Design and experimental investigation of critical components for a multi-functional cassava harvesting machine

To address the current challenges faced by cassava harvesters—including low intact tuber yield, high tuber damage rates, and insufficient automation—this study proposes a design for a multifunctional cassava harvester and conducts experimental evaluations. The machine integrates multiple functions, including digging, conveying, tuber-soil separation, collection, and hydraulic unloading, thereby achieving a complete mechanized workflow. Additionally, based on the physical and mechanical properties of the cassava variety “South China No. 9” and the typical agronomic conditions of Hainan Province, structural designs and analyses were conducted for key components such as the biomimetic digging blade, the chain-type cassava-soil separation device, the separating roller mechanism, and the hydraulic collection and unloading system. The biomimetic digging blade draws inspiration from the shape of a mole's claws, featuring a curved convex surface and an alternating arrangement of long and short teeth. This design reduces soil adhesion and clogging while improving soil penetration and conveying efficiency. The tuber-soil separation device employs a lifting chain arranged at a 26° angle using inclined rods, along with a three-head vibrating wheel operating at a frequency of 10 Hz and featuring an adjustable amplitude of 30 mm with a ±10 mm adjustment range to enhance soil removal while minimizing impact damage to the tubers. A beater roller with three rows of rubber-coated fingers, spaced 60 mm apart and with 18 fingers per row, was designed to gently transfer tubers into the collection box, thereby reducing mechanical damage. The 1.056 m³ collection bin is equipped with a tilting mechanism controlled by a hydraulic cylinder to enable efficient unloading. To evaluate the performance of the proposed machine, field trials were conducted at a cassava plantation in Danzhou City, Hainan Province, using a three-factor, three-level Box–Behnken experimental design. The cassava digging rate, the cassava injuring rate and the cassava-to-soil mass ratio were selected as evaluation indicators, while machine forward speed, digging blade insertion depth, and the conveying line speed of the tuber-soil separation device served as independent variables. Response surface methodology was applied to establish quadratic polynomial regression models for the three evaluation indicators, and the significance of each factor and their interactions was analyzed using analysis of variance. The results indicated that all three models were highly significant and contained no significant model misfit terms, indicating good model fit. Optimization was performed using Design-Expert software, yielding the following optimal operating parameters: forward speed of 1 m/s, digging depth of 224 mm, and conveyor line speed of 1.3 m/s. Under these conditions, the predicted visible tuber rate reached 97.5%, the damaged tuber rate was 2.07%, and the tuber-to-soil mass ratio was 68.9%. In five replicate validation tests conducted under the optimized parameters, the average values for the visible tuber rate, damaged tuber rate, and tuber-to-soil mass ratio were 96.86%, 3.03%, and 67.72%, respectively. The relative errors between the theoretical results and experimental results for the percentage of sound tubers, the percentage of damaged tubers, and the tuber-to-soil mass ratio were all within 5%, indicating that the optimization results are reasonable. This study provides a reference for the design and optimization of multifunctional cassava harvesting equipment.