Analysis, prediction and control of machining dynamics applied to turning processes

The study focuses on the analysis prediction and control of turning dynamics, which impose considerable limitations in machining performance. Two important process outputs are considered, productivity and surface roughness of the workpiece. A two-step approach has been developed to improve machining performance. First, the set points of the controllable process inputs are determined from a powerful predictive theory of turning dynamics. Secondly, an on-line optimization procedure is employed to adjust the process inputs about these set points to further improve machining performance. The theory of turning dynamics developed considers fully the flexibility of both the workpiece and the machine tool structure, the dynamics of which are determined using a comprehensive experimental procedure. Furthermore, the linear model of the cutting process has been enhanced by including previously unmodeled phenomena such as interference forces at the tool-workpiece interface and the variation of the depth of cut due to the relative tool-workpiece motion. The predictive theory has been successfully validated through cutting experiments at various cutting configurations and conditions. The problem of improving the performance of the machining process was formulated in an optimization framework. The objective is to improve productivity quantified by material removal rate and minimize the mean amplitude of the thrust force by adjusting the feedrate and spindle speed subject to chatter constraints, productivity limits as well as the cutting speed, and feedrate bounds. An on-line numerical optimization scheme based on Evolutionary Operation (EVOP) was employed to solve this problem. The control scheme was tested experimentally and was found to be successful in moving the machining process from a chatter state to a stable machining state while improving the material removal rate.