Electro-discharge Precision Truing And Oxide Film States Identification Of Wheel In ELID Grinding Process

ELID technique is one of the ultraprecision grinding methods that has been developed in recent years. This dissertation focuses on the research on the electro-discharge precision truing technology, the influencing factors of the oxide layer state of wheel, and how to characterize, identify and control the states of oxide layer.The electro-discharge precision truing of metal bonded diamond wheel is one of the key techniques in ELID grinding. In this dissertation, the result of pin electrode experiment reveals the action principle of the pulse frequency to the truing process, based on which the choice criteria of pulse frequency is determined. The electric parameters such as current, voltage, duty cycle and non-electric parameters of dielectric and revolution speed are studied and optimized. A real-time control system via a nano-servo device based on LabVIEW is designed and implemented. Using the method of combining ignition delay time detect and average pulse current in pulse discharging, the gap states can be accurately identified, and then the gap controlling and the on-line roundness predicting during truing process according to the arc rate and open circuit rate respectively are carried out, which leads to the increase of truing efficiency and accuracy.An intensive research on the influencing factors and generative mechanism of oxide layer is the key issue in studying oxide layer. In this dissertation the chemical and electrolytic influencing conditions of the oxide layer growth are studied and the properties of oxide layer are analyzed via orthogonal test. So the generative mechanism of oxid layer is revealed and described, the controlling strategies of oxide layer growing are established finally.The characterizing of oxide layer is the prerequisite of identifying and controlling the oxide layer states in ELID grinding process. In this dissertation the surface profile feature, components and growing characteristics in the simulated ELID grinding experiment are analyzed via the structure characterizing by SEM , the signal characterizing of oxide layer thickness with current and strength with grinding force are realized. Based on the above mentioned, two pattern recognition methods of based on fuzzy-neural network and ratio of Fn/I (normal force-to-current ratios) are presented and applied. The former is simulated based on Matlab and the simulation result is consistent with the oxide layer. Based on the latter the identification criterion is established.The two identification and control systems based on fuzzy-neural network and ratio of Fn/I are designed respectively, and the grinding experiments are implemented. The experiment results by active controlling oxide layer states show that the former can accurately identify the oxide layer states and realized the control strategy that is consistent with the states, the latter is implemented via the method that the value of Fn/I is used as the criterions of identifying and controlling the oxide states. Finally, high quality workpieces is obtained by the above two methods.