Modeling and simplification methods for machine tool dynamics prediction in high speed milling

Unstable vibration phenomenon known as chatter is the most limiting factor to the performances of modern milling machine. Stability Lobe Diagram (SLD) is the main tool adopted to avoid chatter, and improve machine tool productivity, since it allows selecting optimal cutting parameters (spindle speed and engagement condition) to ensure a stable operation at the highest material removal rate. This chart can be obtained with simulation, thanks to predictive approaches, or directly by means of experimental tests. The aim of this thesis was to increase the reliability of Stability Lobe Diagram, and to propose and develop new identification techniques in order to support its industrial applications. Different methods have been analyzed both for chatter prediction, and experimental detection. In particular for prediction, research has been principally focused on the main inputs required: tool-tip Frequency Response Functions (FRFs) and cutting force coefficients. Machine tool dynamics has been investigated with the aim at developing methods to quickly and accurately identify tool-tip FRFs, key factor for a reliable chatter prediction. Both full Finite Element models of machine tool, and hybrid experimental-numerical methods have been analyzed and implemented, studying their application limits. On the other hand cutting speed dependence of cutting force coefficients has been investigated in order to improve their reliability in High Speed Milling (HSM). Moreover this work presents an experimental detection technique called Spindle Speed Ramp-up test. Thanks to this technique with few cutting tests Stability Lobe Diagram can be accurately identified without any approximation introduced by predictive approaches. All the proposed methods have been validated and critically discussed. The main goal of this Ph.D. thesis is to improve industrial application of vibrations prediction and detection approaches in milling, proposing simplified methods and easy-to- use systems. In order to do so an extensive critical analysis on advantages and drawbacks of different techniques is here presented.