Electropolishing of A Spherical Surface and the Application of Fuzzy Inference

Various dies and molds involving the spherical surface demand a good surface fmish, in which case either long period of time or expensive equipment for the polishing work is needed. An alternative to the conventionally used polishing method is investigated in the current study. The reported electropolishing can improve the maximum surface roughness to the extent of one order of magnitude within ten seconds of total on-time, and it requires an inexpensive equipment only. The whole-bright condition throughout the polished surface is achieved. The effects of the processing parameters of initial gap width, the electrolyte flow rate, the electrical voltage and the on/off period are studied. The axial stroke motion in phase with the on/off period is found advantageous. For such a nonlinear system of electropolishing, the authors present the use of the fuzzy inference to predict the obtained surface fmish and to design the polishing parameters, where the commonly used linear regression method is of limited use.