Performance Evaluation Method of CNC Units using Torque Curve Patterns

This paper proposes a new performance evaluation method of CNC units to predict the control ability of cutting process, and to localize weak points of CNC functions in complex machine tools. The proposed evaluation method is based on the eight representative torque curve patterns, which are artificially generated corresponding to real machine tools operation and cutting conditions. In order to evaluate CNC units, generally electrical output pulses of CNC units by position command in NC program are counted to compare with theoretical pulse number. But this method is non-realistic and problematic, because this evaluation process is not affected by several physical loads of machine tools. In an evaluation of CNC units, cutting process must be included to consider feedback responses of CNC units and motor amplifier. In the proposed method, the newly introduced velocity error evaluation with eight torque patterns, the step response, and impact response can provide detail information on total control ability of CNC units. The direct application of torque curve patterns without real machine tools has merits such as the static, dynamic and thermal noise loads in machine tools are removed from the evaluation process. In addition, the proposed test method can be helpful in deciding the optimal control parameters in the CNC units to improve cutting surface quality and form accuracy of workpiece in short time.