Models of micro positioning and methods for surface grinding process control.

摘要

In order to achieve active form error control of the surface grinding process, a system model for the process should be established. Gao and Jones have developed a discrete control system model for a transverse grinding process. Not many researchers have investigated a discrete control system model for the surface grinding process. The model for the surface grinding process is presented and it is based on the physical laws associated with the process together with mathematical equations describing the workpiece surface form error profile and flatness error.;The system variables for the surface grinding control include required size reduction, nominal wheel infeed, nominal accumulated wheel infeed, wheel infeed position, and flatness error. Based on the model, the workpiece flatness error decreases when the grinding pass number increases.;The reduction of flatness error through form error control should be fully investigated after establishing a model for the discrete control system. In surface grinding, a high speed workpiece table motion is involved. In order to achieve active form error control for improving the machining accuracy, the control commands will include a substantial amount of high frequency signals that may not be sufficiently implemented by a wheel infeed system. A micro positioning system was designed as an additional unit for providing fast control actions. A composite control scheme was proposed for combining both systems. Models of the discrete control system for the surface grinding process controlled by the micro positioning actuating system, the wheel infeed system, and the composite control were established. The methods which reduce the flatness error effectively were identified.;For micro positioning systems with a piezoelectric actuator used for dynamic active surface grinding control, sinusoidal excitation at a high frequency would be used and would give additional problems in comparison with the commonly used step signals due to the hysteresis effects of the actuator, which require a good modeling approach. Comparisons between models of polynomial regressions with various orders of continuity including single, dual, zero order and first order multiple polynomials, direct mapping with polynomial fit, and direct mapping with numerical analysis were performed.