Abstract: Abstract: Agropyron, a perennial xerophytic grass of the gramineae, is one of the most widespread grass species in arid and semi-arid areas. In artificial grasslands on the dry land, Agropyron is also commonly used in the process of reseeding in grazing areas and spray sowing, due mainly to its drought resistance, as well as cold and grazing tolerance. In addition, the roots of wheatgrass are whisker-like, densely growing, and sand-gathering, particularly for the soil and water conservation, as well as wind-proof and sand-fixing. Therefore, the accuracy of physical parameters is highly demanding in the process of pelleting and coating on Agropyron seeds using a discrete element method. In this study, a novel approach combined physical and simulation test was proposed to calibrate the physical parameters used in an EDEM software. A physical test was conducted to determine the basic physical parameters of Agropyron seeds (dimensions, thousand-grain weight, density, moisture content, Poisson's ratio, elastic modulus, and shear modulus), and the contact parameters (static friction coefficient, rolling friction coefficient, and collision restitution coefficient). The measurement values from the physical test indicated that: In the friction pair of Agropyron seed-Agropyron seed, the collision restitution coefficient ranged from 0.45 to 0.65, the static friction coefficient ranged from 0.5 to 0.7, and the rolling friction coefficient ranged from 0.6 to 0.9, whereas, in the Agropyron seed-steel plate, the collision restitution coefficient ranged from 0.4 to 0.6, the static friction coefficient ranged from 0.2 to 0.4, and the rolling friction coefficient ranged from 0.3 to 0.6. A Plackett-Burman design was carried out to select the physical parameters from the physical test for the later use in the simulation test. The results show that between the Agropyron seed-Agropyron seed, the static friction coefficient, the rolling friction coefficient, and the collision restitution coefficient have a significant effect on the angle of repose. A steepest climbing test was further used to determine the optimal ranges of three parameters. In the Box-Behnken test, the second-order regression equation of the repose angle, and the saliency parameter were established, where an optimal repose angle (30.54°) was used to optimize the model. The best simulation parameters were then achieved: The collision restitution coefficient of Agropyron seed-Agropyron seed was 0.54, the static friction coefficient of Agropyron seed-Agropyron seed was 0.57, and the rolling friction coefficient of Agropyron seed-Agropyron seed was 0.74. In the two-sample T test, there no significant difference in the repose angle from the physical and simulation test (P>0.05). The relative error of repose angle was 1.037%, where the simulation (30.86°) and physical test (30.54°) were under the optimal combination of parameters, indicating the reliability of the simulation test. As such, the optimized parameters that obtained by calibration can be used in the discrete element simulation for the pelletizing and coating process of Agropyron seeds.