Non-linear elasto-dynamic modelling of slender rotors for the condition monitoring of automatic bar feeders

This thesis presents a comprehensive study of the elasto-dynamic behaviour of slender rotating bars, with specific focus on their application in bar feeder–lathe systems. Such systems often suffer from vibration-induced faults on the workpiece surface, when machining long flexible bars. The research aims to develop a physics-based framework capable of describing and mitigating dynamic instabilities that lead to such faults, through combined analytical modelling, numerical simulation, and experimental validation. A finite element model based on Timoshenko beam theory was developed to capture the effects of geometrical nonlinearity, rotating damping, and compliant constraints. These features enable the model to reproduce critical-speed resonances, sub-synchronous whirling, rubbing, and instability phenomena typical of slender rotors. Experimental Modal Analysis and machining tests were performed on a real bar feeder–lathe system under a wide range of operating conditions. The results identified two dominant mechanisms responsible for surface defects: synchronous resonance at critical speeds and asynchronous whirling driven instability fields. Model parameters were updated using experimental data, and the validated model successfully replicated the observed vibration patterns, instability fields, and consequent fault conditions. Furthermore, numerical simulations demonstrated the influence of shaft bow, nonlinear intermediate constraints, and damping on system stability, providing insights into the coupling between modes, nonlinear characteristics of the system and rotes to chaotic response in slender rotors. Finally, the work culminated in the development of an industrial diagnostic and simulation software, integrating the model through analytical and numerical tools for system analysis, fault prediction, and process optimization. The research thus fills the gap between fundamental rotor dynamics and practical industrial application in far feeders technology, offering both physical understanding and predictive capability for improving machining quality and reliability of such systems.