Assessing apple bruise susceptibility using the discrete element method

Abstract

Predicting apple bruise susceptibility is essential for minimising potential damage during postharvest handling. To assess the resistance of apples to mechanical damage, an apple discrete element model was developed at a tissue scale using the particle-filling modelling method. Quasistatic compression experiments, which were designed via the Box–Behnken design, helped determine the optimal simulation parameters. The model accuracy was confirmed by comparing the simulations with the physical tests, which revealed average relative errors of 3.52% for the yield point and 8.55% for the yield force. Subsequently, a dynamic collision test simulation model was established with the linear parallel bond model employed to represent internal particle interactions within the impactor, and a viscoelastic–plastic contact model was integrated to constrain the impactor's motion and limit the force range during loading. The dynamic collisions were simulated at different impactor velocities, and a novel method for calculating the damage volume and damage depth after impact by extracting the crack coordinates was proposed. The resulting damage volume and depth were validated using X-ray computed tomography scans. Finally, the apple's bruise resistance index (BRI) was calculated using data from the simulations, yielding a relative error of 3.27% for the BRI thresholds. Compared with physical tests, calculating the BRI through simulations can be expected to reduce the consumption of experimental materials and time.