A novel experimental-numerical approach to modeling machine tool dynamics for chatter stability prediction

Abstract

Chatter is one of the main limitations to milling performances. Prediction of such unstable phenomenon via stability lobe diagrams requires the measurement of the Frequency Response Functions (FRFs) for each tool and machine tool setup. This paper presents a hybrid FE-experimental approach to identify tool-tip FRFs with only one set of measurements, taking into account tool change without any other experimental test. Machine tool dynamics is modeled using a Finite Element (FE) approach. Machine, spindle and tool-holder are described by a lumped model characterized by frequency-dependent stiffness, while the tool is FE modeled. Lumped model and tool are connected by means of stiffness matrices extracted using the Craig-Bampton dynamic reduction method. The obtained simplified model of machine tool enables chatter prediction by means of stability lobe diagram for different tool without the need for extensive experimentation. Once a new tool is clamped no other measurements are needed, just the new tool FE model. Experimental validation under different conditions is provided, showing accuracy and reliability of proposed approach.