Research On Modeling And Processing Technology Of Micro-Burr In Micro-Milling

In recent years, due to micro-parts with the advantages of small dimensions, light weight, strong function etc., the areas of civil and defense products have grown an increasing demands on micro-parts. Micro-cutting technology with high efficiency, high flexibility and capacity of processing complex 3-D shapes and a variety of materials etc., become an important micro-parts manufacturing technology. As the size of micro-parts is much smaller than conventional components, therefore micro-milling surface quality of the micro-parts has a great impact on performance. Micro-burr as the product of micro-machining, the relative size ratio of micro-burr to work-piece is much greater than conventional cutting conditions, almost equal to the feature size of the work-piece. At the same time, it is relatively difficult to remove the micro-burr. So it is of great importance to study the influence factors on micro-burr, in order to control the surface quality of micro-parts. This paper mainly covers the following three aspects:Firstly, the dynamics model of micro milling is established, by taking into account of the stiffness and radius of micro-cutter, micro-milling cutting thickness which has an essential impact on the burr height. The simulation model is established by using MATLAB Simulink software modules to simulate the impact of dynamic characteristics of micro-milling system on cutting thickness, the path of cutter center and cutting force.Secondly, lead brass and aluminum materials are used in micro-milling experiments. Single-factor method is used to study the influence of depth of cut, feed rate, spindle speed and tool overhang amount on the micro-burr. Micro-burr shapes are studied under different processing parameters. Also, we gain the variation regularity of the micro-burr formation.Finally, by anglicizing the micro milling process, the micro-milling burr prediction model - gray neural network model is established under the impact of single factor. Then the prediction model is tested using experimental data to verify its availability.