An expert troubleshooting system for the milling process

Common problems experienced in milling processes include forced and chatter vibrations, tolerance violations, chipping and premature wear of the tools. This paper presents an expert system which attempts to troubleshoot the source of milling problems by utilising dynamics data coupled with the opinion of the operator and acoustic Fourier spectrum data taken from the cutting process. The expert system utilises a fuzzy logic based process to interpret the signals and information, and recommends possible alterations to the process to achieve high-performance milling operations.Specific inference engines were developed to assess the chatter stability, variation in cutting force coefficient, tool run-out and forced vibration characteristics of the system. Lastly, a stability lobe plot interpretation engine to automate the lobe selection process and recommend new, chatter free cutting conditions, was also developed. The chatter stability inference engine was tested with real cutting data, through acoustic measurements taken from various cutting conditions on an aluminium milling process. The chatter inference engine successfully determined the stability of the system for each sampled cutting condition. The robustness of the troubleshooting system depends on the accuracy of acoustic and frequency response measurements.