Process dynamics models for the control of end milling

Over the last 20 years researchers have attempted to apply a variety of advanced control approaches to the control of end milling operations. These control attempts include various adaptive control approaches, robust controllers, fuzzy and intelligent controllers. These attempts have met with varing degrees of success. Industrial acceptance of these approaches has been weak. The primary difficulties common to all of these control approaches is the development of a suitable model of the system dynamics including the dynamics of the cutting process. The end milling process dynamics include higher order dynamics as well as being nonlinear and non-stationary. For the purpose of controller design, we typically linearize these models and ignore the higher order terms. These simplifications make controller design easier but often result in poor controller performance. This paper considers the various models used for controller design. These models are evaluated based on comparing their predictions to experimental results. The various models are also assessed based on their appropriateness for various control approaches. Based on these evaluations a single unified model is proposed. The application of this model to various classes of controller design is also discussed.