On-Line Identification of End Milling Cutter Runout

Cutter runout has been a target for monitoring and control of machining processes in view of the constraint it places on the achievable productivity. Off-line metrology based on various displacement probes such as dial indicators or proximity sensors provides information regarding the runout characteristics in a non-cutting state. However, during the actual process of machining off-line calibrations often become irrelevant since the cutting parameters and machining configuration significantly affect the behavior of runout.

This paper presents a methodology of in-process identification of cutter runout in end milling based on the analysis of cutting forces. The presence of cutter runout generates cutting force components at one spindle frequency above and below the tooth passing frequency. Through the convolution integration of local chip load the magnitude and angular orientation of cutter runout can be estimated from a frequency domain explicit formula given the in-process measurement of forces. With the use of a windowed discrete Fourier transform the scheme is implemented in an adaptive manner for continuous runout identification. In this paper the analytical basis of the method is formulated, the numerical scheme discussed, experimental set-up described, and resulting data presented in illustration of the methodology.