Chatter Simulation and Detection in CNC Milling

A key feature to maintaining a safe and efficient machining operation is the avoidance of unstable vibrations of the cutting tool, commonly called chatter. This thesis explores chatter simulation and a new chatter detection algorithm developed for use in a CNC milling machine. This algorithm, based on a once per revolution sampling method, samples forces experienced during the milling process, and calculates the variance of the differences between consecutive samples. As a selected level of variance, chatter is assumed. An existing milling simulation program is modified to produce a stability lobe diagram for a given CNC machine, workpiece material, and cutting tool. Stable, unstable, and marginally stable cuts are indicated as a function of spindle speed and axial depth of cut. The chatter detection algorithm is initially verified by simulation and then added to the simulation to auto-generate stability lobe diagrams. A collection of experimental aluminum cuts is run to collect force data that can be analyzed by the detection algorithm and compared to simulation results. Experimental cut stability is determined by observation of the noise, force, and cut surface. Comparing algorithm results to observed results shows the effectiveness of the algorithm in distinguishing between stable and unstable cuts. However, further testing is needed, particularly in determining the variance threshold of the algorithm. The simulation and algorithm are also used to explore the effect of system parameters; specifically, spring constant (k) and damping ratio (ζ). This exploration shows there being a strong connection between the maximum attainable stable axial depth and the parameters