To address uneven coating and agglomeration in wet pelleting of small ellipsoidal rapeseeds, this study proposes a cross-scale paradigm integrating high-fidelity discrete element modeling with response surface optimization. A non-spherical multi-sphere rapeseed model was constructed, and a hemispherical disturbance baffle at the pan bottom was designed to eliminate the central mixing dead zone. A two-stage “Hertz-Mindlin (dry mixing) + JKR adhesion (wet coating)” simulation strategy was adopted. Results show that optimal pan curvature of 35 mm and rotational speed of 250 rpm trigger stable cascading flow, increasing Lacey index to 0.96 and reducing coating uniformity coefficient of variation (CV) to 18.0%. Box-Behnken optimization yields global optimal parameters: 250 rpm, 35 mm curvature, and 15.0 mL equivalent liquid volume. Physical bench tests achieve 92.77% coating qualification rate, with only 0.46% relative error from model prediction. This study reveals the interplay between kinetic energy and adhesion force in multiphase coating flow, providing a theoretical basis for coating equipment design for small irregular seeds.
Zhu et al. (Fri,) studied this question.