EXPERIMENTAL STUDY AND OPTIMIZATION OF A DISCRETE-ELEMENT–BASED MACADAMIA GREEN-HUSK EXTRUSION-PEELING MACHINE

This study established an integrated experimental–simulation framework for optimizing the green-husk dehusking process of macadamia nuts. The objective was to quantify the effects of drum speed, minimum gap, and feed rate on dehusking performance and nut integrity. The nut damage rate was determined by visual inspection, identifying visible cracks or fractures on the nut surface due to the brittleness of the green husk and the hardness of the nut. Mechanical characterization was conducted through static loading tests in three orthogonal orientations to determine the anisotropic fracture behavior of green husks. Discrete element simulations (EDEM) were calibrated using measured friction, restitution, and density parameters and validated through a repose-angle test. A Box–Behnken design with response surface methodology was applied to evaluate interactive effects among process parameters and derive an optimal operating region. The equivalent peeling force of green husks ranged from 0.30 to 2.00 kN. Drum speed was the dominant factor influencing dehusking efficiency (p < 0.01), while minimum gap and feed rate had weaker main effects. The validated model predicted a stable high-performance window at 300–400 r min⁻¹, 9–10 mm minimum gap, and 6–10 kg min⁻¹ feed rate. Under optimal conditions (400 r min⁻¹, 9.5 mm, 6.3 kg min⁻¹), the measured dehusking rate reached 95.4% with a nut damage rate of 4.76%. These results demonstrate that the combined experimental–DEM (Discrete Element Method) approach provides a reliable basis for parameter tuning and structural design of macadamia dehusking equipment, enabling improved efficiency and reduced nut damage