Statistical Identification of the Global Resistive Coefficient in Automatic Machines

This paper addresses the experimental identification of the global resistive coefficient in automatic machinery, with reference to an industrial wrapping machine. The proposed procedure is based on cycle-averaged energy balance under periodic operation. Motor torque and angular speed are acquired during repeated cycles and used to compute the mechanical energy exchanged over a complete working cycle. Since the net variation of conservative energy over a closed cycle is negligible in steady periodic conditions, the cycle energy is attributed to dissipative phenomena. A lumped resistive model is adopted to represent losses as the superposition of a speed-dependent (viscous) term and a direction-dependent dry-friction term. The global resistive coefficient is first evaluated from cycle quantities derived from the measured speed profile. Subsequently, the viscous and dry-friction contributions are estimated through linear regression exploiting datasets collected over multiple cycles and different operating speeds. Results on the case-study machine show stable and repeatable parameter estimates, enabling a concise characterization of the overall dissipation. The identified parameters are suitable for dynamic modelling, efficiency assessment, and condition monitoring of automatic packaging machines.