Abstract: Abstract: A crushing blade has been one of the most important components during harvesting corn for silage in modern agriculture. However, the existing crushing blade cannot fully meet the high requirement of sharpness and slip cutting performance of the silage corn harvester. In this study, a bionic crushing blade was proposed under the mandible of the leaf-cutting ant as a prototype. A kinematic model of the mandible was firstly established to clarify the cutting principle of the leaf-cutting ant. A bionic edge curve was then designed to extract the contour curve of the mandible of the leaf-cutting ant. A reverse engineering was also used to fit the slip-cutting curve. As such, a discrete element model was established in the stalk cutting test using an EDEM software. The evolution of stalk cutting force was then explored under the same parameter setting of bionic, oblique, and straight blade blades. A bench test was carried out to record the stalk cutting deformation using a high-speed camera. The simulation and high-speed camera results show that the stalk cutting deformation was divided into three stages: extrusion, cutting, and through the stage. Specifically, the blade continued to compress the stalk skin in the extrusion stage, until the cutting force increased sharply to the blade's downward pressure exceeding the strength of the stalk skin that caused by the skin rupture, where the cutting force appeared at the first peak after a small decline in the cutting force. The blade was used to cut the stalk with the increase of cutting force, where the maximum cutting force appeared instantly, and then dropped sharply in the cutting stage. The cutting force decreased sharply to a stable state for the stalk cutting in the through the stage, when there was only friction between the blade and the stalk. Correspondingly, the maximum cutting forces of the bionic blade in the simulation were 4.05% and 6.09% lower than those of the angled and straight blade, respectively. By contrast, the maximum cutting forces of the bionic blade in the bench test were 10.53% and 12.82% lower than those of the oblique and the straight blade, respectively. Correspondingly, the relatively relative errors of the maximum cutting force for the three blades in the simulation and bench test were 7.78%, 1.11%, and 0.66%, respectively. More importantly, the better cutting performance of the bionic blade was achieved in the uniform flatness of the stalk cutting surface and the low maximum cutting force of the stalk. An orthogonal test was also carried out in on the stalk cutting bench with the blade type, cross-cutting angle, and cutting angle as the test factors, while the maximum cutting force of stalk as the test index. An optimal combination of parameters was obtained to verify the simulation. Consequently, the blade type and cross-cutting angle posed a highly significant effect on the maximum cutting force of stalk, whereas, there was a significant effect of on the cutting angle. The validation test showed that the maximum stalk cutting force was 623.3 N at the bionic blade, the cross-cutting angle of 90°, and the cutting angle of 20° under the optimal combination of parameters. This finding can also provide a theoretical reference to design the crushing blade in the silage corn harvester.