Development And Application Of A Virtual Simulation System For Peripheral Milling

Peripheral milling is broadly used for the manufacture of profiled components in aerospace, automotive and mould/die industries. The mechanism of milling process is very complex due to occurrence of the periodical and intermittent cutting process, semi-closed-form machining and chip thickness variation. The periodical cutting force excites vibration between the cutter and workpiece. Under certain conditions, the vibration with significant amplitude, or the so-called self-excited chatter, occurs due to the interaction between the cutter and workpiece. Unless avoided, machining with the presence of chatter leads to poor surface finish, low productivity, excessive tool wear even breakage or damage of machine.Cutting parameters in peripheral milling are determined usually based on either experience or reference handbooks rather than a mature virtual simulation system of milling process. As a result, sometimes it leads to low metal removal rate, low productivity and even high cost for the avoidance of chatter and its influence. To solve the problem, it is recommendable to develop an effective and practical virtual simulation system for peripheral milling on the basis of a systemic investigation on the machining dynamics. The system should be capable to demonstrate the machining dynamics of the milling process and predict the vibration between the cutter and workpiece under different cutting parameters and cutter dimensions, and to be used for optimization of the above parameters, such as spindle speed, radial and axial depth of cut, feedrate and cutter dimensions etc.Based on machining dynamics analysis, a virtual simulation system for peripheral milling(VSSPM) is hereby designed and developed using Matlab/Simulink. The main contents and conclusions of this thesis include:1. An improved practical mathematical machining dynamics model is developed with consideration of regenerative chatter theory. In particular, the size effect of uncut chip thickness, the influence of the effective rake angle and cutter runout that are usually missed in most existing models of the same sort, are included in the proposed model.2. Based on the proposed machining dynamics model, a computer simulation model for peripheral milling in time domain using Matlab/Simulink is developed byapplying variable-step numerical integral algorithm (fourth-order Runge-Kutta formula) and recurrence algorithm.3. Integrating experimental modal analysis and ARMAX model, research on system identification and modal analysis theory are conducted on establishment of transfer function models of the relevant milling vibration systems. The transfer function models of the milling vibration systems are presented consequently based on ARMAX identification model. Verified by many milling dynamics experiments and modal analysis tests, accuracy on predicting the vibration of VSSPM based on ARMAX identification model that reflects the dynamic characters of milling process is far better than that of the model based on traditional modal analysis tests.4. A series of computer simulation on the dynamics of milling process, including the dynamic forces and the vibration between the cutter and workpiece are conducted. The visual simulation results that exhibit the same trends as those obtained in actual machining, including dynamic cutting force and vibratory displacement between the cutter and workpiece and their power spectral density, demonstrate the accurate estimating capability of the proposed VSSPM and can be further used to analyze characters of milling process in time and frequency domain.5. Dynamics and phase characters of milling process are discussed on the basis of the analysis of the proposed VSSPM and energy consumed during milling. The optimal process parameters of milling and cutter geometric parameters are designed according to the minimum of amplitude of relative vibratory displacement in cutter-workpiece system.6. An optimal artificial neural networks (ANN) model with lower prediction errors and better performance function is proposed by adopting cu...