Vibration Simulation Of The Grinding Process And The Prediction Of Grinding Temperature

The development of machinery manufacturing is one of the major signs to an industrialized countries. It has to improve the machining accuracy, if a country wants to improve the level of mechanical manufacturing. Grinding is one of the most effective methods to improve the precision. The vibration and temperature during the grinding are the most important factor to the grinding precision.First, analyzing the process of grinding by the methods of dynamics. And the grinding system of the M2110A grinder has been modeled according to grinding principle. Then the system is simulated. The relationship between the nonlinear vibration system and the various kinds factors of the grinding has been found. The measures for weaking grinding vibration are given at the end. Providing a theoretical basis for the actual production. Then finding the factors which affect the grinding temperature by analysising the grinding process. A RBF neural network model for predicting grinding temperature has been established. And the forecast accuracy meet the expected target. Providing technical support for controlling the grinding temperature of the grinding process. The contents of this thesis include the aspects as follows:1. The internal grinding system has been analyzed according to the principle grinding. A five freedom degrees model of the Grinding System's vibration. Considering the impact of nonlinear stiffness factors of specific folder and wheel axis, a mathematical model has been established. Choosing the fourth-order Runge-Kutta method to solve the equation according to the features of non-linear equation. The nonlinear system has been simulated by using the MATLAB software, It has been found that how the nonlinear parameters effect the grinding's vibration. And the simulation results consistent with the actual. It provides theoretical basis for controlling the nonlinear vibration of grinding system.2. In order to further study how the nonlinear parameters effect the grinding's vibration. A single freedom degree vibration model of the workpiece and fixture system has been established. Choosing the multi-scale method for solving the nonlinear equation. The nonlinear vibration characteristics of the grinding process has been analyzed. It has been received how the nonlinear parameters effect the each step of grinding's resonance. And giving the measures to suppress the each step of resonance.3.The RBF neural network model of grinding temperature's prediction has been established by using Matlab neural network toolbox. The optimal network structure has been established by adjusting the spread coefficient of the newrb() function. The prediction achieve the desired accuracy, and the network has great generalization ability. Finally, finding the RBF's superiority by comparing the predictions with the traditional BP network model.In the end of paper, the author give a forecast to the development direction of the vibration rescarch of internal grinding and the application of neural networks in grinding area.