Monitoring a deep hole drilling process by nonlinear time series modeling

Deep hole drilling methods are used for producing holes with a high length-to-diameter ratio, good surface finish and straightness. The process is subject to dynamic disturbances that is classified as either chatter vibration or spiralling. In this work, nonlinear time series modeling is used to setup an on-line modeling approach of the time varying dynamics of the process. An on-line monitoring strategy, based on control charts, is formed to detect chatter vibration. The results show that the proposed modeling approach provides an on-line procedure that can answer questions about the time varying dynamics of the process. The proposed on-line monitoring strategy can detect the start of the transition from stable drilling to chatter vibration and some alarm signals are related to changing physical conditions of the process. r 2008 Elsevier Ltd. All rights reserved.