The Bonded Particle Model (BPM) provides a novel method for simulating chip ejection in wood machining. Besides the structural arrangement of discrete elements, the accuracy of such models relies on proper calibration of the model parameters. This work identifies critical model parameters using Design of Experiments methodology. Material parameters of spheres and bonds, as well as contact model are examined, to assess their effects on chip ejection dynamics. Key findings indicate that controlling the total number of broken bonds and the formation of realistic agglomerates is crucial, because agglomerate shape strongly influences their contact dynamics. The ratios of normal to tangential strength as well as strain to rupture ratios are useful parameters for calibration that mainly influence the median particle speed. The statistical analysis was unable to clearly identify the factors influencing the flight angle of the particles. Visually, the shape of the agglomerates formed also appears to have a significant influence. Overall, most BPM simulations show particle flight angles that nevertheless correspond sufficiently with the experimental data, while the velocity values deviate more strongly from each other. This work enhances understanding of BPM parameters and their impact on chip ejection, highlighting calibration challenges.