Research On CNC System And Breakthrough Detection For High Speed Small Hole EDM Drilling

As one of the electrical discharge machining(EDM)approaches,high speed small hole EDM drilling technology provides the unique capability of drilling ultra-high aspect ratio small hole and has the characteristics of no cutting force,high machining precision and high processing efficiency.High speed small hole EDM drilling is suitable for the manufacture of superalloys,titanium alloys and other hard-to-cut materials.For example,it is commonly used in processing cooling film holes of turbine blades in aircraft engine and gas turbine.CNC(computer numerical control)system and breakthrough detection play important roles in high speed small hole EDM drilling.In order to solve the breakthrough detection problem and improve machining quality and efficiency in high speed small hole EDM drilling,this dissertation investigates several topics,including CNC system,gap discharge status detection,electrode wear prediction and breakthrough detection?In this dissertation,a CNC system for high speed small hole EDM drilling was developed based on the embedded platform and internet.The CNC system consists of several functions such as graphical user interface(GUI),code interpretation and interpolation,motion control,status monitoring and remote monitoring to meet the requirements of intelligent manufacturing on CNC system.After analyzing the gap discharge wave forms collected during the machining,on the basis of microcontroller unit(MCU)and hardware circuits,a gap discharge status monitor system was developed to identify the discharge status of each single discharge pulse.Due to its good performance in data classification and regression analysis,support vector machine(SVM)regression algorithm and classification algorithm are used to build up the electrode wear prediction model and the breakthrough detection model to obtain accurate electrode position and solve the breakthrough detection problem.The input and output of the electrode wear prediction model and breakthrough detection model are chosen by the experimental data analysis.On the basis of training sample,the support vector regression(SVR)model of electrode wear prediction and the support vector classification(SVC)model of breakthrough detection are achieved,and a penetration detection system is constructed for online prediction and identification during machining.The electrode wear on-line prediction results show that the prediction error of electrode wear prediction model is less than 0.04 mm.Experiments of breakthrough detection show that the system can accurately identify the moment when the electrode penetrates the workpiece,and the identification accuracy is 100%.