic contact and associated mechanical parameters were determined through laboratory experiments. Next, based on these parameters, Hertz-Mindlin with bonding V2 contact model and Metaparticle function was used to establish the flexible model for Jerusalem artichoke root-tuber. Single-factor experiment and response surface methodology were used to determine the bonding parameters such as normal stiffness, tangential stiffness, normal strength, and tangential strength between Jerusalem artichoke root particles and between root and tuber particles. Subsequently, the three-point bending and root-tuber tensile simulation experiments were carried out on the determined parameters. The error between the simulation results and the physical experiment values of the root flexural modulus was 4.29%. The error between the simulation results and the physical experiment values of the maximum allowable tensile resistance between the root and tuber was 7.72%. Finally, the calibrated model was used to simulate the Jerusalem artichoke tuber harvesting operation. The effect of drum screen rotary speed on the tuber shedding rate in the simulation and field experiments had a similar trend. The research demonstrated that the method used to establish the model was correct, and the calibration parameters were deemed reliable. As a result, the established Jerusalem artichoke root-tuber flexible model can be effectively employed for further simulation research pertaining to the mechanized harvesting process of Jerusalem artichokes.