Research On Tool Wear Prediction And Compensation Methods In Die-sinking EDM

As one of the advanced manufacturing technologies, electrical discharge machining (EDM) plays an important role in the fields of precision molds, aerospace, medical devices, sensors and so on, for its unique non-contact characteristic and the wide adaptability of material. However, due to the unavoidable tool electrode wear, the machining accuracy, especially for the micro hole and cavity, is reduced seriously. So, an in-depth research focusing on the tool wear in die-sinking EDM is carried out in this paper, both prediction and compensation methods are discussed widely. The main contents are as follows:According to the basic principle of EDM, this paper proposes a tool wear prediction method based on spark discharge simulation for micro electrodes. This method reproduces the processes of tool wear and workpiece material removal, as well as feed movement, by the simulation of the electric erosion in EDM. Therefore a simulation model is established, used to predict tool wear result and morphology features of machined surface. To make the simulation more accurate, cone cave shaped crater is used in this model and the shape of it is adaptive deformable according to different discharge position, which makes the model as close to the actual situation as possible. The experiments show that this model can simulate the micro-hole machining well and obtain accurate tool wear prediction results.However, for the electrode with regular size, the simulation method will produce large amount of calculation, and compared to micro electrode, the accuracy requirement is relatively low. So, this paper actively explores the possibility of tool wear prediction based on the geometric characteristics of the electrodes, and a prediction model based on the surface curvature is built. Research shows that the larger the curvature, the greater the electric field intensity, and the more easy to produce a spark discharge, which results in tool wear. So it is possible to realize the fast wear prediction of curved electrode by obtaining the offset surface of the electrode according to the curvature. In this paper, a NURBS surface is used to approximate the electrode surface or the offset electrode surface and the way how to obtain the NURBS surface will be discussed in detail, especially the calculation of generalized wear coefficient δp and wear transformation matrix K. Verified by the experiments, this method can well predict ellipsoid electrode wear results.In order to reduce or even remove the bad influence of tool wear and improve EDM precision, this paper studies the tool wear compensation methods of EDM on the basis of tool wear prediction models. A reverse-forming compensation model for micro electrode and a surface vector offset compensation model for curved electrode are proposed. These two models respectively use the reverse-forming principle and the general surface offset principle to realize the specific deformation of the initial electrode shape, and as a result, the electrode can just recover into ideal desired shape at the end of the actual machining, then exact copied to the workpiece cavity, in which way realizes the tool wear pre-compensation.Finally, an experimental platform is designed and corresponding experiments are conducted to verify the accuracy and effectiveness of the tool wear prediction and compensation models proposed above. Certainly, error analysis for all models is made to improve the prediction and compensation accuracy.