Abstract: Water chestnut is an important aquatic vegetable, particularly with an annual output of more than 800 000 tons in China. The underground expanded bulb of water chestnut can be eaten fresh or cooked. The edible part of water chestnut has a crisp, sweet, and juicy taste, with many kinds of nutrients. The cuticle of water chestnut needs to be peeled off before eating. A large number of cutting methods can often be adopted to ensure the removal rate of the water chestnut from the peel. However, the complicated shape of the water chestnut can result in very little remaining pulp after peeling, indicating serious loss. There are also serious health and safety hidden hazards in manual peeling. There is no specific peeling device of water chestnut machinery at present. Moreover, the bud and root can be firstly removed for the mechanical peel of water chestnut. In this study, a rotary blade synchronous cutting device was designed for the simultaneous removal of the root and bud of water chestnut, according to the previous peeling, such as corn stalk, cotton, and sugarcane. The positions of the water chestnut were limited by the weight of the water chestnut and the structure of the device. The rotating blade group was also developed to simultaneously cut the bud and root. The non-clamping design was adopted to simplify the working process for the high efficiency of the device. A theoretical analysis was made on the number of holes in the disk, the geometry, and position of the rotary blade, as well as the relationship between the rotary blade and the disk. As such, the water chestnut was fixed in the locating hole. A kind of saw-tooth rotary cutting tool was designed to reduce the cutting force of the bud and root of the water chestnut. The simulation and optimization experiments were carried out using EDEM software, particularly with the response surface test using the Box-Behnken principle. Design-Expert 13 software was also used to fit the experimental data for the quadratic regression. The optimal parameters were achieved as follows. The length of the saw tooth base was 5 mm, the tooth height was 15 mm, and the cutting speed was 0.103 0 m/s. A force measuring device was also selected to verify the reliability of the simulation under the best parameters. The results showed that there was a relative error of 13.3 % between the measured and the simulation. The errors were attributed to the simplified force measuring device was not identical to the experimental one, particularly for the structural differences, and there were inevitable precision errors in the process of machining and installation. More importantly, the optimization experiment of locating the hole was carried out for the clamped and fixed water chestnut, in order to reduce the oblique cut under the resultant force in the actual running. The BBD response surface test was carried out with the minimum difference of oblique height as the response value, while the location-hole parameters and cutting speed as the influencing factors. The optimal parameters were achieved, where the arc depth, arc pitch, and cutting speed were 8.23 mm, 5 mm, and 0.103 0 m/s, respectively. The predictive height difference of the model was 4.1 mm under the optimal parameters, indicating that the test result was consistent with the predicted. Consequently, the cutting force and the height difference of oblique cutting were tested to optimize the working parts of the device. The finding can also provide a strong reference for the cutting device of the root and bud of water chestnut.