Detecting nonlinearities in time series of machining processes

For analysis and supervision of machining processes it is desirable to select a small set of parameter that are suitable to characterize the process. In order to reduce the number of characteristic process parameter one should use nonlinear or linear analysis methods whichever are more appropriate. Detailed analysis on the basis of prediction models and deterministic versus stochastic plots can thus help to develop and select successful monitoring and control strategies. Therefore, investigations on cutting processes with methods of nonlinear time series analysis are made under two aspects. The first aspect is to find out, if nonlinear dynamics components are an essential part of the process. The second aim is to condense and reinforce information that describes the state of the process and the tools used. Measurement signals taken from turning and milling processes are then analyzed under these aspects. The methods described have already proven to be useful for monitoring tasks.