| Aero-engine blade is the core part of the aircraft with high precision.Its machining profile precision and surface quality have great influence on its aerodynamic performance and fatigue life in service environment.How to improve the machining accuracy and surface quality of aircraft blades has always been the focus and difficulty in the field of aviation manufacturing.At present,belt grinding has become a final processing method to ensure the accuracy and surface quality of blade profile,but its characteristics of flexible grinding make it difficult to ensure the precise removal of blade materials.In view of the problem of poor precision and uneven allowance distribution of blades after pre-milling process,belt grinding process cannot guarantee the final dimensional accuracy of blades by controlling the geometric position of cutting tools.Therefore,the machining accuracy prediction of aeroengine compressed air blade is studied in this paper to improve the machining quality of blade milling.The main research work of this paper is as follows:(1)Formulation of machining accuracy prediction scheme for sailing blade milling.Firstly,the machining technology of the aero-hair blade is analyzed and the evaluation index of machining accuracy of the blade milling is determined.Then,a prediction scheme for machining accuracy of blade milling was developed,and a data acquisition system was built based on the scheme.At last,the data processing flow which mainly includes data denoising,data normalization and data set enhancement is established.(2)Establish the prediction model of milling profile accuracy.Firstly,the influence of servo following error on the machining profile accuracy was analyzed,and the following error of shaft position and shaft velocity were determined as the main factors affecting the machining profile accuracy under the normal machining condition.Then the milling experiment was set up and the machine tool following error in the experiment was collected.Finally,the prediction model of contour accuracy is established based on neural network.(3)Establish the prediction model of milling surface quality.Firstly,the milling force model and the milling system dynamics model were established,based on which the internal relations among milling force,spindle vibration and surface quality were explored.Then the milling experiment platform was built and the milling experiment was set for the curved parts and process parameters.The influence of process parameters on the cutting force,vibration and surface quality was analyzed.Finally,a prediction model of milling surface quality was established based on the machining state parameters.(4)Milling experiment and machining accuracy prediction model verification of navigation blades.Firstly,a milling experiment was set for the navigation blade,and the machine servo following error and machining state parameters were collected through the data acquisition system.Then the profile precision prediction model and the surface quality prediction model were used to predict the machining results of the blades.Finally,the validity and stability of the contour precision prediction model and the surface quality prediction model are verified by the performance of the model on the experimental data samples of blade milling,and the optimization and adjustment direction of the model is pointed out. |
