On-line chatter detection and control in boring based on an electrorheological fluid

For on-line chatter avoidance in boring, this paper presents an adaptive co ntrol system which, based on a novel design of the boring bar containing an Electrorheological (ER) fluid, automatically adjusts the dynamics of the b ar according to the information from the monitored sensor signal. For detec ting chatter rapidly and accurately, artificial neural networks are used to decontaminate the sensor signal and recognize the chatter omen. Once chatt er is detected, a strength regulation strategy of the electric field applie d to the ER fluid in the bar is used to suppress chatter. When a certain el ectric field is applied to the ER fluid in the bar it behaves as a viscoela stic spring with nonlinear vibration characteristics. Based on the nonlinea r vibration characteristics of the ER fluid, the strength regulation strate gy is to tune the electric field strength so that the dynamics of the bar i s sensitive to the minute change of the vibration amplitude. Experiments sh ow that chatter in boring can be avoided by using this control system. (C) 2001 Elsevier Science Ltd. All rights reserved.