Modeling of Deep Learning Applications for Chatter Detection in the Milling Process

This study introduced a preliminary investigation of the development of a Digital Twin (DT) model for the milling machining process for the chatter detection phenomenon. Subsequently, chatter has a dynamic interaction in which there is an unstable condition in the material removal process between the cutting tool and work-piece, leading to a decline in surface roughness and tool life, ultimately reducing the quality of machining output. Therefore, this study aimed to develop a chatter detection model using a deep learning application that can identify stable or unstable chatter. The model was built based on the data-driven method where vibration signal data from the milling process is used to train and test various supervised deep learning methods. The result showed that a model with a good level of accuracy was built, and with the help of a chatter detection application, regular operator staff can monitor the machining conditions when no specialist is available.