Chun Li Lei

Thermal Error Modeling and Compensating of Motorized Spindle Based on Improved Neural Network

In a lot of factors, thermal deformation of motorized high-speed spindle is a key factor affecting the manufacturing accuracy of machine tool. In order to reduce the thermal errors, the reasons and influence factors are analyzed. A thermal error model, that considers the effect of thermodynamics and speed on the thermal deformation, is proposed by using genetic algorithm-based radial basis function neural network. The improved neural network has been trained and tested, then a thermal error compensation system based on this model is established to compensate thermal deformation. The experiment results show that there is a 79% decrease in motorized spindle errors and this model has high accuracy.

Modeling and Simulation for the Feed System of CNC Machine

Study the linear and nonlinear stiffness dynamic characteristics of ball screw in feed system. According to the structure and the stiffness of ball screw, considering the influence of damping force, elastic force, friction force, driving force and load, establish the dynamic model of feed system. Use Linz Ted- Poincare (L-P) Method of singular perturbation to solve the model, obtain the quadratic approximate solution of the free vibration, analyze the multiple solution phenomenon of the model, and also conduct the numerical simulation analysis for the model.

Research on the Cooling System of High-Speed Motorized Spindle

In order to control effectively the temperature of the motorized spindle, based on thermodynamics, heat transfer theory and fluid dynamics control theory, the model of motorized spindle with cooling system is established and simulated. Based on the idea of orthogonal experiment and simulation experiment, the comprehensive tests are built, and the optimum matching relation between the heat flux of motor and the flow velocity of cooling liquid is determined in this article. The results show that the flow velocity of coolant can be adjusted according to the heat flux of motor which can control the temperature in the steady range and improves the cooling effect.

A Study of High-Speed Angular Contact Ball Bearing Thermo-Mechanical Coupling Characteristics

Heat generation and deformation of bearing are key factors that influence the rigidity and machining accuracy of the high-speed precision spindle system. Based on heat transfer and thermodynamics, the finite element model of angular contact ball bearing is established for thermal deformation. The contact stress and thermal deformation are analyzed and obtained at a speed of . The results show that the maximum contact stress and thermal deflection appeared at contact region, which is in accordance with actual status. The results provide the reference and the theory basis for research into thermal deformation of bearing.

Structural Dynamic Analysis and Optimization of High-Speed Precision Machining Center Column

The theory of structural dynamic analysis is put forward firstly. Then the dynamic performance of HMC80 high-speed precision machining center column is analyzed by means of finite element method. The dynamic performance of the column is analyzed using structural dynamic optimization theory and variation analysis with the thickness of wall plate and the inner rib plates as the parameters thirdly. Based on the analysis results, the structure optimization scheme of the column is obtained. The structure optimization scheme is analyzed and the analysis results show that the dynamic performance of the column optimization scheme is improved obviously.

The Influence and Suppression of Nonlinear Friction

Direct at NC machine tool feed system, the article analyzed the influence of the nonlinear friction at each motion joint, and points out that the nonlinear friction are the main factors that influence the positioning accuracy of the feed system. The article have discussed the methods for effectively compensation and controlling of the nonlinear friction respectively from two aspects of flutter compensation and predictly controlling the nonlinear friction, and proved the correctness of the compensation and control method that is proposed in this paper.

Thermal Error Robust Modeling for High-Speed Motorized Spindle

To improve the manufacturing accuracy of NC machine tool, the thermal error model based on multivariate autoregressive method for a motorized high speed spindle is developed. The proposed model takes into account influences of the previous temperature rise and thermal deformation (input variables) on the thermal error (output variables). The linear trends of observed series are eliminated by the first difference. The order of multivariate autoregressive (MVAR) model is selected by using Akaike information criterion. The coefficients of the MVAR model are determined by the least square method. The established MVAR model is then used to forecast the thermal error and the experimental results have shown the validity and robustness of this model.

Optimization of Measuring Points for High-Speed Motorized Spindle Thermal Error

According to the location and number of temperature measuring points of the motorized spindle thermal error, a new method for optimizing the locations of thermal key points is proposed. Firstly, temperature measuring points are divided into groups by using fuzzy clustering method. Secondly, grey correlation model is adopted to analyze emphasis of each measuring point to thermal deformation in temperature field distribution of motorized spindle. Finally, temperature measuring points have been optimally selected based on modified coefficient of determination. Comparing to the conclusion of the existed literature, the results show that this method is feasibility and validity. The method can reduce the temperature variables and modeling time, and supply the theoretic support for the engineering experience.

Comparison of Forecasting Methods for Thermal Error on High-Speed Motorized Spindle

In order to reduce the thermal error of the motorized spindle and improve the manufacturing accuracy of NC machine tool, the thermal error forecasting models based on multivariate autoregressive (MVAR) method and genetic radial basis function (GARBF) neural network method are proposed, respectively. According to different representations of generation mechanism of motorized spindle thermal deformation, operation efficiency and curve fit precision of these two models are compared. The studied results show that under the same temperature rise variable conditions, MVAR model and GARBF neural network model have almost the same convergence and operation time and relative errors of two models are less than 3%. The results also show that the MVAR model has higher forecast precision in the prediction former stages; in contrast, the GARBF neural network model has higher forecast precision in the latter stages.

Application of ANN Back-Propagation for Residual Stress in an Alloy Reinforced Ceramics/Metal Composite

Artificial neural network (ANN) back-propagation model was developed to predict the thermal expansion behavior and internal residual strains in reinforced ceramic matrix composites (CMCS).The ANN training model has been used to predict the thermal expansion behavior and internal residual strains, exhibiting excellent comparison with the experimental results. It was concluded that predicted thermal expansion behavior and internal residual strains by the trained neural network model seem more reasonable compared to approximate methods. It is possible to claim that, ANN is fairly promising prediction technique if properly used. Training ANN model was introduced at first. And then the neural network architecture is designed. The performance of system is summarized at last. In order to facilitate the comparisons of predicted values, the error evaluation and mean relative error are obtained. The result shows that the training model has good performance, and the experimental data and predicted data from ANN are in good coherence.