Development of a power prediction model for rotary tillage in paddy loam field using DEM-MBD coupling method

• DEM–MBD coupling predicted PTO power (93.9–87.8%) and drawbar power (90.2–77.4%) accuracy. • 3D reverse engineering of rotary tiller and EEPA contact model enhanced soil–tool interaction fidelity. • Dynamic Time Warping quantified temporal alignment with normalized distances between 0.022 and 0.376. • Framework enables rapid optimization of PTO-driven implement design and power source selection. This study applied a DEM–MBD coupling method to predict the PTO and drawbar power requirements of a rotary tiller and evaluated the predictions against field experiments. The implement geometry was reverse‑engineered via 3D scanning and simulated in EDEM and RecurDyn to model soil–machine interactions. Two operating cases—rotary tillage without prior primary tillage and with prior primary tillage—were examined to predict PTO torque and draft force. Field data recorded at 1 kHz were down‑sampled to 50 Hz for direct comparison. The mean prediction accuracies were 93.9 % and 87.8 % for PTO power and 90.2 % and 77.4 % for drawbar power in the first and second scenarios, respectively. Dynamic Time Warping analysis was also performed to quantify the temporal alignment of dynamic load patterns. These findings offer a practical foundation for power source selection and design optimization of PTO‑driven agricultural implements. While the study focused on a specific sandy clay loam soil (46 % sand, 34 % silt, and 20 % clay), the proposed method shows potential for broader applicability, which should be verified under varying soil and operational conditions.