On-Line Identification of End Milling Process Parameters

Abstract

A method for on-line identification of process parameters relating the feedrate to the machining force in end milling operations is described here. Such on-line identification of changing process parameters is necessary for parameter adaptive force feedback control systems which use feedrate manipulation to maintain a specified cutting force in the presence of varying machining conditions. A simple model form for the complex process mechanics is preferred for the adaptive control application and is proposed here. Accurate low order models for the end milling process are obtained by using the measured feed drive velocity and the machining force as the input and output signals, respectively, and by choosing a sampling interval equal to the tooth delay period. An identification scheme based on the recursive minimization of the sum of the squares of the modeling error is used here, with special emphasis on techniques to keep the estimation scheme alert while avoiding oscillatory parameter estimates. The effectiveness of the proposed process model and estimation scheme under open loop conditions is confirmed by simulation and experimental tests involving changes in the axial and radial depths of cut and in the feedrate. The parameter estimates generated are accurate and responsive to the changing cutting conditions and are steady and accurate under constant cutting conditions because of the simplicity of the proposed model. The estimation scheme described here accommodates the effects of cutter runout, tilt, and interrupted cutting by the multiple flutes more readily and with less detriment to parameter estimates than other schemes reported in the literature. The effectiveness of the estimation scheme under closed loop conditions is also verified by simulation.